Every effort has been made to trace all copyright holders, but if any have

been inadvertently overlooked, the Publishers will be pleased to make the

necessary arrangements at the first opportunity.

Although every effort has been made to ensure that website addresses are

correct at time of going to press, Hodder Education cannot be held

responsible for the content of any website mentioned in this book. It is

sometimes possible to find a relocated web page by typing in the address of

the home page for a website in the URL window of your browser.

Hachette UK’s policy is to use papers that are natural, renewable and

recyclable products and made from wood grown in well-managed forests

and other controlled sources. The logging and manufacturing processes are

expected to conform to the environmental regulations of the country of

origin.

Orders: please contact Hachette UK Distribution, Hely Hutchinson Centre,

Milton Road, Didcot, Oxfordshire, OX11 7HH. Telephone: +44 (0)1235

827827. Email education@hachette.co.uk Lines are open from 9 a.m. to 5

p.m., Monday to Friday.

You can also order through our website: www.hoddereducation.co.uk

ISBN: 978 1 3983 2331 5

eISBN: 978 1 3983 2219 6

© David Bedford, Phil Chaffé, Tim Honeywill and Richard Lissaman 2021.

Selected STEP questions reproduced with the permission of Cambridge

Assessment Admissions Testing.

(Permission to use the material does not imply that Cambridge Assessment

Admissions Testing endorses the content of the book/resource.)

First published in 2021 by

Hodder Education,

An Hachette UK Company

Carmelite House

50 Victoria Embankment

London EC4Y 0DZ

www.hoddereducation.co.uk

Impression number 10 9 8 7 6 5 4 3 2 1

Year 2025 2024 2023 2022 2021

All rights reserved. Apart from any use permitted under UK copyright law,

no part of this publication may be reproduced or transmitted in any form or

by any means, electronic or mechanical, including photocopying and

recording, or held within any information storage and retrieval system,

without permission in writing from the publisher or under licence from the

Copyright Licensing Agency Limited. Further details of such licences (for

reprographic reproduction) may be obtained from the Copyright Licensing

Agency Limited, www.cla.co.uk

Cover photo © Mark Roper - stock.adobe.com

Illustrations by Integra Services Ltd.

Typeset in Pondicherry, India.

Printed in Italy.

A catalogue record for this title is available from the British Library.
Contents

How to use this book

Section 1: Getting started with problem solving

1 An introduction to the Mathematics Admissions Tests

2 Mathematical problem solving in admissions tests

Section 2: Problem solving in multiple choice questions

3 Introductory number and algebra techniques

4 Geometry and coordinate geometry

5 Functions

6 Calculus

7 Graph sketching, identification and transformation

8 Sequences and series

9 Trigonometry

10 Logic and proof

Section 3: Problem solving in extended questions

11 Getting started with longer questions

12 Number theory and combinatorics

13 Further trigonometry

14 Calculus – integration and differential equations

15 Complex numbers

16 Matrices

Answers
How to use this book

Hello and welcome to this book about the mathematics used in admissions

tests. It is designed to grow with you from the start of your post-16

mathematics studies. The content of the early chapters will be accessible to

you even if you have not covered much in your school studies to date. As

you progress through your school course, you will be able to engage with

more and more of the book. The later chapters require knowledge from most

of a two-year course, including some further mathematics. You should

attempt each chapter as soon as you feel confident that you have covered

enough in your school mathematics lessons.

The book is in three sections

Section one describes the main mathematics admissions tests, giving

information about the papers, how they are graded, their syllabus and some

of the universities that require students to sit an admissions test or will lower

their admission requirements for good results in an admissions test.

Admissions tests are meant to be difficult and place a high value on problem

solving skills. Section one concludes with some key techniques for

developing those skills.

Section two focuses on the MAT and TMUA examinations. Students who

will be sitting the STEP examination will also find this part useful,

particularly early on in their studies. The chapters in this section cover some

key topics that frequently occur in admissions tests. There are also worked

examples accompanied by explanations of the thought processes required to

solve them. You should read these examples carefully as there are many

useful techniques within them. There are questions to try as you go along as

well as exercises to reinforce your skills. Each chapter closes with some

questions that are written in the same style as the MAT and TMUA

questions.

Section three covers the longer questions that you find on the MAT papers

and the STEP papers. Chapters 11 to 13 are aimed at students preparing for

either the MAT or STEP examinations. Chapters 14 to 16 are aimed at

students preparing for the STEP papers. Chapter 11 is an introduction to the

algebraic skills required for admissions tests and forms the bridge between

the earlier MAT and TMUA chapters and the more advanced chapters at the

end of the book. Each chapter includes annotated examples and exercises

before concluding with a MAT or STEP style question.

If at any stage you find that you are struggling with any of the problems, put

them to one side and come back to them later. It is often a good idea to let

your ideas settle before coming back to an initially intractable problem.

What this book can and cannot do

There is a very wide variety of questions that can be asked in admissions

tests, even with the specifications aligning with A-level Mathematics and

Further Mathematics (in the case of STEP). It would be impossible to try to

cover everything that could be asked. For that reason, we have concentrated

on topics and questions that allow some key techniques to be strengthened

and for your problem solving skills to develop. The topics in part 3 were

carefully chosen to cover topics that are traditionally popular in the STEP

examinations but still tend to yield poor marks. Working through this book

will develop many of your calculation skills but, more importantly, it will

help you develop the confidence to tackle mathematical problems like those

on admissions papers. It is hoped that by the end of your studies you will be

able to look at the questions on admissions tests and think “that’s

interesting” rather than “oh no!”

There are short solutions to many of the problems at the back of the book.

More detailed solutions to exercises can be found on the website that

accompanies this book at www.hoddereducation.co.uk/step-mat-tmua-

answers. You will also find worked solutions to some additional questions

from past TMUA, MAT and STEP papers that support the topics covered in

this book.

We wish you every success in your future plans.

Phil Chaffé, David Bedford, Tim Honeywill and Richard Lissaman

Section 1 Getting started with problem solving
Chapter 1: An introduction to the Mathematics

Admissions Tests

TMUA MAT STEP

Introduction

Every year universities receive applications from students who will achieve

high grades in their A levels, International Baccalaureate, Higher or other

examinations. These applications are almost always accompanied by

excellent pupil statements and school references.

Yet for some institutions there are always far more applications than places.

Universities with high academic aspirations such as Oxford and Cambridge

therefore need some mechanism to identify those students who have the

potential to thrive on their undergraduate courses. They are looking for

individuals with excellent mathematical knowledge and skills, a high degree

of ability and talent and a strong motivation for studying mathematics.

In order to select students with these qualities, the Universities of Oxford

and Cambridge use admissions tests as part of their selection procedure. By

doing this they hope to ensure that they award places to the most suitable

candidates for their courses.

The University of Oxford uses the Mathematics Admissions Test (MAT) to

identify suitable candidates to invite for an interview. Offers are awarded

following the interview and are based on A level (or equivalent) grades. The

University of Cambridge invites students to interview based on their

application and awards offers (to those successful at interview) that include

a requirement for certain grades in the Sixth Term Examination Paper

(STEP) as well as their A level (or equivalent) grades.

Other universities use admissions tests as either a part of their offer or as a

means to give reduced offers. These are offers requiring lower A level (or

equivalent) grades than the standard offer. Some of these universities use

one or both of the MAT or STEP to do this and some use a different

admissions test, the Test of Mathematics for University Admission (TMUA).

Imperial College, London uses the MAT for all of its undergraduate

Mathematics course applicants. The University of Warwick has offers that

include the MAT, ones that include STEP, ones that include the TMUA and

ones that require no admissions test. Durham University reduces the A level
(or equivalent) requirements for students who achieve a good result in the

TMUA or the MAT.

It is not only mathematics undergraduate courses that may require

mathematics admissions tests. For Computer Science at the University of

Cambridge, the Cambridge TMUA (CTMUA) is used as a pre-interview

admissions test. The CTMUA is exactly the same as the TMUA. The

slightly different title is simply to denote that it is being used for admission

to a University of Cambridge course. A list of courses requiring or using

each admissions test can be found on the Admissions Testing Service

website.

The TMUA and CTMUA

Date

Late October/early November in Year 13.

Papers

Two multiple choice papers each of 75 minutes duration. The papers are

taken consecutively.

Questions

20 multiple choice questions for each paper. Candidates should attempt all

of these.

Grading

A combined score for both papers is awarded.

The lowest grade is 1.0 and the highest is 9.0

Grades for individual papers are recorded but these have no meaning in

terms of university applications.

Description

Paper 1 is the Mathematical Thinking paper and tests your ability to use

your mathematical knowledge to solve a variety of problems.

Paper 2 is the Mathematical Reasoning paper and tests your understanding

of ideas from elementary logic and reasoning. The ideas required for these

questions are described in chapters 9 and 10 of this book.

Syllabus

The mathematical knowledge required for both papers is almost all covered

within the pure mathematics specification of the first year of an AS level in

mathematics and a Higher Level GCSE mathematics course.

The most up-to-date version of the TMUA syllabus can be found on the

Admissions Testing Service website.

The topics covered from AS mathematics are: algebra and functions,

sequences and series, coordinate geometry in the (x, y) plane,

trigonometry, exponentials and logarithms, differentiation, integration, and

graphs of functions.

The topics covered from GCSE are number, algebra, geometry, measures,

statistics, and probability.

The mathematical reasoning required for paper 2 is in three parts; the logic

of arguments, mathematical proof and identifying errors in proofs.

Required/accepted

The University of Cambridge requires the CTMUA for Computer Science

undergraduate degrees. This is taken as a pre-interview test.

The TMUA is used by a growing number of universities as a means of

assessing mathematical talent. At the time of writing this book, the TMUA

was accepted by nine universities for a variety of Mathematics,

Mathematics with Statistics, Financial Mathematics, Data Science,

MORSE and Mathematics joint honours degrees. In all cases, a good score

in the TMUA will result in a lower A level (or equivalent) grade offer.
The most up-to-date list of courses accepting the TMUA can be found on

the Admission Testing Service website.

Additional notes

The MAT and TMUA typically take place on the same day at the same

time.

The TMUA examination is open to all who wish to sit it. There is no

requirement for a student to be applying for any particular course.

Students select which universities receive their TMUA results. This can be

done at any time after entering for the examination and until a few days

after the results are released. Once the universities have been selected, the

results are shared automatically with the selected universities. If no

universities are selected then the results will only be released to the

student.

For more selective universities a grade of 6.5 or higher is considered a

good score and will lead to a lower A level (or equivalent) offer. Durham

University advises all students achieving 4.5 or higher to release their

TMUA result as this will be viewed positively.

The MAT

Date

Late October/early November in Year 13.

Papers

One paper of 2 hour 30 minutes duration.

Questions

There are 7 questions on the paper of which the candidates should attempt

5.
All candidates answer question 1 which consists of 10 multiple choice

problems.

University of Oxford Mathematics or Mathematics and Philosophy or

Mathematics and Statistics, Imperial College and Warwick University

candidates also answer questions 2, 3, 4 and 5.

University of Oxford Mathematics and Computer Science candidates also

answer questions 2, 3, 5 and 6.

University of Oxford Computer Science or Computer Science and

Philosophy candidates also answer questions 2, 5, 6 and 7.

Grading

Each of the ten multiple choice questions is worth 4 marks (0 if incorrect)

and each of the remaining longer questions is worth 15 marks. The total for

the paper is 100 marks.

Students are awarded a mark out of 100.

Description

The MAT tests the depth of your mathematical understanding in pure

mathematics. It aims to be approachable, in terms of content, by all

mathematics students including those without A level Further Mathematics

and those from other educational systems.

Syllabus

The MAT is based on the first year of A level Mathematics, with a few

topics from the fourth term of A level Mathematics which you should have

covered by the time of the test.

The most up-to-date version of the MAT syllabus can be found on the

University of Oxford’s website on the Mathematical Institute page.

Required/accepted
The University of Oxford requires the MAT for Mathematics, Mathematics

and Computer Science, Mathematics and Philosophy, Mathematics and

Statistics, Computer Science and Computer Science and Philosophy.

For the University of Oxford the MAT is taken as a pre-interview test that

is used to decide who to invite for an interview.

Imperial College, London requires the MAT for Mathematics,

Mathematics with Statistics, Mathematics with Statistics for Finance,

Mathematics Optimisation and Statistics, Mathematics with Applied

Mathematics/Mathematical Physics and Mathematics with Mathematical

Computation.

For Imperial College the MAT forms a part of the general offer. Imperial

College rarely interview prospective students.

Some other universities will accept a good MAT score as evidence of

higher attainment and lower their offer accordingly. At the time of writing

this book there was no summary document available detailing universities

who lower the required A level grades based on a good MAT score. You

should check the entry requirements of each university you apply for to see

if there is any mention of the MAT.

Additional notes

The MAT and TMUA take place on the same day at the same time.

The MAT is only available to students applying for relevant courses at the

University of Oxford, Imperial College or the University of Warwick.

For students applying for both The University of Oxford and Imperial

College and/or the University of Warwick, e.g. Mathematics and Computer

Science at Oxford and Master of Mathematics at the University of

Warwick, the Oxford University course choice of questions takes priority.

STEP

Date
June in Year 13.

Papers

Two papers of 3 hours duration: STEP 2 and STEP 3.

Students with an offer to study Mathematics at the University of

Cambridge take the STEP 2 and STEP 3 examinations. Note: An easier

paper, STEP 1 was discontinued in 2020 and it is possible the names of the

remaining papers may be changed from 2022 onwards.

Students applying for Mathematics at Imperial College and who are unable

to sit the MAT may sit the STEP 2 and 3 examinations as an alternative.

For the University of Warwick, the STEP requirement could be met from

either of the STEP papers.

Questions

For each paper students select the questions they wish to answer from three

sections. All STEP questions are marked out of 20. All of the questions

that are attempted are marked. The marks from the best six answers are

used to award a grade. Candidates are rewarded for making good progress

towards a solution, even if the final answer is incorrect. Correct answers

always receive full marks, whatever the method used.

Both STEP 2 and STEP 3 consist of 12 questions: 8 from pure

mathematics, 2 from mechanics and 2 from probability and statistics.

Grading

The grades for each STEP examination are: S (Outstanding), 1 (Very

Good), 2 (Good), 3 (Satisfactory) and U (Unclassified).

For most colleges the mathematics offer is a grade 1 in both STEP 2 and

STEP 3 but this can vary from college to college and you should research

the requirements of the college in which you are interested.

Description
Questions in any STEP paper will test your ability to apply your

mathematical knowledge in novel and unfamiliar ways and will often need

you to consider several different topics. You are expected to show insight,

ingenuity, persistence and the ability to work through substantial sequences

of algebraic manipulation.

Syllabus

The most up-to-date version of the STEP syllabus can be found on the

Admissions Testing Service website.

STEP 2 is based on A Level Mathematics and AS Level Further

Mathematics (with some modifications and additions).

STEP 3 is based on A Level Mathematics and A Level Further

Mathematics (with some modifications and additions).

Required/accepted

The University of Cambridge typically requires a grade 1 in each of STEP

2 and STEP 3 for Mathematics and Mathematics with Physics courses.

This can vary from college to college and you should investigate the

specific requirements for a college on their admissions web pages.

Some other universities will accept a good grade on any STEP paper as

evidence of higher attainment and may accept you even if you don’t quite

achieve the required A level offer. You should check the entry requirements

of each university you apply for to see if there is any mention of the STEP

examinations.

Additional notes

The STEP examinations are open to all who wish to sit them. There is no

requirement for a student to be applying for any particular course.

Formula booklets and calculators

Formula booklets are not provided for any of the admissions tests. You are

expected to know the main formulae to use or be able to derive them.

Calculators are not permitted in any of the admissions tests.

Key ideas

Each admission test has its own key techniques and ideas. It helps to know

these ideas before going in to the test.

The TMUA

The TMUA is difficult to complete within the given time. Students will

often not complete a paper due to spending too long on some questions.

Speed and efficiency are important in the TMUA. It is key to have all of the

required techniques and knowledge at your fingertips. Methods need to be

streamlined and all working should be pared down as much as possible.

Questions in the TMUA often have a long list of distractors. These are

incorrect answers that come from plausible but incorrect working. The

questions are written so that it should be quicker to work out the solution

and then look for it in the options rather than try to eliminate obviously

incorrect answers. Elimination of one or two distractors is often possible but

more often than not, the question needs to be answered by working forwards

through the question. This means that along with the speed and streamlining

of the method, care and accuracy still have a part to play.

The MAT

Multiple choice questions in the MAT often have options that can be

dismissed early if you think in the correct way. Asking yourself ‘Is there a

quick method I need to see here?’ as you go in to every question helps. The

MAT is designed so that you can show your deeper understanding and this

extends to thinking beyond using conventional A to B methods. Some of the

multiple choice questions will need a complete solution to be worked

through but many rely on you spotting a particular property or behaviour of

a mathematical object.

The longer questions in the MAT often start with a relatively simple

calculation but then build on that. At each stage it is important to think

‘Why am I being asked to find this?’ Each stage will build to the next,

usually resulting in a final revelation in the last part of the question. Starting
off with this understanding should help you to look for the best route

through each stage of the question.

STEP

The STEP examinations reward complete solutions over partial bits of

working. It is important that you get as far through a question as possible.

STEP questions, like the longer MAT questions, build from a starting point

so that each stage relates to what has gone before. The first few parts of a

STEP question often give you more help and set you up for the later, more

difficult stages. There is generally less guidance in a STEP question as you

progress through it but you should still be thinking ‘Why am I being asked

to do this?’ and ‘How will this help later on?’

You should make sure you have attempted and completed as many past

paper questions as you can before the actual examinations. That usually goes

without saying, but in the case of STEP, having a good idea of what you like

and do not like doing before you get to a paper helps with question selection.

You should remember that little bits of solutions to several problems doesn’t

usually amount to much so you have to get into the mindset that you will

work through a problem until you get to the end. For some STEP questions

attempting what appears to be half of the question may not result in gaining

half of the marks as the main work may be in the final stages of that

question. Practising will also help you learn to identify when you are likely

to get nowhere with a question and should therefore give up on it at an early

stage.
Chapter 2: Mathematical problem solving in

admissions tests

TMUA MAT STEP

Introduction

At the heart of every question on the TMUA, MAT and STEP papers is the idea of

mathematical problem solving. But what is meant by mathematical problem

solving? It helps to have some sort of definition. If you define a mathematical

problem as one in which the path to the solution is not immediately clear then you

start to see why the questions on admissions tests can be described as problems

and that they are different from A level questions. All of the admissions tests are

designed so that you can demonstrate applying your mathematical knowledge in

novel and unfamiliar ways. They give you the opportunity to show the traits that

universities are looking for, namely, insight, ingenuity, persistence and a good

grasp of mathematical technique.

To be successful in mathematics admissions tests, you need to develop some

problem solving skills. You need to know what you are going to do when a

problem seems impenetrable, how you can get started, how you can find a path

through the problem and how you know you have reached the solution. In this

chapter you will see a number of problem solving techniques that you can apply to

questions from admissions tests. Some of these will seem like common sense and

others will be new to you. You will have the chance to try out these techniques on

some problems that require no more mathematical knowledge than is found in the

first part of an A level course. Once you have an idea of some key problem solving

techniques, try to remember them as you progress through this book and encounter

TMUA, MAT and STEP style questions.

The problem solving process

When you first encounter a problem and cannot see a clear path to its solution, it is

easy to get caught in the trap of just staring at the problem and hoping it will

resolve itself. Experienced problem solvers always make sure they can do

something. Doing nothing will never get you anywhere closer to a solution whereas

doing something, even if it turns out to be the wrong approach, will usually yield

something useful about the problem. You will at least know what approach not to

take!

A model of the problem solving process breaks it up into stages, each with its own

activity and utility. These are illustrated in the diagram.

The problem solving process should start with the process of identification. This

stage is not necessarily about identifying the whole path through the problem, it

may be simply about finding a starting point. In the identification stage an

experienced problem solver will not just read through a problem, they will ask

themselves a number of questions that start the process of finding out what needs

to be done. Every mathematics problem in the TMUA, MAT and STEP

examinations consist of:

• some ‘elements’ with which ‘mathematics’ will be done

• some indication of the mathematical skills that will need to be used – these may

be overt or hidden

• some indicator of the required end product.

Knowing that these elements are there should give you some focus when you ask

yourself questions in the identification stage.

The identification stage is followed by the engagement stage. This is where work

actually starts. During this stage you are testing out your initial ideas by doing the

calculations that you think will lead you towards a solution.

The review stage is often neglected when students are solving mathematical

problems, often at great cost. At various points in working through a problem, you

should stop and review what you have done. There are two key questions that you

should ask yourself at this stage.

• What have I done so far?

• How has it helped?

It is important to decide at an early stage if your initial ideas will take you towards

a solution. One of the traps that inexperienced mathematicians make is to fall in

love with the calculations that they are working with and not want to give them up

no matter how far in the wrong direction they are going. It is particularly easy to

get carried away when you know that each line of calculation is correct. You may

consider yourself to be in the engagement stage of problem solving and plough on

thinking that you will review what you have done at the end but it is far better to

check after a sensible period of time. The review stage should be revisited

regularly. In an admissions test you should always be aware that the first part of a

question should not take that long. If you are taking too long then you are either

taking the wrong approach or there is a better more elegant way of considering the

problem.

If, following a review of your working, you are confident that you are getting

somewhere, you can move on to the completion stage where you finish your

calculations and confirm that you have reached the required end product of the part

of the problem you were working on. At this stage you check your calculations for

any errors. These are often indicated by reaching something close to but not the

same as a given result that you were expected to confirm.

The reflection stage is another part of the process that is often overlooked by

inexperienced problem solvers. Once a solution to a problem has been reached it is

a good idea for you to think about what you have learned from coming up with
your solution. If you still have several parts of the question remaining, you might

like to think about how your work to that point will help as you continue. If you

have reached the end of the question you might like to think about the efficiency,

insight and ingenuity of your method. Could it have been done in a more elegant

way? Is there some underlying mathematical truth that you could have discovered?

The reflection stage is important in preparing your thinking for future

mathematical problems. You should always try to learn something about problem

solving from any solution that you produce.

The final stage of any problem is a check that the problem has been solved and the

required endpoint reached. This final completion stage should involve a detailed

check of presentation and mathematical argument as well as confirmation that you

have satisfied the requirements of each part of the question in full.

The final point that needs to be made about the problem solving process is that

once you are confident, you will pass through each stage without realising that you

are doing so. Confident problem solvers seem to subconsciously go through all of

the stages apportioning their time for each stage as appropriate. Once you have had

enough practice, the problem solving procedures outlined here will not need to be

followed strictly but until you reach that level of confidence it is useful to have

some guidelines.
Worked example

The following worked example illustrates how the stages in the problem solving

model can help to find a path through a problem.

Example 1

Question

In this question the notation is used. For a real number denotes the

largest integer less than or equal to

2 2 2
Write down the values of 1.5 , 2.5 , and 3.5

The sequence T is defined by for n ≥ 1.

(i) Calculate the first 7 terms of the sequence.

The sequence U is defined by

(ii) Write down an expression for un + 1

in terms of n.

The sequence T omits certain integer values.

(iii) Show that the positive integer I is omitted if and only if for some value of n,

(iv) Show that this inequality can be rearranged into the form a < (I ‒ n) ‒ I ≤
2

where are values to be found.

(v) Find the value of (I ‒ n) ‒ I and explain why the sequence T omits the
2

integers that it does.

(vi) Find the value of

Solution

Identification

On a first read through it is obvious that this is a question where each part is

closely linked.
It deals with integers and defines a function that you may not have come across

before. This is the floor function although it is not named in the question.

It is essential in this question to understand what the floor function does before

attempting it.

Some simple examples should help: from

the definition given. The question only deals with positive integers. This is not

stated but a read through confirms this to be the case.

Why are you asked to write down some simple numerical values before you start?

This must be to do with the question requiring numerical calculation where a

calculator is not permitted. The values must be used later in the question

otherwise there is no reason to record them. The words ‘write down’ indicate that

you should know what these values are or that they must be very easy to

calculate. In this case they are straightforward to calculate since

Part (i) of the question asks you to calculate some terms of T using the given

position to term formula. The numerical values calculated may help with this.

This should give you an initial indication of what the question is about.

Engagement

Review
This calculation seems slightly worrying. The presence of seems to imply that

a calculator is needed yet this is a non-calculator question. Is this the correct

approach?

Looking back at the question and what you have done so far, you should see the

reason for calculating the initial values: 2 is less than 2.25 and so is less than

It is also greater than 1. This indicates that and that

This should encourage you to carry on.

Stage completion

Reflection

You have the first seven terms and can see that the process of finding the values

involves discovering two consecutive integer value bounds for The

function will then return the lower bound.

Identification
The next stage looks very easy to do. You will substitute n + 1 into the formula

for n, but there must be some reason for finding an expression for u .
n + 1

Part (iii) talks about the sequence T omitting various integer values. Did you

think about which numbers were omitted in part (i)? These were 1, 4 and 9. It

appears to be the case that square numbers are omitted. It looks like the later

stages are going to be about proving why this is the case.

Once an initial expression for u has been found, can it be rewritten in a more
n + 1

useful form?

Is there a hint in the way the inequality for part (iii) is written? It looks wrong on

the right hand side. Why does it not say

Engagement

Review

It still looks like there is something wrong with the given inequality.

• Why is the right hand side

• Why is it ≤ rather than <?

There must be more to understand about this. You need to think about what have

you just found and what are you trying to find?

You are trying to find the integers omitted by T. That is defined by a related but

different function to U. so something must result from applying the

floor function that causes the change to the inequality.

It is a good idea to focus on the idea of an integer I being omitted. What are the

conditions on n for I to be omitted? The inequality is written with I as the subject

so focusing on I seems sensible.

For an integer to be omitted, u

n
must give a value between I ‒ 1 and I so that
results in I ‒ 1. u
n + 1
must therefore be greater than or equal to I + 1 in order that

I be omitted.
This looks like the way into this.

Stage completion

I is omitted if and only if there is a value of n for which u < I and u ≥ I + 1

n n + 1

If (subtracting 1 from both

sides of the inequality)

So

Review

The inequality turned out to be correct due to un + 1

needing to be greater than or

equal to I + 1.

Identification

Part (iv) is clearly about rewriting the inequality you have just found. The

differences between the two should help you identify what to do.

There are no square roots in the final inequality and there is a squared term which

indicates that at some stage the terms must be squared.

Squaring it as it is written leads to some awkward algebraic manipulation. Can

some simpler manipulation be done first? Is it possible to isolate the on one

side of the inequality and the on the other?

Stage completion

(subtracting from every part of the inequality)

(squaring all parts of the inequality)

Review

Is this valid?

Squaring here will not affect the inequality since n ≥ 1 for all of the terms in the

sequence T.
 2
x is an increasing function for x > 0 so if

Stage completion

Review

You have found the inequality that was asked for but what does it mean for the

value of I?

Identification

What do you know about I and n? They are both integers.

That means that (I ‒ n) ‒ I is an integer.

2

This integer has to be between and there is only one integer in this

interval.

Stage completion

(I ‒ n) ‒ I is an integer since both I and n are integers.

2

The integer I is omitted, if and only if it satisfies the inequality

where n is a positive integer.

So (I ‒ n) ‒ I = 0 giving I = (I ‒ n)
2 2

(I ‒ n) 2
is a square number so the only omitted integers will be square numbers.

Review

You have now proved that the sequence consists of the positive integers with all

of the square numbers omitted.

The final stage may be attempted.

This is the sum of the first 110 integers that are not square numbers.

Stage completion

The sum of the first 110 natural numbers will include ten square numbers.

Reflection

All of the stages have been completed and the earlier stages support the later

ones.

There are some useful ideas to take away from this.

• Inequalities where the variables are defined as integers may give a limited

number (or 1 or even 0) solutions.

• There is very little redundancy. When something is mentioned, it usually has a

use.

• Unfamiliar functions can be used but they will be defined and you are expected

to understand them and use them.

Useful problem solving techniques

The identification stage of the problem solving process is often the most difficult to

manage, particularly if an initial read through sparks a number of different ideas.

There are some useful techniques you can use at this stage to help you identify

sensible places to start.

Technique 1: Simplifying the problem, pattern

recognition and recursion

Some problems seem just too big to manage when you first read them. One way to

put together the ideas you need to solve the problem is to simplify it. If you can

solve a simple related problem you can sometimes see how to deal with the larger

one. This might be by looking at how a pattern of numbers changes as the levels of

complexity are increased or by relating one stage of complexity to the next. The

following example illustrates how this might appear in an admissions test.

Worked example

Example 2

Question

Mountain-range patterns can be generated from a sequence of forward slashes, /,

and backward slashes, \. For example, the sequence for the pattern

is written //\\/\.

Mountain-range patterns are constructed using these rules.

• The ‘empty’ pattern that contains no / or \ is considered to be a mountain-range

pattern and is generated by a sequence of 0 length.

• A pattern always starts and finishes at sea-level.

• A pattern never descends below sea-level.

The length, l, of a mountain-range pattern is equal to the combined number of /

and \ that make up its generating sequence.

(i) Draw the mountain-range patterns represented by the sequences

a) ///\\\

b) /\/\//\\

c) /\//\/\\

(ii) What can you say about the number of / and \ in any valid generating

sequence? Justify your answer.

(iii) List all of the possible generating sequences where l = 6.

You may now assume that every mountain-range pattern of length at least 2 can

be generated uniquely from a sequence of the form / A \ B where A and B are any

valid generating sequences.

(iv) Let Mn be the number of distinct mountain ranges of length 2n. Find an

expression for Mn + 1
in terms of Mn, Mn , Mn
- 1
, …, M0
- 2
(v) How many unique mountain-range patterns are there with generating

sequences of length 16?

Method

This question illustrates the process of building up a recursive pattern by

considering simple cases. As with many longer questions in admissions tests, the

question is structured to allow you to do this. The definition of a mountain-range

pattern is relatively easy to understand. The question that the setter really wants

the answer to is in part (v): How many patterns are there that have generating

sequences of length 16?

If you were asked this question straight away, it would be quite challenging to

come up with an answer without some preparatory work. Trying to draw all of

the possibilities would be time consuming, you don’t know how many there are

supposed to be, there could be a very large number of them. It would be easy to

miss out some correct patterns. As it stands, it is too big to deal with.

In a situation like this, you should simplify the situation. This will allow you to

get a feel for how the patterns work and how you can get from one stage to the

next.

Try out the following procedure. Can you work out how to find M8?

Start with the most basic patterns. Write down the value of M0 (how many ways

to have a sequence of length 0). Now draw the patterns for a length of 2 (to find

M1) and a length of 4 (to find M2).

It is easy to get side tracked by trying to look at the numbers and find the

sequence for those. Questions at this level rarely have sequences of numbers that

are easy to spot at a glance. In this problem it is far better to consider the actual

patterns and their representations as sequences of / and \.

There is a strong indication of how to work with this given in the question: ‘every

mountain-range pattern of length at least 2 can be generated uniquely from a

sequence of the form / A \ B where A and B are any valid generating sequences’.

Look at the patterns for length 2 and length 4. Does this fit with the idea above?

The / and \ come from the length 2 pattern. A and B are the patterns that are

added to make the length 4 pattern.

What are the different possible patterns for A? For each of these, what are the

possible patterns for B?

How many possibilities are there in each case? The sum of these will give the

overall number of possibilities.

Try to write a formula for M2 in terms of M1 and M0. Remember that M0 is the

number of ways to have a pattern of length 0.

Now look at the patterns for mountain ranges of length 6. Can you use the ideas

above to make sure you construct them all? The statement from the question

implies that these can be generated from / A \ B. Do you recognise any familiar

patterns you have seen before when you construct them?

Can you write a formula for M3 in terms of M2, M1 and M0?

Compare the formula for M with the formula for M . What do you notice?
3 2

If you are able to continue to find M8 then continue working before looking at the

worked solution.

Solution

Length 0

M = 1 as there is only 1 way to have a pattern of length 0 (by having no / and \).
0

Length 2

Sequence: /\

M1 = 1 as there is only 1 way to have a pattern of length 2 (by having one pair of /

and \).

Length 4

From / A \ B:

Either A is empty and B is the length 2 pattern giving case 1.

Sequence: /\/\
Or A is the length 2 pattern and B is empty giving case 2.

Sequence: //\\

So M2 = 2

This can be written as a formula in terms of M1 and M0.

For case 1 there are M0 ways to generate A and M1 ways to generate B giving M0

× M possibilities.
1

For case 2 there are M1 ways to generate A and M0 ways to generate B giving M1

× M0 possibilities.

So M = M × M + M × M
2 0 1 1 0

As you are searching for the structure of the patterns, it is sensible to leave this as

it is even though it can be simplified to 2M M .

0 1

Length 6

From / A \ B and based on the possible patterns for A:

Case 1: A is empty and B is a length 4 pattern. There are two possible length 4

patterns.

Sequence: /\/\/\
Sequence: /\//\\

There are M × M patterns.

0 2

Case 2: A is a length 2 pattern and B is a length 2 pattern. There is only one

length 2 pattern.

Sequence: //\\/\

There are M × M patterns.

1 1

Case 3: A is a length 4 pattern and B is empty. There are two possible length 4

patterns.

Sequence: //\/\\
Sequence: ///\\\

There are M × M patterns.

2 0

From the three cases:

You should be able to see a pattern emerging and from this make a prediction for

M4.

This can be confirmed using / A \ B

The possible cases are:

/\/\/\/\, /\/\//\\, /\//\\/\, /\//\/\\,

A is empty and B is length 6 giving M0M3
/\///\\\

A is length 2 and B is length 4 giving M1M2 //\\/\/\, //\\//\\

A is length 4 and B is length 2 giving M M //\/\\/\, ///\\\/\

2 1

//\/\/\\, //\//\\\, ///\\/\\, ///\/\\\,

A is length 6 and B is empty giving M0M3
////\\\\

M = 5 + 2 + 2 + 5 = 14
4

You need to find M8 but it is clear from the work done so far that all of the values

from M0 to M7 will be required to do this.

You would not have wanted to draw all of the patterns of length 16!
Try it out

Now you have seen how to build the formula, try out the whole question. Can you

answer it in full?
Technique 2: Finding a helpful diagram or

representation

The wolf, goat and cabbage problem is a famous river-crossing puzzle that dates

back to at least the eighth century.

A traveller has a wolf, a goat and a cabbage. On the way home, the traveller has to

cross a river by boat. The boat will only hold the traveller and one of the wolf, goat

or cabbage. The wolf cannot be left with the goat nor the goat with the cabbage as

in each case, one would eat the other.

How can the traveller get the wolf, goat and cabbage across the river?

One way to solve this problem is to look at all of the possibilities that do not break

either the wolf-goat rule or the goat-cabbage rule. Each allowable situation can be

represented by a simple code and single moves from one situation to another can

be represented by lines.

One possible representation is to code the traveller, wolf, goat and cabbage as T,

W, G and C respectively and use a dash to show each bank of the river.

Taking this approach, you should be able to see that there are two possible

solutions to the puzzle.

Having a representation in the form of a diagram helps to show what these two

solutions are.
Technique 3: Considering extreme cases

For some problems it is possible to gain some insight into how to solve them by

pushing a variable to an extreme value whilst keeping all other variables constant.

You will see this technique used a number of times in Chapter 7: Graph sketching,

identification and transformation. It is also a useful technique for some geometry

problems. This is particularly the case when some aspect of the situation you are

presented with is not fully specified.

Worked example

Example 3

Question

ABCD is a square with vertex A on the circumference of a circle of radius 4 units.

The side lengths of the square are equal to the diameter of the circle.

The sides AB and AD intersect the circle.

Let X be the area of the intersection between the square and the circle.

What is the maximum value of X?

Solution

This question clearly needs a diagram to be drawn, but how should the sides of

the square intersect the circle? The position for the maximum area has to be

found.

A rough initial sketch of the situation might look like this.

The point A is fixed on the circumference of the circle. It is possible to rotate the

square about this point. The lines AB and AD must intersect the circle.

Looking at the extremes of where the square can be rotated will give some insight

into the problem. Since the side lengths of the square are equal to the diameter of
the circle, one extreme will be when AB lies on a diameter of the circle and the

other will be when AD lies on a diameter.

In each case the area X is half of the area of the circle.

The area X appears to change as the square is rotated from one extreme to the

other. Starting from AB lying along the diameter and rotating for example an

angle of θ° anticlockwise will result in the same value for X as starting from AD

lying along the diameter and rotation through the same angle clockwise. When

the diagonal AC of the square lies along the diameter, the diagram will have

reflection symmetry in the line AC.

The value of X will vary continuously as the square is rotated from one of the

extremes through the symmetrical case to the other extreme. There is a strong

possibility that the maximum value of X will be either at one of the extreme

positions for the square or in the symmetrical position. This cannot be

immediately assumed and some additional work is required. Rotating the square

about vertex A from one position to the next will add some area to X on one side

whilst removing some on the other.

Starting from an initial position:

and rotating anticlockwise to a new position.

The orientation of the square can be measured by the angle between the radii

drawn from the centre to the points of intersection of the circle and the square.

Since the square has a 90° angle at vertex A on the circumference of the circle, it

follows that the two radii form a diameter of the circle.

Angle β is greater than angle α.

The area X will decrease by the difference between the shaded segment for angle

β and the shaded segment for angle α. It will increase by the difference between

the shaded segment for the angle 180° - α and the shaded segment for angle 180°

- β.

If the area of X can be shown to be increasing until the square reaches the

symmetrical position and decreasing after this then the maximum area will be at

the symmetrical position.

Try it out

1 Show that the area X increases as the square rotates from one of the extreme

positions to the symmetrical position.

2 Find the area of X shown in the final diagram. How does it compare to the area

at each of the extremes?

Technique 4: Thinking ahead and testing approaches

This technique is particularly useful in the longer MAT questions and in STEP

questions. A certain amount of intelligent guesswork is required and you will

sometimes find yourself working with only a hazy idea of where you are heading.

Most long MAT and STEP questions are designed to test how you learn as you

work. It often helps to consider these longer questions to be a story that is trying to

lead you to a greater mathematical truth. Like any good story, these questions can

be seen to have a beginning, a middle and an end. If you can take an educated

guess at where the question is leading, you can plot a rough path in your mind

about how to get there. It is a similar experience to trying to guess who committed

the crime in a murder mystery novel except the author of a STEP question will be

providing you with many more useful clues as you go.

It is possible that your initial educated guess is incorrect and what you start doing

doesn’t lead to the solution to the problem you are seeking. This is one of the

reasons why the reflection stages of solving a problem are very important. It is

very easy to get carried away when following through a line of thought. Spending a

little time considering what you have done at any stage can help you avoid

spending too much time following a path that leads to somewhere away from the

solution. Changing direction after following a hunch may feel like you have wasted

time but you will have picked up a feel for the skills needed to solve the problem.

This technique was used in the previous example where the initial approach of

looking at the extremes and considering the rotation of the square between one

position and the next seemed to indicate that the maximum area was at the

symmetrical position. The next step in solving the problem was to confirm that this

was actually the case.

Section 2 Problem solving in multiple choice

questions
Chapter 3: Introductory number and algebra

techniques

TMUA MAT STEP

Introduction

No matter what piece of mathematics you do, algebra skills are likely to play

a key role and all mathematics-based papers will test these skills.

Questions involving algebra in the STEP, MAT and TMUA examinations can

‘look’ quite different to typical school mathematics questions. You will need

to learn some necessary ‘tricks’ in order to realise that you can answer

questions that may seem impossible otherwise. As you work through

problems and read their solutions, do not be too hard on yourself and panic

that you cannot spot what to do. Instead try to understand why and how a

trick works so that you may then become confident at recognising when to

use it in future.
Some questions to think about

Each of the following questions uses a simple idea that you may or may not

have come across. If you are unsure what to do for any of these questions,

read the hint relating to it and then try again. If you still can’t answer some of

the questions, move on to the notes, exercises and reflection sections before

returning to them.

8
1 What is the last digit of (4832 × 2543) ?

66
2 What is the last digit of 1843 ?

3 What is the smallest number that leaves a remainder of 1 when divided by

2, leaves a remainder of 2 when divided by 3, leaves a remainder of 3

when divided by 4, and so on, up to leaving a remainder of 9 when divided

by 10?

4 Find all of the 2-digit numbers such that if you add the number, the number

formed by reversing the digits and the sum of the digits, then you get a

multiple of 96.

5 Show that:

(a) the product of any two consecutive integers must be even.

(b) the product of any three consecutive integers must be a multiple of 6.

(c) the product of any four consecutive integers must be a multiple of 24.

(d) the product of any seven consecutive integers must be a multiple of

1680.

6 Prove that n
2
(n
2
‒ 1)(n ‒ 4) is divisible by 360 if n is a positive integer.
2

7 Prove that n
2
‒ 1 is divisible by 8 if n is an odd integer.

Hints

1 When you multiply two integers together how do you quickly know the

units digit of the answer?

2 Is there a pattern to the units digit as you raise 1843 to the powers 1, 2, 3,

4, 5?
3 If a number has a remainder of 6 when divided by 7 then one more than

that number is a multiple of 7. How can you use this idea for all of the

conditions in the question?

4 Any 2-digit number can be written in the form 10a + b where a and b are

single-digit integers.

5 The two consecutive integers can be written as n and n + 1. If n is an even

integer, what can you say about n + 1? If n is an odd integer, what can

you say about n + 1? A similar idea can be used for parts b, c and d.

6 Can you rewrite the expression as a series of consecutive integers

multiplied by another integer? Is there a helpful way of rewriting n

2
‒ 4?
7 Can you rewrite the expression so that it is in a more helpful form? What

algebraic expression can be used to show that n is an odd integer?

Prime factorisation

Prime factorisation is an incredibly important tool. You may have already

seen how much it helps with finding the highest common factor or the lowest

common multiple of two or more numbers. It is often overlooked when

solving algebraic and numerical problems.

In the following examples, prime factorisation gives an indication of the path

that you have to follow to find an answer.

A zero will be produced for each 2 × 5.

a b
If the prime factorisation of the number can be written in the form 2 × 5 ×

P, where P is the product of the other primes in the factorisation, then there

will be either a trailing zeroes if a < b or b trailing zeros if a ≥ b, one for

each 2 × 5 pair.

Example 1

Question

How many zeroes are there on the end of 2021!?

Solution

In this list there are 404 numbers that are divisible by 5.

404
You now know that 5 is a factor of 2021! but the power of 5 in the full

prime factorisation may be more than this as some of the numbers divisible
2
by 5 will also be divisible by 5 = 25. In fact, a fifth of the numbers

divisible by 5 are also divisible by 25.

 2
80 of the numbers are divisible by 5 .

(404 + 80)
You now know that 5 is a factor of 2021! but some of the 80 that
2 3
are divisible by 5 are also divisible by 5 .

3
16 of the numbers are divisible by 5 .

4
3 of the numbers are divisible by 5 .

There are 404 + 80 + 16 + 3 = 503 trailing zeroes in the expansion of 2021!

What does it mean for a number to have a zero digit on the end?

For a number to end in 0, it must be a multiple of 10

What can you say about the prime factorisation of a number if it does

end in a zero?

The prime factorisation of the number must include 2 × 5

(404 + 80 + 16 + 3)
You now know that 5 = 503 is a factor of 2021! For a

trailing zero to be produced, each one of the 503 fives from the prime

factorisation has to be multiplied by 2. There are 1010 even numbers in the

product for 2021! so this is easily achieved.

Try it out

3 4
1 Use the ideas from the example to find the last non-zero digit of 15 × 14
5
× 13 .

2 Use the prime factorisation of 1200 to find how many factors it has.

3 How many factors of 2160 are multiples of 12?

Example 2

Question

How many multiples of 3927 have 3927 factors?

This initially looks impossible but prime factorisation may help.

Solution

3927 = 3 × 7 × 11 × 17

3927 has 2 × 2 × 2 × 2 = 16 factors

For 16n = 3927 the solution for n is not an integer so there is no number that

3927 can be multiplied by that will result in the answer having 3927 factors.

3927 has 2 × 2 × 2 × 2 factors since any factor of anything of 3927 will be

a b c d
in the form 3 × 7 × 11 × 17 where each of a, b, c, d has the value 0 or

1.

So 3927 has 16 factors.

For 3927 to have 3927 factors, there would need to be an integer n such

that 16n = 3927. There is no integer solution to this equation so no

multiples of 3927 have 3927 factors.

Moving into algebra

At A level you are taught a number of methods that help you to manipulate

algebraic expressions. These methods are very useful for making sure that

you are efficient when manipulating the more awkward expressions and

equations you come across in the STEP, MAT and TMUA examinations. You

need to make sure you are fluent in the following techniques.

The difference of two squares

The difference of two squares, that is is dealt with

more generally in Chapter 11 but it is worth appreciating now how useful a

tool it can be.

Example 3

Question

Expand

Solution

This example could be tackled by most students using methods well

practised at A level. The trick here is to realise that it can be done much

more efficiently by first realising that and then that

Example 4

Question

Prove that if p is a prime number and p ≠ 2, 3 then p

2
‒ 1 must be a
multiple of 24.

Solution

p
2
‒ 1 ≡ (p ‒ 1)(p + 1)
Both p ‒ 1 and p + 1 are even numbers since if p ≠ 2, then p is odd.
p ‒ 1 and p + 1 are consecutive even numbers. One will be a multiple of 2
and the other a multiple of 4 in either order so (p ‒ 1)(p + 1) will have a

factor of 2 × 4 = 8.

p ‒ 1, p, p + 1 are consecutive numbers and so one must be a multiple of 3.

Since p ≠ 3, either p ‒ 1 or p + 1 is a multiple of 3. Hence (p ‒ 1)(p + 1)

will also have a factor of 3.

(p ‒ 1)(p + 1) therefore has a factor of 3 × 8 = 24 and thus p ‒ 1 is a

2

multiple of 24.

This requires both the difference of two squares and some careful thought.

You are told that p ≠ 2, 3. All primes after 2 are odd numbers. On writing

p
2
‒ 1 = (p ‒ 1)(p + 1) you can see that p ‒ 1 is a product of two even
2

numbers. What is more, these numbers will be consecutive even numbers.

One of these numbers will be a multiple of 2 and the other a multiple of 4.

For example if p = 13 then p

2
‒ 1 = 12 × 14 and if p = 23 then p ‒ 1 = 22
2

× 24.

In any set of three consecutive integers (n‒ 1) n(n + 1), one must be a
multiple of 3. Since you are told that p ≠ 3, it follows that either (p ‒ 1) is

a multiple of 3 or (p + 1) is a multiple of 3.

Try it out

2 2
1 Find all positive integer solutions to the equation 360 + x = y

2 Use the difference of two squares to work out the value of

(a) 988 × 1012

2
(b) 503

Hints

1 How can the equation be linked with the difference of two squares?

Remember that y + x and y ‒ x are positive integers since x and y are

2 2
positive integers and y > x so y > x.
 The prime factorisation of 360 should help.

Will every pair of positive integer values for y + x and y ‒ x always lead
to positive integer values for x and y?

2 You are told to use the difference of two squares.

What happens if you let a ‒ b = 988 and a + b = 1012?

What are the values of a and b?

For part b let p = 503. (503 is a prime number.)

What is the relationship between p

2
and (p ‒ 1)(p + 1)?

Example 5

A MAT style question

Question

If a and b are integers then (1 ‒ a) 10b

(2 ‒ a) 9b
(3 ‒ a)
8b
× … × (8 ‒ a) 3b
(9

‒ a) 2b
(10 ‒ a)
b
is negative if and only if

a) a > 10 and b is even

b) a > 10 and b is odd

c) 2 < a < 4 or 6 < a < 8 or a > 10 and b is even

d) 2 < a < 4 or 6 < a < 8 or a > 10 and b is odd

e) none of the above.

1‒ a, 2 ‒ a, 3 ‒ a, etc. will be integers since a is an integer.

Any of the terms 1 ‒ a, 2 ‒ a, 3 ‒ a, etc. will be positive if they are raised

to an even power so

(10 ‒ a) 10b
, (3 ‒ a)
8b
, …, (9 ‒ a) 2b
will all be positive since 2b, 4b, …, 10b

will be even for all integer values of b.

For the terms, (2 ‒ a) 9b

, (4 ‒ a) 7b
, (6 ‒ a) , etc., all except (10 ‒ a) can
5b b

be considered to be (2 ‒ a) 8b
(2 ‒ a) b
, (4 ‒ a) (4 ‒ a) and so on. Each
6b b

one can be written as the product of a number that is definitely positive and

another number that may be positive or negative depending on the values of

a and b.
The original product will therefore be negative if the product (2 ‒ a)b
(4 ‒
a)
b
‒ a) (8 ‒ a) (10 ‒ a) is negative. This can be rewritten as ((2 ‒ a)
(6
b b b

(4 ‒ a)(6 ‒ a)(8 ‒ a)(10 ‒ a)) .

b

If b is even then this will be positive so one condition is that b is odd.

Since this is a multiple choice test, testing values of a for the two options

that state b must be odd will allow you to choose between them.

Solution

b
The expression will be negative if ((2 - a)(4 - a)(6 - a)(8 - a) (10 - a)) is

negative.

If b is even then this is positive so b must be odd indicating options b

and d (or possibly option e if no values for a give rise to a negative

result).

b must be odd.

For

A sketch graph of y = (2 ‒ x)(4 ‒ x)(6 ‒ x)(8 ‒ x)(10 ‒ x) can be used to

show that option d describes the conditions for a precisely.
The y intercept is at 2 × 4 × 6 × 8 × 10 which is a positive value.

Values for a can be tested to see where (2 - a)(4 - a)(6 - a)(8 - a)(10 - a)

is negative. It is better to test one of the regions for option d as a > 10 is

included in both options.

This is enough to indicate that option b is incorrect as it does not

identify any other values of a other than a > 10. As it is possible to find

a negative result, option e is also incorrect leading to option d as the

correct answer.
Inequalities

You may already be familiar with solving linear and quadratic inequalities, as

well as inequalities involving the modulus function. Questions in admissions

tests may appear more difficult, with ‘disguised quadratics’, for example. The

following worked examples illustrate how to give the solution set to a

collection of inequalities.

Example 6

A MAT style question

Question

The inequalities x
2
‒ 3x + 1 < 0 and 5x ‒ x
2
≥ 4 are satisfied by all x in the

region

a)

b)

c)

d)

e)

2
x - 3x + 1 does not factorise easily so an alternative method such as
2
completing the square needs to be used. A sketch graph of y = x - 3x +

1 should make the solution to the inequality easy to find.

If you were given only one of the two inequalities then this would be a

standard school level question, so you can imagine this being two separate

questions instead of one. Look at the solution to each inequality and think

carefully about how to piece these back together; remember that all values

of x in your answer must satisfy both inequalities.

Solution

2
The solution to the inequality x ‒ 3x + 1 < 0 is

2
5x - x ≥ 4 needs to be rewritten in the form q(x) ≤ 0 so that a graph

can be sketched.
The solution to the inequality 5x ‒x2
≥ 4 is 1 ≤ x ≤ 4

making option c the correct answer.

For x to satisfy both inequalities, it needs to lie in both the interval

and in the interval 1 ≤ x ≤ 4. Some further work is

required before the overall solution can be found. You need to be able to

place and in roughly the correct places on a number line in

relation to the numbers 1 and 4.

Since it follows that has a value

between It will be less than 1. In a similar way,

can be seen to be between 2.5 and 3 both of which are greater than 1

and less than 4.

Example 7

A MAT style question

Question

The numbers x and y satisfy the following inequalities.

The greatest possible value of xy is

a)

b) 4

c) 8

d) 9

e) 12

For this question, if you have studied linear programming as part of your

course, you have a slight advantage over students who have not. Your first

thought might be to attempt to manipulate these inequalities as you might

manipulate simultaneous equations. However, you need to be careful.

Each of the inequalities can be rewritten in either the form ax + by ≤ c or

the form ax + by ≥ c. The first step in answering this question is to sketch

a straight line ax + by = c for each inequality which will divide the (x, y)
plane into two areas. The next step is to decide which side of the line the

inequality represents.

Solution

A rough sketch of the lines should indicate what to do. From this you

have to find the greatest value of xy that satisfies all of the given

inequalities. If you let this value be equal to the real number c then you

have the equation xy = c. This can be rewritten as This is the

equation of a rectangular hyperbola with reflection symmetry in the

line y = x. As c increases, the point where the hyperbola crosses y = x

will get further and further from the origin. Thus the maximum value

of xy can be found by considering the value of c for which the point

where the hyperbola crosses y = x is as far away from the origin whilst

the hyperbola still has a point that is still within or on the boundary of

the unshaded region.

The maximum value of xy will be where the line 3x + 4y = 24 is a

tangent to the graph xy = c.

This will be where there is only one real solution when the two

equations are solved simultaneously.

2
There is no need to calculate 24 as several factors can be cancelled.

One real solution when the discriminant = 0

The greatest possible value of xy = 12.

This is actually the point C on the sketch graph above although there is no

need for the point giving the maximum value to be at an actual vertex.

The solution to this problem does not require any of the vertices of the

region to be found. The sketch graph shows that the solution point is where

the line 3x + 4y = 24 is a tangent to the curve xy = c. This sketch graph can

be produced relatively quickly and, once you know what point you are

looking for, the skills needed to actually find it can be applied quickly.

Try it out

For what values of x are the inequalities x

3
‒ 21x + 20 ≥ 0 and 21 + 4x ‒ x 2

> 0 both satisfied?

Hints

A sketch of the curves y = x

3
‒ 21x + 20 and y = 21 + 4x ‒ x2
will help.

• What is the shape of each curve?

• Where does each curve cross the x axis?

You will need to think about the values of x for which both inequalities are

true.
Expanding brackets

In an admissions test it is important to be as efficient as possible when

applying algebraic techniques. There are some ‘tricks’ that are worth learning

that will help you to find the result of multiplying out certain expressions

without the need to do the calculations in full and reducing the opportunity to

make careless errors. These ‘tricks’ can be particularly useful for the shorter

multiple choice questions where you do not want to take too much time

finding the solution.

Example 8

A MAT style question

Question

The power of x which has the greatest coefficient in the expansion of

5
(a) x
6
(b) x
7
(c) x
8
(d) x
9
(e) x

This question could be solved by any patient A level student since it is a

standard, albeit lengthy, process to do the full expansion. You could even

use a trial and error method, finding the coefficients of different terms until

you are sure that you have found the greatest. Surely, however, this is not

the intended method for this question and there must be a more elegant way

of finding the correct option.

Solution
The trick here is to consider the ratio of one coefficient to the next. The

Binomial Theorem can be used to do this.

r
The x term is

n
is the preferred notation at university for Cr.

r
The coefficient of x is

There are several terms that cancel in the fraction that require careful

thought. For laws of indices can be used. The factorial

terms need more care.

as 20 - r is one

more than
Since r is a non-negative integer, r + 1 is positive so both sides of the

inequality can ‘safely’ be multiplied by r + 1.

r is a non-negative integer so when r < 8 the sequence of coefficients in

the expansion is increasing. It follows that the term with the greatest
7
coefficient is the term in x

Example 9

Question

What is the highest power of x in the expansion of

Solution

It goes without saying that the last thing you should be thinking about

doing is expanding all brackets and simplifying! We only care about the

highest power of x.

The term with the highest power of x in

 7 5 5 35
The term with the highest power of x in (3x + 2x + 1) is 3 x , and

this is clearly a higher power of x than you will get in any term from
3
expanding (3x - 2) .

The term with the highest power of x in

Looking at the second half of the expression, the term with the highest
7 6 7 7 49
power of x in (3x - x ) is 3 x , and this is clearly a higher power of x
3 2 3
than you will get in any term from expanding (x - x ) .

This is the same as in the first half of the expression.

You may excitedly be tempted to say that the highest power of x must be
245 2 290
(x ) = x . This is the right idea, but you do need to be careful not to
245
fall in a trap. Thanks to the two x terms being the same, on subtracting
245
there is no longer an x term to square! This is unfortunate and means

that you will need to determine what the next highest power of x is.

203
This is a higher power than x .
You do not have the problem you had before: subtracting will not cause the
240
x term to disappear.

240 2 480
The highest power of x in the full expansion is (x ) = x .

7 5
The expansion of (3x + 2x + 1) is the sum of products of five expressions,
7 7 5
each of which is 3x , 2x or 1. (3x ) gives the highest power of x (the term
7 4
that eventually cancels), the next highest power must come from (3x ) ×
29
2x which gives you an x term. This is still a higher power than anything

you get from the expansion of (3x ‒ 2) 3

.

Looking again at the second half of the expression, the expansion of (3x
7
‒
6 7
x ) can be thought of as the sum of products of seven expressions, each of

which is 3x
7
or ‒x 6
. (3x )
7 7
gives the highest power of x (the one that
7 6
eventually cancels), the next highest power of x must come from (3x ) ×

(‒x 6
) which gives you an x
48
term. This is still a higher power than

anything you get from the expansion of (x

3
‒x
2 3
) .

Try it out

8 5
1 In the expansion of (1 + ax) , the coefficient of x is the greatest of the

coefficients. Use this information to find an interval that a must satisfy.

5 9 2 10
2 Find the coefficient of x y in the expansion of (1 + 2xy + y ) . Give your

answer in prime factorised form.

Hints

y y +1
1 Find an expression for the coefficient of the term in x and x . The ratio

of these coefficients will be greater than 1 if the coefficient is increasing

from one term to the next and less than 1 if it is decreasing. Write an

inequality to show this. Use r = 4 to test the conditions for an increase

4 5
from the x term to the x term and r = 5 to test the conditions for a
 5 6
decrease from the x term to the x . Do the conditions you find for a
5
indicate that no other terms will have a greater coefficient than that of x ?

2 There are two straightforward ways to do this. The first method is to

5 9 2
consider how the x y term can come from combinations of xy and y

terms. From this you can use the number of ways of selecting the xy

terms from the ten brackets that are being multiplied together and then
2
the number of ways of selecting the y from the remaining brackets.

The second method is to consider The

r
terms of the expansion of X will all be of the form

5 9
You can then find the values for r and a that result in the x y term. When

using this second method, it is important to remember the

multiplier as well as the multiplier.

Exercise 1

2
1 How do you know that x = 42 003 has no integer solution?

2 Prove that if you take any 3-digit number and reverse the digits to get a

new number, then the difference between the answer and the original

number will always be a multiple of 99.

3 Prove that n
5
‒ n is divisible by 6 if n is a positive integer.
4 Prove that n
5
‒ n is divisible by 30 if n is a positive integer and n ≥ 2.
5 Prove that 4
n
‒ 1 is divisible by 3 if n is a positive integer.
5 3
6 If the number 259n has 259 factors, how many factors does 18n have?

5 4 3 2
7 The positive integer x is the smallest number such that 6 × 7 × 8 × 9

× 10 × x is a cube number. Find the value of x.

8 The number where a and b are positive integers, is an

integer itself. Although 243 is not a factor of this number, 81 is. What is

the number?

9 What is the highest power of x in the expansion of

10 Are there more, fewer or the same number of zero digits on the end of

the product of the first 50 positive integers compared to the end of the

product of the next 50 integers?

8
11 How many factors does 1.2 × 10 have?

12 What is the smallest possible value of n such that n! contains at least 35

zero digits at the end of the number?

13 Can a positive integer be both a square number and a cube number?

Either prove that this is not possible or try to find at least two examples

where it is.

14 Why are there no pairs of square numbers that have a difference of 270?
15 What is the highest power of x in the expansion of

16 Find all the right-angled triangles whose sides have integer lengths and

such that the length of the shortest side is 11 units.

4 7 2 2
17 The positive integer x is the smallest number such that 9 × 8 × 7 × 6
2
× x is a cube number. Find the value of x.

18 Solve

3
19 If x + 2 is a factor of find the value of a.

20 The number can be expressed as where k is an

7 8
integer that is a multiple of 6 but not of 6 or 7. Find the value of k.

Express your answer as its prime factorisation.

12 9
21 What is the coefficient of y x in the expansion of

where n ≥ 7? Give your answer in terms of n.

Exercise 2

Admissions test multiple choice questions

TMUA style questions

1 The expansion of (a ‒ bx) c

is where a, b, c,

p, q and r are positive real constants.

Find the value of p + q + r.

11 2 3 8
2 The coefficient of x in the expansion of (2 + x + x ) is equal to 28
2 6
times the coefficient of x in (2 + ax) . Find all the possible values of the

constant a.

E ±1

F ±2

3 A region is defined by the inequalities y < x + 4 and x + y < 8.

Consider the four statements

Which of the statements above is/are true for every point in the region?

A 1 only
 B 2 only

C 3 only

D 4 only

E 1 and 2 only

F 1 and 3 only

G 1 and 4 only

H 2 and 3 only

I 2 and 4 only

J 3 and 4 only

K 1, 2 and 3 only

L 2, 3 and 4 only

4 S is the complete set of values x which satisfy both the inequalities

x
2
‒ 4x ‒ 5 < 0 and x 2
+ 4x ‒5<0
The set S can also be represented as a single inequality.

Which one of the following single inequalities represents the set S.

A (x
2
‒ 4x ‒ 5)(x 2
+ 4x ‒ 5) < 0
B (x
2
‒ 4x ‒ 5)(x 2
+ 4x ‒ 5) > 0
C x
2
‒5>0
2
D x + 5 > 0

E x
2
‒ 25 > 0
F x
2
‒ 25 < 0
2
G x < 1

2
H x > 1

2 4
5 The coefficient of x in the expansion of (2 + bx) is 2 times the
3 6
coefficient of x in (1 + bx) . Given that b ≠ 0, what is the value of b?

B
 C

8 6
6 In the expansion of (a + bx) , the coefficient of x is 4 times the

3
coefficient of x . What is the value of

A 1

B 2

C 3

D 4

7
7 Find the coefficient of the x term in the expansion of

A 960

B 320

C 120

D 15

E 15 360

8 Find the value of the constant term in the expansion of (x

9
‒ x‒ 3 12
)

A ‒792
B 792

C ‒495
D 495

E ‒66
 F 66

G ‒220
H 220

9 Find the coefficient of

A 51

B ‒51
C ‒25
D 25

E 1725

10 Find the value of the expression

C 3

E ‒3
F

G 1
MAT style questions

11 If m and n are integers, which of the following statements is true for all

m and n.

(a) m n
2 4
‒m 4
n
2
is a square number.

(b) mn(3n ‒ 3m + 1) + m 3
+ n
3
is a cube number.

4 5
(c) m n is not a square number.

(d) m (m
3 3
‒ 3n 2 2
)+ n (n
2 3
+ m ) is a square number.

2
(e) mn is not a cube number.

12 p and q are positive integers such that p + q = 13. What is the maximum
3
value that p q can take?

(a) 2662

(b) 2916

(c) 3000

(d) 3048

(e) 3401

13 The inequality x (y
2 2
‒ 9) < 4(y ‒ 9) is satisfied when
2

(a)

(b)

(c)

(d)

(e)

14 Let x and y be integers. How many integer pairs (x, y) satisfy the

inequalities

xy(x ‒ 1)(y ‒ 4) ≤ 2y(y ‒ 4) ‒3(x ‒ x ‒ 2), ‒2 < x < 3 and 0 < y < 4?
2

(a) 12
 (b) 16

(c) 20

(d) 24

(e) 28

15 Let a and b be integers. Then

is an integer if

(a) a > 0

(b) a ≥ b

(c) a + b ≥ 0

(d) a + b < 0

(e) b > 0

16 The highest power of x in

is

(a) 75

(b) 120

(c) 126

(d) 360

(e) 378

17 The inequality
 is true for all n ≥ 0. It follows that

(a) k < 5050

(b) k < 10 100

101
(c) k ≥ 2

2
(d) k > 100

(e) k < 6060

18 The inequality

is satisfied precisely when

(a)

(b)

(c)

(d)

(e)

19 The inequality

is satisfied precisely when

(a)

(b)

(c)

(d)

(e)

20 The term with the greatest coefficient in the expansion of

is

60
(a) 67 584x
 27
(b) 101 376x

45
(c) 337 920x

36
(d) 126 720x

54
(e) 132 455x

Full worked solutions can be found to all of the questions from Exercise 2 at

www.hoddereducation.co.uk/step-mat-tmua-answers
Chapter 4: Geometry and coordinate geometry

TMUA MAT STEP

Introduction

For the purposes of this chapter, the word geometry is used to mean the

study of points in two dimensions (the Euclidean Plane). In coordinate

geometry, the positions of points are represented by coordinate vectors

relative to a set of axes. This approach, which enables objects such as lines

and circles to be defined by equations, was not fully developed until the

work of the great 17th century mathematicians René Descartes and Pierre

de Fermat. By contrast, the geometry of the ancient Greeks, which did not

make use of coordinates, is called synthetic or pure geometry.

Points and lines

You will have encountered several forms for equations of lines, such as y =

mx + c for the line with gradient m that crosses the y axis at the point (0, c).

This form does not cover all cases because it excludes vertical lines (for

which the gradient is not defined). The most general form for the equation

of a line is ax + by + c = 0 (where a and b are not both zero). Notice that

vertical lines are defined by equations with b = 0.

The distance from a point P to a line L is the shortest distance from P to any

point on L. If Q is the point on L that is closest to P, then the lines PQ and L

are perpendicular (and intersect at Q).

Clearly vertical lines and horizontal lines are perpendicular. Two non-

vertical lines are perpendicular if and only if the product of their gradients is

−1.

Example 1

Question

(i) Find the point Q on the line L : 3x + y = 19 that is closest to the point

P(3, 0).

(ii) Find the distance from P to L.

(iii) What is the reflection of the point P in L?

Solution

The lines PQ and L are perpendicular and the gradient of L is -3. So the

line PQ has gradient Therefore its equation is of the form

and contains the point P(3, 0).

 (i)

Draw a diagram!

To find the coordinates of Q you solve the simultaneous equations 3x + y

= 19 and

The distance from P to L is the length of the line segment PQ.

(ii)

The reflection of P in L is the other point P on the line PQ which is the

same distance from Q. From P(3, 0) to Q(6, 1) is a translation of

(iii) P′ is (9, 2).

Congruent and similar triangles

It is very easy to forget some of the basic skills that you learned when you

were younger. It is surprising how many students struggle in geometry

questions in admissions tests because they didn’t realise that a situation

included triangles that were congruent or similar. In some cases, an

appreciation of similar or congruent triangles can speed up some

calculations considerably.

Example 2

Question

The line L passes through the point (-4, 1) and is perpendicular to the line
2

L2 : 4x + 3y = 24. Area A is the area enclosed by L1, L2 and the x axis.

Area B is the area enclosed by L1, L2 and the y axis. The ratio A : B is

A 19 : 6

B 47 : 18

C 475 : 12

D 193 : 36

E 289 : 36

Solution
The first thing to do is to draw a diagram.

The temptation here is to opt for an entirely coordinate geometry driven

approach, working out the equation of the line L and its intersection with
2

both axes and line L1. Whilst its equation and intersection with the y axis

are going to be useful, there is no need to find its intersection with the x

axis or with line L . In this situation the question can be done quickly by
1

realising that triangles A and B are similar to the triangle that line L1

makes with both axes.

There is one more useful triangle that is similar to all of the others and

that is the triangle formed by line L2 and the two axes:

Similar triangles can now be used. The length of PQ is 10 units as it is

simply an enlargement of a 3, 4, 5 right-angled triangle.

Using the intersections of the lines is a perfectly acceptable way of doing

this. The numbers are deliberately unfriendly in this question so the

intersection of L1 and L2 gives awkward fractions for both the x and the y

coordinate. In an admissions test, the choice is yours as to what path to

take so you should weigh up the effort required and the opportunity to

make errors when selecting your approach.

Circles

A circle is a set of points which are equidistant from a fixed point. The

familiar equation for the circle with centre (a, b) and radius r is:

The expanded form of the equation of a circle is:

To convert from the expanded form to the centre–radius form you effectively

complete the square on both the x and y terms as follows:

Clearly for some choices of A, B and C in the expanded form there will be

no solutions for x and y.

Try it out

Show that the expanded form of the equation of a circle has solutions if
2 2 2 2
and only if A + B ≥ 4C. Describe the ‘circle’ in the case when A + B =

4C.

Example 3

Question

A ladder is placed vertically upright against a wall.

(i) Show that the midpoint of the ladder traces an arc of a circle as the

base slides away from the wall.

(ii) What if the ladder toppled backwards away from the wall instead?
(Assume there is no friction and the ladder moves smoothly in each case.)

Solution

Set up coordinate axes so that the floor is the x axis and the wall is the y

axis. Let the length of the ladder be L. The base of the ladder stays at the

origin while the midpoint starts at

Before you can use coordinates you need to set up some axes. The

obvious choice would be to use the floor and the wall as the x and y

axes respectively. You could choose units so that the ladder has length

1 (or even length 2 so that the midpoint starts at (0, 1)).

If you denote the coordinates of the midpoint of the ladder by (x, y)

you may be able to find a formula for y in terms of x. It often helps to

draw a picture!

(i) Let the midpoint of the ladder have coordinates (x, y). The two

triangles in the diagram are congruent since they have the same
 interior angles and equal hypotenuse lengths. Since the upper triangle

has base x and the lower triangle has height y it follows by

Pythagoras’ Theorem that and hence that the path

traced by the centre of the ladder is a quarter circle of radius

It is tempting to use a similar approach in part (ii) but there is a

much simpler solution based on the synthetic definition of a circle.

(ii) The midpoint of the circle is always distance from the origin and

hence traces the path of a quarter circle of radius

In fact every point on the ladder moves along the arc of a circle in (ii) but

not in (i).
Circle theorems

You are expected to know and be able to apply the following facts about

circles.

The perpendicular from the centre to a chord bisects the chord.

The tangent at any point on a circle is perpendicular to the radius at that

point.

The angle subtended by an arc at the centre of a circle is twice the angle

subtended by an arc at any point on the circumference.

The angle in a semicircle is a right angle.

Angles in the same segment are equal.

Opposite angles in a cyclic quadrilateral add up to 180°.

The angle between the tangent and the chord at the point of contact is equal

to the angle in the alternate segment.

Example 4

Question

In this diagram, all of the circles are touching and the sides of the outside

triangle are tangents to the largest circle.

The radius of the smallest circle is 1 unit.

Find the side length of the triangle.

Solution

For this question a sketch of the smallest circle and the three

surrounding it is a good starting point.

Let the length of the radii of the middle sized circles be r.

There is a lot of symmetry that can be used to your advantage in this

question. Joining the centres of the three circles surrounding the central

one will produce an equilateral triangle. The line joining one of the

vertices of that equilateral triangle to the centre of the small circle will

bisect one of the angles of the triangle. You now have some angles that

you know and some lengths to work with. Joining the centre of the small

circle to the midpoint of one of the equilateral triangle sides creates a

right-angled triangle. One of the circle theorems can tell you why this is –

do you know which one?

The hypotenuse of the right-angled triangle formed will be the sum of the

radius of the small circle and the radius of one of the middle sized circles.

Some simple trigonometry can be used.

The radius of the large circle is the sum of the diameter of one of the

middle sized circles added to the radius of the small circle, i.e. 1 + 2r.

Joining the centre of the large circle to one of the vertices of the

equilateral triangle and to the midpoint of one of the sides gives a right-

angled triangle that is similar to the one that has just been used.
Example 5

Question

2 2
The diagram shows the circle x + y - 6x - 8y + 21 = 0. Two tangents to

the circle pass through (0, 1).

The angle between the two tangents at (0, 1) is θ.

Show that

Solution

The centre and radius of the circle will help to establish some lengths.

Marking in some useful radii will help as well as labelling some points to

help with the explanation.

The circle has centre (3, 4) and radius 2.

The radii connecting the centre of the circle to the point of

intersection with a tangent create a right angle. Joining the centre of

the circle to the point (0, 1) will bisect the angle θ.

By Pythagoras’s Theorem:

You now know the lengths AB, AC, AD, BC and CD (and that ADC is

congruent to ABC) which should be enough to find cos θ. One way to find

cos θ is to find and use the double angle formula cos 2A =

2 2
cos A - sin A. If this does not come easily to mind, similar triangles and

the cosine rule can be used to find the result. The length of the line

segment joining B to D is needed for this. A diagram of the triangle ABD

will help you see how to find this.

By the cosine rule:

If you used the double angle formula, you would need to state that

and
Coordinate geometry and other curves

In an admissions test, you may find yourself having to work with some

curves that are not usually part of A level mathematics (although they may

be part of further mathematics). If that is the case you are not expected to

know anything in particular about the curve in question. What you are

expected to be able to do is use the equation of a curve to find points of

intersection with a line by solving the equations of the line and curve

simultaneously usually by substitution. You may also have to find where two

curves intersect. If this is the case then the equation resulting from

whichever substitution you use will be something that you should be able to

solve without using anything from beyond the syllabus.

Example 6

Question

2 2
The ellipse defined by 4x + y = 4 is intersected at two points A and B by a

straight line with gradient m that passes through the point (−0.5, 0). The

point M is the midpoint of the straight line segment that joins A to B.

What is the equation of the locus of the point M as m varies between −∞

and +∞?

Solution

2 2
4x + y = 4

2
x = 0, y = 4 so y = ±2. The ellipse crosses the y axis at (0, 2) and (0, −2).

2
y = 0, x = 1 so x = ±1. The ellipse crosses the x axis at (1, 0) and (−1, 0).

A sketch of the ellipse would definitely help here. The intersections

with the axes can be found by setting x = 0 and then y = 0. The

2
equation of an ellipse looks a little like the equation of the circle x +
2
y = 4 so perhaps that can help with the shape.
 2 2 2 2
4x + y = 4 ⇔ (2x) + y = 4

2 2
The ellipse can be found by transforming the circle x + y = 4 by a stretch

in the x direction with a scale factor of

The line will pivot about the point (-0.5, 0) and will go from vertical

through a complete 180° rotation to vertical once more. It is possible

to imagine the shape that the point M describes and this can help

confirm the final result. It would help to have an equation for the line

through A and B (in terms of m) so that the points of intersection

with the ellipse can be found.

Line through A and B: y = mx + c

Passing through (−0.5, 0): x = -0.5, y = 0

 2 2
Substitute into 4x + y = 4

Rather than actually calculate the x values from this to find the midpoint,

you can speed this process up by remembering that the x coordinate of the

midpoint will be half of the sum of the x coordinates of A and B. From a

2
quadratic equation of the form ax + bx + c = 0, the sum of the roots is

equal to Since the roots of the equation (*) will be the x coordinates

of A and B, their sum can be read directly from the equation.

There is a potential division by 0 at this last stage. Substituting m = 0 into

(1) and (2) results in x = 0, y = 0 so the locus includes (0, 0).

Substituting into (2)

This could be substituted into either equation (1) or equation (2).

Equation (2) looks the slightly simpler equation of the two.

This is the equation of the locus of M and therefore the answer to the

question. With a little manipulation of this equation, you can use what

you know about graphical transformation to sketch it. Note that you are

not asked to do this so the next piece of working is simply to see how the

curve could be sketched.

 2 2
This can be compared to the original equation of the ellipse 4x + y = 4

with a transformation and some sort of stretch applied. It must also be an

ellipse. To sketch it the points where can be considered. You already

know that it passes through (0, 0).

The ellipse will have a line of symmetry at
Exercise 1

1 The points (−3, 1) and (1, −2) are two vertices of a square. Find the

equation of the line on which all of the possible centres of the square

lie.

2 2
2 Circle C1 and circle C1 are defined by the equations x + y − 4x − 2y −

2 2
20 = 0 and x + y + 6x + 8y + 16 = 0 respectively.

(i) Find the distance between the point P1 on circle C1 and the point

P2 on circle C2 where P1 and P2 are as close to each other as

possible.

(ii) Find the distance between the point P1 on circle C1 and the point

P2 on circle C2 where P1 and P2 are as far away from each other as

possible.

2 2
3 The parabola intercepts the circle x + y − 4y = 16 at the

points A, B, C and D.

Find the area of the trapezium ABCD.

4 The straight line y = mx + c, where m < 0, intercepts the y axis at point A

and the x axis at point B. It also intercepts the curve xy = 1 at points C

and D. Given that the y coordinate of A is greater than 0, the x

coordinate of B is greater than 0 and that the x coordinate of D is

greater than the x coordinate of C, show that the length of AC is equal to

the length of BD.

5 The parabolas in this diagram have a common tangent at x = 2.

 Given that one parabola has a vertex at (0, 2) and the other has a vertex

at (3, 5), find the equation of each curve and the equation of the

common tangent.

6 Find the distance between the parallel lines with equations

A circle C touches both of these

lines at the points A and B. Given that where O is the

origin, find the coordinates of A and B and the equation of the circle.

7 ABC is an equilateral triangle.

P is a point on AB such that AP : PB = 1 : 2

Q is a point on BC such that BQ : QC = 1 : 2

R is a point on AC such that RC : RA = 1 : 2

The lines AQ and BR intersect at point X.

The lines BR and CP intersect at point Y.

The lines CP and AQ intersect at point Z.

Find the ratio of the area of triangle XYZ to triangle ABC.

Exercise 2

Admissions test multiple choice questions

TMUA style questions

2 2 2 2 2
1 The circles with equations (x - r) + (y - 2) = r and (x + r) + (y + r) =
2
4r touch precisely once when

2 The shortest distance between the parabola and the line

B 1

3 The maximum distance between a point on the circle

and a point on the circle

B 13

E 14
4 Find the value(s) of a such that the turning point of the parabola

is closest to the origin.

E a = 0

5 One tangent to the circle passes through the point (0,

8) and the positive x axis. What is the value of x at the point where it

crosses the x axis?

A 4

D 8

2 2
6 A circle with equation x + y − 14x − 2y + 34 = 0 has a regular octagon

drawn inside it so that every vertex of the octagon touches the circle.

What (in square units) is the area of the octagon?

A 8

C 16

E 32

G 48

2 2
7 The line y = 2x + c is such that it intersects the circle x + y = 9 at two

points A and B. M is the midpoint of the chord AB of the circle. The

equation of the locus of M as c varies between is

 2 2
A x + y = 2

2
D y = x

E y = x

F y = −x

2 2
8 The tangent to the circle x + y + 6x + 2y + 2 = 0 at the point (−5, 1)

passes through the point (−3, 3). The other tangent to the circle that

passes through this point touches the circle at the point

B (−3, 1)

D (−1, 1)

E
MAT style questions

2 2
9 The parabolas with equations y = k − (x + 2) and x = k − (y + 2) touch

(i.e. meet tangentially) at a single point. It follows that k is equal to

(a) 1

(b)

(c)

(d) -1

(e)

10 An A4 sheet of paper has side lengths in the ratio The paper is

folded as shown in this diagram.

If the shorter side is said to have length 1, what is the perimeter of the

trapezium that is created?

(a)

(b)

(c)
 (d)

(e) 2

11 The circle encloses the point (1, 1) precisely

when

(a)

(b)

(c)

(d)

(e)

12 Which of the following is a tangent to the circle

(a)

(b)

(c)

(d)

(e)

2
13 Given that f(x) = x − 3x + 9, what are the coordinates of the turning

point of the curve y = 3 − 2f(x + 3)?

(a)

(b)

(c)

(d)

(e)

14 The reflection of the point (a, a) in the line where a ≠ 0 is

 (a)

(b)

(c)

(d)

(e)

15 ABCD is a square with vertex A on the circumference of a circle of

radius 4 units. The side lengths of the square are equal to the diameter

of the circle. The sides AB and AD of the square intersect the circle as

shown in this diagram.

Let X be the area of the intersection between the square and the circle.

What is the maximum value of X?

(a) 8(1 + π)

(b) 2(4 + π)
 (c) 16(1 + π)

(d) 8(2 + π)

(e) 4(2 + π)

2 2
16 The circle C has equation x + y = 1 and is intersected at (0, 1) and (0,

-1) by two circles of radius r, one with a centre at (a, 0) and the other

with a centre at (−a, 0). This is shown in the diagram.

The value of a that results in C having three distinct regions of equal

area satisfies the equation

(a)

(b)

(c)

(d)

(e)

Note: this question assumes that you will be working in degrees rather

than radians as radian measure is not a part of the MAT syllabus.

Full worked solutions can be found to all of the questions from

Exercise 2 at www.hoddereducation.co.uk/step-mat-tmua-answers
Chapter 5: Functions

TMUA MAT STEP

Introduction

The modern definition of function was first given by the German

mathematician Peter Dirichlet in the mid 19th century. A function is the

relation between two sets and maps one element in the domain to precisely

one element in the range. Functions may be one-to-one mappings such as x

2
2x or many-to-one mappings such as x x . At A level you only need a

basic understanding of functions. Function notation provides a convenient

way of expressing relationships and allows some calculations to be

expressed concisely.

The TMUA specification states that you need a qualitative understanding

that a function is a many-to-one (or sometimes just a one-to-one) mapping.

You are also expected to be familiar with the properties of common

functions, including which always means the ‘positive square

root’ and

The MAT specification does not make any reference to functions

specifically. However, a look at past MAT papers, including the most recent

ones, makes it clear that you are expected to understand the idea of a

function and be able to use and interpret function notation. You will find

examples of function notation being applied in various places throughout

this book.

In this chapter you will see how functions are used in the way that both

TMUA and MAT questions are presented. You will see how to find the

greatest or smallest value of a function, how to find composite functions and

how recursive functions are defined. You will revise the factor theorem and

be introduced to the remainder theorem for polynomial functions. Finally,

you will see some standard functions that you are expected to know and use

as a matter of course in admissions examinations.

Some key concepts
Domain

The domain of a function is the constraint on the values that can be input

into the function. You will often see phrases such as ‘where x ∈ ’ (this

means that x is a member of the set of real numbers) or ‘for x ∈ , x > 1’

(this means that x is a member of the set of integers and also greater than 1).
Range

The range of a function is the complete set of all possible resulting values

after substituting in all of the possible values in the domain. For the
2
function defined by f : x x over the domain x ∈ , the range of f is f(x)

∈ , f(x) ≥ 0.

There are a number of ways of describing a domain or range correctly. Set

notation is often used for real numbers ( ), integers ( ), rational numbers (

), irrational numbers ( ) and natural numbers ( ). These are sometimes

+
further defined by a + or − superscript for example, would indicate

positive integers (which could be written x ∈ , x > 0).

Inequalities are often used so x ∈ , −1 < x < 1 would indicate real

numbers between (but not including) −1 and 1. An alternative notation uses

round brackets for strict inequalities such as x < 1 and square brackets for

non-strict inequalities such as x ≤ 1. The interval −1 ≤ x < 1 could

therefore be written as [−1, 1) which is more compact.

Notation

You will see two common ways to define functions at A level and in
2
admissions tests. One uses mapping notation, e.g. f : x x , x ∈ , and the
2
other uses a notation akin to a formula, e.g. f(x) = x , x ∈ . You can treat

these two notations as equivalent when working with functions but there is a

subtle difference between the two.

2
The notation f : x x , x ∈ defines a function called f that maps an

input value to the square of that input value for the domain of the real

numbers. f(x) is actually referring to the output number itself so, technically
2
f(x) is a part of the image set of the mapping. The fact that f(x) = x is

written simply shows how you would calculate that output value given an

input value x. A common and not particularly serious error is to say ‘the

function f(x)’ when you mean ‘the function f’.

In practice, f(x) is a convenient notation to use, particularly for problems

involving differentiation.

The letter f is commonly used for functions although for several different

and unrelated functions other letters, usually g and h, are used. The letter p

is often used for polynomial functions.

Functions can include more than one variable. For example, the function f

defined by f(x, y) = xy, x, y ∈ would map points on a Cartesian grid to a

single numerical value.

When differentiating a function the notation f ′(x) is used for the first

derivative. This is called prime notation and would be read as ‘f prime of x’.

This notation will be further explained in the next chapter.

n
Families of related functions can be defined using suffix notation. fn(x) = nx

2 3
defines the ‘family’ of functions f1(x) = x, f2(x) = 2x , f3(x) = 3x and so on.

You will see this notation used later in this chapter as it is often used in the

MAT and STEP examinations.

Inverse functions

An inverse function is a function that ‘undoes’ another function. If a

function f applied to an input x results in a value y, then the inverse function

-1 -1 -1
f applied to y results in the value x. The notation f or f (x) is used for

the inverse function to f.

Two questions to think about

1 As x varies over the real numbers, what is the largest value taken by the
2 2
function (9 sin x + 12 cos x - 8) ?

2 The function is defined on the interval 0 ≤ x ≤

3. Find the least value of the function on this interval.

Try these two problems. Do not worry of you cannot answer them straight

away. Read through the hints and then try again. If you still cannot answer

them, there are worked examples and explanations in the next section.

Hints

1 Is there an identity relating sin x and cos x that will help here? Can you

manipulate the expression into a form where you can easily see how to

find the greatest (and least) value that the function can take?

2 Are the greatest and least values found by substituting in the boundary

values? What shape could the graph of y = f(x) have in the interval 0 ≤

x ≤ 3?
The greatest or smallest value of a function

The two questions you have just seen are examples of a type that commonly

occur in the multiple choice sections of admissions tests. They could be

based on the ideas in either of the questions above or they could be based

around finding the greatest or least in a set of expressions.

Example 1

Question

As x varies over the real numbers, what is the greatest value taken by the
2 2
function (9 sin x + 12 cos x - 8) ?

Solution

could be used to write in terms of

This example is the first question you were asked to think about.

Calculus is not required for this question.

This would be easier if the term was positive. This can be done by taking

out a factor of −1.

This a quadratic function in cos x and completing the square helps in

establishing minimum or maximum values.

Since -1 ≤ cos x ≤ 1 it follows that -5 ≤ 3 cos x -2 ≤ 1 and 0 ≤ (3 cos

2 2
x - 2) ≤ 25. From this -5 ≤ (3 cos x - 2) -5 ≤ 20 and 0 ≤ ((3 cos x -
2 2
2) - 5) ≤ 400.

The greatest value is 400.

Example 2

Question

The function is defined on the interval 0 ≤ x ≤

3. Find the least value of the function on this interval.

Solution

This example is the second one you were asked to think about. If you

have not thought about this before, you might be tempted to just substitute
into the function the values at each end of the interval for x. f(0) = 4, f(3)

= 27 − 9 − 9 + 4 = 13

From this you might wrongly conclude that the least value is 4.

3
is a cubic function with a positive coefficient of x . This

might be always increasing but only if there are no turning points;

between turning points the graph would be decreasing.

Where are the turning points?

 3
gives the local maximum (the coefficient of x is positive).

gives the local minimum.

The function is

defined on the interval 0 ≤ x ≤ 3 so it will be decreasing between x = 0

and will be less than f(0) and f(3).

Although this is not a particularly easy number to compare to f(0) = 4 and

f(0) = 13, you can have confidence that it is less than both as you know
the function decreases between x = 0 and and then increases

again.
Composition of functions

A composite function results when two or more functions are applied in

succession. For example, given the functions f(x) = 2x − 1, x ∈ and g(x)

2
= x , x ∈ then applying function f to an element x in the domain will
2
result in 2x − 1. Applying function g to this result gives (2x − 1) . The
2
composite function ‘apply f then apply g’ is (2x − 1) .

This can be found by g(f(x)) which is shortened in its notation to gf(x). It is

important to remember that the function on the right is the one that is

applied first.

2
gf(x) = (2x − 1)

2 2
gf(x) = f(x) = 2x − 1

The domain of the composite function gf is the set of all elements, x, in the

domain of f for which f(x) is in the domain of g. For f and g above, the range

of f is entirely within the domain of g so the domain of gf is x ∈ .

For the functions and g(x) = 5 - x, x ∈ , x > 0,

the composite function

The domain of fg will be the set of x values in the domain of g for which

If Since the domain of g is , this means

that the domain of fg is .

In admissions tests, questions involving composition of functions often

involve applying the functions several times.

Example 3

Question
The function P is defined for real numbers by P(x) = 2x and the function Q

is defined for non-zero real numbers by .

The function P is applied p times and the function Q is applied q times, in

some order, to produce the function

It is possible to deduce that

(a) p = 4 and q = 1.

(b) p is odd and q is even.

(c) p is even and q is odd.

(d) p and q are both odd.

(e) None of the above.

Solution

It is unlikely that you will have seen anything like this before in your

studies. When this happens it is a good idea to ‘play’ with the functions

to get a feel for them.

There are some key things for you to notice here. QP(x) is not the same as
2
PQ(x). QQ(x) = x means that Q (x) leaves x unchanged (we could call this

n
the identity function). This means that if n is odd and Q (x)

= x if n is even. For the function R to be , Q has to be applied an odd

number of times. This eliminates option b. If Q is applied an odd number

of times then q = 1, 3, 5, 7,…

A look at the odd number cases may help.

If P is applied m times, then Q is applied once, then P is applied a further n

n m
times this gives P QP (x) where m and n are both integers and m, n ≥ 0

(it may be that you don’t apply P either before or after Q so m = 0 or n = 0

are allowed).

For q = 1

m is odd ⇔ n is odd, m is even ⇔ n is even.

m + n is even.

This eliminates option d as this is a counter example.

n = 0, m = 4

n = 1, m = 5

n = 2, m = 6

and so on.
There are an infinite number of possibilities.

This eliminates option a which may well be a possibility but certainly

isn’t unique. You can’t deduce that p = 5 and q = 3 only that they could be

those values.

Applying function P will multiply whatever it is applied to by 2. This will

increase the power of 2 in the multiplier by 1.

For q = 3

k, l, m and n are integers that are either positive or 0.

The number of times P is applied is therefore k + l + m + n

Working through the composite function in order:

m + k - l - n is an even number.

This means that either all four numbers k, l, m, n are odd or that

there are an even number of odd numbers (treating 0 as an even

number).

Thus k + l + m + n is an even number.

An alternative way to show this involves a nice ‘trick’.

You know that is an even number so adding an even number

to this will result in an even number.

is an even number.

Thus is an even number.

Since q is odd, all of the possibilities can be written in the form

where there will be an odd number of Q and

a
an even number of P functions (one more than the number of Q).

a
Since there are an even number of P functions, will

sum to an even number.

c is the correct option.

Recursively defined functions

You have possibly come across recursive sequences before, in which terms

are calculated by performing operations on terms earlier in the sequence.

Recursively-defined functions are exactly the same idea. There is a

seemingly infinite array of possible functions that could be presented to you,

so you need to be prepared to invest time in the examination playing with

the function you are given to make sure you know what it does.

Example 4

Question

The function f(n) is defined for positive integers n, where 1 ≤ n ≤ 100,

according to the rules

The least value of f(n) is

(a) 1
−6
(b) 2
−14
(c) 2
−30
(d) 2
−62
(e) 2

Solution
This needs a few values to be calculated to get a feel for the function.

Calculating the first few terms up to f(6) seems to indicate that there

are no values where f(n) < 1 so option a is the correct option. Going

as far as f(7) shows that it is possible to have values that are less than

1, so option a is eliminated.

Since f(2n) = 2f(n), even values of f(2n) will be greater than values of

f(n) so the highest value will not come from an even n.

The only opportunity for a value to be lower than that from which it

is calculated is for it to be from

f(63) stems from f(3) = 1 through 2n + 1 being repeatedly applied to n.

This gives the n-values by applying the

formula four times.

The next case where f(n) = 1 is f(13). This would give

This is only three applications of four times.

Any later cases where f(n) = 1 will not result in more than four

applications of

−30
The correct answer is therefore 2 , option d.

The next value that can be calculated from f(63) is f(127) but 1 ≤ n

≤ 100 so this need not be considered.

It seems that option d is the correct answer but it is important to be

convinced that there are no values that could be lower.

Try it out

The function f(n) is defined for positive integers n by

The value of

(a) 16

(b) −16

(c) 50

(d) 63

(e) 0
Hints

• Calculate some values for f(n).

• What values will f(n) return?

• Be careful, at first it may seem like all of the terms cancel. This is not the

case.

• Where do the changes in the values occur? Can you spot any pattern to

this?
Exercise 1

1 The functions a(x) and g(x) are such that a(1), a(2) and a(3) are the first

three terms of an arithmetic sequence and g(1), g(2) and g(3) are the

first three terms a geometric sequence.

Which, if any, of the following two sequences show three consecutive

terms of a geometric sequence?

(a) ag(1), ag(2) and ag(3)

(b) ga(1), ga(2) and ga(3)

(c) Neither of these sequences

(d) Both of these sequences

2 The smallest value of the function in the

interval 1 ≤ x ≤ 2 is

(a) −15

(b)

(c)

(d) −14

(e) −18

3 The maximum value achieved by the function

in the interval 1 ≤ x ≤ 2 equals

(a) −2

(b) −6

(c) −10

(d) −14

(e) −18

4 As x varies over all real values, what is the greatest value which the

function

5 This question requires the use of radians which are on the TMUA

specification but not the MAT specification.

 In the interval 0 ≤ x ≤ 2π, the equation

has

(a) 0 solutions.

(b) 1 solution.

(c) 2 solutions.

(d) 3 solutions.

(e) 4 solutions.

6 The greatest value that the function

takes, over all real values, equals

(a) 1

(b) 3

(c) 5

(d) 7

(e) 9

7 The function f(n) is defined on positive

integers by and, for n ≥ 1,

Find the value of

8 For an integer n where n ≥ 2, the greatest prime factor of n is denoted

by gpf(n) For example

The function

The value of

(a) 401

(b) 223

(c) 123

(d) 119

(e) 53

9 The function f(n) is defined for positive integers and for

n ≥ 2 by
 How many numbers n in the interval 1 ≤ n ≤ 50, satisfy f(n) ≤ 10?

(a) 8

(b) 16

(c) 29

(d) 58

(e) 72

10 The function f is defined for real numbers by and the

function g is defined for non-zero real numbers by

How many different ways are there to generate the function 8x as a

composition of the functions f and g if neither function is to be applied

more than three times?

(Note that ffg(x) and fgf(x) are considered to be different

compositions.)

(a) 45

(b) 23

(c) 18

(d) 12

(e) 10
The remainder theorem and the factor theorem

The factor theorem is an incredibly useful tool that you will have already

used often, namely that (x - a) is a factor of the polynomial p(x) if and only

if p(a) = 0, and (ax - b) is a factor of p(x) if and only if

It is important that you know both the factor theorem and the remainder

theorem and that you understand why they work.

When dividing a polynomial p(x) by (x - a) for some number a, the result is

a polynomial of degree less than p(x) with a remainder that is a number,

possibly zero. For example, using long division of polynomials or otherwise,

you may want to check you agree that the result of

2
is 2x + 3x remainder 2.

An algebraically equivalent way of writing this is

In general, when

remainder r, this can be rewritten as p(x) = q(x)(x - a) + r.

If you understand this, then, by substituting x - a, it is clear that p(a) = r, the

remainder.

This is essentially the remainder theorem.

The factor theorem is then just the remainder theorem with the realisation

that when (x - a) is a factor of p(x) then the remainder, r = 0.

A full understanding also allows you to adapt the above slightly. For

example, if dividing p(x) by (2x − 1) then p(x) = q(x)(2x − 1) + r for some

polynomial q(x) of degree less than the degree of p(x), and remainder r. For

this example, the remainder is . You just think about what value of x

would make (2x − 1) equal zero and substitute this into p(x) to find the

remainder.
Whilst it is true to say that if is a factor of p(x), it

is also the case that (2x − 1) is a factor of p(x) and this is often more useful

– especially when working with polynomials with integer coefficients.

Example 5

Question

The polynomial is defined for integer values of n where n ≥ 2 by

The reminder when p (x) is divided by p (x) is

n n −1

(a) 1

(b) 2n − 1

(c) n
2
(d) n

(e) n(2n − 1)

Solution

The appearance of this question may make it seem worse than it is. It

seems more manageable when you realise that p (x) is a linear

n

polynomial for any positive integer value of n.

1 + 3 + 5 + 7 + … + 2n - 1 is the sum of the first n odd numbers. You

2
may already know that this is equal to n or you may have to find this

result by considering it as an arithmetic series with first term 1 and

common difference 2.
This can be used to find p (x).
n −1

To use the factor theorem you need to know when this is equal to 0.

Fortunately n ≥ 2 so we are can divide the whole equation by n − 1.

Option c is the correct answer.

This result could also have been found by inspection.

Using

Equating coefficients of x gives

Equating the constants gives

So which leads to the same result.

Example 6

Question

2
If n is a positive integer, then x + 2 is a factor of
for

(a) no values of n

(b) n ≥ 2

(c) odd n

(d) even n

(e) all n

Solution

2
This looks a little different because x + 2 does not factorise into linear

factors. You can, however, apply exactly the same reasoning that is used
2
to justify the familiar version of the factor theorem to state that (x + 2) is
2
a factor precisely when substituting x = −2 into the polynomial gives an
2
answer of zero. If you find this hard to grasp, let y = x , say, to obtain the
n n n 2 n
polynomial p(y) = (y + 4) (3 - y) - y (y + 1) . Then p(−1) = 0 exactly
2
when (y + 2) = (x + 2) is a factor.
Logarithms

In typical school mathematics, questions involving logarithms are mostly

procedural. In admissions tests the questions are designed to test that you

fully understand the definition of a logarithm, namely that

, and how to invert a logarithm:

It is assumed that you have seen and know

the laws of logarithms.

If you struggle to remember the relationship between logarithm and index

form it can help to learn a simple example such as The only

way to arrange the numbers 2, 10 and 100 in the form is as

It is easy to be bogged down in algebra and not appreciate how incredibly

useful logarithms are, particularly as modern calculators are able to give the

logarithm of a number to any base (provided it exists). Admissions tests are

non-calculator papers so being able to manipulate logarithms is an

important skill. The following example illustrates one of the useful things

you can find using logarithms.

Example 7

Question

0.97
Given that log10 5 = 0.698 97 to five decimal places and that 10 > 9 it is

possible to deduce that

1000
(a) 5 begins in an 8 and is 697 digits long
1000
(b) 5 begins in a 9 and is 698 digits long
1000
(c) 5 begins in an 8 and is 699 digits long
1000
(d) 5 begins in a 9 and is 699 digits long
1000
(e) 5 begins in an 8 and is 700 digits long.

Solution
This looks very different from any school logarithms question, but it must

be that with a school-level knowledge of logarithms you have the tools

necessary to answer the question. You have two challenges to address

here, using logarithms to determine both the number of digits in a number

and its leading digit. Each should be dealt with in turn.

For the number of digits, you can imagine numbers increasing and think

about the points at which the number of digits increases. What do you

know about the logarithm (with base 10) of these numbers.

For the leading digit, you can use a similar process, but this time focus on

when the leading digit changes from a 9 to a 1, and again from a 1 to a 2.

Once again, you need to think about the logarithm (with base 10) of these

numbers.

If you have never thought about this before, you may like to try this

question for yourself before looking at the solution.

The smallest number with 1 digit is 1, log10 1 = 0

The smallest number with 2 digits is 10, log10 10 = 1

The smallest number with 3 digits is 100, log10 100 = 2

Using this idea, if for a whole number n, k ≤ log10 n < (k + 1) then n

is a k + 1 digit number e.g. 3 ≤ log 12344 < 4 since 1234 is a 4-digit

10

number.
To find the first digit, standard form can be used. If the number can
b
be expressed in the form a × 10 where 1 ≤ a < 10 then the first digit

of the answer will be the units digit of a. Since there are 699 digits,
1000 698
5 can be expressed in the form a × 10 (there will be 698 digits

following the first digit).

The first digit is 9. Option d is correct.

The other piece of information given in the question now helps. Since 1
0.97
≤ a < 10 and you were given 10 > 9, it follows that 9 ≤ a < 10 so the

first digit must be 9.

Try it out

Example 7 used the fact that log10 A < log10 B < log10 C ⇒ A < B < C.

• Is f(A) < f(B) < f(C) ⇒ A < B < C true for any function f(x)?

• If it is not then why is it true for log10 x?

• It may help to sketch the graph of y = log10 x.

Can you find a function, g(x), such that g(A) < g(B) < g(C) ⇒ A > B > C?

Example 8

Question
Given real numbers x, y, z where x, y, z > 0 how many solutions are there

for the simultaneous equations

Solution

As with much simpler equations, the goal is to obtain an equation in just

one variable. Each of the equations may be written in an equivalent form

that may be easier to manipulate.

This can be used to simplify the three equations.

This is not straightforward to solve, but you just want to know how many

solutions there are and not what they explicitly are. A sketch of the

2
graphs of Y = X and for X > 0 shows that they only intersect

2
once: Y = X crosses the vertical axis at the origin and is increasing for all

X > 0 whereas crosses the vertical axis at 1 and is decreasing

for all X > 0.

Let X = y
So, although you do not know what it is explicitly, you do know that there

is precisely one positive value of y that satisfies both equations, and this is

enough.

Since , this also leads to a unique positive value of x. Since

is specified at the start this is the only solution for x, y and z.

There is one solution for X in X > 0 so there is one solution for y.

Since from (B), there is one solution for x.

There is one unique solution for x, y and z.

The modulus function

A modulus function, written returns the absolute value of a number or

variable. For the domain of real numbers the modulus function can be

defined by

The output from the function for non-zero input is always positive.

Example 9

Question

Find the value of

Solution

To sketch the graph, each part of the function can be considered

separately. will return a negative value for . The effect of the

modulus function is to make this value a positive value of the same

magnitude. This means that values of for can be calculated by

. For , values of will be the same as those of . For

, the critical value where the change from using to is at

. For the critical value is but for this is used for

and is used for . For the critical value is .

The integral gives the area under a graph so a sketch of the graph of

will help.
These calculations have the same result from the equation on the left of

the point as they do from the equation on the right.

Two new functions

Admissions tests occasionally introduce you to functions that you have not

seen before. The aim of this is to test your ability to read and interpret a

mathematical definition. Two functions that are often used are the floor and

ceiling functions. Although you are not expected to know anything about

these two functions for any of the admissions tests, it does help to have seen

them and the way in which new functions are presented.

The floor function, written , returns the greatest integer less than or

equal to x. It can also be called the greatest integer function. In an

admissions test it would be introduced with its definition for example it

might say ‘In this question denotes the floor function and returns the

greatest integer less than or equal to X’. You would be expected to

understand this definition and then use the floor function in the question.

When you are given a definition of a new function you should always think

thoroughly about what it does. For the floor function it is easy to see that

and even that but this might lead you to think that it

truncates an input (i.e. removes anything after a decimal point). This would

be an incorrect conclusion. Imagine the number −5.84. The greatest integer

less than this is −6 and hence

The notation [x] can also be used for the floor function. Some people

consider this notation to be incorrect but it was used by one of the great

historical mathematicians, Carl Friedrich Gauss. Whichever notation is used

in an admissions test, the function will be defined in the question as it is not

a part of the specification.

The ceiling function, written , returns the least integer greater than or

equal to x. It is also known as the least integer function.

Example 10

Question
In this question denotes the greatest integer less than or equal to x.

Find the value of

Solution

The integral gives the area under a graph so a sketch of the graph of

will help.

For 0 ≤ x < 1, will return the value 0 since 0 is the greatest integer

less than or equal to any value in that domain. This means that for 0

≤ x < 1 you will be sketching y = 2x + 0 = 2x.

For 1 will return the value 1. Sketch y = 2x + 1.

For 2 will return the value 2. Sketch y = 2x + 2.

For 3 will return the value 3. Sketch y = 2x + 3.

Exercise 2

1 Put these numbers in order from least to greatest. Do not use a

calculator.

2 Put these numbers in order from least to greatest.

3 Put these numbers in order from least to greatest.

4 Place these in order of size from the least to the greatest.0

5 Use the facts that and to find an

interval for and hence an interval for

6 The equation

(a) 4 real solutions.

(b) 3 real solutions.

(c) 2 real solutions.

(d) 1 real solution.

(e) 0 real solutions.

7 The equation

(a) 4 real solutions.

(b) 3 real solutions.

(c) 2 real solutions.

(d) 1 real solution.

 (e) 0 real solutions.

8 The function f is defined for by

Sketch the graphs of on the same axes.

Find the total area enclosed by the two curves.

9 In this question denotes the least integer greater than or equal to x.

The value of is

(a)

(b)

(c)

(d)

(e) 8

10 In this question denotes the greatest integer less than or equal to

X.

The value of is

(a)

(b)

(c)

(d)

(e)
Exercise 3

Admissions test multiple choice questions

TMUA style questions

1 The function f is defined on the positive integers as follows.

What is the value of

A 523

B 535

C 546

D 560

E 563

F 600

G 603

H 640

2 The function f is defined on the positive integers as follows.

The function g is defined on the positive integers as follows.

What is the value of f(99) + g(99)?

A 288
 B 300

C 388

D 400

E 488

F 500

G 588

H 600

3 The sequence of functions is defined as

follows.

where is the first derivative of with respect to x.

What is the value of

4 When is multiplied by and the resulting

product is divided by x + 2, the remainder is 182.

What is the value of a?

A −9

B −5

C 5

D 9
 E

5 The function p is defined by .

a, b and c are integers such that and .

The remainder when P(x) is divided by (x − 1) is R.

The remainder when P(x) is divided by (x + 1) is S.

Given that S − R = 4 and that S = 2R, what is the value of c?

A 1

B −1

C −2

D −3

6 Find the greatest sum, x + y, of the real values x and y that satisfy the

simultaneous equations

C 16

D 20

E 32

G 120

H 256

7 Find the non-zero solution to the equation

B
 C

D 2

8 The solution of the simultaneous equations

is x = a and y = b. Find the value of a + b.

9 You are given

where x, y and z are real numbers greater than 1.

What is the value of a?

A 1

B x

C y

D z

E There is not enough information to determine if a is any of the other

options.

10 The curve passes through the points

where p is a positive real number.

 Find the value of p.

D 1

E 4

F 16

G 32

11 Evaluate .

C 0

D 1

12 x and y satisfy . What is the

greatest possible value of

A 21

B 10.5

C 6

D 2

E 1.5
MAT style questions

13 The maximum value achieved by the function

equals

(a) −3

(b) 2058

(c) 2916

(d) 20

(e) 2401

14 As x varies over the real numbers, the least value taken by the function

f defined by equals

(a) 2

(b) 4

(c) 17

(d) 289

(e) 1764

15 Which is the largest of the following numbers?

(a)

(b)

(c)

(d)

(e)

16 Which is the least of the following numbers?

(a)
 (b)

(c)

(d)

(e)

17 Let n be an integer such that denote the greatest prime

factor of n and d(n) denote the final digit of n. For example

Which of the following statements is true?

(a) If cannot be 3.

(b) If cannot be prime.

(c) If cannot be 9.

(d) If cannot be 3.

(e) If cannot be 9.

18 The function f(x) is defined for x ∈ and satisfies the equation

f(3) is equal to

(a) 10

(b)

(c) 5

(d) 2

(e) 1

19 The function S(n) is defined for positive integers n by S(n) = the sum

of the digits of n.

For example, .

The value of
 (a) 855

(b) 900

(c) 945

(d) 990

(e) 1035

20 The function F(n) is defined for all positive integers, n by

and for n ≥ 2

if 2 is a factor of n but 3 is not a factor of n.

if 3 is a factor of n but 2 is not a factor of n.

if both 2 and 3 are factors of n.

if neither 2 nor 3 is a factor of n.

The value of F(15) equals

(a) 527

(b) 1051

(c) 2103

(d) 6311

(e) 12623

21 The function f(n) is defined for positive integers n by f(1) = 1, f(2) = −1

and then by the identities

(a) 88

(b) 80

(c) 60

(d) 50

(e) −22

22 Let n be a positive integer.

2
x + 2 is a factor of
 (a) odd n

(b) even n

(c) no n

(d) all n

(e) n ≥ 5

23 a and b are non-zero integers. When the polynomial is

divided by x − b, the remainder is 1. The polynomial has

2x − a as a factor.

It follows that a + b equals

(a) −5

(b) −3

(c) −1

(d) 0

(e) 5

24 The polynomial p is defined by

n

Given that n ≥ 2, what is the remainder when Pn(x) is divided by

Pn−1(x)?

(a)

(b) n

(c) −1

(d) 1

(e)

25 Given that x, y and z are positive real numbers, the equations

(a) have no solutions for x, y and z.

(b) have a unique solution for x but not for y and z.

 (c) have unique solutions for x and y but infinitely many solutions for z.

(d) have unique solutions for x and z but infinitely many solutions for y.

(e) have a unique solution for x, y and z.

26 How many real distinct solutions does the following equation have?

(a) 0

(b) 1

(c) 2

(d) 3

(e) 4

27 Let p, q and r be positive real numbers and . The equation

has a repeated root when

2
(a) q = 4r

(b)

(c) qr = 1
2
(d) r = q

(e)

28 For a real number x we denote by the largest integer less than or

equal to x.

The value of

(a) 4

(b) 8

(c) 10

(d) 16
 (e) 20

29 For a real number x we denote by the largest integer less than or

equal to x.

Let n be a natural number. The integral equals

(a)

(b)

(c)

(d)

(e)

Full worked solutions can be found to all of the questions from Exercise

3 at www.hoddereducation.co.uk/step-mat-tmua-answers
Chapter 6: Calculus

TMUA MAT STEP

Introduction

Calculus, that is, differentiation and integration, lies at the heart of seemingly

every aspect of pure mathematics. Knowing rates of change, finding maximum

and minimum values of a function and finding the area under a curve are just

some of the many applications.

Differentiation

The MAT specification (but not the TMUA specification), requires you to know

about differentiation from first principles, namely that for a function f(x) as shown

in the diagram, the gradient of the tangent to y = f(x) at the point A(x, f(x)) is

approximated by the chord AB. The approximation becomes increasingly accurate

as B moves towards A, or equivalently as h tends to 0.

Thus .

In this way, f ′(x) is a function that computes the gradient of the tangent to a curve

given the x coordinate of the point on the curve. The gradient of the normal can

then also be found by using the result that if m1 and m2 are the gradients of two

perpendicular lines (neither being vertical) then m1m2 = -1.

f ′(x) is a function that computes the gradient of the tangent to a curve given the x

coordinate of the point on the curve. The function f ′ is called the derivative of f

because it is derived from the function f. This notation was introduced by Joseph-

Louis Lagrange with second and third derivatives denoted by f ′′ (x) and f ′′′ (x).

For even higher derivatives, the primes (dashes) are often replaced by numerals in
(4)
brackets, e.g. f (x) for the fourth derivative.
Leibnitz used the notation in place of f ′ (x) when y = f(x). Using this notation

it is not always clear that is a function; it looks more like a fraction. When

evaluating the derivative at, for example, x = 2, Lagrange’s notation gives f ′ (2)

whereas when using Leibnitz’s notation we either write when x = 2, , or

we write

There are some advantages to Leibnitz’s notation. If, in the above diagram, you

denote h by ∆x, and let ∆y = f(x + ∆x) - f(x), so that ∆y is the small change in y

caused by the small change ∆x in x, then is the limit of as ∆x → 0. So,

although is definitely not a fraction, it is defined in terms of the limit of a

fraction. This explains why appears to behave like a fraction in, for example,

the chain rule, where if z = f(y) and y = g(x), then, provided all of the derivatives

exist, . The corresponding result using Lagrange’s notation is the

rather cumbersome (fg)′ = f ′(g(x))g ′(x).

You need to be familiar with both Lagrange’s and Leibnitz’s notations. In

addition, be aware that is really a shorthand version of . This is why the

second derivative is written as and not .

You are expected to know and use the results that for any

rational n (MAT and TMUA) and that for any real k (MAT only).

Try it out

You should already know how to differentiate some functions from the work you

have covered in school. Make sure you know how you would deal with each of

the questions below.

For the MAT and TMUA

In questions 1 and 2, differentiate each of the following functions with respect to

x.

3 The function f(x) is defined for x > 0 by .

Find the value of f ′(9).

4 Find the gradient of the normal to the curve at the

point where x = 16.

5 The normal to the curve is drawn at the point on the curve, P, where x

= 1. This line cuts the x axis at A and the y axis at B. Simplify the ratio AP :

PB.

2
6 The line y = mx - 15, where m < 0, is tangent to the curve y = 2x - 5 at the

point (p, q).

What is the value of p?

3 3
7 The curves with equations y = x + c and y = x touch (that is, meet

tangentially) at a single point. Find all the possible values of c.

For the MAT only

Differentiate each of the following functions with respect to x.

x 3
1 (e )

3
Stationary points

The derivative of a function is used to calculate the gradient of the tangent at any

point on the graph of the function. Of particular interest is when the tangent is

horizontal, and therefore the derivative is zero, since this is true at any local

maximum or local minimum on the graph. (The word ‘local’, sometimes omitted,

is used to clarify that the maximum or minimum is achieved just in a

neighbourhood around the point and not necessarily on the whole domain of the

function.) The points on the graph of a function where the derivative is zero are

called stationary points. Assuming the curve is not a horizontal line, stationary

points are characterised by whether the gradient of the tangent is positive or

negative on either side, as shown in the diagram.

The number of roots of a polynomial equation

There are various methods for finding the number of solutions of a given

equation, and some are discussed elsewhere in this book.

In the case of a polynomial, thinking of its graph and the location of its stationary

points allows you to determine how many real roots it has.

Example 1

Question

(i) Show that

(ii) How many real roots does the equation

Solution

(i)

Want to show

i.e.

which is > 0.

Hence

5 4 3 2
For part (ii) let f(x) = 4x + 35x + 60x + 130x − 600x + 300. Since the
5
coefficient of x is positive, you know that as x → -∞ then f(x) → -∞ and as

x → ∞ then f(x) → ∞. You may also know that a degree 5 polynomial

curve has at most four stationary points (since f ′(x) = 0 is a quartic

equation).

(ii) Let .
Since adding the coefficients of the left-hand side gives zero, you can

quickly spot that (x - 1) is a factor of the quartic polynomial.

Now the reason for the first part of the question starts to become clear:

since it is equivalent to substituting x = -6 into the cubic factor you

have just found.

3 2 2 3
(-6) + 8 × (-6) + 17 × -6 + 30 = 8 × 6 + 30 - (6 + 17 × 6)

It follows that (x + 6) is a factor of the cubic giving

has no real roots since its discriminant is negative.

x - 1 and x = -6 are the only real roots of the equation f ′(x) = 0 and x = -6.

The curve will have precisely two stationary points.

Stationary points at x - 1 and x = -6.

The second derivative can be used to find the nature of these stationary

points.

x = 1,

(1, -71) is a local minimum point.

This is equivalent to the second result that you proved in the first part

of the question.

f ′′(-6) < 0 from part (i)

There is a local maximum at x = -6.

When x = -6, y > 0

f(x) = 0 has three real roots.

Even though you haven’t found the actual y value when x = -6, you

know enough to roughly sketch the curve and find the number of roots.
 Although the sketch gives a not-to-scale simplification of the actual shape

of the graph, all that matters is where the gradient is positive, negative and

zero.

Just from the sketch of the curve you have established that the equation f(x)

= 0 has three real roots.

Example 2

Question

How many real roots does the equation have?

Solution

9 3 3 6 3 3 3
It is easier to consider f(x) = 9x - 9x = 9x (x - 1) = 9x (x - 1)(x + 1)

first.

It is clear that f(x) = 0 has three real roots, x = 0, x = 1 and x = -1 but you

actually have f(x)+ 2 = 0 here so you need to know about the stationary

points of the graph of y = f(x).

The graph of y = f(x) has three stationary points and these can be used to

sketch the graph of the function.

The stationary points have to be, from left to right, a local maximum, a point of

inflection and a local minimum, as this is the only way to draw a continuous

curve through both the x intercepts and the stationary points.

has three real roots.

The key observation here is that (as ). Translating the graph 2

units parallel to the y axis to give y = f(x) + 2 will not raise the point
 above the x axis so the graph of y = f(x) + 2 will cross the x axis in

three distinct places.

Try it out

By using first derivatives and considering long run behaviour, sketch the graphs

of the following functions and hence determine the number of roots in each case.

Hint

For number 3 be very careful when you are finding the stationary points.
The second derivative

can be thought of as the gradient function of the function y. It is a

function in its own right and so the process of differentiation can be applied again:

. This cumbersome notation is abbreviated to the more convenient

form . If y = f(x) then this may also be written as f ″(x). For

, the first derivative and the second

derivative .

The second derivative can possibly be used to determine the nature of stationary

points. In particular, if, at a point on the curve, and then the

stationary point must be a local minimum, whereas if and then it

must be a local maximum. If then all you can conclude is that you

have a stationary point and not what type of stationary point it is. For example,
3 3 4 4
the curves y = x , y = -x , y = x and y = -x all have zero first and second

derivatives at the origin.

Higher derivatives such as and can be computed and do have

applications such as approximating functions, but this is not something you will

encounter on MAT or TMUA papers.

Exercise 1

1 How many real roots does each of the following equations have?

(a)

(b)

(c)

(d)

(e)

(f)

2 How many stationary points do each of the following functions have?

(a)

(b)

(c)

(d)

(e)

(f)

3 2 2
3 Find the value(s) of k for which the function y = (k - 8) x - 8 (3k + 6k + 8) x

+ 56 (2 - k) is a quadratic function with a maximum turning point at x = -2.

4 For what value(s) of the real number k does the graph

have exactly one real stationary point?

5 has a turning point that is a

minimum when x = -1, precisely for

(a) all values of k.

(b) .

(c)

(d) .

(e) .

(f) all values of k except for 0.

6

has a turning point, that is a maximum, when x = 1, precisely for

(a) -4 < k < 7

(b) k = 1 or 7

(c) k > 7

(d) k = -4 or 7

(e) k < 1

(f) k = -4 or 1

(g) k < -4

(h) k = -4, 1 or 7

4 3
7 How many possible stationary points could the graph of y = 3x + 4(k − 2)x +
2
12(2 − k)x - 96x + 11 have?

Give the value or values of k that result in each number of stationary points.
Integration

In your school course, you will have been introduced to both indefinite integration

(sometimes called anti-differentiation), that is reversing the differentiation

process, and definite integration, that is evaluating the integral at given limits and

finding the difference.

An example of indefinite integration is

For an indefinite integral, c can be any constant and it is referred to as ‘the

constant of integration’.

An example of definite integration is

One application of definite integration that you meet in school is that

computes the signed area between the curve y = f(x), the x axis and the lines x = a

and x = b. The area is signed in the sense that it is positive when the curve is

above the x axis and negative when it is below.

Consider the definite integral . This is shown by the shaded area in the

diagram.
The definite integral equates to 0.

The area is clearly .

This happens because (the area is above the x axis) and

(the area is below the x axis) so adding these gives zero. If you are interested in

the actual total area between a function and the x axis, you need to split the

integral into different sections according to when the graph is above or below the

x axis.

For the example above the area is equal to

Note that this can also be written as . Swapping the limits for the

integration changes the sign of the definite integral.

Try it out
Find the area enclosed by the curve , the x axis and the lines x

= -2 and x = 2.

Hint

You need to know where the curve is above or below the x axis. Sketch the curve

to see what the areas look like.

The area between two curves

In addition to using definite integration to calculate the area between a curve and

some straight lines, you may also need to find the area between two or more

curves. For two curves this involves using the integral of the difference between

the two functions.

The area between y = g(x) and y = f(x) from x = a to x = b is

The order of subtraction in the integrand does affect whether the answer is

positive, so if the graph of y = f(x) is above the graph of y = g(x) then the area

would be .

A good question to ask would be whether this formula for the area between two

curves is also valid if the graphs of one or both of the functions lie all or partly

below the x axis. The answer is yes, as the following diagram shows for one

particular case.
Just as care is needed when finding the area between the x axis and a curve that

intersects the x axis, care is needed when finding the area between two curves that

intersect.

Try it out

3 2
Find the area between the curve y = x - x and the line y = 2x from x = -1 to x =

2.

Hint

This is clearly not the area. A sketch of both graphs will show you what you

need to calculate.
Example 3

Question

By means of a sketch show that

for any real numbers and and any f(x) which is defined on the

interval .

Find the area enclosed by the curve , the line and the x

axis.

Solution

The graph of y = f(x - c) is a translation of the graph of y = f(x) by c units

parallel to the x axis. If the limits, x = a and x = b are also translated, the two

definite integrals correspond to the same area.

The prime factorisation 962 = 2 × 481 = 2 × 13 × 37.

 is a translation of one unit to the right.

The straight line y = x - 31 has positive gradient and passes through the

points (0, -31) and (31, 0). It intersects the curve at a single point. The first

thing to do is to find that point of intersection.

This seems to give two values of x but you need to appreciate that only one of

them is valid. If x = 37 then and but if x =

26, and . Squaring both sides of an equation

often introduces new solutions that are not valid in the original equation.

2 2 2 2
In general, A = B implies that A = B but A = B does not imply that A = B. It

is true that but it is also true that

. x = 26 is a solution of .

From the work in the first part of the question, taking , a = 0, b =

36 and c = 1 it follows that .

Example 4

Question

The area enclosed by the graphs of the functions

where a is a real number is 32 square units.

Find all possible values of a.

Solution
If the graphs intersect, the x coordinates of any points of intersection must
2 2
satisfy x + 6ax + a = a − 2x

If a > 0, the enclosed area is

If a < 0, the enclosed area is

For a > 0
For a < 0

For there to be a positive enclosed area, a = 0. Since the coefficient of x is

positive for one quadratic and negative for the other, you know that, for the
2 2 2
enclosed area, the graph of y = a - 2x is above that of y = x + 6ax + a.

The only possible values of a are a = 2 or a = -2.

Simultaneous integrals

For any function f(x) defined on the interval ,

This can be thought of in simple terms as ‘the area from x = a to x = c is the same

as the area from x = a to x = b added to the area from x = b to x = c.’

If you recall that then you realise that this result holds

whether or not b lies in the interval [a, c]. This result lends itself to a specific type

of question that has appeared on past MAT papers.

Example 5

Question

The function f(x), defined on the real numbers, satisfies the condition that f(-x) =

f(x).

It is given that

Calculate the value of

Solution
Why have you been told that f(-x) = f(x)? It must be for a reason. What

does it mean?

For every point on the curve, (x, f(x)), there is a corresponding point, (-x, f(x)),

that is a reflection in the y axis. Since this is true for each point on the curve, it

follows that the graph of the function is a reflection in the y axis. In terms of

the integrals, this means that

Since , this means that

You know from the question and have just found . All that

remains is . This will be the same as

Piecewise functions

A piecewise function is a function that is formed by piecing together different

functions, each defined on different intervals of the real number line. For example,

consider

For this example, the function is continuous in the sense that there are no ‘jumps’

when sketching the graph, but this need not always be the case.

A sketch graph of y = f(x) is shown in the diagram.

The graph of y = f(x) defined above.

The graph with each of the functions used to define f(x).

Piecewise functions are integrated by piecing together integrals just as you piece

together functions.

Try it out

Use the sketch of y = f(x) above to find the value of the following definite

integrals.

5
Hints

2
For 4 and 5, think about what type of function x is. What will that tell you
2 2
about the features of the graph of y = f(x )? At what values will x be equal to 0,

2 or 3?
The trapezium rule

The trapezium rule provides an approximate value for a definite integral for which

there is no analytical solution. It is part of the A level course but is often taught in

school after the date of the TMUA or MAT examinations. You may wish to do

some preparatory work for it on your own.

To approximate the definite integral the interval on the x axis,

, is broken into n segments of equal width (calculated by ).

Trapezia of equal width are formed by drawing vertical segments from the axis to

the curve and then joining successive points.

The area of a trapezium is calculated using the formula

At A level you use the trapezium rule formula

where

You need to be able to apply this formula but also to appreciate two key points.

Increasing the number of trapezia usually gives a more accurate estimate of the

actual area.

One trapezium
Ten trapezia

The gradient of the graph of the function can indicate whether the trapezium rule

provides an underestimate or an overestimate of the actual area.

The gradient of the curve is increasing.

The trapezium rule overestimates the area.

The gradient of the curve is decreasing.

The trapezium rule underestimates the area.

Increasingly accurate overestimates will decrease in value towards the actual area

and increasingly accurate underestimates will increase in value towards the actual

area.

All areas in the diagrams have been drawn above the x axis. If any of the graph is

below the x axis then the trapezia below the x axis will have negative area. The

trapezium rule provides an estimate of the definite integral whether the

graph of y = f(x) is above or below the x axis.

Example 6

Question

The trapezium rule is used to estimate the integral by dividing the

interval into N intervals of equal width. Find a simplified expression

for the estimate.

Solution
Writing 3 + 2 instead of 5, allows the 2 to be associated with the

terms giving a geometric progression with

first term 2, common ratio and N terms. The sum of this progression is

simpler than the sum of the alternative geometric progression

that has first term , common ratio

and N - 1 terms.
Exercise 2

1 Find the area enclosed by the graphs of the following functions.

(a)

(b)

(c)

(d)

(e)

2 For a particular function, f(x), the trapezium rule with three ordinates (i.e. two

equal width trapezia) approximates the definite integral as 8. With

five ordinates (four trapezia) the approximation is 8.4.

What would the trapezium rule approximation with two ordinates (one

trapezium) of be?

3 Using the trapezium rule with five trapezia of equal width provides an

overestimate for both and . Prove that using the trapezium

rule with five trapezia of equal width also provides an overestimate for

What can you say, if anything, about the trapezium rule estimate, using five

equal width trapezia again, for each of the following definite integrals?

(a)

(b)

(c)
When is a calculus question not a calculus

question?

In both the MAT and TMUA examinations there are questions presented using the

language of calculus, which obviously require you to understand how the language

relates to areas and gradients, but which are, in fact, best answered by not

employing calculus techniques at all.

Example 7

Question

The expression polynomial of

what degree?

Solution

2 3 5 2 3 5 11
The highest power of x in (x - 1) (2x - 1) is (x ) × (2x) = 32x .
10 9
Differentiating this once gives a term in x and a second time gives a term in x .

2 5 5 10
The highest power of x in (2x - 1) is (2x) which gives a term in x .
9
Differentiating this gives a term in x .

9
As both are x terms, it is worth checking that they don’t cancel each other out.

11 10 9
For 32x , differentiating gives 352x and differentiating again gives 3520x .

10 9
For 32x , differentiating once gives 320x .

9
The x terms will not cancel so the expression is a polynomial of degree nine.

This only requires the highest power of x in the expression above, so it

would be a waste of time to actually perform the differentiation.

Example 8

Question
Find the area bounded by the graphs of the functions

Solution

The main clue to the fact that this question does not require you to integrate is

that you don’t know how to integrate . There must be an alternative

way to think about the question.

is a semicircle, centred at (3, 0) with radius 3, for which y ≥ 0.

The straight line crosses the x axis at (6, 0), passes through a

point on the circumference of the semicircle and then meets the y axis at

. It has a negative gradient. The area bounded by the two graphs

will be the area of a minor segment of the semicircle.

There is no need to find the points of intersection of the two graphs. The

gradient of the line is so the line makes an angle of 60° with the x axis.

This means that there is an equilateral triangle with vertices that are the centre

of the circle, the point (6, 0) and the point where the line intersects the

semicircle.

square units

Example 9

Question

This example requires radians and so would not appear on a MAT paper.

Which of these integrals has the largest value?

You are not expected to calculate any of the integrals exactly.

A

Solution

You are not expected to calculate any of these integrals, so your options for

dealing with a question like this are limited. You should think about the graphs

of the functions and decide which has the greatest area above the x axis

between x = 0 and .

All of the integrals A to D can be thought of as integrals of functions of cos x.

passes through (0, 1) and so E < C. This can be shown on a

sketch graph.
 n
Reflecting the graph of gives sin x and vice versa (since

). This means that and

For

This means that

Since it follows that B is the largest value.

This answer has been reached without integrating at all. Very little working was

needed, just an appreciation of the symmetry of the graphs of the functions

being integrated.

Example 10

Question

This example requires radians and so would not appear on a MAT paper.
Which of the following is true?

A T = 0

B T < 0

C T > 0

Solution

The definite integral, , is positive since,

between and x = π, there is twice as

much area above the x axis as below.

In a similar fashion, a quick sketch of the graph of y = sin x shows that the

definite integral, , is negative, since more of the identified area is

below the x axis than above it.

y = cos 3x is a one-way stretch of the graph of y = cos x, with scale factor

parallel to the x axis.

You only need to know whether or not the value of T is positive, negative or

zero. You are therefore only interested in whether the graphs are above or

below the x axis between the limits of the integral.

The graph shows that is negative, since more

of the area is below than above the x axis.

T is the product of one positive and two negative numbers, therefore T must be

positive.

Example 11

Question

This example requires radians and so would not appear on a MAT paper.

How many real solutions does each of the following equations have in the

interval 0 < x ≤ 2π?

(a)

(b)

Solution

(a) As x varies between 0 and 2π, cos x varies between -1 and 1. For y = cos(cos

x) you will be finding the cosine of values between -1 and 1.

Since π is just a little bit larger than 3, 1 radian is just a little smaller than

radians so cos 1 is just greater than .

cos(cos x) is an even function that must be bounded by cos 1 and 1. When x

= 0, π, 2π, etc.

cos(cos x) = cos 1 and when etc. cos(cos x) = cos 0 = 1.

 You don’t know how to integrate cos(cos x), so there must be an

alternative approach. It is important that you are working in radians.

As this is an unusual function a sketch will help.

For part a the equation is written as .

To appreciate the meaning of this, change the x axis to a t axis and consider

the area between the t axis and the graph of y = cos(cos t) between t = 0

and t = x, where x can be any value in 0 < x ≤ 2π.

The integral is positive for all values of x in 0 < x ≤ 2π so

there are no solutions to

has no real solutions.

Once this is clear, the solution to part a comes with relative ease.

(b)

For part b consider the area of the rectangular region below the graph from

t = 0 to t = 2π. A height of 0.5 can be used as this is just below the graph.
Since this rectangle has an area of , it follows that the area

is greater than π.

is 0 when x = 0 and a value greater than π when x = 2π. The

area will always increase as x moves from 0 to 2π so it follows that

for precisely one value of x. The equation has precisely

one solution in the interval 0 < x ≤ 2π.

has one real solution.

You can answer this question without actually saying what the solution is.
Differential equations

A differential equation is an equation that involves derivatives such as

as well as x and y. You are likely to be more familiar with these by

the end of your school studies, but you do need to be able to handle basic

questions on the TMUA and the MAT. The TMUA explicitly states that you need

to be able to solve differential equations of the form . If you see an

unusual looking differential equation on a MAT or TMUA paper, it is likely that it

can be solved using simple differentiation or integration or even by considering a

graph.

Example 12

Question

The function y is a function of x that satisfies the conditions that y = 45 when x =

8 and

Solution
Even if you have not studied differential equations before, you might realise that

in order to solve this problem you need to reverse the differentiation process; in

other words, you need to integrate. A common challenge when differentiating

or integrating on TMUA and MAT questions is the need to rewrite the function

before you can begin the differentiation or integration.

Example 13

Question

kx kx
The MAT syllabus requires you to know that the derivative of e is ke .

kx
For precisely which values of k does the function y = e satisfy the differential

equation

Solution

Even if you have studied differential equations before, you are unlikely to have

seen one presented like this. Do not be put off by questions of this sort, instead

think about what you do know and what you are expected to know.

kx kx kx
The derivative of e is ke . Starting with y = e and differentiating it

to find and again to find is really the only thing you can do. Once

you have these values, they can be substituted into the equation.
 2kx 2
e is never 0 so this means that you are looking for values that satisfy 2k − k −

6 = 0

The equation is satisfied when .

Example 14

Question

For all real numbers x, the function f(x) satisfies the equation

Find the value of .

Solution

Since you do not even know what the function f(x) is, this may initially seem to

be an impossible question. The first thing to appreciate is that

are the same value. The variable t was just introduced as

x was already in use.

 will give a value. If you let then the equation becomes 8 -

4
f(x) = 3f(-x) + Kx which is already looking more friendly.

This ‘new’ mathematical statement says that the function 8 - f(x) is an

4
identical function to 3f(-x) + Kx ; they give the same output value for any

real value of x, they have the same graphs and, in particular, they give the

same answer when integrated between the same two limits. Why integrate?

4
You want to find so integrating both sides of 8 - f(x) = 3f(-x) + Kx

between -1 and 1 with respect to x will result in at least one term that is

At this stage you are close but you have both .

Since the graph of y = f(-x) is a reflection in the y axis of the graph of y =

f(x), it follows that . You also know that

Exercise 3

1 This question is suitable for the TMUA.

Is the value of T where

zero, positive or negative?

2 What is the degree of the polynomial

Find the coefficient of the highest power of x.

3 For what values of k does the function satisfy the differential equation

4 This question is suitable for the TMUA only.

Which of the following integrals has the largest value?

5 Find the area bound by the line x + y = 7 and

the curve .

6 How many positive value solutions for x does

 the equation have?

7 The function f(x) satisfies the equation

What is the value of ?

Exercise 4

Admissions test multiple choice questions

TMUA style questions

1 Find the value of .

A 0

B 1

C 2

D 3

F 12

2 A curve has equation y = (2a − x)(2a + x) where a is a constant.

A different curve has the equation y = (2a − x)(3a + x).

Find the value(s) of a for which the maximum point on the first curve is

closest to the maximum point on the second.

A 0

D ±0.8

E ±1

3 A function f is such that

The value of is
 A 2

B 3

C 5

D 8

E 11

F 12

4 The trapezium rule approximation for

using four trapezia is equal to

F 30

3 2
5 Find the complete set of values of k for which the cubic equation 2x − 3x −

36x + k = 0 has three distinct real solutions.

A x < 44

B x < 81

C −81 < x < 44

D −44 < x < 81

E x > 44

F x > 81
MAT style questions

4
6 The equation x − 4000x = 40 000 has

(a) no real solutions.

(b) exactly one real solution.

(c) exactly two real solutions.

(d) exactly three real solutions.

(e) an infinite number of real solutions.

2
7 If f(x) = 4x − 12x + 18, what are the coordinates of the minimum point of

(a)

(b)

(c) (4, 3)

(d) (6, 3)

(e)

2 2
8 The diagram shows the ellipse 4x + y − 24x = 32.
 The point P moves on the ellipse. What is the greatest possible distance of P

from the origin?

(a)

(b)

(c)

(d)

(e)

9 Which of these gives the greatest value?

(a)

(b)

(c)
 (d)

(e)

10 Sketch the graph of . ( denotes the largest integer

less than or equal to x.)

(a) Use your graph to find when

(i) a = 4

(ii) a = 2.5

Show that .

(b) Find .

Show that

if k and n are both positive integers.

Find the exact value of .

Give you answer in the form

where a and b are positive integers.

Full worked solutions can be found to all of the questions from Exercise 4 at

www.hoddereducation.co.uk/step-mat-tmua-answers
Chapter 7: Graph sketching, identification and

transformation

TMUA MAT STEP

Introduction

The specifications for the TMUA, MAT and STEP examinations all include

specific sections about the graph sketching skills required. Curve sketching

and identification questions appear frequently on all of the papers. In some

cases, the sketch of the graph is a means to an end whilst in others, the sketch

of a graph illuminates the path through the problem.

Curve sketching is a popular choice of question for university interviews as

they are a good way for students to show how they think.

In this chapter you will see a number of examples of the sort of graph

sketching questions that are found in the TMUA and the MAT multiple

choice section. Some of the examples will focus on the key skills involved in

graph sketching whilst others will encourage you to think about efficient ways

to identify the equation that matches a certain graph. You will learn how to

decide from the list of things you are told to do at A level to sketch a graph,

which are important in a particular situation and which are not.

Two questions to think about

1 Find the fraction of the interval 0 ≤ x ≤ 180° for which the inequality

2sin(2x - 90°) cos(45° - x) ≥ 0 is satisfied.

This question does not state that you have to sketch a graph and you may be

able to find the answer without a sketch. Drawing a sketch for a question like

this will help you to focus on the key information that is needed to answer it.

If you can’t answer this question now, return to it later.

Hints

• What graphs would you sketch to help you see what to do?

• Would you sketch one graph or more than one?

• Does anything you know about graphical transformation help you to

sketch your graph(s)?

• The question asks you for what fraction of the interval the product of the

two trigonometric functions is ≥ 0. What do you know about two

 numbers that multiply together to give a result that is ≥ 0?

2 Sketch the graph of

Show the values of any points where the graph crosses either axis in exact

terms.

Give the equations of any asymptotes.

If you can’t answer this question now, return to it later.

Hints

• What graphs could you sketch to start with?

• Is there a part of the function that you can rewrite in a more helpful form?

• What does this new form tell you about the final sketch graph? Does it

indicate any symmetry? Does it tell you anything about where the graph

exists?

• Does the graph have any vertical asymptotes?

• What happens to the shape of the graph as x → +∞ and as x → -∞?

• What happens to the shape of the graph as it approaches the vertical

asymptotes?
Key methods for sketching graphs

A level tends to encourage you to follow a set procedure for sketching a

graph.

1 Find the coordinates where the graph intercepts the x axis by setting y = 0

and the y axis by setting x = 0.

2 Differentiate and equate the derivative to 0 to find the x coordinates of the

maximum and/or minimum points. Substitute this back into the original

equation to find the y coordinates for these points.

3 Differentiate once more and substitute in the x values at the turning points

to ascertain their nature.

4 Identify any vertical asymptotes by looking for values of x that would result

in a division by 0.

5 Investigate what happens for very large positive and very large negative

values of x by either substituting in a large value or by considering what

happens as x → ±∞.

There is nothing wrong with any of these steps and for some graphs you may

have to do all of them to get a complete idea of the shape of a graph. Using a

step by step approach that is a series of instructions does not encourage you to

think about the behaviour of a function. When you are sketching a graph, it is

important to ask yourself ‘Why am I sketching this graph?’ and ‘What am I

hoping to get from a sketch?’ Curve sketching is an incredibly useful skill for

certain types of mathematical problem solving. At A level you rarely get to

see this as questions tend to be centred on testing whether you know the key

techniques for graph sketching rather than testing what useful information you

get from it. The purpose of graph sketching is to get some key information

about the function you are dealing with; its behaviour at extremes, its

stationary points, where it can and can’t exist and so on. Going through the

graph sketching process is really investigating how a function works.

For the graph sketching and identification questions in admissions tests you

will have to apply some of the techniques above but the important thing is

that you should be choosing which ones give you the best information for

what you want to do. You should always be thinking ‘Which one of these will
tell me the most about the function I am dealing with?’ In the rush to apply a

set procedure some key ideas about a function from other areas of A level can

get forgotten. There are also some techniques that are not a part of A level

that you can understand with what you know from A level. It doesn’t hurt to

have an idea of what is meant by an oblique asymptote or what a point of

inflection shows.

There are several tools at your disposal when you sketch a graph. Each of the

following properties or techniques can be used to identify the key features of

a graph.
Symmetry

Symmetry is an incredibly useful tool when sketching a graph. If you know

that the graph has reflection symmetry about a line then you only need to

identify what half of the graph looks like to be able to sketch the whole thing.
Even functions

A function f of a real value x is an even function if, for the domain of f, f(-x)

= f(x).

The graph y = f(x) of an even function has reflection symmetry in the line x =

0 (the y axis).

There are some obvious examples of even functions that are easy to spot, for
2 2 2
example x for . For all x in the real domain, (-x) = x . In fact, x raised

to any even power will be an even function. cos x is another well-known

example of an even function since cos(-x) = cos x. Well-known even functions

2
such as x and cos x can be taken to be even without requiring any formal

verification. Some functions are obviously even functions but in admissions

tests you will often come across some more disguised examples, for example

for domain is an even function.

If then

since

4 4 2 2
cos(-x) = cos x, (-x) = x and (-x) = x .

So f(-x) = f(x).

It is an even function.
Odd functions

A function f of a real value x is an odd function if, for the domain of f, f( ‒x) =
‒f(x).
Substituting a negative value of x into an odd function will give a result of the

same magnitude as substituting the equivalent positive value but the sign will

be changed (either from + to - or from - to +). Similarly, substituting a

positive value for x will give a result of the same magnitude as the equivalent

negative value but with the sign changed.

The graph of y = f(x) for an odd function f has rotational symmetry (180°)

about the point (0, 0) (the origin).

3
There are obvious examples of odd functions, for example x for . For
3 3
this domain, (-x) = -x . In fact, x raised to any odd power will be an odd

function. sin x is another well-known example of an odd function since sin(-

x) = -sin x.

Odd functions can be quite hard to spot. For example the function is

an odd function. If then

Since f(‒x) = ‒f(x) it is an odd

function.

It is not necessary for a function to be an even or an odd function. Many

functions are neither even nor odd.

Combining even and odd functions

You can discover some important properties of even and odd functions by

trying out examples using known even and odd functions.

Try it out

Decide whether each of the following results in an even function, an odd

function or neither.

1 The sum/difference of two even functions

2 The sum/difference of two odd functions

3 The sum/difference of an even and an odd function

4 The product/quotient of two even functions

5 The product/quotient of two odd functions

6 The product/quotient of an even function and an odd function

Periodic functions

A periodic function repeats its values at regular intervals. For a periodic

function f(x + p) = f(x) for some p. The trigonometric functions sin x, cos x

and tan x are all periodic. sin x and cos x repeat every 2π radians (360°) and

tan x repeats every π radians (180°). The periodic functions you will meet in

admissions tests will have some part that is defined by a trigonometric

function.
Other symmetry

A function may not be odd or even but still have some reflection symmetry.

For example, the graph defined by has reflection

symmetry in the line x = 2.

Rewriting the quadratic expression in the function by completing the square

shows why this is the case.

2
y = (x - 2) + 1 has reflection symmetry in the line x = 2 so

will also have reflection symmetry in x = 2.

Reflection symmetry is not always in a vertical or horizontal line.

The graph of xy(x + y) = 2 has reflection symmetry in the line y = x.

Swapping the x and y terms in the equation results in an identical equation so

the graph has reflection symmetry in y = x.

Try it out

For each of the following equations, identify all of the symmetries of the

corresponding graph.

5
6
Existence

An important step before sketching a graph is to identify its domain and range

as far as possible. For some functions there are some values for x or y for

which the graph cannot exist.

The graph of does not exist for values of x where x < 3. Its

domain is therefore x ≥ 3.

It also does not exist for values of y where y < 0.

The graph of would start at x = 3, y = 0 and exist only in the first

quadrant.

The domain of a graph you are sketching does not necessarily have to be of

the form x ≥ a or x ≤ a. It is possible for a graph not to exist between two

values or not be defined for a value. For example, the graph of does

2
not exist at x = 0 as is undefined for x = 0 and the graph of y = log10 (x -

2x - 3) does not exist for -1 ≤ x ≤ 3.

2
You can see why by considering the domain of y = log10 (x - 2x - 3), which is

2
x > 0, and the function x - 2x - 3 which can be factorised to (x - 3)(x + 1). (x -

3)(x + 1) is less than or equal to 0 for -1 ≤ x ≤ 3.

2
The graph of log (x - 2x - 3) therefore does not exist for -1 ≤ x ≤ 3.
10
Try it out

Sketch the graph of .

Asymptotes

An asymptote of a curve is a line such that the distance between the curve and

the line approaches zero as one or both of the x or y coordinates tend to

infinity.

There are two types of asymptotes encountered in A level mathematics;

vertical asymptotes and horizontal asymptotes. It is also possible for an

asymptote to be oblique. Oblique asymptotes are not a part of the A level

specification but it is useful for you to be able to find them as they do give a

good idea of the way in which a function behaves at its extremes.

Vertical asymptotes

The graph of has three vertical asymptotes. These asymptotes

can all be found by considering where the denominator of the expression is 0.

The values x = 0, x = -1 and x = 1 result in a division by 0.

The three vertical asymptotes are therefore the vertical lines x = 0, x = -1 and

x = 1. Since these are values for which the graph doesn’t exist, the graph does

not cross any of these asymptotes.

When vertical asymptotes have been identified, it is important for you to

consider how the graph approaches them.

Rewriting the equation of the graph as helps to do this.

2 -
As x → -1 from the left (i.e. x takes values less than -1), x(x - 1) → 0
2
(approaches 0 from below) since x will be negative and x - 1 will be positive.

This means that will be a large negative number so you can

conclude that y → -∞ (y will tend to negative infinity).

2 +
As x → -1 from the right (i.e. x takes values greater than -1), x(x - 1) → 0
2
(approaches 0 from above) since x will be negative and x - 1 will be negative
2
(when -1 < x < 1, the size of x will be less than 1). From this you can

conclude that as x → -1 from the right, y → +∞ (y will tend to positive

infinity).

2
As x → 0 from the left, y → +∞ since x will be negative and x - 1 will be

negative.

2
As x → 0 from the right, y → -∞ since x will be positive and x - 1 will be

negative.

2
As x → 1 from the left, y → -∞ since x will be positive and x - 1will be

negative.
 2
As x → 1 from the right, y → +∞ since x will be positive and x - 1will be

positive.

Combining this information with the information about any horizontal

asymptotes gives a good idea of what the graph looks like.

Many graphs do not have vertical asymptotes.

Horizontal asymptotes

Horizontal asymptotes are found by considering the limiting value of y as x

approaches -∞ and +∞.

For the graph of , dividing both top and bottom of the fraction

3
by x (the highest power of x in the denominator) will put the fraction in an

equivalent form that is easier to consider for large positive and negative x.

The graph approaches y = 1 and this is a horizontal asymptote. It is important

to consider from which side of y = 1 the graph approaches. This can be done

by considering the equation in the form

For large positive x, all the terms will be positive so y > 1 and the graph

approaches the asymptote from above.

2
For large negative x, x(x - 1) will be negative so y < 1 and the graph

approaches the asymptote from below.

Unlike vertical asymptotes that are found by looking for values for which the

graph does not exist, it is possible for a graph to cross its horizontal

asymptote. In this case there are no solutions to the equation

so the graph will not cross its horizontal asymptote.

From the horizontal and vertical asymptotes you can begin to build a sketch

of the graph.
There are two useful pieces of information that will help you complete the

sketch.

• Knowing that the final graph will not cross any vertical asymptotes.

3 3
• Both x - x + 1 and x - x are cubic polynomials so any horizontal line

drawn on the final graph will cross it up to three times but no more

(rearranging will result in a cubic equation).

Using that information and joining up the ends of the part-curves that have

already been added gives something like this.

The maximum and minimum points are not known so the graph may not look

exactly like it should but it is showing a lot of the information about the

function.
Oblique asymptotes

The graph of has an oblique asymptote.

As x → ∞, .

This is easier to see if the equation is rewritten as .

2
As x → ∞, subtracting 1 from x has very little effect. The curve therefore

approaches the line y = x. This is an asymptote as it is a line that the graph

approaches as both x and y approach infinity. It is not parallel to the x or y

axis, it is an oblique asymptote.

3 3
Since for large positive x, x - x is less than x , the graph will approach the
3 3
asymptote from below. For large negative x, x - x is greater than x and the

graph will approach the asymptote from above.

You might be tempted to think that every oblique asymptote is of the form y =

mx where but it is quite possible to have oblique asymptotes of the

form . The oblique asymptote of the graph

is one such example.

It can be rewritten as which indicates that there are two

vertical asymptotes, x = 0 and x = 1.

The oblique asymptote can be found by considering the equation in the form

2
Dividing both the top and the bottom of the fraction by x (the highest power

of x in its denominator)

gives .
As x → ∞, .

y = x - 1 is the oblique asymptote.

Try it out

Find out how the graph approaches y = x - 1 for large positive and large

negative x.

Find out how the graph approaches each of the vertical asymptotes.

Sketch the graph of . You do not need to mark the values of

the intercepts with the axes.

Standard graphs

The TMUA specification states that you need to be able to recognise and be

able to sketch the graphs of common functions and these include lines,

quadratics, cubics, trigonometric functions, logarithmic functions,

exponential functions, square roots and the modulus function.

The MAT specification states that you need to know the graphs of quadratics

and cubics, as well as sin x, cos x, tan x, and loga x.

You are expected to be able to solve equations and inequalities with graphs.

Knowing standard graph shapes can help you to answer some admissions test

questions quickly and efficiently.

The following two examples show how standard graphs can be used to answer

graph identification and graph sketching questions.

Example 1

Question

Which of the following shows a sketch of the graph with equation y =

2
log10(x − 4x + 5)?

E
F

Solution

2
y = x − 4x + 5

2
y = (x − 2) + 1

2
The correct graph can be identified by considering the graphs of y = x

− 4x + 5 and y = log10 x

2
The graph of y = x − 4x + 5 is a parabola with minimum point (2, 1).

Two features of this graph allow for three of the options to be

eliminated immediately: its symmetry and intercept.

2
The graph of y = log10(x − 4x + 5) will have reflection symmetry in x =

2
2 because y = x − 4x + 5 has reflection symmetry in x = 2.

This eliminates options C, D and E.

2 2
When x = 0, x − 4x + 5 = 5 so log (x − 4x + 5) = log 5. This means that
10 10

the intercept with the y axis will be the point (0, log10 5).
log10 5 is a positive value since log10 x is an increasing function and log10 5

> log10 1.

The remaining options are B and F. To choose between these, the graphs of
2
y = x − 4x + 5 and y = log10 x can be considered together.

This eliminates options A and G (and E) where the y intercept is

negative.

2
Moving to the right, away from x = 2 on the graph of y = x − 4x + 5,
2
values of x − 4x + 5 increase. The reflection symmetry means that this
2
will also happen in the other direction too. As x → ∞, x − 4x + 5 → ∞.

The graph of y = log10 x also tends to infinity as x tends to infinity. This

2
will result in the graph of y = log (x − 4x + 5) approaching ±∞ as x →
10

±∞.

2
The gradient of y = log x approaches 0 as x → ±∞. The gradient of y = x
10

2
− 4x + 5 as x → ∞ increases. This will result in the graph of y = log10(x −

4x + 5) ‘flattening off’ more rapidly than the graph of y = log x. It will

10

have a positive gradient that tends to 0 as x → ∞.

Option B shows a graph where the gradient approaches 0 as x → ∞.

Option G shows a graph where the gradient increases as x → ∞.

2
Option B shows the graph of y = log10(x − 4x + 5).
Try it out

2
Are there any other features of the graphs of y = x - 4x + 5 and y = log x
10

that allow you to choose between option B and option F?

Example 2

Question

Sketch the graph of .

Solution

y = cos4x is an even function.

y = sinx is an odd function.

To sketch this graph it is obviously helpful to have a sketch of both y =

cos 4x and y = sin x. To make this as simple as possible, the x axes of

the two graphs should line up as much as possible.

The product of an even and an odd function will be an odd function so

will have rotational symmetry about (0, 0).

The two sketch graphs also show where the graph of

crosses the x axis.

y < 0 everywhere else.

The graph of y = Acos 4x is the transformation of the graph of y = cos 4x by

a stretch factor of A applied parallel to the y axis. The result of this is that

the graph of y = Acos 4x is constrained by the lines y = A and y = -A. This is

because the range of cos 4x is between +1 and -1

The graph of y = cos 4x sin x will be constrained within the curves y = +sin

x and y = -sin x. The multiplier of sin x can be treated as a variable stretch

factor.
To finally sketch the graph of y = cos 4x sin x, some more information

is required.

The final sketch graph could have been drawn by considering the graph of

y = sin 4 being constrained by y = +cos 4 x and y = -cos 4 x but this is more

difficult to imagine. (See if you can.)

Try it out

Sketch the graphs of

Graphical transformation

The MAT specification states that you need to know the relations between the

graphs y = f(ax), y = af(x), y = f(x - a), y = f(x) + a and the graph of y = f(x).

The TMUA specification expects knowledge of the effect of simple

transformations on the graph of y = f(x) as represented by y = af(x), y = f(x) +

a, y = f(x + a), y = f(ax) with the value of a. positive or negative, together

with compositions of these transformations.

The following two examples illustrate how you may use graphical

transformations to sketch or identify a graph.

Example 3

Question

Sketch the graph of

Solution

x2 - 2x + 5
The equation y = 2 can be rewritten in a more convenient form

by first completing the square with its quadratic part and then by

using laws of indices, in the form y = af (x - b).

 x2
If f(x) = 2 then the graph being asked for is y = 16f(x - 1)

x2
This can be sketched by transforming the graph of y = 2 by first

translating it by and then stretching it by a scale factor of 16

parallel to the y axis.

x2
The graph of y = 2 can be sketched by considering the graphs of y =
x 2
2 and y = x .

2
x
From these two graphs it is clear that the graph of y = 2 has

reflection symmetry in the y axis. As x → ∞, y → ∞.

Translate then stretch parallel to y axis by a scale factor of 16.

This graph is translated by and stretched by a scale factor of 16

x2 - 2x + 5
parallel to the y axis to get y = 2 .

Example 4

Question

The graph of y = cos 2x is reflected in the line y = 1 and then in the line

The resulting graph has equation

(a)

(b)

(c)

(d)

(e)

Solution

This example is really a test of how well you can apply your knowledge of

graphical transformation.
A sketch of the graph of y = cos 2x and the lines y = 1 and is a

sensible starting strategy.

After sketching, apply each of the transformations in turn.

You can draw the results using the same axes but for this example the

diagrams are separated for clarity.

Reflecting the graph in y = 1:

Reflecting in :
This is clearly a translation of y = cos 2x by 2 units parallel to the y axis.

The correct answer is therefore option (d) y = 2 + cos 2x.

Try it out

The graph of the function can be can be obtained from the

graph of by

(a) A stretch parallel to the x axis followed by a translation parallel to the y

axis.

(b) A stretch parallel to the y axis followed by a translation parallel to the y

axis.

(c) A translation parallel to the x axis followed by reflection in the y axis.

(d) Reflection in the y axis followed by a translation parallel to the y axis.

(e) A translation parallel to the x axis followed by a stretch parallel to the x

axis.

(f) A translation parallel to the x axis followed by a stretch parallel to the y

axis.
Turning points

Turning points may be expected when you are asked to sketch a graph or they

may simply be useful to know.

Calculus is not always required to find turning points and it is useful to know

how to do this by non-calculus methods.

The following two examples show how non-calculus methods can be used to

find turning points

Example 5

Question

3 2
Without using differentiation, find the turning points of y = x - 3x - 9x + 2.

Solution

The wording makes it clear that there are turning points so it is safe to

assume that the graph will follow a typical cubic graph shape with two
3
turning points rather than one inflection. The coefficient of x is positive so

the graph will look something like the sketch shown.

The horizontal straight line y = k will have 1, 2 or 3 intercepts depending

on the value of k.

The local maximum and local minimum points on the graph occur where y

= k has two intercepts.

 3 2
The solution to x - 3x - 9x + 2 = k will have two roots if k is equal to the y

coordinate of the maximum or minimum point.

The y coordinates of the turning points can be found by considering

3 2
the roots of the equation x - 3x - 9x + (2 - k) = 0.

For the turning points, the line y = k touches the graph so the turning points
3 2
correspond to a repeated root of x − 3x − 9x + (2 − k) = 0.

3 2
Let the repeated root of x − 3x − 9x + (2 − k) = 0 be a and the other root be

b.

3 2 2
Hence x − 3x − 9x + (2 − k) ≡ (x − a) (x − b)

2
Equating coefficients of x gives −b − 2a = −3 so b = 3 − 2a

(A)

2
Equating coefficients of x gives a + 2ab = −9

(B)

2 2
Equating the constants gives −a b = 2 − k so k = 2 + a b

(C)

Substituting (A) into (B) gives

This gives the x coordinates of the turning points x = 3 and x = -1.

a = 3 or a = -1

From (A), when a = 3, b = -3 and when a = -1, b = 5

Substituting a = 3 and b = -3 into (C) gives k = 2 − 27 = −25

Substituting a = -1 and b = 5 into (C) gives k = 2 + 5 = 7

The turning points are (-1, 7) the maximum and (3, -25) the minimum.

Example 6

Question

Sketch the graph of

Solution

Using techniques covered earlier in this chapter, an initial sketch of the

graph looks as shown in the diagram.

You may wish to make sure you are able to get to this stage yourself.
It is clear that the graph does have turning points. These can be found

by considering the straight line y = k.

(A)

At the turning points this equation will have only one solution. Provided k

≠ 1, (A) will be a quadratic equation and its discriminant will be 0 when

the graph meets the line y = k at one point only i.e. at the turning points. (If

k = 1, (A) will be linear and there will be only one solution which

corresponds to the curve crossing the horizontal asymptote).

The y coordinates of the turning points are .

The values of k can be substituted into equation (A) to find the x

coordinates of the turning points.

This seems like a challenging process but it can be simplified as, for those

values of k, the equation will have a repeated root and can therefore be
2
written in the form (x - a) = 0.

Working in terms of k and leaving the substitution to the end of the

calculation will make life easier for this case.

If

then the value of a can be found by equating either the coefficients of x

or the constants.
Although the graph looks horizontal to the naked eye as x tends to -∞, the

graph passes through the line y = 1 and then moves back up towards it. It

must therefore have a turning point.

Try it out

Sketch the graph of .

Hint

This can be rewritten as .

Example 7

Question

Which of the following shows the graph defined by the implicit equation
2 2
xy(x + y ) = 5?
(a)

(b)

(c)
(d)

(e)

Solution

What are the features of each graph that makes them differ from

another graph?

Can any graphs be ruled out?

Considering symmetry: The graph has reflection symmetry in y = x

Swapping x and y in the equation gives the same equation.

Options (b) and (c) can be eliminated immediately.

For x = 1 and y = 1 in

LHS = 1 + 1 = 2

RHS
The curve does not pass through (1, 1) so option (a) can be eliminated.

2 2
If x > 0, so y(x + y ) > 0

2 2
y > 0 since (x + y ) > 0 for all real non-zero x and y.

The correct graph must be option (d).

There is an alternative way to eliminate options (b) and (c).

Rewriting the equation as indicates that the graph contains no

points for which x = 0 or y = 0. The graph does not exist for x = 0 and y =

0.

Both (b) and (c) pass through two points where x = 0.

There are three graphs remaining. The key differences are:

• Curve (a) is the only one that passes through the point (1, 1).

• Curve (e) is the only one that exists in the second and fourth

quadrants.

Testing the point(s) on the graph where x = 1 and y = 1 in the equation

will show if option (a) is correct or can be eliminated.

The final test is to see if the graph should exist in the fourth quadrant.

This will show if option (e) is correct or can be eliminated.

2 2 2 2
x + y must be positive so y must be positive for the product y(x + y ) to

be positive.

The graph does not exist in the fourth quadrant so option (e) can be

eliminated.
Example 8

Question

Sketch the graph clearly showing any intercepts with the axes.

An initial approach does not require calculus.

Solution

The graph is defined for all real x.

Does the graph exist for all real values of x?

Does the graph have any obvious symmetry?

is an odd function.

The graph has rotational symmetry about the origin.

The graph passes through the origin.

When x = 0, y = 0.

3
When y = 0, x - x = 0

2
x(x - 1) = 0

x = 0, x = 1 or x = -1

As

so y = x is an asymptote.

For x < -1, y < 0

For -1 < x < 0, y > 0

For 0 < x < 1, y < 0

For x > 1, y > 0

The graph also passes through the points (-1, 0) and (1, 0).

The next step is to consider the shape of the graph for large positive values

of x.

-
Since the graph has rotational symmetry about (0, 0), as x → ∞, y → x

so (y = x) is an asymptote in that direction too.

What is the shape of the graph either side of where it crosses the x

axis?

There is enough information to draw an initial sketch of the graph.

This sketch is good enough for most purposes, it gives a good idea of the

general shape and behaviour of the function.

Using differentiation to improve the sketch

In some aspects the sketch above does not give all of the detail that it

could.
It seems to imply that the local maximum point is at and the local

minimum point is at .

Using differentiation allows for a much better sketch.

For the MAT and TMUA examinations, the chain rule for differentiation is

not required so it seems that the maximum and minimum points cannot be

found. When it seems like you are being asked to do something that is not

on the specification, there is usually an alternative method you can find that

uses what you are expected to know. For this example it is possible to use a

different function to find the maximum and minimum points. The cube

root function is an example of an increasing function i.e. the cube root of a

value will be greater than the cube root of a lower number. You are
3
working with the function in this example. A maximum value of x
3
- x will result in a maximum value of and a minimum value of x -

x will result in a minimum value of . For this reason, the function

inside the cube root can be used to find the maximum and minimum

points.

For a student who knows the chain rule an even better shape for the graph

can be found.
From

3 2
As x → -1, (x - x) → 0. The gradient of the graph approaches that of a

vertical line.

3 2
As x → 0, (x - x) → 0. Again, the gradient of the graph approaches that of

a vertical line.

3 2
As x → +1, (x - x) → 0. Once again, the gradient of the graph approaches

that of a vertical line.

Putting the new pieces of information with what was found before leads to

the shape shown in the diagram.

From the shape of the graph the maximum occurs when

and the minimum when .

Exercise 1

1 Identify all of the symmetries of the graphs of the following equations.

For equations written in the form y = f(x), state whether f(x) is an even

function, an odd function or neither.

For reflection symmetry state the equation of the line of reflection

symmetry.

For a graph with multiple parallel lines of reflection symmetry give a

general equation that describes all of those lines in terms of n where n ∈

For rotational symmetry give the centre of rotation as a coordinate.

For multiple points of rotational symmetry give a general equation that

describes all of the points in terms of n where n ∈ .

For the graph of a periodic function, give the period of the function.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

2 Use the formula for differentiation from first

 principles, , to show that the derivative of an

even function is an odd function.

Show that the derivative of an odd function is an even function.

Hint

You are starting with an even function so f(-x) = f(x)and you need to

show that

f ′(-x) = f ′(-x).

3 Find the domain for each of the following functions and hence state where

the graph of y = f(x) exists in each case.

(a)

(b)

(c)

(d)

(e)

4 For each of the following equations, identify any vertical, horizontal

and/or oblique asymptotes that the graph will have. State how the graph

approaches each asymptote.

(a)

(b)

(c)

(d)
 (e)

5 For each region in the Venn diagram below, find an example of a graph of

a rational function, where p(x) and q(x) are both polynomials in x.

represents the set of all rational functions.

A represents the set of rational functions with vertical asymptotes.

B represents the set of rational functions with horizontal asymptotes.

C represents the set of rational functions with oblique asymptotes.

6 Sketch the following graphs.

(a)

(b)

(c)

(d)

(e)
7 For each of the following pairs of graphs, write down two linear

transformations that when applied in order will transform graph A to the

graph B.

(a)

(b)

(c)

(d)

(e)

8 Without differentiating, find the turning points for the following graphs.

(a)

(b)

(c)

(d)

9 Sketch each of the following graphs showing

• any intercepts with the axes

• any asymptotes, giving their equations

• any stationary points, giving their coordinates.

(a)

(b)

(c)

(d)

(e)

Hints
 Try to be as efficient as possible.

Think about what you need to do and the best order in which to do those

things.

10 Solve the equation .

11 Solve the equation .

12 How many real solutions does the equation have?

13 Solve the equation .

14 How many real solutions does the equation have?

Exercise 2

Admissions test multiple choice questions

TMUA style questions

1 f(x) is a quadratic function in x.

Given that the graph of y = f(x) passes through the point (-1, -3) and has a

vertex at (1, 5), which of the following is an expression for f(x)?

2 Find the complete set of values of x, with −π ≤ x ≤ π, for which (1 − 2

cos 2x)(1 + 2sin x) ≤ 0.

3 Find the fraction of the interval

for which the inequality

is true.

A
 B

4 It is given that for

The complete set of values for which y is negative is

5 Find the number of solutions of the equation

A 0

B 1
 C 2

D 3

E 4

6 How many solutions does the equation have in the

interval ?

A 0

B 1

C 2

D 3

E 4

F 5

G 6

7 How many real roots does the equation

have?

A 1

B 2

C 3

D 4

E 5

8 Find the complete set of values of the constant k for which the quartic

equation has four distinct real solutions.

G
MAT style questions

9 The equation in x, has three real solutions

when

(a)

(b)

(c)

(d)

(e)

10 The equation has three real roots (including possible

repeated roots) for

(a) all real values of a

(b)

(c)

(d)

(e)

11 Into how many regions is the plane divided when the following curves

are drawn?

3 2 3 2
y = x − x , y = x − 4x, y = x

(a) 11

(b) 10

(c) 9

(d) 8

(e) 7

12 Which of the graphs below shows a sketch of

(a)

(b)

(c)
 (d)

(e)

This question requires radians so would not appear on a MAT paper. It

is still a useful exercise in graph identification.

13 Which of the following is a sketch of the graph of

(a)
 (b)

(c)

(d)

(e)

4
14 Which of the following shows a sketch of the graph of y = log (x −
10

16)

(a)

(b)

(c)
 (d)

(e)

15 The graph of all the points (x, y) in the xy-plane that satisfy the

equation sin(x + y) = 1 is shown in

(a)

(b)
(c)

(d)

(e)
This question requires radians so would not appear on a MAT paper. It is

still a useful exercise in graph identification.

Full worked solutions can be found to all of the questions from Exercise 2 at

www.hoddereducation.co.uk/step-mat-tmua-answers
Chapter 8: Sequences and series

TMUA MAT STEP

Introduction

It is not possible to compute the exact value of some numbers, such as π, e or

sin (2). You can, however, compute values that are as close as you wish to

them. Doing this almost always involves working with sequences and series.

Being able to work with sequences and series is essential to all areas of

mathematics.

The specifications for the TMUA, MAT and STEP examinations all include

specific sections about sequences and series. Questions about sequences and

series appear frequently on all the papers.

Sequences and series can be defined in many ways. In questions about

sequences and series it is often helpful to do some calculations to get a feel

for the values involved and any patterns in their properties and behaviour.
A question to think about

This question assumes knowledge of the usual notation for sequences and

series. This is covered briefly elsewhere in this chapter.

n + 1
Let an be the sequence given by an = (−1) n.

Let .

For which value of n is b = 1000?

n

A 1999

B 2000

C 1000

D 999

E 501

F 499

Hints

• What are the explicit values of the first few terms in the sequence a ?
n

• Make sure you are happy with the way that the sequence b is defined in
n

terms of the sequence a . Ask yourself what is b in terms of values in the

n n

sequence an for example?

• What are the explicit values of the first few terms in the sequence b ?
n

• Can you spot any patterns? Have you calculated enough values to be able

to spot any patterns and feel confident that they persist for the entire

sequence?

As with many questions on sequences and series the notation for this question

may make it appear more tricky than it really is. Often when you remove the

notation in questions like this and start writing down explicit terms it

becomes much easier to spot patterns.

If you can’t answer this question now, return to it later.

Key methods for sequences and series

Many A level questions on this topic are about arithmetic and geometric

sequences and series. These are defined in a certain way with a very clear

pattern and, over time, you become familiar with them. Although these

examples remain important, in TMUA, MAT and STEP you may need to look

at sequences and series that you have not necessarily considered before. You

will often need to calculate some values to get a feel for any patterns or

behaviours.

Admissions test questions on sequences and series provide great opportunities

to get candidates exploring and pattern spotting. With enough practice you

should be able to start to spot questions where this approach is likely to be

needed.

th
There are essentially two ways to define a sequence. The first is where the n

term of the sequence is given in terms of n itself. Here are some examples (in
th
each case where for a positive integer n, a is the n term of the sequence).
n

th
The other way is to describe how you can find the n term in the sequence

using earlier terms in the sequence that you already know. To describe a

sequence in this way you need to be told the values of suitable terms early in

the sequence in order to start this process off. A simple example of this is a
1

= 1 and then an + 1
= 2an for all integers n ≥ 1.

This is enough information to, in theory, calculate any term in the sequence.

On the face of it this means that, in the example above, if you want to know

a1000 you need to know what a999 is, and for a999 you need to know what a998

is… and so on. Proceeding this way you would calculate all of a1, a2, …, a999

before you found a .

1000

Actually, for the sequence defined by a = 1 and then a = 2a for all

1 n +1 n

integers n ≥ 1, it is pretty straightforward to write down a in terms of n.

n
Check that you can do it.

The Fibonacci sequence is a more complicated example. The first two terms

are 1 and 1 and then any other term is the sum of the two previous terms (for

example a = a + a ). This can be written as a = 1, a = 1 and a = a

6 5 4 1 2 n + 2 n + 1

+ a for all integers n ≥ 1.

n

It is worth noting is that if you are asked about a term a long way into a
th
sequence which is defined recursively, for example a2020 (the 2020 term if

the sequence begins with a ) then there’s a good chance that there is a pattern
1

in the terms of the sequence that will enable you to work it out.
Sigma notation for series

This has already been used in the question to think about.

Consider the summation

This is a geometric series with 10 terms. It is slightly easier to see this if you

write the summation out without the sigma notation:

Remember that in a summation of the form the values you are

summing could be independent of the ‘dummy’ variable r which then only

acts as a ‘term counter’.

For example in there are n terms but every one of them is 6. So

where there are n sixes on the right hand side.

This means that .

A summation like can be split into separate sums. It is

worth thinking about why this can be done. Here is the detail for this

example.
Summing finitely many terms from arithmetic

and geometric sequences

For a positive integer n:

• the sum , which in sigma

notation would be written as , is equal to

• the sum , which in sigma notation would be

written as , r ≠ 1, is equal to . For r > 1, the equivalent

form may be preferred.

Note that the formulae above deal with finitely many terms. You will come

back to consider what it means to sum infinitely many values later.

Try it out

Let an be the sequence defined by a1 = 1 and . Find a

formula for a in terms of n only.

n
Connections between summations and algebraic

identities

There are some really useful connections between sums of geometric

progressions and algebraic identities. For example for any positive integer n

the following is an identity:

This means that for any positive integer n and any x ≠ 1 the following will be

true.

Setting x ≠ r and multiplying through by a gives the formula that you saw

earlier for the sum of n terms in the geometric progression with first term a

and common ratio r (where x ≠ 1).

In other words

Try it out

How does the identity connect

to the formula for the sum of certain geometric progressions?

More on notation (and getting rid of it)

Some sequences that initially look tricky can be simplified by changing the

way they are written.

Consider the sequence given by where n

≥ 1. At first glance it is not obvious how the sequence “works” but with a

little investigation, calculating its terms becomes a straightforward exercise.

When n = 1, 3n = 3 which is odd, 3n + 1 = 4 which is even and 3n + 2 = 5

which is odd. This means that .

When n = 2, 3n = 6 which is even, 3n + 1 = 7 which is odd and 3n + 2 = 8

which is even. This means that .

When n = 3, 3n is odd, 3n + 1 is even and 3n + 2 is odd and so a3 = -24.

When n = 4, 3n is even, 3n + 1 is odd and 3n + 2 is even and so a4 = 24.

In fact, when n is odd, an = -24 and when n is even, an = 24.

n
So actually this sequence could have been described as an = (-1) × 24.

Notice how the sequence feels much more manageable when written like this.

Try it out

1 The terms in each of these sequences can be expressed in a different and

arguably more straightforward way. How?

(a)

(b)

(c)

This question requires radians so would not appear on a MAT paper.

2 Find a formula in terms of n for each of the following.

(a)
(b)

(c)
Properties of sequences
Increasing and decreasing sequences

A sequence (an) is said to be increasing if an +1

≥ an for all integers n ≥ 1

and strictly increasing if an +1

> an for all integers n ≥ 1.

So for a sequence to be strictly increasing, infinitely many inequalities: a2 >

a1 and a3 > a2 and a4 > a3 and so on must be satisfied.

For example the sequence in which for all n ≥ 1 is a strictly

increasing sequence (and therefore also an increasing sequence). Again, it is

easier to see this if you write out the terms ‘long hand’.

You should be able to write down the definitions of a decreasing and a

strictly decreasing sequence by writing out the above again and making the

appropriate small change in each case.

Try it out

Write down an example of sequence a which is both increasing and

n

decreasing (it is possible!).

Why is it not possible to find a sequence an which is both strictly increasing

and decreasing?
Periodic sequences

A sequence a1, a2, a3, … is said to be periodic with period k if there is a

positive integer k such that an + k

= an for all positive integers n and there is no

positive integer m with 0 < m < k such that an + m

= an for all positive integers

n.

Suppose that you had a periodic sequence where the positive integer k in the

statement above was 3. This would mean that a = a for all positive
n + 3 n

integers n and that it is not true to say that an + 1

= an for all positive integers

n, or that an + 2
= an for all positive integers n.

So the sequence would satisfy a1 = a4 and a2 = a5 and a3 = a6 and a4 = a7 and

a = a and so on forever.
5 8

This means that the sequence

a1, a2, a3, a4, a5, a6, a7, …

actually looks like this

a1, a2, a3, a1, a2, a3, a1, …

Try it out

1 Show that the sequence defined by a1 = x and for n ≥ 1 is

periodic with period 2 for all but one non-zero value of x.

2 Show that the sequence defined by a = x, a = y and for n

1 2

≥ 1 where x ≠ 0, y ≠ 0 is periodic with period 6 for all but one value

of x.

Example 1

Question
Assume x ≠ 2 throughout this example.

Show that the sequence defined by a = x and for n ≥ 1 is

1

periodic with period 2 for all but two values of x.

State the two values of x for which the sequence is not periodic with period

2.

Solution

To ease yourself into this question, it is certainly worth calculating the

first few terms in terms of x.

So a3 is the same as a1. Since the relationship between a4 and a3 is the same

as the one between a2 and a1 it follows that a4 = a2. Similarly it follows that

a5 = a3 and so on.

Therefore x = a1 = a3 = a5 … and and the

sequence is periodic with period 2 unless .

Assuming x ≠ 2
Limits of sequences

When the terms in a sequence get closer and closer to a particular fixed value

you say that the sequence converges. The phrase ‘closer and closer’ needs to

be defined more rigorously and that is the concern of Analysis in university

level mathematics.

For example, consider again the sequence given by where n ≥ 1.

The terms are listed below in order.

Here are those values plotted, the subscript n is the x coordinate and the value

of a is the y coordinate.
n

You can see that the terms get closer and closer to the value 1. In fact they get

as close as you care to specify.

For example the distance between and 1 is and all of the

subsequent terms a100, a101, a102 are within a distance of 1 too.

Then the distance between and 1 is and all of the

subsequent terms a1000, a1001, a1002,… are within a distance of 1 too.

You can say that the limit as n → ∞ of an is 1 and write limn → ∞

an = 1.
Now consider the sequence for n ≥ 1.

Looking at graphs of and will help with this.

You can see that an appears to be a decreasing sequence. If you can, use a

graphics package to look at the graphs above for larger values of x. You will

see that they are very close. This makes sense as it is possible to imagine, for

example, that and are very close to one another.

In fact, there is a way to see that the values in this sequence are very small.

By the difference between two squares formula you have

for any integer n ≥ 1.

This gives that for any integer n ≥ 1.

With a little thought, you should be able to see that will take

very small positive values for large values of n.

This means that tends to 0 as n → ∞.

Try it out

1 Without using a calculator, prove that .

2 Prove that .

Hence deduce that the sequence for n ≥ 1 tends to 0

as n → ∞.
Partial sums and infinite series

A series with infinitely many terms such as is only assigned a value

if the sequence of partial sums:

gets ‘closer and closer’ to a particular value b (as stated earlier ‘closer and

closer’ will be defined more rigorously in university level mathematics).

If it does then you can say .

This means that the sum which has infinitely many terms

can only be assigned a value if the sequence gets closer and closer

to a particular value.

This only happens when -1 < r < 1, in which case gets closer and

closer to .

Therefore you can say that

If r ≤ -1 or r ≥ 1 then you say that does not

converge (and it can’t be given a value).

Example 2

Question

What is the sum of the first 2n terms of the following series?

Given that the series converges, calculate its limit.

Solution

Looking carefully at the series reveals it to be formed by alternately

interleaving terms from the two geometric series:

Each of these has a common ratio of .

The sum of the first 2n terms of the original series can be found by adding

the sums of the first n terms of each of the geometric series to give:

You are given that the original series converges so its limit is
Sequences and induction (STEP only)

Suppose you have a statement P(n) about the natural number n. For example

P(n) could be

n
‘The value 10 - 1 is divisible by 9 for all natural numbers n.’

This is really infinitely many statements, one for each value of n. In fact it is a

sequence of statements!

1
P(1) is the statement ‘10 - 1 is divisible by 9’.

2
P(2) is the statement ‘10 - 1 is divisible by 9’.

3
P(3) is the statement ‘10 - 1 is divisible by 9’.

With a little thought you will probably realise that all three of these

statements are true and that P(4), P(5) and so on are true too.

However, just feeling very confident about this is not sufficient for you to

consider it proven. There are examples where an apparent pattern observed

for part of a sequence of statements like this does not continue throughout the

statements, i.e. it does not continue ‘forever’. So you need to be careful.

There are various ways that you might prove a statement like this and here is

one of them. This will be presented informally first of all and then formalised

later.

1
It is clear that P(1) is true because P(1) is the statement ‘10 - 1 is divisible

by 9’ i.e. ‘9 is divisible by 9’. Nobody can argue with that.

2
Let’s look at the second statement ‘10 - 1 is divisible by 9’. That’s just ‘99 is

divisible by 9’. OK so this one is pretty indisputable too. But that would be

missing the point in the proof method that is going to be used here. In this

method a connection between the truth of P(1) and the truth of P(2) is

needed. The same connection should exist between the truth of P(2) and the

truth of P(3) and between the truth of P(3) and the truth of P(4) and so on.

Let’s return to P(1) which says ‘9 is divisible by 9’.

P(2) then says ‘99 is divisible by 9’.

Starting from the fact that P(1) is true you can argue that P(2) is true using

the following steps. (Note that this will feel pretty strange when you first read

it, just keep going.)

1 P(1) is true.

2 In other words 9 is divisible by 9.

3 Multiplying by 10 therefore gives a value that is also divisible by 9.

4 In other words 90 is divisible by 9.

5 Adding 9 gives a value which is also divisible by 9.

6 In other words 99 is divisible by 9.

7 Therefore P(2) is true.

This feels like a strange way to prove that 99 is divisible by 9. Strange until

you realise the same key steps (in bold above) show you that if P(2) is true

then P(3) is true.

1 P(2) is true.

2 In other words 99 is divisible by 9.

3 Multiplying by 10 therefore gives a value that is also divisible by 9.

4 In other words 990 is divisible by 9.

5 Adding 9 gives a value which is also divisible by 9.

6 In other words 999 is divisible by 9.

7 Therefore P(3) is true.

Hurray! One more for good measure.

1 P(3) is true.

2 In other words 999 is divisible by 9.

3 Multiplying by 10 therefore gives a value that is also divisible by 9.

4 In other words 9990 is divisible by 9.

5 Adding 9 gives a value which is also divisible by 9.

6 In other words 9999 is divisible by 9.

7 Therefore P(4) is true.

You could go on. Can you imagine writing out another 96 versions of this

which end with ‘Therefore P(100) is true.’?

This example shows the principle of mathematical induction and how the

statement P(n) can be proved for any natural number n.

Writing it out in full as you have just seen is not really necessary and the

whole process can be simplified by:

1 Demonstrating that an initial (basis) case is true. This is usually showing

that a statement is true for n = 1.

2 Assuming the statement is true for some particular case where n = k.

3 Showing that if the statement is true for the n = k case then it is also true

for the n = k + 1 case.

4 Concluding the argument by a statement that makes it clear that, if in

general, the case when n = k is true means that the case when n = k + 1 is

also true, as the case when n = 1 has been demonstrated to be true then the

case when n = 2 is true, so the case when n = 3 is true, so the case when n

= 4 is true, and so on.

n
The proof for the statement P(n), 10 - 1 is divisible by 9 for all natural

numbers n, could be written:

1
P(1) = 10 − 1 = 9 so P(1) is true.

k
Assuming that P(k) for some particular k , is true i.e. 10 - 1 is divisible

by 9.

k
Let 10 - 1 = 9m where m

k + 1 k
For P(k + 1): 10 - 1 = 10(10 − 1) + 9 = 10 × 9m + 9 = 9(10m + 1)

k + 1
Since 10m + 1 is a natural number P(k + 1) is also true and 10 - 1 is

divisible by 9.

Conclusion: If P(k) is true then P(k + 1) is also true. Since P(1) is true then,

by induction, P(n) is true for all natural numbers n.

The following example shows another proof by induction.

Example 3

Question
Prove the result

for all natural numbers n by mathematical induction.

It is worth getting a feel for what this is saying by substituting a

number for n.

Solution

n = 1

The statement is true for n = 1.

Assuming the statement is true for a particular n = m then

You would usually assume the statement is true for n = k next but as k

has already been ‘used’ in the statement, a different variable name is

needed.

For n = m + 1
The n = m + 1 can be found by using the sum for the n = m case and

adding the ‘next’ term. If this can be rearranged to the correct form

for the result then the statement is true for n = m + 1.

so the statement is true for n = m + 1

Conclusion: If is true for n = m, it is also true for

n = m + 1. As it is true for n = 1 then, by induction, it is true for all natural

numbers n.
Exercise 1

1 What is the sum of the first 100 odd integers?

2 Let an for n ≥1 be the sequence defined by

How many times does each positive integer appear in this sequence.

Prove your result.

3 Show that the sequence defined by a = x, a = y and for n

1 2

> 1 where x ≠ 0 and y ≠ 0 is periodic.

What is the period of this sequence?

4 Find necessary and sufficient conditions on k for the sequence

, x ≠ -k to have period 4.
1

2
5 A sequence is defined by a = sin (1), a = 4a (1 - a ).
1 n + 1 n n

Conjecture and prove by induction, a formula for an in terms of n.

6 An arithmetic sequence has third term 32a + 19b and tenth term 18a +
th
12b where a and b are constants. What, in terms of a and b, is the 16

term.

7 Observe that

1 = 1

1 − 4 = −(1 + 2)

1 − 4 + 9 = 1 + 2 + 3

1 − 4 + 9 −16 = −(1+ 2 + 3 + 4)

Guess the general law and state it carefully.

 Prove it by induction.

8 Observe that

Guess the general law and state it carefully.

Prove it by induction.

9 Guess the general law which simplifies the product

and state it carefully.

Prove it by induction.

n n n
10 Let an = 101 − 99 and bn = 100 .

n n
Expand (1 + x) − (1 − x) .

Hence or otherwise prove that a > b for all n > 50.

n n
Exercise 2

Admissions test multiple choice questions

TMUA style questions

1 The sequence t is given by:

n

a
The value of t is 10 where a is equal to
100

2 The sequence tn is given by:

As n → ∞, the value of tn tends to

F
3 S is a geometric sequence with a common ratio r where r > 1.

The sum of the first 8 terms of S is equal to four times the sum of the

first 4 terms of S.

The common ratio of S is

4 An arithmetic sequence has first term a and common difference d.

The sum of the first 100 terms is one hundred times the sum of the first

10 terms.

Which of the following describes the relationship between a and d?

A a = 10d

B d = 10a

C a + d = 90

D a - d = 90

E a = 2d

F d = 2a

5 The geometric progression A has a common ratio r where −1 < r < 1 and

terms a1, a2, a3, a4, …

The geometric progression B has terms

The sum to infinity of B is denoted by B and the sum to infinity of A is

∞

denoted by A .
∞

Given that , find the common ratio, r, of A.

A
B

E
MAT style questions

6 A sequence tn has the property that

for every n ≥ 2.

Given that a1 = 3 and a2 = 12, what is a2020?

(a) 1.5

(b) 3

(c) 6

(d) 12

(e) 24

7 A sequence an has the property that

a = 2ra
n + 1 n

for n ≥ 2 where r is a constant.

Given that the first term a = 2, what is the product of the first 20 terms
1

of the sequence?

200 190
(a) 2 r

20 190
(b) 2 (2r)

210
(c) (2r)

190
(d) (2r)

190
(e) 2r

8 For a real number x we denote by the largest integer less than or equal

to x.

A sequence a is defined by
n

The sum of the first 45 terms is

(a) 285
 (b) 420

(c) 855

(d) 1125

(e) 2125

9 The function F(n) is defined for positive integers n by

F(n) = the sum of the squares of the digits of n.

2 2 2
For example F(234) = 2 + 3 + 4 = 29.

is equal to

(a) 1650

(b) 1710

(c) 1725

(d) 1750

(e) 1800

10 The sum of the first 2n terms of the sequence

1, 64, 2, 32, 4, 16, 8, 8, 16, 4, …

is

n 7 - n
(a) 2 (1 + 2 )

7 7 - n
(b) 2 - 2

9 9 - n
(c) 2 - 2

n 7 - n
(d) 2 - 2 + 127

n 9 - n
(e) 2 - 2 + 511

Full worked solutions can be found to all of the questions from Exercise

2 at www.hoddereducation.co.uk/step-mat-tmua-answers
Chapter 9: Trigonometry

TMUA MAT STEP

Introduction

The TMUA syllabus states that you are expected to know the sine and cosine

rules and the area of a triangle in the form sin C. Radian measure and

the formulae for arc length and area of a sector and segment of a circle are

expected to be known. You need to know the values for sine, cosine and

tangent of 0°, 30°, 45°, 60° and 90° as well as the graphs for each function,

their periodicity and symmetries. You are expected to know two identities,

tan and You are expected to be able to solve

trigonometric equations in a given interval for which there may be multiple

solutions.

The MAT syllabus states that you are expected to be able to solve

trigonometric equations, know the periodicity of the functions and the sine

and cosine rules. You are expected to know the same two identities as well as

the relationship between sine and cosine: sin(90° - x) = cos x.

The key difference between the two examinations is that the TMUA could

give questions that use either degrees or radians as a measurement of angle,

whereas all questions on the MAT use only degrees. The only trigonometric

functions that should feature are sinx, cosx and tanx.

Some standard values

You need to be able to recall these very quickly. Neither examination allows a

calculator to be used. It may help to remember the pattern that the values of

sinx read as for n = 0, 1, 2, 3, 4, …, then the values of cosx thanks to the

identity sin(90° - x) = cosx, are the values of sinx in reverse. The values of

tanx can be quickly found from the identity . You should be

confident at sketching the graphs of the three functions and be able to use

them to quickly find the sine, cosine or tangent of any angle that is a multiple

of 90° (or radians) added to any of the angles above. For example,

(the radian equivalent of this is ).

Try it out

Sketch the graphs of y = sinx, y = cosx and y = tanx.

Use your sketches to work out the exact values of the following. (Do not use

a calculator.)

4
5

Using radians:

Example 1

Question

Find the largest of the following list of numbers.

(a)

(b)

(c)

(d)

(e)

Solution
The largest value is option (b) cos(-45°).

You are not expected to know the exact value of cos(75°) but you should

know from a sketch of y = cosx that cos(75°) < cos(45°).

Example 2

Question

Find the smallest of the following numbers.

(a)

(b)

(c)

(d)

This example uses radians which are not on the MAT syllabus.

Solution
 and since log9 x is an increasing function,

The smallest is option (d) .

The smallest of options (a), (b) and (d) is (d), . You do not

need to be able to work out the exact value of , you just need

to determine if it is greater or less than .

Trigonometric equations

You will solve a number of trigonometric equations as part of your school

mathematics course. In doing this you will often make use of the identities

. Some of the questions that have

appeared on the MAT and TMUA do resemble what you might encounter in

school. Sometimes, however, a question may only ask for the number of

solutions in a given interval and not what they are explicitly. The easiest

approach might be just to solve the equation as you would do normally, but be

aware that there are times when the best approach is to use what you know

about the graphs and ranges of the three trigonometric functions. You should

always spend a minute or two deciding which will be the best approach when

you are given a trigonometric equation in an admissions test.

Try it out

For each of the following equations, work out how many solutions there are

in the interval .

Example 3
Question

2 2
How many solutions does the equation cos x - 3sinx + 3sin x = 0 have in the

interval

The standard ‘school’ approach is probably best here.

Solution

There are three solutions in .

Example 4

Question

How many solutions does the equation

have in the interval

There is a little bit of work to do here before this looks like a school

question. The approach to solve trigonometric equations is normally to

gather all the terms on one side and then factorise, but it is not

immediately obvious how to factorise the left-hand side of the

equation. When you are not sure what to do on this type of question,
remember that there are only two identities you might have to use, and

helps here.

Solution

Using gives

Each of these has two solutions in the interval .

There are four solutions in total.

At this stage a quick sketch graph will help.

Trigonometric inequalities

All the skills you have learned for manipulating inequalities can be applied to

inequalities of trigonometric functions.

Example 5

Question

What fraction of the interval satisfies the inequality

This does not factorise as it is written; you need to use the identity

Solution

Just as the inequality has the solution or ,

the inequality has the solution or

.
A sketch will help to solve the inequality .

of the interval satisfies the inequality.

On the interval , precisely when or

Example 6

Question

For what fraction of the interval is the inequality

Gathering all terms on the one side is a sensible starting point.

Solution
You are looking for a multiplication that gives a negative result or 0.

For a negative result, either is negative and is positive

or is positive and is negative. A sketch graph of

and on the same axes will help with this. Some

care is needed as they are both transformations of known graphs and

you need to be able to find the key points where each function changes

from one sign to the other. Where either of the graphs crosses the x

axis, the result will be 0.

is negative for , ,

and .
Exercise 1

1 Put the following values in order from least to greatest.

sin(20°) cos(40°) sin(60°) cos(80°)

2 Note: This question uses radians.

Which of the following has the largest value?

(a)

(b)

(c)

3 How many solutions does the equation have in the

interval ?

4 How many solutions does the equation

have in the interval ?

5 For what fraction of the interval is the inequality

satisfied?

6 Solve the inequality

7 Solve the inequality

8 for exactly one third of the interval . Find the

value of c.

9 The inequalities and , where k is a real number, are

both satisfied by some values of x in the interval . What is

the smallest possible value of k?

10 Solve the inequality on the interval .

Exercise 2

Admissions test multiple choice questions

TMUA style questions

1 Find the fraction of the interval for which the inequality

is satisfied.

2 Find the value of .

A 0

3 Find the number of solutions of the equation

in the interval .

A 0
 B 1

C 2

D 3

E 4

F 5

4 x, satisfies the simultaneous equations

Find the sum of the possible values of x.

A 45°

B 90°

C 135°

D 180°

E 225°

F 0°

5 ABCD is a trapezium with side AB parallel to side CD. The lengths of

sides AD and BC are both 6 units. The diagonal AC is of length 8 units

and makes an angle of α with both of sides AB and CD as shown in the

diagram.
Given that the area of triangle ACD is five times the area of triangle ABC,

find the value of cosα.

F
MAT style questions

6 How many solutions does the equation

have in the interval ?

(a) 0

(b) 1

(c) 2

(d) 3

(e) 4

7 How many solutions does the equation

have in the interval where n is a positive integer?

(a) one for all n

(b) two for all n

(c) two for odd n only

(d) two for even n and one for odd n

(e) two for even n and three for odd n

8 What is the maximum value of

in the interval .

(a)

(b)

(c) 2
 (d)

(e) 4

9 In triangle ABC below, angles α, β and γ are such that

Which of the following statements must be true?

(a)

(b)

(c)

(d)

(e)

10 The line y = x intersects the circle at the point A

where the x coordinate and y coordinate of A are both less than 1. The

point B is on the circumference of the circle and directly above its centre,

X. The point C is such that ABC forms a right-angled triangle. This is

shown in the diagram.

From this diagram, 67.5° is equal to

(a)

(b)

(c)

(d)

(e)

Full worked solutions can be found to all of the questions from Exercise

2 at www.hoddereducation.co.uk/step-mat-tmua-answers
Chapter 10: Logic and proof

TMUA MAT STEP

Introduction

Just as you learned to talk without being explicitly aware of things such as

verbs, nouns and grammar, so you first learned mathematics without much

concern for its language and syntax. This is fine if you only need to use

maths as a tool but in order to understand mathematics you need to

understand its language. The admissions tests all require you to understand

and present mathematics written in a rigorous way.

Mathematics is concerned with statements expressed in precise and

technical terms, where a statement is an assertion which is either true or

false. The following are examples of statements.

• The Earth is hotter than the Sun.

3 3 3 3
• 9 + 10 = 12 + 1

• Every even integer greater than 2 is the sum of two primes.

The first statement is false whilst the second is true. The third statement is

Goldbach’s Conjecture which has been verified for all even integers up to 4
18
× 10 but is still unproven. Even though it is currently unknown whether

Goldbach’s Conjecture is true it is still a statement because it is either true

or false.

Mathematics also uses assertions containing variables which denote

arbitrary elements of a set (such as the set of real numbers or the set of all

primes). Strictly speaking, these are not statements but they become

statements when the variables are replaced by particular elements from a

set. Such assertions are called predicates but are often referred to as

statements about the elements of a particular set. For example, the following

are statements about real numbers x, y and z.

• x ≤ 5

2 3 4
• x - y = z

• -1 < xy < 1

The statement x ≤ 5 is true when x = 2 but false when x = 10.

Quantifiers

The following joke is much loved by mathematicians.

A cheese sandwich is better than nothing.

Nothing is better than eternal happiness.

Therefore, a cheese sandwich is better than eternal happiness.

The humour, such that it is, derives from the fact that ‘nothing’ is being used

as a noun in the first statement and a quantifier in the second. In everyday

speech the context tells us which interpretation to use but in mathematics

the aim is to eliminate such ambiguities.

A quantifier is a word or phrase which is used before a noun to indicate the

quantity of the object. Examples of quantifiers in everyday speech include

‘some’, ‘all’, ‘a few’, ‘several’.

In mathematical logic there are two quantifiers, the universal quantifier and

the existential quantifier.

Although mathematics uses just two quantifiers, there are different ways of

expressing each of them.

These quantifiers are used so often in mathematics that they have their own

symbols, for ‘for all’ and for ‘there exists’. You do not need to know

these for the admissions tests but if you study mathematics further then

you will encounter them.

The following (true) statements are equivalent ways of expressing the

universal quantifier.

2
• For all real numbers x, (x - 1)(x + 1) = x - 1

2
• For any real number x, (x - 1)(x + 1) = x - 1

2
• For every real number x, (x - 1)(x + 1) = x - 1

The following (true) statements are equivalent ways of expressing the

existential quantifier.
 2
• There exists a real number x, such that x = 2

2
• For some real number x, x = 2

2
• There is at least one real number x, such that x = 2

Notice that it is also true that there exists a real number x, such that (x - 1)(x
2
+ 1) = x - 1. You can prove this statement by showing that it is true for a

particular value of x such as x = 1. It should be clear that it is not possible to

2
prove that for every real number, x, (x - 1)(x + 1) = x - 1 by substituting

values for x as there are infinitely many values to try.

And, or, not and conditional statements

Statements may be combined to form compound statements by using and

and or.

It is useful to use P and Q for arbitrary statements in the same way that you

use x and y for arbitrary real numbers. If P is the statement 2 > 1 and Q is

the statement 3 > 2, then (P and Q) is the statement (2 > 1 and 3 > 2).

Intuitively you know that if P and Q are statements then for the statement (P

and Q) to be true you require that both P is true and Q is true.

In mathematics or is always inclusive, that is it includes the possibility ‘or

both’. Hence, for the statement (P or Q) to be true you require that at least

one of P and Q is true.

Examples of compound statements:

1 2 > 1 and 3 > 2

2 2 > 1 and 1 > 2

3
3 5 = 100 or 7 - 3 = 4

3
4 5 = 125 or 7 - 3 = 4

3
5 5 = 100 or 7 + 3 = 4

Statements 2 and 5 are false, the rest are true.

The definitions of and and or can be illustrated by means of truth tables.

Truth tables are not required for any of the admissions tests but they are a

useful aid in understanding statements and logic.

The truth tables for and and or are as follows.

P Q P and

F F F

F T F
T F F

T T T

P Q P or Q

F F F

F T T

T F T

T T T

The truth tables tell you the truth value (true or false) for the statements P

and/or Q for all of the possible truth values of P and Q. The final line of P

or Q confirms that in mathematics or is always inclusive.

Given a statement P, not P, is called the negation of P. The negation of P is

true if and only if P is false. The following tables show that the negation of

(P and Q) is logically equivalent to the statement (not P) or (not Q).

P Q P not(P

and and Q)

F F F T

F T F T

T F F T

T T T F

P Q not not (not P) or

P Q (not Q)

F F T T T

F T T F T

T F F T T

T T F F F
Try it out

Show that the statements not(P or Q) and (not P) and (not Q) are logically

equivalent.

Notice that each of the possible truth values assigned to P, Q, the truth

values of not(P and Q) and (not P) or (not Q) are the same. This is what it

means to say that the two statements are logically equivalent.

The idea of negation appears simple but is often a source of confusion. For

example, Goldbach’s Conjecture may be expressed concisely as:

For all integers n ≥ 2, there exist primes p and q such that p + q = 2n.

Which of the following statements is its negation?

1 For all integers n ≥ 2, there exist primes p and q such that p + q ≠ 2n.

2 For all integers n ≥ 2, there do not exist primes p and q such that p + q =

2n.

3 There exists an integer n ≥ 2 and primes p and q such that p + q ≠ 2n.

4 There exists an integer n ≥ 2 such that for all primes p and q, p + q ≠

2n.

A conditional statement takes the form ‘if P, then Q’, or ‘P implies Q’, or ‘P

⇒ Q’. Informally such a statement may be read as ‘assuming P is true, does

it follow that Q is true?’.

When considering statements involving variables, you need to know the

range of values the variables can take. Often this is not stated explicitly but
2
is given by the context. For example, the statement ‘if x = 2, then x = 4’ is
2
true because knowing that x = 2 allows you to determine that x = 4. What
2
about the statement ‘if x = 4 then x = 2’? Assuming x is allowed to be any
2
real number then knowing that x = 4 is not enough for you to determine

that x = 2 as it is possible that x = -2 instead. If, however, the statement is

restricted to positive (or non-negative) values of x, then the statement is true.

The following conditional statements are all true.

2
• For all real numbers x, if x = 2, then x = 4.
• For all integers n, if n is a prime greater than 3, then n = 6k + 1 or n = 6k -

1 for some integer k.

2
• For all integers n, if n is odd, then n - 1 is divisible by 8.

2
• For all real numbers x and y, if y = x - 5x + 6, then y = (x - 2)(x - 3).

The following conditional statement is false.

• For all integers n ≥ 2, if n is odd, then n is prime.

The truth table for conditional statements requires a truth value to be

assigned to ‘P ⇒ Q’ for all possible truth values of P and Q. This forces you

to consider the truth or otherwise of statements when P is false. The

convention adopted is that when P is false, the statement ‘P ⇒ Q’ is true

regardless of whether Q is true or false. Here is the truth table for

implication.

P Q P ⇒ Q

F F T

F T T

T F F

T T T

Try it out

What is the smallest value of n ≥ 2 for which the statement ‘if n is odd,

then n is prime’ is false?

Students are often uncomfortable with the first two lines in the table above

and this unease comes from the informal interpretation of P ⇒ Q as

‘assuming P is true, does it follow that Q is true?’. So when confronted by a

statement such as ‘for all integers n ≥ 2, if n is odd, then n is prime’ it is

natural to interpret this as a statement about odd integers greater than 2 even

though it is explicitly stated n ranges over all integers greater than 2. Each of

the following statements must have a truth value.

• If 2 is odd, then 2 is prime.

• If 3 is odd, then 3 is prime.

• If 4 is odd, then 4 is prime (and so on).

In particular, if you correctly answered that 9 is the smallest value of n ≥ 2

for which the statement ‘if n is odd, then n is prime’ is false, then you are

stating that each of the above statements is true! It follows that your

deduction was consistent with the table above since ‘if 4 is odd, then 4 is

prime’ is an example of P ⇒ Q where P and Q are both false; ‘if 2 is odd,

then 2 is prime’ is an example of P ⇒ Q where P is false and Q is true; and

‘if 3 is odd, then 3 is prime’ is an example of P ⇒ Q where P and Q are both

true. You kept going until you found an integer n ≥ 2 such that P is true (n

is odd) and Q is false (n is not prime).

Implication is a good example of why more precision is needed in the use of

language in mathematics. In everyday speech, conditional statements of the

form ‘If …, then …’ normally involve some causality; for example ‘if it

rains tomorrow, then I will stay indoors’. Whereas something like ‘if the

moon is made of cheese, then I am the prime minister’ would be considered

nonsense (and hence neither true nor false). In mathematics there is no

assumption of underlying causality, so the statement ‘if the moon is made of

cheese, then I am the prime minister’ is true simply because there is no

explicit counter-example (assuming the moon is not made of cheese).

Converse, contrapositive and negation of P ⇒ Q

Consider the implication P ⇒ Q. The reverse implication Q ⇒ P (sometimes

denoted P ⇐ Q) is called the converse of P ⇒ Q.

2 2
For example, the statement x = 1 ⇒ x = 1, has converse x = 1 ⇒ x = 1. Note

that in this case the first implication is true whereas the second is false

(assuming that x is any real number). This shows that a conditional

statement and its converse are not logically equivalent.

The converse should not be confused with the contrapositive of P ⇒ Q,

2
which is the statement not(Q) ⇒ not(P). The contrapositive of x = 1 ⇒ x =
2
1 is x ≠ 1 ⇒ x ≠ 1.

It is important to understand that the contrapositive is logically equivalent to

the original implication in that they are either both true or both false. This

fact will be very important when proving conditional statements as it may

be much easier to prove its contrapositive instead.

A different type of example may help to clarify this. Suppose Liverpool are

about to play a football match and consider the following statements.

• Liverpool will win.

• Liverpool will score (at least one goal).

• Liverpool will not score.

• Liverpool will not win (meaning they lose or draw).

The statement Liverpool win ⇒ Liverpool score is true because if you know

that Liverpool have won a match then you can deduce that they must have

scored at least one goal.

The converse of Liverpool win ⇒ Liverpool score is Liverpool score ⇒

Liverpool win. This statement is not true as it is possible for Liverpool to

lose 2-1 for example.

The contrapositive of Liverpool win ⇒ Liverpool score is Liverpool do not

score ⇒ Liverpool do not win. Both of these statements are true, if

Liverpool do not score a goal then they cannot win the match and if

Liverpool win the match then they must have scored. These are essentially
two ways of saying the same thing – in other words the original statement

and its contrapositive are logically equivalent.

What is the negation of P ⇒ Q?

It is tempting to think that the negation of ‘if P is true, then Q is true’ must

be ‘if P is true, then Q is false’. This is incorrect. One way to see this is to

use the truth table for P ⇒ Q.

P Q P ⇒ Q not(P ⇒ Q)

F F T F

F T T F

T F F T

T T T F

First notice that the negation of an implication cannot be another

implication because the four rows of an implication must contain the truth

value ‘true’ 3 times and ‘false’ once. Its negation, therefore, contains ‘false’

3 times and ‘true’ once and hence is not another implication. Since there is

only one row which is ‘true’ you can see that not(P ⇒ Q) is logically

equivalent to P and not(Q).

The negation of ‘if Liverpool win, then Liverpool score’ is ‘Liverpool win

and Liverpool do not score’. You can think of this as what would have to

happen in order to conclude that the original statement is false.

Try it out

Consider the following conditional statement about a real number x.

2
x > 4 ⇒ x > 2

Write down the converse, contrapositive and negation of this statement.

Determine whether each of the four statements you have just written is true

or false.
Necessary and sufficient

If P ⇒ Q, then P is sufficient for Q, and Q is necessary for P. The idea here

is that if P ⇒ Q then for Q to be true it is enough that P is true so P being

true is sufficient information for you to deduce that Q is true. Alternatively,

if P ⇒ Q and Q is false then it must also be the case that P is false, so Q

must be true for there to be any possibility that P is true, so Q is necessary

for P.

The words ‘necessary’ and ‘sufficient’ are often used the wrong way round.

You may find the football example useful again. Given the true statement: if

Liverpool win, then Liverpool score you can say that winning is sufficient

for scoring and scoring is necessary for winning.

The different ways of using these words can also cause confusion. The

following two statements are not logically equivalent – in fact, the first is

true whereas the second is false.

• For S to be a quadrilateral, it is sufficient for S to be a square.

• If S is a quadrilateral, then S is a square.

Example 1

Question

Let S represent the statement ‘There is snow’ and N represent the

statement ‘There is a north wind’. You are given that the following

statements are true.

• If there is no snow, then there is no north wind.

• There is a north wind.

(a) Express what you are given in terms of S, N and symbols.

(b) Is it possible to determine the truth value of S?

(c) You are also given that if it snows, then the robin hides its head

under its wing (poor thing). What can you deduce about the weather

and wind if the robin does not have its head hidden under its wing?
Solution

(a) N and (not(S) ⇒ not(N))

For (a), if there is no snow, then there is no north wind translates

to ‘not(S) ⇒ not(N)’. There is a north wind so N is true.

(b) Since N is true and N ⇒ S is true it follows that S is true.

There is snow.

For (b), the contrapositive of ‘not(S) ⇒ not(N)’ is ‘N ⇒ S’.

(c) Let H represent the statement ‘the robin hides its head under its wing’.

Given S ⇒ H and not(H) is true

Contrapositive not(H) ⇒ not(S)

For (c) writing ‘the robin hides its head under its wing’ each time

you want to refer to it is not sensible.

not(S) is true

not(S) ⇒ not(N)

not(N) is true

So there is no snow and there is no north wind.

You are given that ‘S ⇒ H’ and not(H). The contrapositive of ‘S ⇒

H’ is ‘not(H) ⇒ not(S)’. Since not(H) is true and ‘not(H) ⇒ not(S)’

is true, it follows that not(S) is true. From part (a) ‘not(S) ⇒

not(N)’, so not(N) is true.

Biconditional (if and only if) statements

If P ⇒ Q and Q ⇒ P (or P ⇐ Q), then you can write P ⇔ Q and say that P is

necessary and sufficient for Q (or that Q is necessary and sufficient for P), or

P if and only if Q, meaning that P and Q are logically equivalent. The

following statements are all true.

• n is odd ⇔ n + 1 is even

• x + y = 0 is necessary and sufficient for x = -y

• xy = 0 if and only if x = 0 or y = 0

The use of ‘P if and only if Q’ leads to the consideration of the statements

‘P if Q’ and ‘P only if Q’. Each of these is logically equivalent to either P ⇒

Q or Q ⇒ P but which is which?

‘P if Q’ means that Q is sufficient for P, hence ‘P if Q’ is equivalent to Q ⇒

P.

‘P only if Q’ means that Q is necessary for P, hence ‘P only if Q’ is

equivalent to P ⇒ Q.

Try it out

Determine the truth values of the following statements.

1 x > 2 if x > 4

2 x > 2 only if x > 4

3 S is a square if S is a rectangle.

4 S is a square only if S is a rectangle.

5 ab = ac if and only if b = c

2
6 x < x if x < 1

2
7 x < x only if x < 1

2
8 x < x if and only if -1 < x < 1
Exercise 1

1 The cards shown below are from a pack in which every card has a letter

on one side and a number on the other.

It is claimed that behind a vowel there is always an odd number.

Determine the least number of cards which must be turned over to

verify this claim.

2 A joke has it that army recruits used to be instructed ‘If it moves, salute

it. If it doesn’t move, paint it.’

Assume that this instruction has been carried out completely in the local

universe, so that everything that doesn’t move has been painted.

(i) A recruit encounters something which is not painted. What should

they do, and why?

(ii) A recruit encounters something which is painted. Do we know what

they should do? Justify your answer.

3 Siân: ‘I won’t go for a walk if it’s not dry or not warm.’

Max: ‘It’s dry and warm, let’s go for a walk!’

Will what Max has said convince Siân? Justify your answer.

4 If it is daytime and the car headlights are on, then it is raining.

If the dashboard lights are dimmed, then the car headlights are on.

It is daytime.

It is not raining.

What can you deduce?

5 Identify which of the following deductions are correct. For those that are

not correct, show why they are not.

(a) If it is raining, then you must put up your umbrella. It is raining.

Therefore you must put up your umbrella.

(b) If it is raining, then you must put up your umbrella. You have put

up your umbrella.

Therefore it is raining.

(c) All cows have three legs. Ermintrude is a cow.

Therefore Ermintrude has three legs.

(d) All cats purr. This animal purrs.

Therefore this animal is a cat.

(e) If it is very windy, we will not ski. It is not very windy.

Therefore we will ski.

Proof

Informally, a mathematical proof is an argument that convinces you (and

anyone else) that a mathematical statement is true.

Formally, a proof consists of a list of statements, the final one being the

statement you want to prove. Each statement must be either something that

is already known or assumed to be true (for example, an axiom, a definition,

an assumption or a theorem) or a direct consequence of earlier statements in

the list following the application of logical principles (for example, if both P

is true and P ⇒ Q is true then Q is true), or rules for algebraic manipulation

(such as x(y + z) = xy + xz).

In practice, you do not need to spell out every single step in a proof but the

argument should be sufficiently detailed to allow the reader to fill in all of

the intermediate steps.

In trying to prove a statement, experimenting with examples may well help

in gaining understanding. However, unless there are finitely many cases to

consider, no amount of examples will ever constitute a proof.

Methods of proof

How would you prove the following statement about an integer n?

2
• n is odd if and only if n is odd.

An ‘if and only if’ or biconditional statement can be considered as two

conditional statements combined into one. Here these statements are:

2
• If n is odd, then n is odd.

2
• If n is odd, then n is odd.

Note that these statements, although both true, are not saying the same thing

and therefore need different proofs.

2
Statement: If n is odd, then n is odd.

2 2
Proof: If n is odd, then n = 2k + 1 for some integer k. Hence n = (2k + 1)
2
= 2(2k + 2k) + 1, which is also odd because it is of the form 2m + 1 where
2
m = 2k + 2k is an integer.

The above is an example of a direct proof in that you proceed directly from
2
the hypothesis that n is odd to the conclusion that n is odd. You may be
2
tempted to write 4(k + k) + 1 and state that this is odd (which it is) but
2
writing 2(2k + 2k) + 1 instead allows you to appeal directly to the definition

of an odd integer.

It is not clear how to prove the second statement directly since it is not clear
2
how knowing that n is odd can lead directly to the conclusion that n is odd.

This is an example of a conditional statement where it is easier to prove its

contrapositive instead. Recall that a conditional statement and its

contrapositive are logically equivalent, meaning that they are either both

true or both false.

2
Statement: If n is odd, then n is odd.

2
Proof: The contrapositive of ‘if n is odd, then n is odd’, is ‘if n is even,
2 2
then n is even’. If n is even, then n = 2k for some integer k. Hence n =
2 2 2
(2k) = 2(2k ), which is even because it is of the form 2m where m = 2k is

an integer.
Another indirect method of proof is proof by contradiction. This method

relies on the fact that a statement is either true or false so if you can prove

that the negation of a statement is false then the original statement must be

true. Two classic examples are the standard proofs that is irrational, and

that there are infinitely many primes.

Statement: is irrational.

Proof: Suppose that the statement is false and hence where p and q are

integers. In addition, assume the fraction is in its lowest terms (so p and q

have no common factors other than 1). Then and hence

2 2 2 2
p = 2q . It follows that p is even but if p is even, then p is even. Writing p
2 2 2 2 2 2
= 2k for some integer k, gives 2q = p = (2k) = 4k and hence q = 2k so q

must also be even. So the assumption that where the fraction is

in its lowest terms has led to the conclusion that p and q are both even. This

gives a contradiction because if p and q are both even then the fraction is

not in its lowest terms. The only logical conclusion is that the original

assumption is false and hence is not rational.

It is important to understand that the above is a logically correct proof of the

statement ‘if , then p and q are both even’ so this statement is true.

It is another example of a statement of the form ‘if P, then Q’ where P is

false and so the conditional statement is true.

Statement: There are infinitely many primes.

Proof: The first four primes are 2, 3, 5, 7 (so you know that primes exist).

Now suppose that there are finitely many primes p ,p ,…, p . In other
1 2 k

words, assume that there are exactly k prime numbers (where k could be

very large). Define n = p p …p + 1. Clearly n has remainder 1 on division

1 2 k

by each of the primes p ,p ,…, p . It follows that n is not divisible by any of

1 2 k

the primes p1,p2,…, pk. But any integer greater than 1 is either a prime or is

divisible by a prime. This gives a contradiction since n is greater than each

of the primes p1,p2,…, pk (and hence n is not a prime) but not divisible by

any of p ,p ,…, p . It follows that the assumption that there are finitely
1 2 k

many primes is false.

Disproof by counter-example

To prove that a statement P is false, it is sufficient to find a counter-example,

that is find an example for which the statement P is not true.

The great French mathematician, Pierre de Fermat, made the following

conjecture in 1650:

2n
For any positive integer n, 2 + 1 is prime.

2n
Fermat knew that 2 + 1 is prime for n = 1,2,3,4. In 1732 the great Swiss
32
mathematician Leonhard Euler showed that 2 + 1 = 641 × 6 700 417, and

hence Fermat’s conjecture is false for n = 5. Since Fermat’s conjecture stated

2n
that 2 + 1 is prime for any positive integer n it is not necessary to examine

any further cases to prove that the statement is false.

Recall that you were asked to consider the negation of Goldbach’s

Conjecture:

For all integers n ≥ 2, there exist primes p and q such that p + q = 2n.

You were given the following four alternatives to choose from.

1 For all integers n ≥ 2, there exist primes p and q such that p + q ≠ 2n.

2 For all integers n ≥ 2, there do not exist primes p and q such that p + q =

2n.

3 There exists an integer n ≥ 2 and primes p and q such that p + q ≠ 2n.

4 There exists an integer n ≥ 2 such that for all primes p and q, p + q ≠

2n.

As Goldbach’s conjecture is a statement about all integers n ≥ 2, it would

only take a counter-example to prove it wrong. A counter-example would be

an integer n such that there are no primes p and q such that p + q = 2n. In

other words, for all primes p and q, p + q ≠ 2n so the negation of

Goldbach’s Conjecture is statement 4.

Logical structure of proofs

Read the following proof that argues that the difference between the squares

of two consecutive integers is equal to the sum of those integers.

Let n be an integer.

n and n + 1 are consecutive integers.

2 2
The difference between the squares of n + 1 and n is (n + 1) − n .

2 2
This is equal to n + 2n + 1 − n = 2n + 1.

This can be rewritten as (n + 1) + n.

So the difference between the squares of two consecutive integers is equal to

the sum of those integers.

This is an example of a simple direct proof. For the moment you are going

to consider the structure of the proof, the statements used and the

relationship between those statements.

The proof is of the form:

If statement A is true then statement B is true.

If statement B is true then statement C is true.

If statement C is true then statement D is true.

Conclusion: If statement A is true then statement D is true.

In the example above, you can consider statement A to be that n is an

integer.

Statement B is that n + 1 is the next consecutive integer.

If n is an integer then n + 1 is the next consecutive integer.

Statement C is that the difference between the squares of the two integers is

2n + 1. Some simple algebraic manipulation following from A and B shows

this to be the case.

Statement D is that 2n + 1 is equivalent to the sum of the consecutive

integers n and n + 1 and again some simple algebraic manipulation shows

this to be the case.
Extraneous solutions

Consider the following solution to the equation .

Squaring both sides of gives .

Rearranging this equation gives , which factorises to give

Hence the solutions are and .

Substituting these values back into the original equation shows that whereas

x = 5 is a solution, x = -2 is not. What has gone wrong?

The answer may surprise you in that nothing has gone wrong but the steps

are not reversible. The very first step may be expressed as

This statement is true, the

converse, however is false and, more generally is true but

is false. Unless every step is reversible (that is an ‘if and

only if’ statement), an argument like that above will only produce candidate

solutions. It is important to know that the argument above has proved that

or and that this statement is true. The

extraneous solution x = -2 was introduced in the very first step.

Exercise 2

2n
1 Prove that for any positive integer n, if 2 - 1 is prime then n is prime.

Is the converse also true?

2 Prove or disprove each of the following statements about real numbers a

and b.

(a) If

(b) If
Exercise 3

Admissions test multiple choice questions

TMUA style questions

1 Alice and Marcus Praeter are twins who make life very difficult for their

parents. Alice tells the truth on Monday, Tuesday, Wednesday and

Sunday but lies on every other day of the week. Marcus lies on Monday,

Tuesday and Wednesday but tells the truth on all other days. Their

mother asks both of them if they were lying yesterday. ‘I was!’ says

Alice. ‘So was I!’ says Marcus.

What day of the week is it?

A Monday

B Tuesday

C Wednesday

D Thursday

E Friday

F Saturday

2 You are given that a, b and c are positive real numbers with a < b and c

≠ 1.

Which of the following statements must be true?

D There is not enough information to determine the relative sizes of

and

3 Five empty jars are labelled A to E. £2 is placed into one jar at random,

£5 into another, £10 into a third, £20 into a fourth and £50 into the

remaining jar. Clues to the amounts in some of the jars are written on

signs that are attached to the jars. The signs read as follows:

Jar A: ‘Precisely one of the signs is not telling the truth and that jar

contains £2.’

Jar B: ‘Jar C contains £10.’

 Jar C: ‘This jar does not contain £50.’

Jar D: ‘Jar E contains £20.’

Jar E: ‘Jar A contains £5.’

Given that the statement on jar A is true, which jar contains £50?

4 a, b and c are real numbers.

Given that a(b - c) = a(c - b), which of the following must be true?

I a = 0

II a = 0 or b = c

III b = c

IV a = 0, b = 0 and c = 0

A none of them

B I only

C II only

D III only

E I and II only

F I and III only

G I and IV

H II and III only

I II and IV only

J III and IV only

5 Consider the following statement about integers p, q and r.

The condition ‘q and r are factors of p’ is

A necessary but not sufficient for (*).

B sufficient but not necessary for (*).

C necessary and sufficient for (*).

D not necessary and not sufficient for (*).

6 a, b and c are integers such that 0 < a ≤ b ≤ c.

If , which of the following statements must be true.

I a = 3, b = 3 and c = 3

II b = 3 or b = 4

III a > 1

A none of them

B I only

C II only

D III only

E I and II only

F I and III only

G II and III only

H I, II and III

7 a, b and c are real numbers with < a < b < c

Which of the following statements must be true?

2 2 2
I 0 < a < b < c

2
II a(b + c) > a

III

A none of them

B I only

C II only

D III only

E I and II only

F I and III only

G II and III only

H I, II and III
8 A student attempts to solve the simultaneous equations

The student’s attempt is as follows.

Which of the following best describes this attempt?

A It is completely correct.

B It is incorrect and the first error occurs on line (I).

C It is incorrect and the first error occurs on line (II).

D It is incorrect and the first error occurs on line (III).

E It is incorrect and the first error occurs on line (IV).

F It is incorrect and the first error occurs on line (V).

G It is incorrect and the error is that some of the solutions are not

consistent with both equations.

Full worked solutions can be found to all of the questions from Exercise 3

at www.hoddereducation.co.uk/step-mat-tmua-answers
Section 3 Problem solving in extended questions
Chapter 11: Getting started with longer questions

MAT STEP
Introduction

The STEP specification states that for the problems set, ‘solutions will frequently

require insight, ingenuity, persistence, and the ability to work through substantial

sequences of algebraic manipulation’. This means that when you are solving any

problem, you may have to deal with some challenging algebraic manipulation. You

will be expected to use the algebraic skills from A level accurately and efficiently but

with expressions and equations that are much less friendly than you are used to

working with. You may have to work with more variables or with more general

forms of equations.

The same level of algebraic dexterity is needed for the longer MAT questions. A lot

more is expected of your algebraic manipulation techniques than is expected at A

level.

So, what level of skill is needed? Read through the following example. Try to follow

what is going on. At several stages, you will be given the start of some algebraic

manipulation and expected to complete the calculations.

Example 1

Can you complete the calculations?

In the cubic equation where p and q are distinct real

numbers, use the substitution

to show that the equation reduces to where a and b are to be expressed

in terms of p and q.

Cambridge Assessment Admissions Testing STEP 1 2018 Question 7 (part)

Before starting this it is important to reflect on the statement in the question that ‘p

and q are distinct real numbers’. You should reflect on why this condition is

necessary. If p and q are not distinct then p = q and the substitution would become

x = p. Substituting this into the original cubic equation would result in the equation

0 = 0 which is not particularly helpful.

This already looks different to the typical sorts of question you are likely to have

met. You have a more general form of a cubic equation which is expressed using
two parameters p and q.

With any substantial algebraic manipulation you should try to be as efficient as

possible, choosing techniques to reduce the opportunity to make simple errors.

There is not a great amount of additional thinking about what to do for this

question as you are told to substitute into the equation

Try it out

Here is the first stage of the substitution:

At this stage, multiplying every term by seems like a sensible option.

You have the target Multiplying by will give you a cubic

2
equation in z and it is feasible that the z and z terms will cancel to 0. Seeing this

should give you confidence that you are on the right track and encourage you to

continue.

Multiply out the brackets for your equation. You can be more efficient here by
3 2
considering which multiplications give z terms, which give z terms, which give z

terms and which give terms that are independent of z.

If you are very confident in expanding terms such as you might be able

to do this directly. If you are less confident and wish to be careful then a sensible

next step is to expand first. A binomial expansion

using the third row of Pascal’s triangle (1 3 3 1) will speed up this step.

Remember to focus on the required form for the answer, when

simplifying your working.

If you were careful with your working, you should have found that the equation can

be written as p(p ‒ q) 2 3
z + q(p ‒ q)
2
= 0 giving a = p(p ‒ q) 2
and b = q(p ‒ q)2
. If

you did not get these two expressions for a and b, look back through your

calculations and try to find where you have made any mistakes. If you do not think

you have made any mistakes, see if you have the same answers written in a different

form.
Note that the final values for a and b would both be 0 if p and q are not distinct.

This question should give you a good idea of the level of algebraic manipulation

required in the STEP examinations and the longer MAT problems. None of the

techniques is beyond anything taught in school but the level of effort and care

required is much greater.

A question to think about

The expressions 7x
2 2
+ 14y ‒ 28xy − 16x + 24y and
can both be written in the form

For each expression, find the values of the constants a, b, c, d and e.

Find all of the real solutions for x and y for the simultaneous equations

This question requires a number of algebraic techniques that are also used at A

level. You may be able to answer it straight away, in which case, have a go!

Read through the following hints which will give you an idea of how you will need

to think to solve the problem. If you still can’t answer the question, move on to the

exercises and reflection sections before returning to it.

Hints

• Is multiplying out the brackets on the right-hand side of the identity the most

efficient way of finding a, b, c, d and e?

• Can you see anything you recognise in the simultaneous equations?

• Can the original identity help with both of the later equations?

• The question asks you to solve the simultaneous equations. The question makes it

clear that you have to find all of the real solutions. How many solutions are there?
Key algebraic methods

At A level you are taught a number of methods that help you to manipulate algebraic

expressions. These methods are very useful for making sure that you are efficient

when manipulating the more awkward expressions and equations you come across in

the STEP and MAT examinations. You need to make sure you are fluent in the

following techniques.
Completing the square

This is an often overlooked but incredibly useful skill. Completing the square allows

the graph of a quadratic function, including its turning point, symmetry and any

roots to be easily obtained.

Completing the square is the first case of a class of transformations introduced by

Ehrenfried Walther von Tschirnhaus in 1683. Tschirnhaus transformations simplify

polynomials by using an algebraic substitution. A general quadratic equation

can be rewritten by completing the square as

In terms of Tschirnhaus transformations, the original equation has been transformed

to the simpler form where k is a real number.

The substitution that is used to do this is

Solving for t is much more straightforward than solving

Quadratic equations are so commonly encountered nowadays that we think nothing

of solving them and have a number of tools at our disposal for doing so.

Seventeenth-century mathematicians were very interested in these sorts of

transformation and held mathematical duels, using the transformations to develop

formulae for solving cubic and quartic (degree 4) equations.

In undergraduate mathematics they form part of the introduction to Galois theory.

You have already seen an example of a Tschirnhaus transformation in this chapter.

In Example 1 at the start of the chapter, the substitution was used to

reduce the equation to the simpler cubic equation

The STEP and MAT longer questions often allude to things that will be taught as

part of a university undergraduate course and often show you how the A level skills

you know will be developed in later study.

The factor theorem

In Chapter 5: Functions, you saw several examples of questions on the factor

theorem and remainder theorem.

When dealing with the sort of algebra found in the STEP and MAT examinations,

the factor theorem can be used to deal with some quite awkward looking

expressions; for example the expression looks

quite unusual and not much like the algebra you usually meet in school.

If you look at each part of the expression, you should be able to see that substituting

x = y will result in the expression evaluating to 0.

This means that (x ‒ y) is a factor of

Looking at each part of the original expression carefully once more, you should be

able to see that x = ‒z will also result in the expression evaluating to 0.

This means that (x + z) is a factor of

Try it out

Can you complete the factorisation of

Equating coefficients

Equating coefficients from two identically equal expressions is often an efficient way

to rearrange an algebraic expression into an equivalent form. It can be used when

decomposing algebraic fractions into partial fractions, dividing one polynomial by

another, rewriting nested radicals, testing linear dependence in a system of equations

and comparing real and imaginary parts of complex numbers.

For frequently encountered situations it is possible to learn the results from equating

coefficients, for example when dividing a cubic polynomial by a linear polynomial.

The result of dividing the cubic polynomial by the linear one

x ‒ 3 can be found by equating coefficients in this way.

Defining

Using the factor theorem shows that (x – 3)

is a factor of Dividing by the linear (x

– 3) will result in a quadratic expression.

This is equivalent to saying that

3
From this, equating coefficients of x gives a = 2, and equating the constants gives

These steps are straightforward and consistent and most students will start by writing

rather than starting from the division.

2
The coefficient k can be found by equating coefficients of x which in this case gives

This shows that

The important thing for you to realise is that you will frequently find yourself

dividing a cubic polynomial by a linear one and that the process will always be the
2
same. You will always find yourself equating coefficients of x . For example the

result of dividing can be found directly by

observation.
 3
By comparing the 6x terms to the 2x term you can see that the first term of the
2
resulting quadratic polynomial will be 3x .

Comparing –12 to –3 gives a +4 constant at the end

The x term will come from solving the equation 3k – = –11 giving k = –1.

You can apply this idea to any situation in which a polynomial of degree greater than

1 is divided by a linear polynomial.

Starting with

and

equating coefficients gives

With the exception of the first and last terms of the result, each successive coefficient

can be found using the result from the previous line. You can see that to find β you

need the value of α, to find γ you need the value of β and so on.

As a result of this, it is possible to find the result of a division of this type by

observation.

For example the result of can be found by

observation in the following way.

It is not necessary to be able to divide one polynomial by another by observation but

by practising the thinking skills necessary to do it in this way you will improve your

ability to manipulate complicated algebraic expressions.

It is also possible to find the same results by using the more formal method of

algebraic division taught in most schools. This method is an acceptable alternative to

equating coefficients but it does have the two disadvantages in that it takes a lot

longer and has several stages of calculation giving plenty opportunities to make

mistakes.

The method of equating coefficients can be applied even when an algebraic division

results in a quotient and remainder.

Try it out

Use the method of equating coefficients to find the values of a, b, c, d and e for the

following identity.
Key algebraic expressions to recognise

There are some standard algebraic results that you should be able to recognise as and

when they appear.

The difference of two squares

The difference of two squares identity, is easy to apply to

cases such as or even

You will need to be able to spot more disguised examples of these such as

For this you would need to notice that the terms

This shows that which can then be factorised to

You will also need to stay alert to examples where the squares are less obvious. For

the expression you would need to spot that

Hence

Try it out

Factorise as far as possible.

The sum of two squares

The two familiar identities and

can be rearranged to:

These two identities are so familiar that they are easy to miss when working through
2 2
a question. It is important to remember that there are two identities for a + b as it is

possible that one will be more helpful than the other for a particular question.

These two rearrangements can be particularly useful when thinking about the

minimum values that some expressions can take.

For example you can show that sin x cos x by starting with the identity

and rewriting it as

This rearranges to

Since for all real x, it follows that and

hence that
The sum and difference of two cubes

The sum of two cubes identity is

The difference of two cubes identity is

The difference and sum of two numbers raised to the

power n

The difference can be factorised for natural n to give

The sum can also be factorised but even values of n will include complex

numbers.

For odd values of n

For even values of n

Try it out

2 2
(a) Use the identity for a – b to find the prime numbers that can be expressed as

one less than a square number.

(b) Find the prime numbers that can be expressed as one less than a cube number.

(c) Find the values of a such that a

n
‒ 1 is a prime number.
Further factorisation

In admissions tests you will often come across some quite awkward looking

expressions that turn out to factorise comparatively easily.

can be factorised by careful grouping of the terms. A partial

factorisation of pairs of terms shows how the whole expression can be fully

factorised: This gives the final factorisation as

The more experience you have at manipulating algebraic expressions, the better

equipped you will be to spot how to factorise a variety of commonly encountered

expressions.

Try it out

Multiply out

Use your result to factorise

Exercise 1

1 Express in the form where p, q and r

are written in terms of α, β and γ.

Use your result to factorise each of the following to the product of three linear

factors.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

2 Multiply out

Use your result to factorise the following to the product of three linear factors.

(a)

(b)

(c)

(d)

3 Factorise each of the following as completely as possible. You will have to use a

number of techniques. Try to be as efficient as you can.

(a)

(b)

(c)

(d)
(e)

(f)
Exercise 2
MAT style questions

1 The equation has three non-zero roots α, β and γ.

(i) Express in terms of α, β and γ.

(ii) Explain why the equation where p is an integer

cannot have three distinct integer solutions.

(iii) Show that the equation cannot have three positive

integer solutions.

(iv) Given that p and q are integers, find all of the pairs of integer values for p

and q such that the equation has three positive

integer solutions.

(v) Given that p, q and r are integers, how many distinct equations of the form

have four positive integer solutions?

Hints

(i) Can you write a cubic equation that has roots α, β and γ and for which the
3
coefficient of x is 1?

What is the relationship between b and α, β and γ and between c and α, β and

γ ? Can you write each of b and c in terms of α, β and γ ?

(ii) Does the relationship between c and α, β and γ help?

(iii) Does what you found in part (i) help in any way? If α, β and γ are positive

integers, what does this tell you about the ratio It will help you to

consider the possibilities if you consider There will only be a

finite number of possibilities.

(iv) You should be able to use all of the ideas you found in parts (i) to (iv) to

answer this.

2 Given that a, b, c and k are non-zero real numbers

(i) Multiply out

(ii) Find two conditions on a, b, c and k such that

can be written in the form

where A is a function of a, b and c and B is a function of k.

(iii) Show that is a factor of

 (iv) Factorise into three quadratic factors.

(v) Find all the real roots of the equation

Hints

(i) You are explicitly told to multiply out the expression so that is what you

should do. How can you do this efficiently in order to avoid any errors?

(ii) You should have an expression for each coefficient. You should know the

values of some of the coefficients from the form and these

should give you the two conditions you are asked for. In the language used in

Chapter 10: Logic and proof, these are necessary and sufficient conditions for

the expression to be written in the required form.

(iii) The expression does not factorise and equating to 0 does not help

(unless you wish to use complex numbers). Is it possible to simply divide

by without having a remainder in the result?

(iv) If proves to be a factor, how can the other factors be found? Can

you use anything you’ve found from the earlier parts of the question to help?

(v) If you have managed to factorise into three quadratic factors,

you should be able to find all of the real roots.

A STEP question

3 You have already seen a worked solution to the first part of this question at the

start of this chapter. Here is the question in full so you can see how it develops.

(i) In the cubic equation where p and q are

distinct real numbers, use the substitution

to show that the equation reduces to where a and b are to be

expressed in terms of p and q.

(ii) Show further that the equation where c and d are non-

zero real numbers, can be written in the form

where p and q are distinct real numbers, provided

(iii) Find the real root of the cubic equation

(iv) Find the roots of the equation and hence show how

the equation can be solved in the case

Cambridge Assessment Admissions Testing STEP 1 2018 Question 7

Hints

(i) Why do p and q have to be distinct real numbers? You are told to substitute

into the equation. How can you make sure you don’t make any

mistakes when you do this? You have a target expression to aim for

so what does that tell you about some of the terms you get

when you are simplifying your expression after using the substitution?

(ii) Can you relate c and d to p and q? Can you use this relationship to write a

direct relationship between c and d by eliminating p and q? For what

conditions will this relationship give real values for c or d?

(iii) Can you find the values of c and d for the given equation? Do the values fit

the required condition to be able to use the substitution from part (i)? What

values of p and q do you get from c and d? Can you use what you know from

part (i) to write a simple cubic in z and solve that? How can you use this

solution to find the solution for x?

(iv) Is there an ‘obvious’ value for x that makes equal to 0 (the

factor theorem). Can this be used to factorise into linear

factors? What values of x result? Can these be written in terms of c and d?

Can the condition be used to write the values of x in terms of, for

example, c? Based on your final expressions for the values of x, can the

equation be solved?
A STEP question

4 (i) The number α is a common root of the equations

(that is, α satisfies both equations).

Given that show that

Hence, or otherwise, show that the equations have at least one common root

if and only if

Does this result still hold if the condition is not imposed?

(ii) Show that the equations and have

at least one common root if and only if

Hence, or otherwise, find the values of b for which the equations

and have at least one common

root.

Cambridge Assessment Admissions Testing STEP 3 2010 Question 4

Hints

(i) If a number is described as a root of an equation, what happens when that

number is substituted into the equation? α is a root of both of the given

equations. Can you use this to find two equations in α? Can you solve these

equations for α ? Now you have α in terms of a, b, c and d (and it is given in

the question), can you get the required result from both of the original

equations? What is meant by ‘if and only if’ in this case? What happens if a

= c? What does this tell you about b and d? What does this mean for the

original equations?

(ii) Can you adapt the approach used in part (i)? How do you get two equations
3
that start with α ? Can you use your equations to find a quadratic equation in
2
α ? Can you use any other quadratic equation to eliminate α ? Which values
 do you know from a, b, q and r? The condition has to hold, so what do you

do with these values? How can you use this to find the values of b?

Full worked solutions can be found at www.hoddereducation.co.uk/step-mat-tmua-

answers
Chapter 12: Number theory and combinatorics

MAT STEP
Introduction

Number theory is something that you are not likely to have come across

formally in school mathematics. It is a branch of Pure Mathematics that you

may choose to study in far more detail at university. Much of what you will see

in this chapter will be using skills that you have practised elsewhere in this

book, but there will be some key differences.

Consider the two equations .

The equation has a unique solution for x but the equation

has infinitely many solutions, namely all x and y such that (x, y) is a point on the

line given by this equation. To get a unique solution, you need a second equation

in x and y that corresponds to a line that is not parallel to the first line. However,

what if you only had the original equation but were only interested in values of x

and y that are positive integers. The smallest value x can take is 1, and if x = 1

then . Moving systematically through the possible values of

then

but this is not a

valid solution since y is not a positive integer. The infinitely many solutions of

have become just three valid solutions simply by imposing the

constraint that x and y must be positive integers.

Note that (0, 7) is not an acceptable solution as positive integers are required.

In this chapter, you should expect to find equations which, without any

constraints, do have infinitely many solutions and therefore would not normally

be given to you. You will realise how powerful the constraints can be and

develop a systematic approach to be sure that all solutions are found.

Try it out

1 Find all positive integer solutions for x and y of the equation

2 Find all the negative integer solutions for x and y of the equation

3 Find all positive integer values of a and b such that is an integer.

4 Find all positive integer solutions of

5 Find all positive integer solutions of that satisfy

Hints

1 This is similar to the example given in the introduction.

2 Do you see why you are asked for negative integer solutions instead of

positive integer solutions? Does the equation have any

positive integer solutions?

3 If is an integer, what can you say about the number

4 It may help to focus your attention on the larger numbers, so, here, the value
c
of 5 . There are only a small number of possible values of c, so go through

each in turn.

3
5 Notice that rewriting the inequality using the fact that x is an increasing

function gives These are all cube numbers, positive

multiples of which are added together to give 150. There are only a small
3 3 3
number of possibilities for the values of a , b and c such that
Writing expressions as products

How many positive integer solutions does the equation

have?

The expression on the left-hand side does not factorise, but you can find two

brackets that expand to give all but the constant term:

This means that you can at least rewrite

the equation as In school you

are rightly told to make sure you have zero on one side of the equation before

you factorise. Here, however, because you are only interested in integer

solutions, this is not a problem. There are infinitely many pairs of real numbers

that multiply to give 14 but only finitely many pairs of integers.

x − 5 y − 8 x y

1 14 6 22

14 1 19 9

2 7 7 15

7 2 12 10

−1 −14 4 −6

−14 −1 −9 7

−2 −7 3 1

−7 −2 −2 6

There are five pairs of values of x and y that satisfy the original equation. Notice

that x and y each need to be positive, but this does not mean that this has to be

true of A common mistake would be to miss the solution x = 3

and y = 1, ironically the simplest of the solutions!

Try it out

Find all positive integer solutions of the equation

Variables in the denominator

A useful trick when looking for integer values that result in a rational expression

also having an integer value is to rewrite that expression so that any variables

appear only in the denominator.

Consider this question:

How many positive integer values of n give a positive integer value of

The (n ‒ 13) factor in the numerator indicates that only values of n

greater than 13 need be considered.

If the expression was then you would simply realise that 2n + 1 needs to

be a factor of 20. The actual numerator, 20(n ‒ 13), will changes as n changes. It
is much simpler if the numerator is a constant, so the key step is to realise that,

by long division of polynomials or by inspection, the fraction can be rewritten

as:

So for the above expression to have an integer value, 2n + 1 has to be a factor of

270. Also remember that the question requires n to be a positive integer, so 2n +

1 must also be an odd positive integer other than 1. You may have learned that

knowing the prime factorisation of a number greatly speeds up the process of

finding all of its factors. factors, and 4 × 2

= 8 of them are odd, seven of them after discounting 1.

2n + 1 n

27 13 0

45 22 4

135 67 8

There is one further trap to avoid. Read the question carefully. The value of

must be a positive integer, and this constraint reduces the number

of possible values of n from seven to two: n = 22 and n = 67 are the only valid

solutions.
Irrational numbers

An irrational number is any number that cannot be written as a fraction (a

rational number), in other words, that cannot be written in the form where a

and b are integers and b ≠ 0. Equivalently, it is any number the decimal

expression of which never terminates and has no repeating pattern (for, if it did,

you may have seen how to write this as a fraction). Integers, m, are not

irrational, since they can be written as The first irrational numbers you

come across are surds. The square root of any positive integer that is not a

square number is irrational. You have already seen the proof that is

irrational in Chapter 10: Logic and proof. You can use a similar approach (proof

by contradiction) to prove that the square root of any positive integer that is not

a square number is irrational.

Let n be a positive integer that is not a square number. Suppose that is

rational, so that where a and b are integers and b ≠ 0. Squaring both

sides gives and from this

Now, the trick is to spot that, since square

numbers are precisely those whose prime factorisation only contains even
2 2
powers of primes. Therefore the prime factorisations of a and b consist of only

prime numbers raised to even powers, whereas at least one prime number in the

prime factorisation of n is raised to an odd power. Pick one of those primes and
2 2
call it p. Then p is raised to an odd power in nb but to an even power in a . Two

numbers are equal if and only if their prime factorisations are the same.
2 2
Therefore nb and a cannot be equal and the initial assumption, that is

rational, cannot be true and therefore must be irrational.

Try it out

Try to adapt the proof above to prove the more general result that the rth root

of any positive integer that is not an rth power is also irrational. You can then

conclude that any root of an integer is either an integer or an irrational number.

Two commonly occurring irrational numbers are π and e. It is unlikely that you

have proved that these are irrational and you need to read any question carefully

to know whether you can assume this or not.

Example 1

Question

Solve the equation for rational numbers x

and y.

Solution

Equating terms:

(A)

Equating other terms:

(B)

Substituting (A) into (B)

The only way that the left-hand side can equal the right-hand side is if

the coefficient of on both sides is equal and the non- terms on both

sides are equal.

There is a warning that needs to come with this problem and its solution. The

solution is unique if and only if x and y are rational numbers as

stated in the question. If this were not the case then this solution would not be

unique. You might like to check for yourself that is a

solution to the equation.

So, why does restricting x and y to the rational numbers allow you to equate the

coefficients of separately? Suppose that a is an irrational number and m, n, p

and q are rational numbers such that then you can

show that but this would contradict the fact that a is irrational, since

a rational number divided by a non-zero rational number must also be rational.

So m has to equal n and, subtracting ma = na from the original equation gives n

= q. So
Digits

You may have been under the ‘spell’ of a mathematical magic trick such as the

following.

• Write down two integers from 1 to 9 inclusive, possibly the same, and keep

them secret.

• Take your first number and multiply by 3.

• Add 1 to your answer.

• Multiply your answer by 4.

• Add your second number.

• Subtract 1 from your answer.

• Multiply your answer by 5.

• Subtract 9.

• Add your second number.

• Divide your answer by 6.

• Subtract 1.

What do you notice? You should find that you have the two-digit number formed

by placing your first number in the tens column and your second number in the

units column (this is called concatenation).

Whilst this can surprise people who have never seen it before, if you let your

first number be x and your second number be y then you should finish with the

expression 10x + y and this has the effect of placing x in the tens column and y

in the units column.

In a similar fashion, any three-digit number can be expressed as

for some integers, x, y, z, from 0 to 9 inclusive You can then prove

some surprising results using algebra.

One of the best-known mathematical magic tricks was popularised by David

Acheson in his book 1089 and all that.

The 1089 problem goes as follows.

• Write down any three-digit number whose first digit differs from the last digit

by at least 2.
• Reverse the digits to form a new number and then find the difference, D,

between these two numbers.

• Finally, take the number formed by reversing the digits of D and add it to D.

For example, if you start with 835, reversing the digits gives 538 and the

difference between these numbers is Reversing the

digits of D gives 792 and adding this to D gives 792 + 297 = 1089.

What surprises people is that the answer is 1089 no matter what three-digit

number you begin with, so long as you heed the requirement that the first and

last digit must differ by at least two. Now why does this work?

Let the three-digit number be where x, y and z are integers

from 0 to 9 inclusive, except that Reversing the digits of

gives and the difference between these

numbers is

Even though there are seemingly very many choices for the starting number, the

value of D can only be 198, 287, 396, 495, 594, 693, 792 or 891, since is

an integer satisfying

Now suppose that Knowing that D must belong to the list

of numbers given above, you know that b = 9 and a + c = 9. Adding D to the

number formed by reversing the digits gives

The magic is, in fact, mathematics!

As an aside, if you start not with any three-digit number but with any seven-

digit number, still with the requirement that the first and last digits differ by at

least 2, and follow exactly the same instructions as before, then, with a little

algebra and possibly a spreadsheet to help, you find that there are only eight

possible answers. Granted, ‘the 9900099, 9901089, 10008999, 10890990,

10891089, 10998900, 10999890 or 10999989 problem’ does not have the same

ring to it as ‘the 1089 problem’, but which do you find more impressive?
Exercise 1

1 Suppose that a and b are distinct irrational numbers and that m is a non-zero

rational number. Work out whether each of the following numbers is

rational, irrational or could be either.

(a) ma

(b) ab

(c) m + a

(d) a + b

2
(e) a

2 How many pairs of positive integers x, y satisfy the equation

(a) 0

(b) 1

(c) 2

(d) 19

(e) ∞

3 Find all integer solutions of the equation where a can

be any integer in the interval

4 A two-digit number, x, is written down and a new two-digit number, y, is

formed by reversing the digits. Prove that the difference between x and 4y is

always a multiple of 3. What is a necessary and sufficient condition for the

difference to be a multiple of 6?

5 A two-digit number is squared. A new two-digit number is also formed by

reversing the digits of the original two-digit number and this new number is

also squared. Show that the difference between the two square numbers is

always a multiple of 99. If the difference is not a multiple of 198, what can

you deduce about the digits in the original two-digit number?

6 Start with a three-digit integer, x, the last digit of which is non-zero. Create a

new three-digit number in the following way: move the last digit to the first

digit, the first digit to the second digit and the second digit to the last digit.

For example, if x = 123 then y = 312. Prove that is always a multiple

 of 9 and find a necessary and sufficient condition for to be a multiple

of 99.

7 I write down three distinct non-zero single-digit integers that sum to 14. I

write down the six two-digit numbers that can be formed by concatenating

two of the three integers. What is the sum of the six numbers I have just

written down?

8 Prove that the sum of any four-digit integer and the number formed by

reversing its digits is always a multiple of 11. Give a necessary and

sufficient condition for the sum to be a multiple of 110.

9 Let x be a positive integer of any number of digits. Let y be the number

formed by chopping off the end digit from x and subtracting five times this

end digit from the resulting number.

For example, if x = 1 234 567 then y = 123 456 – (5 × 7) = 123 421.

Prove that x is a multiple of 17 if and only if y is a multiple of 17.

Hint

Since you do not know how many digits x has, you cannot have a variable for

each digit. However, you do not need this: you remove and use the last digit,

and the other digits are all kept together. It therefore makes sense to write x as

10a + b where b, the last digit, is any integer from 0 to 9 inclusive. Then show

that y is a multiple of 17 if and only if 10y is a multiple of 17.

Notice that this gives a test for divisibility by 17. The same thinking can give

tests for divisibility by any prime, in principle.

Combinatorics
Introduction

Combinatorics is the area of mathematics concerned with counting

arrangements of a finite collection of objects subject to certain constraints. You

often need to either list, or imagine a set of objects listed, in a systematic way

that enables you to count how many objects there are. It is sometimes helpful to

partition the objects according to one or more properties and consider separate

cases. It is important to ensure that your cases are exhaustive so you do not miss

any. Two of the main tools of combinatorics are permutations and combinations;

in particular the binomial coefficients have a central role. Admissions test

questions may require you to pick out a coefficient from a binomial expansion or

use an expansion in a novel way.

Combinatorics can easily appear to be a collection of disconnected techniques

and methods but there are some underlying principles which help to make sense

of the subject. These principles are not always referred to explicitly but you have

most likely seen examples.

Counting and conditioning

Conditioning refers to the process of partitioning the elements of a set into

subsets according to whether they satisfy certain conditions.

You may have already met an example of conditioning when deriving a

recursive formula for (sometimes denoted In choosing a set

of size k from a set of size n (with 0 < k < n) you can condition on a particular

element, say x. The number of sets of size k that include x is and the

number that do not include x is Since every set of size k either contains

x or not, it follows that:

This, together with the boundary conditions

explains why the recursive construction for Pascal’s triangle generates the

binomial coefficients.

This is an example of a combinatorial proof in that it involves counting the

elements of a set in two different ways to form an identity.

You can also establish the identity algebraically by using the fact that

Applying this formula to the right-hand side of (*), combined with some

algebraic manipulation, eventually gives you the expression for the left-hand

side of (*).
Example 2

Question

315728 is a six-digit number whose digits sum to

(i) Show that there are 21 six-digit numbers whose digits sum to 52. You

should explain your reasoning clearly.

(ii) How many six-digit numbers are there whose digits sum to 49?

Solution

(i) Conditioning on the number of 9s gives: 999997 (starting with five 9s) and

999988 (starting with four 9s).

Part (i) is a ‘show that’ and there are only 21 six-digit numbers to

find. You could start writing examples in any order until you have

found all 21 but this approach may not help you with the second

part. Instead you should think about how to set out your working in

part (i) so that you can generalise to part (ii).

It is not possible to have a digit sum of 52 using three (or fewer) 9s.

All possible six-digit numbers with digit sum 52 are obtained by

rearranging the digits in these two numbers.

 For 999997 there are six positions for the 7 and the remaining digits must

be 9s giving 6 possible arrangements.

For 999988 there are positions for the 8s and the remaining digits

must be 9s giving 15 possible arrangements.

No further solutions are possible so there are 15 + 6 = 21 six-digit

numbers whose digits sum to 52.

Part (ii) can be done by using the working in part (i) as a guide, you

could start by listing all examples with digits in decreasing order

and then consider how many ways there are of arranging the digits

in each case.

(ii) Conditioning on the number of 9s gives:

There are six positions for the 8 and, for each of these, five positions

for the 5.
 There are six positions for the 6 and, for each of these, there are

positions for the 8s.

There are six positions for the 7 and, for each of these, there are

positions for the 9s. Note that this is the same as considering six

positions for the 7 and, for each of these, positions for the 8s.

Try it out

1 A pay and display car park accepts 10p, 20p, 50p and £ coins. It costs £1.50

to park for up to one hour.

Show that you can make the £1.50 fee in 22 ways.

Sanaya arrives at the car park with four 10p coins, four 20p coins and two

50p coins in her pocket. She picks coins out at random and feeds them into

the parking ticket machine one at a time. What is the probability that she

makes the exact £ fee in five coins or fewer?

2 Find an algebraic proof and a combinatorial proof of the following identity

for

3 By repeatedly applying show that for n > k > 0:

Can you find a combinatorial proof of this result?

Hints
1 What can you condition on? How can you make £1.50 with five coins or

fewer using the coins in Sanaya’s pocket? What is the probability of the

correct coins being put in in one particular order? How many combinations

of the correct coins are there?

2 For the combinatorial proof, given n players how many ways are there to

select a squad of size m, and then a team of size k? Can you think of how to

do this in two distinct ways?

3 For the combinatorial proof try to express each of the terms as a selection as

you did for question 2.

Recurrence relations

A recurrence relation is an equation that generates a sequence once one or more

initial terms are given; each further term of the sequence is defined as a function

of previous terms.

A familiar example of a sequence of numbers generated by a recurrence relation

is the Fibonacci sequence 0, 1, 1, 2, 3, 5, 8, 13, 21,….

It is usual to denote the nth term by Fn (starting with F0), in which case the

boundary conditions and recurrence relation are:

Sometimes you can find a closed form (that is a formula involving just n) for the

nth term of a sequence generated by a recurrence relation. Usually this requires

you to guess (or conjecture) a formula based on a pattern formed by the first few

examples. You can then either argue directly that the formula is correct or verify

that it is correct by substitution.

Example 3

Question

Given the recurrence relation

(i) Calculate u3, u4, u5.

(ii) Conjecture a formula for u .

n

(iii) Verify that the sequence generated by your formula in (ii) satisfies the

recurrence relation.

Solution

(i)
 2
The first six terms satisfy un = n so it is a reasonable conjecture that

2
un = n for all integers n ≥ 0.

2
(ii) Conjecture: un = n for all integers n ≥ 0

2
This argument shows that un = n for all n ≥ 0. It is essentially a proof

by induction.

2 2 2
If un = n then substituting (n − 1) for un − 1
, (n − 2) for un − 2
and

2 2
(n − 3) for un − 3
into 3un − 1
− 3un − 2
+ un − 3
should result in n .

(iii)

Try it out

In music, a motif is a set of notes that repeats during a piece. The recurrence of

the notes in a motif is called a motif pattern. The length of a motif pattern is

the number of notes that are played before the music moves on to a different

motif. Each note in a motif pattern is identified by a number defined as the first

position in the motif pattern where that note occurs, thus, a motif of five

different notes would be written 12345 and a motif of three notes where the

first and last notes are the same but the middle one is different would be

written 121. This structure is called the motif pattern.

There are five possible motif patterns of length three: 111, 121, 112, 122 and

123.

For this question, the number of different motif patterns for a motif of length l

is denoted by M(l). Thus, M(3) = 5.

(i) Write down the value of M(1) and M(2).

(ii) List all of the different motif patterns of length 4 and write down the

value of M(4).
(iii) If N(l ,n) denotes the number of different motif patterns of length l that

use n different notes, write down the values N(4,1), N(4,2), N(4,3) and

N(4,4).

(iv) Write down the values of N(l,1) and N(l,l) for all lengths l.

(v) Explain how N(5,3) can be found from the values of N(4,3) and N(4,2)

and hence find the value of N(5,3).

(vi) There are eight notes in an octave (not including sharps and flats). How

many motifs of length 8 are possible?

Exercise 2

1 How many distinct numbers greater than 5000 can be formed from the digits

3, 4, 5, 6 and 0, each digit being used at most once in any number? The first

digit may not be 0.

One of these numbers is chosen at random. What is the probability that it is

divisible by 3?

2 How many ways can the integers 1, 2, 3, 4, 5 be arranged so that there is

only one integer that is immediately followed by a smaller integer?

How many ways can the integers 1, 2, 3, …, n with n > 1 be arranged so that

there is only one integer that is immediately followed by a smaller integer?

3 In a multiple choice test with 20 questions, candidates score 3 marks for a

correct answer, 0 marks for abstaining and −2 marks for an incorrect

answer. If the pass mark is 42 marks, how many ways can a candidate pass

the test in terms of correct answers, C, incorrect answers, I and abstentions,

A? (The order in which these occur is not important.)

In another test, the same marking system is used. If there are 50 questions in

this test and the pass mark is 105, how many ways can a candidate pass the

test in terms of correct answers, C, incorrect answers, I and abstentions, A?

4 This diagram shows nine small triangles arranged to make a large triangle.

In how many ways can two circular counters be placed such that each

counter is in the centre of a smaller triangle and each triangle contains at

most one counter (so the counters are not stacked)?

If reflections and rotations (and combinations of these) are considered to be

identical arrangements, how many different arrangements are there?

For example
 and

would only count as one arrangement as rotating one gives the other.

In a similar way,

and

would only count as one arrangement as they are reflections of each other.

5 Pogo the clown has a red left sock, a blue right sock, a yellow left boot and a

green right boot. He wears two gloves on each hand, an inner glove and an

outer glove. If Pogo is to achieve his trademark look, he must put the socks
and boots on the correct feet with the boots over the socks and the outer

gloves over the inner gloves. Needless to say a left-handed glove must go on

the left hand.

(a) How many different ways can Pogo achieve the correct look when

putting his socks, boots and gloves on?

Pogo finds that he is unable to put his socks, boots or inner gloves on once

he has put either of his outer gloves on.

(b) How many ways can Pogo achieve the correct look now?
Exercise 3
A MAT style question

There are two ways of dividing a quadrilateral into two triangles using one

diagonal line.

The two are considered different arrangements as a different pair of vertices is

joined in each case.

The notation Tn is used to denote the number of ways in which a convex (n +

2)-sided polygon can be split into n triangles using (n – 1) diagonal lines that

do not cross. A diagonal line is one that joins non-adjacent vertices. The

example above shows that T2 = 2.

(i) Explain why T1 = 1.

(ii) Draw all of the arrangements for T3.

One of the arrangements for T4 is

(iii) By considering all of the possible triangles that contain (1, 2) and the

effect of removing edge (1, 2), find a formula for T4 based on T3, T2, T1

and T0. (One possible triangle removal is shown below.)

(iv) Calculate the value of T4.

(v) Explain why

where we define T0 = 1

(vi) How many ways are there to divide an octagon into six triangles using five

diagonal lines that do not cross?

A full worked solution can be found to the question from Exercise 3 at

www.hoddereducation.co.uk/step-mat-tmua-answers
Chapter 13: Further trigonometry

MAT STEP
Introduction

The trigonometry requirements of the MAT syllabus are detailed in Chapter 9:

Trigonometry. There is not too much that you are expected to know and the

absence of radians on the MAT paper means that it is unlikely that there will be

any calculus involving trigonometric functions. It would be easy to imagine that

the questions requiring trigonometry in the MAT are going to be straightforward

and comparatively simple. This is not actually the case. The MAT exam can

contain some questions that really push your understanding, particularly in the

longer questions.

The STEP specification has far more content. You are expected to be able to use

the sine and cosine rules and the formula for the area of a triangle.

You are expected to work in radians and know various circle area and arc length

formulae. You are expected to know the small angle approximations for sinθ,

cosθand tanθ. The functions secθ, cosecθ and cotθ are used and you should make

sure you know the relationships between these and cosθ, sinθ and tanθ. Formula

sheets are not allowed in the STEP examinations so you will have to learn the

wide range of trigonometric identities including double angle formulae, addition

formulae, the sum and difference of two sines or two cosines and the

substitution formulae. You are also expected to be able to use trigonometric

functions in differentiation, integration and the solution of differential equations.

Trigonometry in the MAT

Trigonometry in the MAT tends to focus on using what you know in a problem

solving way rather than on simply going through a mechanical procedure. You

should work through problems looking for relationships that you recognise. You

will not be set something that is not in the syllabus.

Example 1

Question

For this question you may assume Heron’s formula for the area of a triangle

with side lengths a, b and c:

(a) The diagram shows three similar right-angled triangles PRQ, PTR and RSQ.

Find the lengths PR, QR, SR, QS, RT and PT in terms of θ.

Find an identity for cos2θ in terms of cosθ and sinθ.

(a) Triangle ABC has side lengths a, b, and c and angles A, B and C. Use the

sine rule, the cosine rule and the formulae for the area of a triangle to show

the following relationships.

 (i)

(ii)

Solution

(a)

The vertical dotted line is a clue to what to do here. cos2θ will be .

(b)

For part (b) a diagram will help. At this stage a path to the required

answer is not clear. You are told to use the sine rule, cosine rule and

triangle area formulae.

If you let this be equal to a constant r then you can write a, b and c in

terms of sinA, sinB and sinC.

a + b + c is the sum of the side lengths, in other words the perimeter of the

triangle. This approach has at least given a geometrical appreciation of

what you are trying to find. The sine rule has helped and indicates that it

is connected to Heron’s formula that you are given at the start of the

question. The question does say to use triangle formulae implying that you

may wish to use more than one formula.

Nothing is jumping out from this yet so it is worth leaving it there and

coming back to it. The cosine rule hasn’t been used yet so that should be

the next stage. Once a relationship has been established using one angle

e.g. A, two more can be written directly from the result by changing the

angle and relative position of the side lengths in the formula.

The right-hand side of the expression for part (i) includes terms in

The work from part (a) should help as

can be interpreted as

Each term in the brackets on the right-hand side is the difference of two

squares.
This now looks like the result of Heron’s formula so you can rewrite it in a

slightly more compact way.

You are trying to show that this is the same as

The question says to use the triangle area

formulae and so implies that you should use more than one area formula.

That took a lot of effort but having established this result you should expect the

other results to be easier to find. This can be done much more quickly when

you know and are allowed to use addition formulae.

Try it out

Finish the question. You have most of the results you need already.

Hints
For part (ii) establish an expression for in terms of cos A using the

identity you found in part (a). You can find similar expressions for and

Try to write the expressions you get as the difference of two squares.

Find an expression for and take the square root.

You should also have enough information to write down the result of cos A +

cos B + cos C. Don’t forget that you can also use the results you know from the

area formulae. It will still take a lot of algebraic manipulation to get to the

required result.
Exercise 1

2 2
1 A triangle ABC has sides AB = n − 1, BC = n − n + 1 and AC = n(n − 2).

Find angle

2 For a triangle ABC with side lengths BC = a, AC = b and AB = c,

3 Point D on the perimeter of triangle ABC is the point at which the line from A

that is perpendicular to BC hits BC. Point M is the mid-point of BC. Given

that AB = c, BC = a, AC = b and AD = BD, show that

(i)

(ii)

4 For a triangle ABC with side lengths BC = a, AC = b and AB = c, use the

formulae

2 2 2
2bc cos A = b + c − a and to show that the triangle area

can also be written as

Hint

You should square both of the given formulae.

Trigonometry in STEP

Trigonometry in the STEP papers can be very challenging. You are expected to be

fluent in using a number of identities and, as there is no formula sheet, you are

expected to know them. In addition to the usual identities you use in your school

mathematics lessons, it is also useful to know some additional ones.

The sum and difference of two sines
The sum and difference of two cosines
The formulae

Using the substitution is particularly useful for solving equations of

the form a cos x + b sin x = c, where a, b and c are constants. It is also useful as a

substitution in some integration problems as it replaces trigonometric functions

with a simple variable. You will see this in Example 2 in Chapter 14: Calculus –

integration and differential equations.

Example 2

Question

Prove that

Find the least value of 2 sin x sin (x + α), where α is a constant, as x varies.

Solve the equation

Solution

Using

There is more than one way to do the first part. You should use something

you are expected to know. The addition formulae should give this result

fairly directly.
There must have been a reason for proving this result. The next part of

the question will use it to good effect.

The final part of the question must use the same relationship once again.
No other values will give answers in 0 ≤ x ≤ 2 π after is subtracted.
Exercise 2

1 Triangle ABC has sides AB = c, BC = a and AC = b. Point M is the mid-point

of AB.

If , show that

2 cot θ = cot B − cot A

and also that

where T is the area of the triangle ABC.

2 2 2
Given that a , b and c form the first three terms of an arithmetic

progression, describe the relationship between cot A, cot B and cot C.

2 Show that

(i)

(ii)

Use your second answer to give an exact value for

3 If A, B and C are the three internal angles of a triangle, prove that

(i) sin 2A + sin 2B = 2 cos(A − B) sinC

(ii) sin 2A + sin 2B + sin 2C = 4 sin A sin B sin C

(iii) sin 2A + sin 2B − sin 2C = 4 cos A cos B sin C

4 Use the substitution to solve the equations, giving your answers in

the form 2 arctan a + 2nπ

(i)

(ii) 3 sin x − 4 cos x = 4

(iii)
Exercise 3

1 (a) Find the fraction of the large semicircle that is shaded in the diagram.

2 (b) In the diagram, the two shaded areas are equal.

If the angle between the diameter AB and the chord PB is θ, show that
A STEP question

2 Given that cos A, cos B and β are non-zero, show that the equation α sin(A −

B) + β cos(A + B) = γ sin(A + B)

reduces to the form

(tan A − m)(tan B − n) = 0,

2 2 2
where m and n are independent of A and B, if and only if α = β + γ .

Determine all values of x, in the range 0 ≤ x < 2π, for which

(i)

(ii)

(iii)

Cambridge Assessment Admissions Testing STEP 2 2007 Question 4

Hints

This is a question where you should get started as soon as possible rewriting

sin(A − B), cos(A + B) and sin(A + B) in terms of products of the sines and

cosines of A and B. It is difficult to see how to progress unless you have done

this. Putting all the terms on one side and factorising as much as possible will

also help. Remember that you were told that cos A, cos B and β are non-zero.

This is not redundant information. You should eventually be able to get to a

form that is based entirely on tan A and B This can be compared to (tan A − m)

(tan B − n) to find some relationships between m, n, α, β and γ. The question

uses the words if and only if. Have you done enough to show this?

For parts (i), (ii) and (iii), you should use the relationship you have just

established. The first two of these are very similar. The last part takes a little

more effort. You should be careful and make sure you have all of the solutions

in the required range.

Full worked solutions can be found to all of the questions from Exercise 3 at

www.hoddereducation.co.uk/step-mat-tmua-answers
Chapter 14: Calculus – integration and differential

equations

STEP
Introduction

You will have studied some integration techniques and some methods for solving

differential equations. Many of the skills you need are either taught in A level

Mathematics or A level Further Mathematics. What they don’t prepare you for is the

wide variety of ways in which you can be presented with an integration or with an

unusual looking differential equation. The calculus on STEP papers often does not

resemble the calculus that you have seen before. When confronted with a difficult

looking integration or differential equation, it is often the best strategy to start by

diving in and applying what you do know, keeping in mind that you will not be asked

anything that is beyond what is stated on the specification. You will often find that

what you are dealing with resolves itself to a much nicer and more recognisable form.
Integration

The STEP 2 specification states that you need to be able to evaluate improper integrals

where either the integrand is undefined at a value in the range of integration or the

range of integration extends to infinity. You need to know how to integrate functions

of the form and and be able to choose trigonometric

substitutions to integrate associated functions. You are expected to know how to

rewrite algebraic fractions as partial fractions (including those with quadratic factors

in the denominator) and integrate them. You should also know how to derive and use

reduction formulae.

The STEP 3 specification adds integration of hyperbolic functions and their use in

integrating functions of the form and . You need to be able to

choose suitable hyperbolic substitutions to integrate some functions.

Different routes to equivalent results

At A level you will have been used to looking up integrals from a formula sheet. This

tends to give the impression that an integral has only one correct form. In the STEP

examination you may well be expected to use a slightly different method (e.g. an

unusual substitution) to find an integral you would normally look up on a formula

sheet.

The result you get for that integral may not even resemble the one on a formula sheet.

It will be equivalent but may well have a form that proves more useful for the later part

of the question.

Consider the indefinite integral

An A level formula sheet will tell you that the integral of this is ln(sec x + tan x) + c.

One alternative way to integrate this is to first consider the integral of cosec x and the

double angle formula for sin2A.

This can be rewritten by multiplying both top and bottom of the integrand by

This is now in the form since .

Now, how does this help with ?

Since

the previous result can be used, substituting for

A third, and final way (for this chapter) to do the same thing is to start with

and then use the substitution

so

The integral becomes

Using partial fractions this becomes

Try it out

The three results and

are all equivalent.

Can you show that they are?

Integration by substitution

You will have been taught how to integrate using a substitution. In the STEP

examinations, the substitutions you are asked to use are often more unusual than the

ones you meet at A level. It is worth remembering the A level methods you have been

taught as they do still work, even with some of the substitutions you are asked to

make.

Example 1

Question

Use the substitution to show that

Find the value of

This doesn’t look that much like an A level question but starting in the same way

should give you some idea of what to do.

Solution

This has, at least, given the limits you are looking for. There are two

possibilities for the next stage. You could differentiate u with respect to x or

rearrange or make x the subject and differentiate x with

respect to u. In a STEP examination you would probably try one and if it

worked, stick to it and if it didn’t, try the other. A better approach is to

consider the result you are trying to obtain. From the substitution, if you
2
differentiate u with respect to x, you will have a lot of x and x terms to deal

with. In this case, it is probably better at a first glance to rearrange the

substitution to make x the subject and then differentiate with respect to u.

To find requires the quotient rule for differentiation.

This actually looks like a part of what you are aiming for so you should feel

encouraged to continue. The elements of the substitution are all ready to be

used.
The second part is clearly about applying what you have just found to find the

actual value of an integral. For this integral if the substitution

then .

The fraction will need to be expressed in partial fractions to enable a

known integration method to be used.

Let

Let

Let
As so and
Trigonometric substitutions

In the same way that the algebraic substitutions you are given are often more unusual

than you are used to, if you are asked to use a trigonometric substitution it will often

be a little different than you see at A level. As with all of the questions in the STEP

examination, you should keep it in mind that you will not be expected to use any

trigonometric identities that you haven’t seen before. You will need to learn some as

you do not have a formula sheet in the STEP examination. Sometimes the substitution

can seem quite straightforward and it is the amount of manipulation that is required

that makes the question difficult.

Example 2

Question

Use the substitution where to show that

where c is a constant of integration.

(i) Find the value of

(ii) Use the substitution to show that

(iii) Evaluate

Solution
The initial part of the question is asking for the derivation of a standard result that is

usually given.

Part (i) must link in to this in some way but it is not apparent until you actually start
2 2
work. You can exploit the identity sin x + cos x = 1 by using the substitution u =
2
2sin x. This will make the denominator of the fraction a + u which goes some way

to giving something like the integral you have just found.

(i) Let u = 2sin x

Part (ii) gives some idea of what to do in terms of saying the substitution that

you need to use. The temptation, particularly with the way the question is

written, is to try to substitute for u in the integral on the left-hand side. You

should recall, however that the substitution allows and to

be expressed in terms of t. These t formulae are easy to forget but they have

turned up on several STEP papers. If you can’t remember what they are, at least

remember that they exist and how you can derive them from the double angle

identities sin2A = 2sin A cos A and cos2A = cos2 A − sin2 A.

Hence, if

Hence, if
(ii) Let

At this stage you should not be put off. Although the integral looks unpleasant

at the moment, some careful manipulation will make it look far more

reasonable. Since you are trying to end up with something like the left-hand

side integral, you do at least have a target to aim for.

The final part of the question looks horrendous but it is worth thinking about

what you have done so far. Part (iii) must be related to everything in some way.

You need to look for clues and perhaps make a leap of faith based on a hunch.

The first clue to spot is that the lower limit is and the upper limit is 0.

Compare this to the limits of the integral from part (ii). They have been shifted

to the left by . The substitution may make the integral more

recognisable. It is a bit of a leap of faith but worth trying.

(iii) Let

It is a good idea to separate out some of the working to make things

manageable and to help avoid making simple algebraic slips.

At this stage, working term by term will help maintain accuracy.

This is times the integral from part (ii) which in turn was of the integral

from part (i).

That was a very long and drawn out question but it does serve a point. STEP questions

are often written so the parts gradually build to contribute to solving a difficult

problem at the end. In this case the awful looking final integral was rendered

manageable by appreciating how it was related to the earlier integrals. Make sure that

you understand how the result at each stage was reached. You should realise on

reflection that there is not much to understand beyond the fact that the integrals were

linked. The majority of the work was in manipulating algebraic expressions to look

like something else. STEP questions often require this level of algebraic dexterity.
Exercise 1

1 Use the substitution to find the indefinite integral

2 Use the substitution t = tan x to evaluate the following definite integrals.

(a)

(b)

2
3 By using the substitution x = sin u, evaluate the definite integral

4 Use the substitution to evaluate

(a)

(b)

Hint

In both of these there is something to spot. (It is easier to spot in part (a) what

makes the integration easier than expected.)

5 Use the substitution u = cosh x to evaluate

6 Use the substitution u = sinh x to evaluate

Differential equations

For both STEP papers you will be expected to be able to evaluate the analytical

solution of simple first order differential equations with separable variables, including

finding particular solutions.

For STEP 3 you need to be able to find and use an integrating factor to solve first order

differential equations. You will be expected to be able to solve some second order

differential equations and understand the language and concepts of some simple

methods including the terms auxiliary equation, complementary function and

particular integral.

Differential equations are often found in the STEP mechanics questions.

First order differential equations

You have two methods at your disposal that you can use to solve first order differential

equations: separating the variables and using an integrating factor. These methods may

be obvious from a first glance or may become apparent after you use a given

substitution.

You are expected to be confident enough with both methods that you are able to deal

with differential equations that are a little more unusual than those found at A level.

Example 3

Question

Given that y = 0 when x = 0, solve the differential equation

for the case where k = 2.

Solution

This is similar enough to a standard problem with the only real difference

being that you have a general index k applied to y and an index of applied to

k
(1 − y ). This is simply subterfuge as you are told which value of k you will be

using straight after the equation. As usual, it is a good strategy to start by

doing what you can and seeing where you get to.

This now becomes a case of finding a suitable substitution to be able to work

with the left-hand side. Since you have in the denominator then using
2 2
the substitution y = sin u will give 1 − sin u which is equal to cos u.
To find c, the information that y = 0 when x = 0 can be used.

Example 4

Question

The curve C is defined for x ≥ 0 by the differential equation

Given that when x = 1, show that

Solution

The first stage is to separate the variables. That is comparatively

straightforward.
The right-hand side is easy to integrate. The left-hand side looks challenging.

The trick here is to see whether writing the top of the fraction in the form

will simplify the integration.

The two parts of the integral on the right-hand side are both of the form

and so can be integrated by sight.

The constant can now be found using the condition when x = 1.

The final stages will be to put in the value of the constant and rearrange the

expression to the required form.

Substitution

Just as a carefully chosen substitution can greatly simplify some integrals, some

differential equations can be simplified by making a suitable substitution.

Example 5

Question

Use the substitution y = vx where v is a function of x and x > 0, to find the general

solution of the differential equation

Solution

The first step will be to find in terms of v, x and then substitute this for

and vx for y in the equation. You need to remember that v is a function of x

so differentiating y = vx with respect to x to get will involve the product rule.

2
You can divide through by x here since you are told that x > 0 in the question.
If C is a constant then so is In C (provided C > 0). It will make the

simplification at this stage much easier.

Example 6

Question

(i) Use (*) to find the solution to the differential equation

that satisfies y = 1 and

(ii) Use (*) to find the solution to the differential equation

Solution

The first part will involve both implicit differentiation and the product rule for

k
differentiation. z is the product of y and Each of these will have to be

differentiated implicitly with respect to x.

There is a helpful target to aim for. You can see that for the required form of

the answer, factors of can be taken out of the expression. The

required result can be reached relatively quickly.

(i)
For part (i) you are clearly going to use the expression you have just

found. The left-hand side corresponds in part to the part

of

To make the left-hand side of the equation the same as , the equation

can be multiplied by

1/2
The left-hand side is now and the right-hand side is z

The conditions y = 1 and when x = 0 can be used to find c at this

stage.

The solution isn’t over yet as the equation is given in terms of x and y. It

should be possible to write the solution in terms of x and y too.

(ii)

For part (ii) the first two terms of the equation correspond to parts of with

k = ‒2 and n = 3. The term can be removed by adding y

5/3
to both sides of

the equation.

To make the left-hand side into , both sides of the equation need to be

multiplied by

Try it out

You should be able to complete the solution of this equation. Be careful as the

conditions that this equation has to satisfy are slightly different than for part (i).
You should see that you have once again on the left-hand side of the equation.

How can you write the right-hand side in terms of z?

The differential equation in x, y and that you eventually reach will need a little

manipulating before you are able to solve it using conventional A level methods.
Second order differential equations

The methods you have been taught to solve second order differential equations at A

level are expected to be understood for the STEP 3 paper. This includes finding the

auxiliary equation and the complementary function for all forms of the solution of the

auxiliary equation. You are expected to be able to choose a suitable test function to

find a particular integral. As with first order differential equations, you may be given a

substitution to use to change something that looks unusual into a familiar looking

differential equation.

Example 7

Question

z
Use the substitution x = e to find the general solution to the differential equation

Solution

The first task for this question is to use the chain rule to find expressions for

and

can be found using the product rule.

You now have the terms you need to be able to make the substitution.

The substitution has resulted in a recognisable second order differential

equation that can be solved using conventional methods. The auxiliary

equation factorises easily.

For a repeated real root, the complementary function is

The next step is to find the particular integral that solves the differential
2z
equation. The right-hand side of the equation is ze so the particular integral
2z
will be based on this. Since both e and z form a part of the complementary

function, the particular integral will have to be of the form

These can now be substituted into the differential equation to find the value of

a.

At this stage, many of the terms will conveniently sum to 0.

The general solution to the differential equation is the sum of the

complementary function and the particular integral.

This is the general solution in terms of y and z. The original equation was in
z
terms of x and y so the final stage of the solution is to substitute x for e and ln

x for z.
Exercise 2

1 Given that where k is a constant, show that

2 Show that is a general solution of the equation

3 Use the substitution to find the general solution of the differential equation

4 Use the substitution v = sin y to find the general solution of the differential

equation

5 Show that is a solution to the differential equation

6 Use the substitution to solve the differential equation

given that y = 1 when x = 1.

7 Use the substitution x = tan z to solve the differential equation

for the conditions that when x = 0.

8 By using the substitution , find the general solution to the equation

 9 By using the substitution , solve the equation

for the conditions that when x = 0.

10 The equation of motion of a particle in orbit under a force of F towards the centre

of the object it is orbiting is given by the differential equation

In this equation z is the reciprocal of the distance of the particle from the centre

of the object it is orbiting and k is a constant.

Solve the equation given that

2
K, k and d are constants and K > k .
Exercise 3
STEP questions

1 (i) Show that

(ii) By means of the substitution , show that

where f is any function for which these integrals exist.

Hence evaluate

(iii) Evaluate

Cambridge Assessment Admissions Testing STEP 2 2015 Question 6

Hints

(i) It may be better to start with . Is there an identity that will

allow you to write this in terms of cos(f(x))? You should be able to spot a very

simple trigonometric relationship that you can use if you do this.

You should use the relationship to integrate . You should know the

2
integral of sec x with respect to x.

(ii) You are told the substitution to use. Work with the left-hand side and show that

it can be manipulated to the expression on the right-hand side.

You should be able to write in a better way.

 You can split the integral you get into two integrals, one of which you should

recognise.

Once you have established that the relationship is true, you can apply it to

since this is of the form xf(sin x). This means that .

Can you see how the parts of this question slot together?

(iii) This needs a little manipulation before you can use what you have found. The

hint to how to integrate the function is at the start of part (ii).

Split the integral into two parts and

Do these integrals respond to using the substitution ? Start with

and think of this as

This will give you something that can be split into four related integrals. You

should see many things you recognise in these (including

which you can piece together to make the integral you need.

Once you have simplified the various pieces, you should see how to evaluate the

given integral.

2 (i) By writing , where u is a function of x, find the solution of the

equation

for which .

(ii) Find the solution of the equation

for which .

(iii) Give, without proof, a conjecture for the solution of the equation

for which y = 1 when x = 0, where n is an integer greater than 1.

 Cambridge Assessment Admissions Testing STEP 2 2008 Question 7

Hints

(i) This is a fairly direct application of standard methods. You can think of

as a product and use the product rule to find an expression for .

You are eliminating y so substitute and your expression for into

the original equation. You should now be able to solve the differential equation

by separating the variables and integrating each side of the equation you get.

(ii) This is very similar to part (i) in the method you use. Looking ahead at part (iii)

should help you to realise that you want the final answer to look as similar as

possible to the answer to part (i) so that you are able to make the conjecture for

part (iii).

Full worked solutions can be found to all of the questions from Exercise 3 at

www.hoddereducation.co.uk/step-mat-tmua-answers
Chapter 15: Complex numbers

STEP
Introduction

The STEP 2 syllabus (2020) requires you to be able to solve any quadratic, cubic

or quartic equations with real or complex coefficients. You need to know how to

add, subtract, multiply and divide complex numbers in the form x + iy with x and

y real. You are expected to understand and use the complex conjugate and know

that non-real roots of polynomial equations with real coefficients occur in

conjugate pairs. You need to be able to use Argand diagrams and construct and

interpret loci in the Argand diagram including those expressed as and

arg(z − a) = θ. You should be able to convert between Cartesian and modulus-

argument form (using radians) and to multiply and divide complex numbers in

modulus-argument form.

The STEP 3 syllabus (2020) adds that you should understand de Moivre’s

theorem and use it to find multiple angle formulae and sums of series. You need

to know and use the definition and the form . This

includes finding the nth roots of a complex number and understanding their

geometrical interpretation on an Argand diagram. You should be able to use

complex numbers to solve geometric problems.

This content mirrors what you should have encountered or will encounter if you

are studying Further Mathematics at A level but, as this is an admissions test, you

can expect the level of algebraic manipulation and understanding of what you are

doing to be much greater. As you work through the examples and exercises in this

chapter, you should try to understand why everything works. As much as possible

attempt to see how what you are doing would be interpreted geometrically. It is

assumed that you have an introductory understanding of the ideas above so, if you

have not covered some of the content, you may find it helpful to look it up before

working through this chapter.

The different forms of complex numbers

There are three ways that you will see complex numbers presented.

• Rectangular form x + iy

This form is useful when identifying the real and imaginary components of a

complex number. It is also a convenient form to use when adding and

subtracting complex numbers.

• Polar form r (cosθ + i sinθ)

This may also be referred to as modulus-argument form as you can read the

modulus and argument directly from r and θ. This form is not only useful for

identifying the modulus and argument of the complex number, it also provides

the link between complex numbers and trigonometry.

• Exponential form

This is another modulus-argument form; the values can be read directly from

the way it is presented. This form is convenient for multiplying and dividing

complex numbers. More importantly it provides the link between complex

numbers and hyperbolic functions.

How you choose to present complex numbers when tackling particular questions

can affect how easy it is to spot the path to the solution. You need to be confident

with all forms and be able to move between them as quickly and efficiently as

possible.
Properties of conjugates

Every complex number has another complex number that is associated with it; its

conjugate. For a complex number expressed in the form x + iy the conjugate of

that complex number is the complex number x − iy. If the complex number is

expressed in the form r (cosθ + i sinθ) then the complex conjugate would be

written r (cosθ − i sinθ). For the form , the conjugate is . The algebraic

notation used for the conjugate of the complex number z is z*.

On an Argand diagram, the conjugate of a complex number is the reflection of

that complex number in the real axis.

A complex number and its conjugate have the same modulus.

A complex number and its conjugate have arguments of the same magnitude but

of opposite sign.

When solving polynomial equations with real coefficients, complex roots occur in

conjugate pairs. Any polynomial equation of this sort of odd degree must

therefore always include at least one real root.

Multiplying a complex number by its conjugate results in a real number equal to

the square of the modulus of that complex number: .

Example 1
Question

By considering the complex numbers z = a + bi and z = p + qi simplify the

following expressions by removing the brackets. Give your answers in terms of

z and z .
1 2

(a)

(b)

(c)

(d)

Sketch each of the above on an Argand diagram choosing suitable values for a,

b, p and q. Interpret parts (a) and (b) of the above geometrically.

Solution

(a)

(b)

(c)

(d)

Using z = 3 + 2i and z = −2 + i to illustrate each one:

1 2

(a)

z + z represents a forward movement from the origin to the point z

1 2 1

followed by a forward movement from z through z . The conjugate of this

1 2

is a reflection of this in the real axis, hence

z * + z * represents a forward movement from the origin to the point z *

1 2 1

followed by a forward movement through z2* hence

(b)

z * − z * represents a forward movement from the origin to the point z *

1 2 1

followed by a backward movement through z *. The conjugate of this is a

2

reflection of this in the real axis, hence

z1 − z2* represents a forward movement through z1 followed by a backward

movement through z *. The real component of z is the same as the real

2 1

component of z * and the real component of z * is the same as the real

1 2

component of z2. The imaginary component of z1 is the same magnitude but

opposite sign to the imaginary component of z1* and the imaginary

component of z * is the same magnitude but opposite sign to the imaginary

2

component of z .
2

Hence
You should notice the equivalence between the use of complex numbers on the

Argand diagram and vectors in two dimensions.

de Moivre’s theorem

When you multiply two complex numbers together, you get a new complex

number in which the modulus is the product of the moduli of those numbers and

the argument is the sum of the arguments of the numbers.

Using polar form, if you have

and

then, knowing the effect of multiplying the two together on the modulus and

argument of the result,

This implies that multiplying the complex number by itself

gives the result

Multiplying this result by z once more should give

This suggests the result that is known as de Moivre’s theorem after Abraham de

Moivre, who, although he never published it explicitly, clearly applied it in many

of his papers in the early eighteenth century.

de Moivre’s theorem considers a complex number with modulus 1 and states that

Try it out

1 Prove de Moivre’s theorem for positive integers n by induction.

2 Prove de Moivre’s theorem for n = 0 by substitution.

3 Prove de Moivre’s theorem for negative integers n by letting m = −n.

Applications of de Moivre’s theorem

The most common application of de Moivre’s theorem is in finding identities for

k k
cos kθ, sin kθ, cos θ and sin θ.

For cos kθ and sin kθ, a binomial expansion is used.

Example 2

Question

Express cos4θ in terms of cosθ.

Solution

This can be expanded using the fourth row of Pascal’s triangle: 1, 4, 6, 4,

1.

To simplify the notation, c can be used for cosθ and s for sinθ.

2 3 4
Since i = −1, i = −i and i = 1, this can be rewritten in a more helpful way.

Equating real parts

This is not entirely in terms of cosθ. The identity can be

used to rewrite the and terms.

is the real part of so the expression for can be

found from the real part of .

It is worth noting that you can also find an expression for by equating the

imaginary parts of the expansion.

For and the approach is to use from which, by de

Moivre’s theorem,

and

(since and )

Adding these cancels the imaginary parts to give

Subtracting cancels the real parts to give

Example 3

Question

Express in terms of multiple angles.

Solution

Let

The binomial expansion can be done using the sixth row of Pascal’s

triangle . For the right-hand side, and since ,

.
 5 −5 3
This doesn’t look that hopeful but pairing the terms in z with z , z with
−3 −1
z and z with z should give something familiar

Using means that can be rewritten as ,

as and as .

This is particularly useful when you wish to integrate expressions like or

Try it out

Find the value of

Hint

Let and use to rewrite in terms of multiple

angles first.

Example 4

Question
Show that if then

Solution

At an initial glance this does not appear to be anything to do with complex

numbers. The indicator here is that substituting into the left-

hand side of the expression would result in a series of terms, the

highest power of which is . This can be linked to the on the

right-hand side as that can be written as terms of powers of . This will

include a term in which must be reduced to the required by

dividing by .

Using

Now you have an expression for , you can simplify the right-hand

side. This will still involve a lot of algebraic manipulation. Using c for cosθ

should make this easier to work with.

Let c = cosθ

This can either be simplified by algebraic division or by equating

coefficients.
Equating coefficients

After all that work you are nearly there. The last stage is to show that this

is equal to if x = 2cosθ.

Is this equal to ? It will be easier to multiply this out in

order to compare.
The roots of unity

In 1608, mathematician Peter Roth’s publication Arithmetica Philosophica

included the conjecture that a polynomial equation of degree n with real

coefficients has solutions. Many famous mathematicians, including d’Alembert,

Euler and Laplace, attempted to prove this was the case. The first really rigorous

proof was published by Jean-Robert Argand in 1806, a publication that also

included the first statement of the fundamental theorem of algebra that included

polynomials with complex coefficients.

The polynomial equation has n roots. For these roots are

and . For is obviously a root but what are the other

two?

If is a root of then, by the factor theorem, is a factor of

It follows that for some value of . By equating

coefficients of it can be seen that .

This means that the equation can be rewritten as

. The other two roots must come from .

Using the quadratic formula

The other two roots are the conjugate pair of complex numbers

and .
Since these are the roots of the equation which can be written as

, they are referred to as the cube roots of unity.

The diagram shows the three roots plotted on an Argand diagram.

The roots of are all a distance of 1 from the origin and form an

equilateral triangle when connected.

For , the modulus and the argument

For , and

For , and

You should know that multiplying a complex number ω by a complex number z

results in a complex number with modulus and argument .

The geometrical effect on the Argand diagram is that ω is enlarged by a scale

factor of z and rotated anticlockwise by .

2
Consider . To find it can be multiplied by z to get z and then by z once

more. The first multiplication will enlarge it by a scale factor of 1 and rotate it 0

radians anticlockwise giving . Multiplying it by z once more, will enlarge it

by a scale factor of 1 and rotate it 0 radians anticlockwise giving .

Geometrically can be seen to be a cube root of 1.

Consider . Applying the same process means that the first

multiplication will enlarge it by a scale factor of 1 and rotate it radians

2
anticlockwise. z therefore has a modulus of and an argument of

. Multiplying by z once more will not change the modulus but the

3
argument of z will be . It is on the real axis and so in this

case and can be seen to be a cube root of 1.

Try it out

Show that the same is true if .

Cubing and should result in 1.

Find expressions for and and show that they are

both equal to 1.
 n
z - 1 = 0

Every root of the equation must have unit modulus. If this were not the

case then the modulus of zn would not be 1. Each argument must be such that

when multiplied by n, the result is an integer multiple of .

Example 5

Question

Find the sixth roots of unity and sketch them on an Argand diagram.

Solution

The modulus of each root is 1. One root must be . All of the other

roots must have an argument arg(z) such that where k is an

integer. Using the convention that the principal argument is given in the

range , the values of k used will be and 3.

This is in exponential form which is very convenient for roots of unity.

They could also be written in modulus argument form as 1, ,

and .
From this the answers could be put into x + yi form as

and .

On an Argand diagram, the points will be arranged as the vertices of a

regular hexagon centred on the origin with the first vertex at (1, 0).

If an integer such as had been used, the result that this returns,

would need to have been interpreted as the equivalent principal argument of

.
Properties of the roots of unity

The roots of can all be written in the form

where k is any integer.

In the example of finding the sixth roots of unity, negative integers were used to

ensure that the value of each argument was between and .

If integers from 0 to are used for , the roots can be written as

and from this some useful properties can be

found.

If we let the root with the smallest positive argument be i.e. then

, and so on to . Geometrically this can be

interpreted by representing the rotation to get from one root to the next.

Multiplying 1 by moves from the root to the root .

Multiplying this by moves from the root to the root and so

on until the last root is reached.

This means that the nth roots of unity can be written as

For the sixth roots of unity the roots can be labelled on the diagram like this:
The sum of the nth roots of unity form a geometric

series with first term 1, common ratio and n terms.

For this it can be seen that

Multiplying by times gives the last root of unity. Multiplying by once

more will result in one more turn of giving hence

The sum of the nth roots of unity is zero.

In admissions tests you are expected to be able to find roots of unity very quickly

and use their properties as a matter of course.

Roots of complex numbers

The nth roots of any complex number can be found by a similar method.

Example 6

Question

Find the cube roots of and represent these on an Argand diagram.

Solution

This is the same as finding the roots of the equation

Whatever the cube root of is, it has to be such that the cube of its

modulus is equal to the modulus of .

The argument of has to be such that when it is applied three times, it

reaches the same value as the argument of or a complete rotation

from that argument.

The cube roots of are , and .

This final value is not the principal argument. The value

should be used.

The roots are arranged in an equilateral triangle. This should not be a surprise

as the same process has been followed as for finding the cube roots of unity.

The nth roots of any non-zero complex number will form the vertices of a

regular n-gon centred at the origin.

Using the nth roots of unity to find the nth roots of any

complex number

The nth root of any complex number can be found using the nth roots of unity.

Compare the cube roots of unity to the cube roots of . Each root has been

enlarged by a scale factor of 2 and rotated an angle of anticlockwise (i.e.

clockwise).

So to find all of the cube roots of , the cube roots of unity can all be

multiplied by the complex number . This is the first cube root of

that was found.

If you let then the cube roots of are and .

In general, the nth roots of any complex number are

The sum of the nth roots of any complex number is

Example 7

Question

Find all the roots of the equation . Give your answers in the form

where and θ are real numbers.

Solution

The roots are and where .

The roots are

Some of these need to be rewritten in terms of principal arguments.

Example 8

Question

Solve the equation

Solution

gives so is a factor.

There is often an easy factor to spot and in this case satisfies the

equation.

Equating coefficients

This will use the square roots of unity 1 and ω where .

The roots of are

, , , and where
The roots are where .
Exercise 1

1 By considering the complex numbers and simplify

the following expressions by removing the brackets. Give your answers in

terms of z1 and z2.

(a)

(b)

(c)

(d)

2 Given that and where , and are non-

zero complex numbers, show that .

3 Given that is a root of , prove that

4 Show that if then and find all of the

solutions to

5 Solve the equation

6 Solve the equation

7 Given that and are complex numbers and that , find

8 Given that and are complex numbers and that , find .

9 Given that where show that where

is a positive real number to be found.

10 Prove that if and are complex numbers then

11 Express as a polynomial in .

12 The complex numbers and form the vertices of an equilateral

triangle on the Argand diagram. Show that

.
Exercise 2
STEP questions

1 If

and , find the possible values of u. Hence find the possible values

of z. (Do not try to simplify your answers.)

Show that, if z satisfies , then

Hence write the solutions of in the form for suitable

real r and θ.

Deduce that

Cambridge Assessment Admissions Testing STEP 2 1996 Question 5

Hints

−1
• What can you divide every term in (*) by to give the terms z and z as part

of the equation?

• What are the other terms in the equation when you do this? Can they be

related to ?

• Substitute your value(s) for u back into to find z. Can you see

why this can be rearranged to make a quadratic equation?

• The question states that you should not simplify your answers so the values

you find for z may not look that friendly.

• Compare the equations and . The first

clearly has as a solution whilst the second does not. has five

roots and has four roots. If z satisfies both then the

equations must have four identical roots with as the ‘extra’ root that is

only for . How would this help you to factorise ?

• The solutions to are the fifth roots of unity. How does this help

you write the solutions to in the form

 ?

• Where have you seen and or their equivalent

expressions before?

• is in the first quadrant of an Argand diagram. Which of your earlier

answers does it relate to?

2 The point P in the Argand diagram is represented by the the complex number

z, which satisfies

Here, r is a positive real number and .

By writing as , show that the locus of P is a circle, C,

the radius and the centre of which you should give.

(i) The point Q is represented by ω, and is related to . Let the

locus of Q be .

Show that is also a circle, and give its radius and centre.

If C and are the same circle, show that

and that either a is real or a is imaginary. Give sketches to indicate the

position of C in these two cases.

(ii) Suppose instead that the point Q is represented by , where . If

the locus of Q is C, is it the case that either a is real or a is imaginary?

Cambridge Assessment Admissions Testing STEP 3 2019 Q6

Hints

• You are told what to do for the first part of this question: start with

.
• Expand the expression on the right-hand side of this using what you know

about conjugates of complex numbers. What expression do you get?

• How does it relate to ?

• It should give the locus of P in a form that you recognise is a circle and that

you can read the centre and radius from.

• For part (i), if you write z in terms of ω and substitute it into

what happens?

• Can you use what you know to write it in the same form as you used for the

locus of P?

• If two circles are identical they have the same centre and radius. Can you use

the centres and radii that you have found for C and to show that

• You are asked to give two sketches, one for real a and one for imaginary a.

These will be circles. Where will you put the centre in each case?

• For part (ii) you are using rather than . How does this change

what you did for part (i). Do you still get a circle? If you do, what is its

centre and radius?

• What is the value of now?

• What does this mean for the value of a?

Full worked solutions can be found to all of the questions from Exercise 2 at

www.hoddereducation.co.uk/step-mat-tmua-answers
Chapter 16: Matrices

STEP
Introduction

The STEP 2 syllabus (2020) requires you to be able to add, subtract and multiply

conformable matrices and multiply a matrix by a scalar. You are expected to

understand and use zero and identity matrices. You need to know how matrices

are used to represent linear transformations in two- and three-dimensions. You

need to be able to find invariant points and lines for linear transformations. You

should be able to calculate determinants of 2 × 2 matrices and interpret these as

scale factors. You are expected to understand and use singular and non-singular

matrices and properties of inverse matrices and to calculate and use the inverse of

a non-singular 2 × 2 matrix.

The STEP 3 syllabus (2020) requires you to be able to calculate determinants of 3

× 3 matrices and interpret these as volume scale factors. You are expected to be

able to calculate and use the inverse of non-singular 3 × 3 matrices and to solve

three linear simultaneous equations in three variables by matrix methods. You

need to interpret geometrically the solution and failure of solution of three

simultaneous linear equations.

For this chapter, it will be assumed that you have an initial understanding of

matrices and their addition, subtraction and multiplication by a scalar and of

multiplication of more than one matrix. It will also be assumed that you have an

understanding of 2 × 2 matrices and how they are used to represent linear

transformations.
Invariant lines and lines of invariant points

An invariant point is a point which maps to itself under a transformation.

The transformation represented by the matrix is a reflection in the

line .

Under reflection, any point on the mirror line will map to itself. Any point on the

line will be an invariant point. The line can be described as

a line of invariant points.

Every matrix transformation will have at least one invariant point (the origin)

since

To find an invariant point or a line of invariant points, the mapping

can be used.

For the reflection matrix above

Both of these rearrange to showing that it is a line of invariant points.

An invariant line is not necessarily a line of invariant points and it is important to

draw a distinction between the two.

An invariant line is one in which any point on the line is mapped to another (or

the same) point on that line.

For a reflection, any line perpendicular to the mirror line is an invariant line.

To find an invariant line or a set of invariant lines, the mapping

can be used.

For reflection in

gives
Substituting the result from (A) into (B) gives

In this equation m and c are constants. The only variable is x.

If a line is an invariant line then every single point on that line will satisfy this

equation.

So setting x = 0 (or 1 or –1 or 2, etc.) will allow the values of m and c to be

found.

x = 0 gives

This is true if or if c = 0.

If , becomes 0x = 0.

If c = 0 then equation (*) becomes .

Either x = 0 or .
You should realise that one of these is the mirror line and the other is

perpendicular to it.

The two invariant lines are and .

Of these two, we know that is more precisely a line of invariant points.

Example 1

Question

Find the invariant lines or lines of invariant points for the transformation with

the matrix .

Solution

It is easier to find any lines of invariant points first.

This is true for any 2 × 2 matrix transformation. There is no line of

invariant points for this transformation. The next step is to try to find any
invariant lines.

Substituting (C) into (D)

There are two invariant lines y = x and .

You may well have been taught about eigenvalues and eigenvectors.

For the given matrix , the line y = x corresponds to the eigenvector

and the line to the eigenvector .

An alternative method for this question would be to find the eigenvalues and

from these the corresponding eigenvectors.

λ = −7 or λ = 2 (These are the eigenvalues.)

For λ = −7

For

Try it out

By either using or by finding the eigenvectors of A,

identify any invariant lines for the following matrix transformations.

Find any invariant points or lines of invariant points if they exist.

1

3
Determinants

One of the main applications of matrices is in the solution of systems of linear

equations. Any university linear algebra course will focus heavily on the use of

matrices.

Consider the system of simultaneous equations

The solution for x can be found by multiplying equation (A) by b2 and equation

(B) by b1 and then eliminating y.

This gives .

The solution for y can be found by multiplying equation (A) by a and equation
2

(B) by a1 and then eliminating x.

This gives which can be rewritten as .

You can see that is the denominator of both fractions. If

then the equations either have no solutions, for example,

or they have an infinite number of solutions, for example,

In the first example, the graphical representation of the equations is a pair of

parallel lines. In the second both equations define the same straight line.

The simultaneous equations

can be written in matrix form as

since the operation of matrix multiplication would result in the two equations.

The expression is the product of the cross-multiplication necessary to

bring about the elimination of x or y.

For the matrix , the determinant .

For a general square matrix, the determinant can be found by extending the cross-

product idea.

The system of equations

by elimination will result in each of x, y and z having the denominator

This can be rewritten as

In terms of matrices, the equations can be written as the matrix equation

The determinant of the matrix is

You can see that the determinant of a 3 × 3 matrix can be calculated using the

determinants of three 2 × 2 matrices.

For the 4 × 4 square matrix , the determinant is

You will not be expected to find the determinant of a 4 × 4 matrix in the STEP

examinations but from this you can see how the cross-product method of finding

a determinant generalises.

In the STEP examination you may be expected to find the determinant of a 2 × 2

or a 3 × 3 matrix. You need to be able to do this quickly and accurately.

Example 2

Question

Find the determinant of the matrix A where

Each value in the top row needs to have the appropriate 2 × 2 determinant

associated with it.

It is important to remember that the associated signs are +, −, +.

Solution

Try it out

Find the determinants of the following matrices.

1

3
Inverse matrices

You should know that the inverse of the 2 × 2 matrix A where is

You should also know a procedure for finding the inverse of a 3 × 3 matrix which

involves several stages. Modern calculators can give the inverse of a 3 × 3 matrix

without the need for intermediate calculation. In the STEP examinations you are

not allowed a calculator so it is important that you learn an efficient way of

finding the result quickly.

This is the standard procedure.

1 Find a matrix of cofactors.

2 Change the signs of the elements by where + means that an

element’s sign is maintained and − that it is changed.

3 Find the transpose of this new matrix. This is known as the adjugate of the

original matrix.

4 Multiply the adjugate by where ∆ is the determinant of the original matrix.

If you do not know what this means, you will need to look through a Further

Mathematics textbook to see the process described in detail.

The mathematical steps in this process cannot be avoided but it is possible to go

through it in a slightly more memorable way.

To find the inverse of the 3 × 3 matrix

1 Consider the three columns of the matrix as column vectors a , b and c.

2 Find the cross-product b × c and enter the three values as the first row of a new

3 × 3 matrix. Next, find the cross-product c × a (in that order, the vector cross-

product is not commutative) and enter that as the second row of your new

matrix. Find the cross-product a × b and enter that as the third row. You now

have the adjugate of the original matrix.

3 Multiply the adjugate by where ∆ is the determinant of the original matrix.

You do not save any calculation time at all by considering the calculation in this

way as finding each cross-product involves calculating three 2 × 2 determinants. It

does, however, present it in a more memorable way and highlights the connection

between the determinant of a 3 × 3 matrix and the cross-product of two vectors.

If you have found the adjugate of a 3 × 3 matrix and you are sure you have

completed all of the steps without making any errors then there is a fairly quick

way to find the determinant.

For the 3 × 3 matrix M, where ∆ is the determinant

of M.

Example 4

Question

Find the inverse of the matrix A where .

Solution
Start with b × c, i.e.

These need to be entered as the rows of the adjugate matrix.

To find the determinant, the first element in the top row of adj(A) A can be

used.

Try it out

Find the inverse of the following matrices.

For question 3, which values of k would result in C being a singular matrix?

Example 5

Question
Find the determinant of the matrix A where

Solution

There must be a trigonometric identity or two that can help here. sin2x +

sin2y and sin2x − sin2y are the most likely candidates to be replaced by

something more helpful.

Using the identities

gives

Using the identities

Example 6

Question

The matrices A and B are given by

−1 −1 −1 −1 −1
Find A , B , B A and AB, (AB) .

State the value of the determinant of AB.

Solution
 At this point you are expected to know that .

Finally, to find the determinant of AB it is enough to know that the

−1
determinant of A is 1 since A = adj(A) and the determinant of B is 1
−1
since B = adj(B). The determinant of AB is therefore 1.

−1
If you had tried to find (AB) from AB, the calculation is quite lengthy and has a

lot of opportunity for simple algebraic errors.

In any admissions test it is important to weigh up the effort required before

embarking on a lengthy process. You should ask yourself the question ‘Do they

really expect me to spend that much time on this?’. In this case, the question asks
−1 −1 −1 −1 −1
for A , B and B A before asking for (AB) . There was another efficiency

that could have been made in realising that matrix B was the transpose of matrix
 −1
A (albeit with p, q and r replacing a, b and c). The inverse B could have been
−1
written directly by considering A .
Solutions to systems of equations

The simultaneous equations

Can be solved by using the matrix equation

In this case the solution is

In two dimensions, the solution to this system of equations can be interpreted as

the point of intersection of two straight lines. If then the two lines

intersect and there is a unique solution to the equations. If , i.e. A is

singular, then the two lines are coincident, that is they have all of their points in

common, or they are distinct parallel lines. If the lines are coincident then there

are an infinite number of solutions and they can be described using a single

parameter (the parametric equation of a straight line). If the two lines are distinct

parallel lines then the equations are described as being inconsistent.

The simultaneous equations

Can be solved by using the matrix equation

In this case the solution is

In three dimensions, the solution to this system of equations can be interpreted as

the intersection of three planes.

If then the three planes have a single common point and there is a unique

solution to the equations. The equations would be described as consistent and

independent.

If , i.e. A is singular, then the equations either have infinitely many

solutions or are inconsistent and have no solutions. If there are infinitely many

solutions then the equations are described as consistent and dependent.

Infinitely many solutions

Unlike the example in two dimensions, there is more than one arrangement of

planes that gives rise to infinitely many solutions. The most obvious one is three

planes that coincide.

If this is the case then the best that can be done is to describe the solutions as the

equation of one of the planes. This is often done using two parameters.

The other case with infinitely many solutions is the known as a sheaf. In this

case, all of the planes coincide along one straight line.

In this case the solutions are given as the equation of the straight line along which

all three planes meet. This is often done using one parameter.
No solutions

There are four possible arrangements of the three planes that result in no

solutions.

The three planes are parallel and distinct.

Two of the planes are coincident and the third is parallel but distinct.

Two parallel and distinct planes are crossed by the third plane.
The three planes form a triangular prism.

Example 7

Question

For what values of k are the equations

(a) inconsistent?

(b) consistent but dependent?

Interpret the situation geometrically in each case.

Solution

In both cases, the determinant of the 3 × 3 matrix of the coefficients of x, y

and z will be 0 as there is no unique solution. The determinant calculation

can be simplified by adding row 3 to row 1.

Let z = t

This seems to indicate that one value of k will result in the equations being

inconsistent and the other will result in the equations being consistent but

dependent. The two can be distinguished by writing the solution of each

pair of equations using a parameter. This will be the line of intersection of

the pair of planes in each case.

Intersection of planes P1 and P2

Substituting into P2

The parametric equations of the line of intersection of P1 and P2 are

Doing the same for P and P gives

2 3
For P1 and P3 it gives

Case 1: k = −1

Intersection of P1 and P2 x = 3 − 4t, y = 4t − 5 and z = t

It may seem like some effort to get the lines of intersection of the planes in

terms of k but, having these, the two values can be checked relatively

quickly.

Intersection of P and P x = −5 − 4t, y = 4t + 3 and z = t

2 3

Intersection of

(a) When the three lines are parallel and distinct so the equations are

inconsistent. The three planes form a triangular prism.

Case 2: k = 3

Intersection of

The other case, where k = 3, is likely to be consistent but dependent (as

this is implied by the question). If this is the case then either all three

planes are coincident or a sheaf is formed. As it is clear that, when k = 3,

the equations give three different planes, you can aim towards showing

that the planes are a sheaf in which case the parametric equations of the

three lines of intersection should all be the same line.

Intersection of

Intersection of
(b) When k = 3 the three planes form a sheaf so the equations are consistent but

dependent
Exercise 1

1 Find the invariant lines or lines of invariant points for the transformations

given by each matrix below.

(a)

(b)

(c)

(d)

2 Find the determinant of each of the following matrices.

(a)

(b)

(c)

(d)

(e)
 (f)

3 Find the inverse of each of the following matrices where it exists. (Do not use

a calculator.)

(a)

(b)

(c)

(d)

(e)

(f)

4 For what values of are the following systems of equations

(i) inconsistent?

(ii) not independent?

 (a)

(b)

5 Solve the following system of equations where p and q are constants.

Give a geometrical interpretation of the following cases.

(a)

(b)

6 Solve the equations

given that they have a unique solution and a, b and c are constants. Give your

answer in terms of a, b and c.

Exercise 2
A STEP question

The matrix A is given by

(i) You are given that the transformation represented by A has a line of

invariant points (so that each point on is transformed to itself). Let

be a point on .

Show that .

Show further that .

What can be said about if does not pass through the origin?

(ii) By considering the cases and separately, show that if

then the transformation represented by has a line of

invariant points. You should identify the line in the different cases that

arise.

(iii) You are given instead that the transformation represented by has an

invariant line (so that each point on is transformed to a point on )

and that does not pass through the origin. If has the form ,

show that .

Cambridge Assessment Admissions Testing STEP 3 2019 Question 3

Hints

This question seems, on the face of it, to be quite straightforward but in the

year it appeared it was one of the four least successful questions on the whole

paper for the candidates.

• What is meant by a line of invariant points?

• Is there a simple matrix equation you can form to find the line of invariant

points?

• When you appear to get to a stage where you can see the expression that you

want, what things do you need to make clear so that you aren’t making any

fundamental errors such as dividing by 0.

• How do you get to the two expressions that they are asking for from what

you have found?

You need to justify your answers fully for this question.

For part (ii) you are told to consider the cases b ≠ 0 and b = 0 so you should

do both separately. There is more than one thing that can happen when b ≠ 0.

Make sure you cover every possibility. Don’t lose sight of what you are dealing

with. Remember that the question is about lines of invariant points.

For part (iii) remember the difference between an invariant line and a line of

invariant points. You can do a similar calculation to the one you did for part (i)

but consider how it should be different from part (i).

A full worked solution can be found to the question from Exercise 2 at

www.hoddereducation.co.uk/step-mat-tmua-answers
Answers

For every question with an answer than can be simply expressed, the

answers can be found in this section. For more involved questions, please

refer to the full worked solutions on the website.

Section 1
Chapter 2

Try it out

1 See the full worked solutions.

2 8(π + 16)
Section 2
Chapter 3

1 6

2 9

3 2519

4 79, 97, 71, 17, 62, 26, 53, 35, 44, 88 (you have to decide if you are

allowing 80 and 08)

5 See the full worked solutions.

6 See the full worked solutions.

Try it out

1 2

2 30

3 18

Try it out

1 There aren’t any where both x and y are positive

2 (a) 999 856

(b) 253 009

Try it out

2 80 640

Exercise 1

1 Squares of integers only end in 0, 1, 4, 5, 6 and 9

2 See the full worked solutions.

3 See the full worked solutions.

4 See the full worked solutions.

5 See the full worked solutions.

6 528

7 1225

8 20 250

9 85

10 The same

11 160

12 145

13 e.g. 64 and 729

14 See the full worked solutions.

15 54

16 11, 60, 61

17 588

18 0 < x < 1

19 2

10 7
20 2 × 3

21

Exercise 2

1 B

2 E

3 L

4 G

5 B

6 B

7 A

8 G

9 C
10 D

11 d

12 c

13 c

14 a

15 c

16 e

17 b

18 a

19 b

20 d
Chapter 4

Try it out

Starting with and completing the square for the

x and y terms gives . The radius of the

circle is .

For the circle to exist which is equivalent to

When , the radius is 0 so the ‘circle’ is actually a point.

Exercise 1

1 8x − 6y + 5 = 0

2 (i) (0 the circles intersect)

(ii)

3 36 square units

4 See the full worked solutions.

5
6

7 1 : 7

Exercise 2

1 D

2 A

3 C

4 D

5 C

6 F

7 B

8 D

9 c

10 b

11 a

12 e

13 d

14 b

15 d

16 b
Chapter 5

Two questions to think about

1 400

Try it out

Exercise 1

1 b

2 b

3 d

4 81

5 b

6 e

7 −1

8 b

9 c

10 b

Try it out

It is not true for any function.

It is true for log10 x as it is an increasing function for all x > 0

A simple function for which would be

g (x) = 10 − x
Exercise 2

5 See the full worked solutions.

6 c

7 a

8 1

9 b

10 a

Exercise 3

1 E

2 C

3 H

4 C

5 D

6 D

7 A

8 E

9 D

10 G

11 D

12 B
13 d

14 b

15 d

16 d

17 d

18 a

19 b

20 d

21 a

22 a

23 c

24 c

25 e

26 b

27 d

28 c

29 c
Chapter 6

Try it out: For the MAT and TMUA

Try it out: For the MAT only

Try it out

1 3 roots

2 1 root

3 2 roots

Exercise 1

1 (a) 1

(b) 3

(c) 1
 (d) 3

(e) 4

(f) 2

2 (a) 0

(b) 1

(c) 2

(d) 2

(e) 0

(f) 3

3 −2

4 1

5 d

6 f

7 1 if , 2 if k = 4 or −4, 3 otherwise

Try it out

17.5 square units

Try it out

square units

Try it out

1 0

2 6

3 4.5

5 6
Exercise 2

1 (a)

(b)

(c)

(d)

(e)

2 4.4

3 (a) underestimate

(b) overestimate

(c) cannot tell

Exercise 3

1 negative

2 35, 36

3 −3,1

4 C

6 2

Exercise 4

1 A

2 A

3 D
4 A

5 D

6 c

7 b

8 c

9 d

10 (a)

(i) 15

(ii) 5

(b)
Chapter 7

Two questions to think about

1 E

Asymptotes: x = – 4 – 3 , x = – 4 + 3

Intercepts: (– 4 – , 0), (– 4 + , 0)

Try it out

(a) The sum/difference of two even functions is even.

(b) The sum/difference of two odd functions is odd.

(c) The sum/difference of an even and odd function is neither even nor

odd, unless one of the functions is equal to zero.

(d) The product/quotient of two even functions is even.

(e) The product/quotient of two odd functions is even.

(f) The product of an even function and an odd function is an odd

function.

Try it out

(a) Reflection symmetry in , periodic with

period
(b) Reflection symmetry in x = 0

(c) Reflection symmetry in y = x

(d) Reflection symmetry in x = –1

(e) Reflection symmetry in x = 0

(f) Rotation symmetry about (0, 0)

Try it out

Try it out

The graph approaches y = x − 1 from above for both large positive and large

negative x.

As the graph approaches x = 0 from the left, y approaches

As the graph approaches x = 0 from the right, y approaches

As the graph approaches x = 1 from the left, y approaches

As the graph approaches x = 1 from the right, y approaches

Try it out
Try it out

(f) A translation parallel to the x axis followed by a stretch parallel to the y

axis.

Try it out

Exercise 1

1 (a) : Neither odd nor even. Reflection symmetry in x = 3

(b) : Odd. Rotation symmetry about (0, 0)

 (c) : Neither odd nor even. Reflection symmetry

in x = 3

(d) y = sin(cos x): Even. Periodic with period 2π. Reflection symmetry

in . Rotation symmetry about

(e) y = cos(sin x): Even. Periodic with period π. Reflection symmetry in

. Rotation symmetry about

(f) : Odd. Rotation symmetry about (0, 0)

(g) : Reflection symmetry in y = x

(h) : Neither even nor odd. Rotation symmetry

about (1, 2)

(i) : Even. Reflection symmetry in x = 0

(j) : Odd. Periodic with period π. Reflection symmetry

in . Rotation symmetry about

2 See the full worked solutions.

3 (a) : Exists for

(b) : Exists for

(c) : Exists for

(d) : Exists for

 (e) : Exists for

4 See the full worked solutions.

5 See the full worked solutions.

6 (a)

(b)

(c)
 (d)

(e)

7 Note: there is more than one possibility in each case. You may wish

to check using a graph drawing app if you have a different answer.

(a) Reflection in the x axis followed by translation

 (b) Reflection in y = x followed by translation

(c) A stretch parallel to the y axis with scale factor 2 followed by

translation

(d) A translation of followed by a stretch parallel to the y axis with

scale factor 4

(e) A translation of followed by a stretch parallel to the y axis with

scale factor 3

8 (a) Max at

(b) Max (−1, 9), Min (2, −18)

(c) Max (0, 3), Min (−2, −13), Min (2, −13)

(d) Max

9 (a)
(b)

Intercepts: and (– 3, 0)

Asymptote: x = – 2
(c)

Intercept: (1, 0)

Asymptote: x = 0

(d)
 Intercepts: (– 2, 0), (1, 0) and (0, – 2)

Minimum:

(e)

Intercept:

Minimum:

Asymptotes: x = – 1, x = 3, y = 0

10 x = 0

11 x = 0

12 No real solutions.

13 x = 90

14 1

Exercise 2
1 F

2 B

3 F

4 C

5 E

6 C

7 C

8 C

9 b

10 e

11 b

12 d

13 e

14 a

15 b
Chapter 8

A question to think about

Try it out

for

Try it out

2
By the identity if r ≠ −1, a − ar + ar −

Try it out

1 a) an = 2, b) , c)

2 a) n,

Try it out

See full worked solutions.

Exercise 1

1 10 000

2 See the full worked solutions.

3 Period 5

4 See the full worked solutions.

6
7

10 See full worked solutions.

Exercise 2

1 B

2 E

3 D

4 F

5 A

6 d

7 b

8 a

9 d

10 d
Chapter 9

Try it out

5 1

Try it out

1 2

2 4

3 0

4 4

5 2

6 0

7 0

Exercise 1

1
2 b

3 2

4 8

6 30° ≤ x ≤ 45° or 90° < x ≤ 150°

7 60° < x < 90° or 270° < x < 300°

10

Exercise 2

1 B

2 B

3 E

4 D

5 D

6 c

7 c

8 d

9 c

10 b
Chapter 10

Try it out

n = 9

Try it out

See full worked solutions.

Try it out

1. True

2. False

3. False

4. True

5. False

6. False

7. True

8. False

Exercise 1

1 2

2 (i) If it is not moving, paint it. If it is moving salute it.

(ii) If it is not moving, paint it again. If it is moving, salute it.

3 Yes

4 Nothing without further information

5 See the full worked solutions.

Exercise 2

See the full worked solutions.

Exercise 3

1 D

2 C

3 E

4 C

5 B

6 G

7 E

8 G
Section 3
Chapter 11

Try it out

See the full worked solutions.

A question to think about

x = , y = or x = , y =

or x = , y = or x = , y =

Try it out

(x − y)(x + z)(y + z)

Try it out

a = 2, b = 3, c = 1, d = 3, e = −2

Try it out page 131

2 2 2
(x – y + z)(x + y – z)(x + y +z – 2yz)

Try it out

(a) p = 3

(b) p = 7

n
(c) a = 2, p = 2 − 1

Try it out

3 2
x + (a + b + c)x + (ab + bc + ac)x + abc

(x + 2p)(x + q)(x + 3)
Exercise 1

1 (a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

2 (a)

(b)

(c)

(d)

3 (a)

(b)

(c)

(d)

(e)

(f)

Exercise 2

1 (i)
 (ii) See the full worked solutions.

(iii) See the full worked solutions.

(iv) p = −9, q = −27 or p = −10, q = −32 or p = −11, q = −36

(v) 13

2 (i) See the full worked solutions.

(ii) See the full worked solutions.

(iii) See the full worked solutions.

(iv)

(v) x = −2

3 (i) See the full worked solutions.

(ii) See the full worked solutions.

(iii)

(iv) x = p or x = −2p

4 See the full worked solutions.

Chapter 12

Try it out

1 x = 1, y = 2

2 x = −3, y = −4; x = −6, y = −2

3 (a, b) = (1, 1), (1, 2), (1, 4), (2, 2), (1, 10), (2, 8), (3, 6), (4, 4), (5, 2)

4 (a, b, c) = (1, 5, 2), (6, 4, 3), (8, 2, 1)

5 a = 2, b = 3, c = 4, p = 4, q = 2, r = 1

Try it out

x = 6, y = 2

Try it out

See full worked solutions.

Exercise 1

1 (a) irrational

(b) either

(c) irrational

(d) either

(e) either

2 E

3–9 See the full worked solutions.

Exercise 2

n
2 26, 2 − n − 1
3 27, 115

4 36, 9

5 (a) 2520

(b) 360

Exercise 3

See the full worked solutions.

Chapter 13

Try it out

See the full worked solutions.

Exercise 1

2–5 See the full worked solutions.

Exercise 2

1 See the full worked solutions.

2 . See the full worked solutions for more.

3 See the full worked solutions.

4 (i)

(ii)

(iii)

Exercise 3

1 (a)

(b) See the full worked solutions.

2 (i)

(ii)

(iii)
Chapter 14

Try it out

See the full worked solutions.

Exercise 1

2 (a)

(b)

4 (a) 0

(b)

Try it out

See the full worked solutions.

Exercise 2

See the full worked solutions.

Exercise 3
1 (i)

(ii)

(iii)

2 (i)

(ii)

(iii)
Chapter 15

Try it out

Exercise 1

1 (a)

(b)

(c)

(d)

2 See the full worked solutions.

3 See the full worked solutions.

5–12 See the full worked solutions.

Exercise 2

2 See the full worked solutions.

Chapter 16

Try it out

Try it out

1 12

2 (c − b)(a − b)(a − c)

3 −2

Try it out

C is singular if k= –2 or k = 8

Exercise 1
1 (a) y = x is a line of invariant points, y = – 3x is an invariant line

(b) y = − and y = 3x are invariant lines

(c) y = x and y = −2x are invariant lines

(d) y = is a line of invariant points. All points on the line y = −

map to the origin.

2 (a) −98

(b) 352

3 2
(c) −x + 17x − 43x + 27

(d) 0

(e) 8(2 − 3x)

2 2 2 2
(f) x(x + y)(x − y) (x + y (x + xy +1))

3 (a)

(b) Does not exist

(c)

(d)

(e)
 (f)

4 (a) (i) consistent for all k

(ii) independent for all k

(b) (i) consistent for all k consistent for all k

(ii) independent for all k

5 See the full worked solutions.

6 See the full worked solutions.

Exercise 2

See the full worked solutions.