Mathematical Statistics for Economics

Lecturer: Dr Ioannis (Yiannis) Karavias

University of Birmingham

October 11, 2018

Summary of today’s lecture

1. What is Statistics?
2. Set Theory
What is Statistics?

I Webster’s New Collegiate Dictionary: "A branch of

mathematics dealing with the collection, analysis,
interpretation and presentation of masses of numerical data".
I "The science of of basing inferences on observed data and the
entire problem of making decisions in the face of uncertainty"
(Freund and Walpole (1987)
I Statistics is a theory of information, with inference making as
its objective (Wackerly et al.)
What is Statistics?

I Population: The large body of data that is the target of our

interest.
I Sample: A small subset of the population which is available to
us.
I Billions are spent each year in gathering data: government,
industry, academia.
I Large sample means more information. More information
means results of higher quality
What is Statistics?

The objective of statistics is a) to make inference about a

population based on information contained in a sample from that
population and b) to provide an associated measure of goodness
for the inference.
I The European Trasport Safety Council wants to estimate the
proportion of car tires which are faulty coming out of the
Michelin factory in Dundee during the current production year.
I Apple wishes to determine whether Iphone 4’s average length
of life is greater than 4 years.
I A political scientist wants to determine whether May will win
the next elections.
Graphical Methods
I An individual population or sample can be characterized by a
relative frequency distribution which can be represented by a
relative frequency histogram.
Graphical Methods
Graphical Methods
25 families: number of quarts (1quart=1.66 pints) of milk purchased in
the previous week:

I What is the number of pints that purchased by the largest

proportion of families?
I What proportion of the 25 families purchased more than 2 quarts of
milk?
I What proportion purchased more than 0 but fewer than 5 quarts.
Numerical Methods

I Statistics can be separated in two categories: Descriptive

statistics and Inferential statistics. Descriptive statistics are
good for some purposes but not for most.
I I.e. many histograms can be made with the same data.
I We prefer rigorously de…ned quantities for summarizing
information.
I Numerical descriptive measures:
I Measures of Central Tendency: Mean, Mode, Median
I Measures of Dispersion or Variation: Variance, Standard
Deviation, Range
Numerical Methods

De…nition
The mean of a sample of n measured responses y1 , y2 , ..., yn is
given by
1 n
ȳ = ∑ yi .
n i =1
The corresponding population mean is denoted µ.

De…nition
The variance of a sample of measurements y1 , y2 , ..., yn is
n
1
1∑
s2 = (yi ȳ )2 .
n i =1

The corresponding population variance is denoted by the symbol

σ2 .
Numerical Methods
Probability and Statistics

I Probability: It is when you know the structure of the

population of interest: Coin toss
I Statistics: we use probability for the reverse: Is a coin fair?
Theory and Reality

I Theories are conjectures proposed to explain phenomena in

the real world. As such, they are approximations to reality.
I These models are explanations of reality based on some
assumptions.
I If the mathematical model is a good approximation, then we
can arrive to correct conclusions for the process itself.
I This course is concerned with the theory of statistics and
hence with models of reality.
To summarize:

I The objective of statistics is a) to make inference about a

poulation based on information contained in a sample from
that population and b) to provide an associated measure of
goodness for the inference.
I Population and Sample
I Relative frequency distribution which can be represented by a
relative frequency histogram.
I Numerical descriptive measures
I This course is concerned with the theory of statistics and
hence with models of reality.
To summarize:

1. We talked about population and sample but weren’t very

speci…c about it.
2. These notions are described by set theory.
3. In this lecture we will review some basic elements of set theory.
Set Theory

I We use capital letters A, B, C ... to denote sets of points.

I If the elements in the set A are a1 , a2 and a3 we will write
aj 2 A and
A = f a1 , a2 , a3 g .
I Let S denote the set of all elements into consideration. S is
the universal set.
I For any two sets A, B, we will say that A is a subset of B
(A B )if every point in A is also in B.
I The null or empty set is denoted as ∅. Thus ∅ is subset of
every set.
I The union of A and B is the set of all points in A, in B, or
both, denoted by A [ B.
Set Theory

I The intersection of A and B, denoted by A \ B (or AB) is

the set of all points in both A and B.
I If A is a subset in S, then the complement of A, denoted as A
is the set of points that are in S and not in A
I Two sets are disjoint or mutually exclusive if A \ B = ∅.
I Some basic laws:
I A \ (B [ C ) = (A \ B ) [ (A \ C )
I A [ (B \ C ) = (A [ B ) \ (A [ C )
I (A \ B ) = A [ B
I (A [ B ) = A \ B
Set Theory
Set Theory
Set Theory
Set Theory
Set Theory
Set Theory

I Suppose a family contains two children of di¤erent ages, and

we are interested in the gender of these children. Let F
denote that a child is female and M that the child is male and
let a pair such as FM denote that the older child is female
and the younger is male. There are four points in the set S of
possible observations:

S = fFF , FM, MF , MM g

Let A denote the subset of possibilities containing no males;

B the subset of possibilities containing two males and C the
subset containing at least a male. List the elements of
A, B, C , A \ B, A [ B, A \ C , A [ C , B \ C , B [ C and C \ B
Set Theory

I Suppose A = fa, b, c g and B = fb, c, d g. List all the

elements of A [ B.
Set Theory

I Suppose A = fa, b, c g and B = fb, c, d g. List all the

elements of A [ B.
I From the law (A [ B ) = A \ B we have

A[B = A\B
= fb, c g
A Probabilistic Model for an Experiment: The Discrete
Case

I We will use the term experiment to include observations

obtained from completely uncontrollable situations (such as
observations on the daily price of a particular stock) as well as
those made under controlled laboratory conditions.

De…nition
An experiment is the process by which an observation is made.

De…nition
When an experiment is performed, it can result in one or more
outcomes, which are called events.
A Probabilistic Model for an Experiment: The Discrete
Case

I Some events associated with a single toss of a balanced die

are these:
I A: Observe an odd number. B: Observe a number less than 5.
C : Observe a 2 or a 3.
I E1 : Observe a 1. E2 : Observe a 2........E6 : Observe a 6.
I Events A,B, C which can be decomposed are called compound
events. In contrast, the events E are called simple events.
A Probabilistic Model for an Experiment: The Discrete
Case

I We shall use set theory to describe the outcomes of an

experiment.

De…nition
A simple event is an event that cannot be decomposed. Each
simple event corresponds to one and only one sample point. The
letter E with a subscript will be used to denote a simple event or
the corresponding sample point.

De…nition
The sample space associated with an experiment is the set
consisting of all possible sample points. A sample space will be
denoted by S.
A Probabilistic Model for an Experiment: The Discrete
Case

De…nition
A discrete sample space is one that contains either a …nite or a
countable number of distinct sample points.

De…nition
An event in a discrete sample space S is a collection of sample
points— that is, any subset of S.
A Probabilistic Model for an Experiment: The Discrete
Case

In the die tossing example: The experiment is the tossing of the

die. The sample space S consists of 6 sample points.
A Probabilistic Model for an Experiment: The Discrete
Case
I A probabilistic model for an experiment with a discrete sample
space can be constructed by assigning a numerical probability
to each simple event in the sample space S.
De…nition
Suppose S is a sample space associated with an experiment. To
every event A in S (A is a subset of S), we assign a number, P (A),
called the probability of A, so that the following axioms hold:
Axiom 1: P (A) 0.
Axiom 2: P (S ) = 1.
Axiom 3: If A1, A2, A3, ... form a sequence of pairwise mutually
exclusive events in S (that is, Ai \ Aj = ∅ if i 6= j),then
∞
P (A1 [ A2 [ A3 [ ...) = ∑ P ( Ai )
i =1
A Probabilistic Model for an Experiment: The Discrete
Case

I For discrete sample spaces, it su¢ ces to assign probabilities to

each simple event.
I For the die tossing:
I P (Ei ) = 1/6 for i = 1, ..., 6, by Axiom 1.
I P (S ) = P (E1 ) + ... + P (E6 ) = 1
I P (A) = P (E1 [ E3 [ E5 ) = P (E1 ) + P (E3 ) + P (E5 ) = 3/6,
by axiom 3.
Summary

I In this lecture we reviewed set theory

I We showed how a real life experiment can be described
mathematically.
I Now we can use this knowledge to compute probabilities of
more complicated events.