CHAPTER 2 (PART 2)

CONFIDENCE INTERVAL

SLIDE | 1
 CONTENTS
2.1 Estimate, Estimation, Estimator
2.1.1 Definition
2.1.2 Properties of Good Point Estimator
2.1.3 Confidence Interval

2.2 Confidence Interval For Population Mean

2.3 Confidence Interval For Difference Between Two Population Means

2.3.1 Independent Samples
2.3.2 Dependent Samples

SLIDE | 2
 CONTENTS

2.4 Confidence Interval For Population Proportion

2.5 Confidence Interval For Difference Between Two Population Proportions

2.6 Confidence Interval For Population Variance And Population Standard Deviation

2.7 Confidence Interval For Ratio Of Two Population Variances

SLIDE | 3
 OVERVIEWS
CONFIDENCE INTERVAL COVERED IN CHAPTER 2

POPULATION

ONE TWO

MEAN PROPORTION VARIANCE MEAN PROPORTION VARIANCE

(SUB-CHAPTER 2.2) (SUB-CHAPTER 2.4) (SUB-CHAPTER 2.6) (SUB-CHAPTER 2.3) (SUB-CHAPTER 2.5) (SUB-CHAPTER 2.7)

SLIDE | 4
 2.3

CONFIDENCE INTERVAL FOR

DIFFERENCE BETWEEN TWO
POPULATION MEANS

SLIDE | 5
 LEARNING OUTCOMES

Estimate the confidence interval for the difference between two population means of
independent samples when the population variances are known or unknown.

Estimate the difference between two population means of dependent samples when the
population variance of the differences is known or unknown.

SLIDE | 6
 ONE-SIDED CONFIDENCE INTERVAL
CONFIDENCE INTERVAL FOR DIFFERENCE BETWEEN TWO POPULATION MEANS

LOWER ONE-SIDED UPPER ONE-SIDED

• SCENARIO 1: POPULATION VARIANCE IS KNOWN • SCENARIO 1: POPULATION VARIANCE IS KNOWN

• SCENARIO 2: POPULATION VARIANCE IS UNKNOWN AND

EQUAL, AND SAMPLE SIZE • SCENARIO 2: POPULATION VARIANCE IS UNKNOWN AND
EQUAL, AND SAMPLE SIZE

IMPORTANT NOTE
Since there is some measurement unit that is not allowed for negative values, therefore you are advised to use zero instead of for the
100% upper one-sided confidence interval.

SLIDE | 7
 EXERCISE 2.3.1 (Page 99)
QUESTION 1

Find the 95% confidence interval for the difference of population means of children’s sleep time and adult’s sleep
time if given that the variance for the children’s sleep time is 0.81 while for the adults is 0.25. The mean sample
sleep time for 30 children is 10 hours while for 40 adults is 7 hours. Assume that the children’s and adults’ sleep
times are normally distributed. Give a comment on the parameter estimate.

SLIDE | 8
 EXERCISE 2.3.1 (Page 99)
SOLUTION

Based on the question, we know that ,

Information Children Adults

hours hours

hours
Interpretation: We are 95% confident that the difference of population means of children’s sleep time and
adult’s sleep time lies within 2.6426 hours and 3.3574 hours.

SLIDE | 9
 EXERCISE 2.3.1 (Page 99)
QUESTION 2 (MODIFIED)

The burning rates of two different solid-fuel propellants used in aircrew escape systems are being studied. It is
known that both propellants are normally distributed and have the same population variances. Two random
samples of 26 specimens for each propellent are tested. The sample means and standard deviation of burning
rates for both propellants are summarised in the table below.

Type of Propellent Mean (cm/s) Standard deviation (cm/s)

Propellant 1 18.312 3.110

Propellant 2 24.412 2.708

Construct a 90% confidence interval for the difference in population means for propellent 1 and propellent 2.
Interpret the resulting answer.

SLIDE | 10
 EXERCISE 2.3.1 (Page 99)
SOLUTION

Based on the question, we know that

Information ,
Propellent 1 Propellent 2

cm/s 24.412 cm/s

cm/s cm/s

cm/s

Interpretation: We are 90% confident that the difference in population means for the propellent 1 and propellent
2 lies within -7.4553 cm/s and -4.7447 cm/s.

SLIDE | 11
 EXERCISE 2.3.1 (Page 99)
QUESTION 3

The mean of sleep time for 50 IPTS students is 7 hours with a standard deviation of 1 hour. The mean sleep time
for 60 IPTA students is 6 hours with a standard deviation of 0.7 hours. Assume that the sleep time for the IPTS
and IPTA students are normally distributed, find the 99% confidence interval for the different population means of
sleep time between the IPTS and IPTA students. Interpret the resulting answer.

a) Assume the population variances are the same.

b) Assume the population variances are different.

SLIDE | 12
 EXERCISE 2.3.1 (Page 99)
SOLUTION (a)

Based on the question, we know thatIPTS

Information , IPTA

hours 6 hours

1 hour 0.7 hour

hours

Interpretation: We are 99% confident that the difference in population means of sleep time between the IPTS
and IPTA students lies within 0.5811 hours and 1.4189 hours.

SLIDE | 13
 EXERCISE 2.3.1 (Page 99)
SOLUTION (b)

Based on the question, we know that ,

Information IPTS IPTA

hours 6 hours

1 hour 0.7 hour

hours

Interpretation: We are 99% confident that the difference in population means of sleep time between the IPTS
and IPTA students lies within 0.5677 hours and 1.4323 hours.

SLIDE | 14
 EXERCISE 2.3.1 (Page 100)
QUESTION 4

Two groups of students are given a problem-solving test, and the test marks are compared.

Mathematics majors Computer Science majors

Find the 98% confidence interval for the difference in population means of test marks between the two groups of
students. Assume the population variance test marks are not the same for both groups. Give a comment on the
parameter estimate.

SLIDE | 15
 EXERCISE 2.3.1 (Page 100)
SOLUTION

Based onInformation
the question, we knowMathematics
that , majors Computer Science majors

79.2

Interpretation: We are 98% confident that the difference in population means of test marks between the two
groups of students lies within 1.5892 and 7.2108.

SLIDE | 16
 ADDITIONAL EXERCISE
QUESTION

Let and respectively represent the blood volumes in millimetres for a male who is a paraplegic and participates
in vigorous physical activities and a male who is able-bodied and participates in normal activities. Assume that is
and is By using the following seven observations of and ten observations of , find a 97% one-sided lower
confidence interval for with assuming that the population variances for both samples are not equal.

Observation 1 2 3 4 5 6 7 8 9 10
1612 1352 1456 1222 1560 1456 1924
1082 1300 1092 1040 910 1248 1092 1040 1092 1288

SLIDE | 17
 ADDITIONAL EXERCISE
QUESTION

Based onInformation
the question, we know that ,

mm

Interpretation: We are 97% confident that the minimum difference population means of blood volume between
the two groups of males greater than 189.9959 mm.

SLIDE | 18
 EXERCISE 2.3.2 (Page 102)
QUESTION 1

Adding computerised medical images to a database promises to provide great resources for physicians.
However, there are other methods of obtaining such information, so the issue of efficiency of access needs to be
investigated. An experiment was conducted in which eight computer-proficient medical professionals were timed
(in seconds) both while retrieving an image from a library of slides and while retrieving the same images from a
computer database with a Web front end and the data are given as follows.

Subject 1 2 3 4 5 6 7 8
Slide 30 35 40 25 20 30 35 62
Digital 25 16 15 15 10 20 17 16

Assume that the time retrieving the image is normally distributed, find the 95% confidence interval for the true
mean difference (keywords) between slide retrieval time and digital retrieval time. Interpret your result.

SLIDE | 19
 EXERCISE 2.3.2 (Page 102)
SOLUTION

Subject 1 2 3 4 5 6 7 8
Slide 30 35 40 25 20 30 35 62
Digital 25 16 15 15 10 20 17 16
Slide-Digital 5 19 25 10 10 10 18 46

Based on the question, we know that ,

seconds

Interpretation: We are 95% confident that the true mean difference between slide retrieval time and digital
retrieval time lies within 6.9524 seconds and 28.7976 seconds.

SLIDE | 20
 EXERCISE 2.3.2 (Page 102)
QUESTION 2

A sample of 38 diesel lorries was run for both hot and cold engines. A study is conducted to estimate the
difference in fuel economy. The mean difference (keywords) of fuel mileage between hot and cold engines is
0.250 litre per km and the sample variance of the differences is 0.013 litre per km. Assume that the fuel mileage
is normally distributed, find the 98% confidence interval for the population mean difference (keywords) of fuel
mileage between hot and cold engines. Give a comment on the parameter estimates.

SLIDE | 21
 EXERCISE 2.3.2 (Page 102)
SOLUTION

Based on the question, we know that litre per km; litre per km; ;

litre per km

Interpretation: We are 98% confident that the population mean difference of fuel mileage between hot and cold
engines lies within 0.2070 litre per km and 0.2930 litre per km.

SLIDE | 22
 ADDITIONAL EXERCISE
QUESTION

Twenty-four 9th- and 10th-grade high school girls were put on an ultraheavy rope jumping program. The following
data give the time difference for each girl-“before program time” minus “after program time”-for the 40-yard dash.

0.28 0.01 0.13 0.33 -0.33 0.07 -0.18 -0.14

-0.33 0.01 0.22 0.29 -0.08 0.23 0.08 0.04

-0.30 -0.08 0.09 0.70 0.33 -0.34 0.50 0.06

Find a one-sided 95% confidence interval which is a lower bound of true mean difference and interpret your
results.

SLIDE | 23
 ADDITIONAL EXERCISE
SOLUTION

Based on the question, we know that ; ; ;

Interpretation: We are 95% confident that the minimum population time mean difference before and after the
program is greater than -0.0273.

SLIDE | 24
 2.4

CONFIDENCE INTERVAL FOR

POPULATION PROPORTION

SLIDE | 25
 LEARNING OUTCOMES

Estimate the confidence interval for a population proportion.

Estimate the sample size by using the concept of confidence interval for population proportion.

SLIDE | 26
 DETERMINATION OF THE SAMPLE SIZE
MARGIN OF ERROR

Confidence for One Population Proportion

Proof:
In general, we know that , therefore
You all may compare with the following equation (Slide 10-Part 1)

STATISTICAL TABLE
(Page 8)

SLIDE | 27
ONE-SIDED CONFIDENCE INTERVAL
CONFIDENCE INTERVAL FOR ONE POPULATION PROPORTIONS

LOWER ONE-SIDED UPPER ONE-SIDED

SLIDE | 28
 EXERCISE 2.4 (Page 106)
QUESTION 1

The cure rate for the standard treatment of a disease is 45% . Dr Amani has introduced her new treatment which
she claims is much better. She tested the new treatment on 50 patients with the disease and cured 30 of them.

a) Find the sampling distribution of the cure rate of disease (This content has been removed from your syllabus,
you all merely assumed as the additional knowledge).
b) Construct a 99% confidence interval on the cure rate of disease for the new treatment. Give a comment on
the resulted confidence interval.
c) If no estimation of the sample proportion is available, how large is the samples of patient would be if we are
90% confident that the margin of error is 5%?

SLIDE | 29
 EXERCISE 2.4 (Page 106)
SOLUTION a)

Based on the question, we know that

Therefore, the sampling distribution of the sample proportion (cure rate of a disease) is @ .
Proof:

SLIDE | 30
 EXERCISE 2.4 (Page 106)
SOLUTION b)

Based on the question, we know that ; ; ;

Interpretation: We are 99% confident that the true cure rate of a disease for the new treatment lies within
0.4215 and 0.7785.

SLIDE | 31
 EXERCISE 2.4 (Page 106)
SOLUTION c)

Based on the question, we know that ; ;

patients (round-up)

SLIDE | 32
 EXERCISE 2.4 (Page 106)
QUESTION 2

In a random sample of 85 automobile engine crankshaft bearings, 12% have a surface finish that meets the
roughness specifications.

a) Construct a 98% confidence interval of bearings that do not meet the roughness specification (Beware of
the trick). Give an interpretation of the parameter estimate.
b) How large the sample size would be if we wish to be at least 95% confident that the error in estimating the
proportion of the bearings in the population exceeds the roughness specification (Beware of the trick: do
not meet the roughness specifications) is 0.02?

SLIDE | 33
 EXERCISE 2.4 (Page 106)
SOLUTION a)

Based on the question, we know that ; ; ;

Interpretation: We are 98% confident that the true rate of the bearings that do not meet the roughness
specification lies within 0.7980 and 0.9620.

SLIDE | 34
 EXERCISE 2.4 (Page 106)
SOLUTION b)

Based on the question, we know that ; ;

bearings

SLIDE | 35
 ADDITIONAL EXERCISE
QUESTION

A machine shop manufacturer’s toggle levers. A lever is flawed a standard nut cannot be screwed onto the
treads. Let equal the proportion of flawed toggle levers that the shop manufacturers. If there were 24 flawed
levers out of a sample of 642 that was selected randomly from the production line. Find a one-sided 95%
confidence interval for that provides an upper bound for .

SLIDE | 36
 ADDITIONAL EXERCISE
SOLUTION

Based on the question, we know that ; ; ;

Interpretation: We are 95% confident that the maximum true proportion of flawed toggle levers is less than
0.0497.

SLIDE | 37
 2.5

CONFIDENCE INTERVAL FOR

DIFFERENCE BETWEEN TWO
POPULATION PROPORTIONS

SLIDE | 38
LEARNING OUTCOME

Estimate the confidence interval for the difference between two population
proportions.

SLIDE | 39
 ONE-SIDED CONFIDENCE INTERVAL
CONFIDENCE INTERVAL FOR DIFFERENCE BETWEEN TWO POPULATION
PROPORTIONS

LOWER ONE-SIDED UPPER ONE-SIDED

Should be -1 due to the difference in the sample proportion can be a negative value.
(Please try the additional exercise)

SLIDE | 40
 EXERCISE 2.5 (Page 109)
QUESTION 1

An experiment was conducted to estimate the brace force for a compression web brace. In a sample of 380 short
test columns and 394 long test columns, there were found that 304 and 360 of them have overestimated forces,
respectively. Construct a 97% confidence interval for the difference in population proportions between the
overestimated forces for long and short columns. Interpret your answer.

SLIDE | 41
 EXERCISE 2.5 (Page 109)
SOLUTION

Information
Based on the question, we know that , Short Long

304 360

Interpretation: We are 97% confident that the difference in population proportions between the overestimated
forces for long and short columns lies within 0.0596 and 0.1678.

SLIDE | 42
 EXERCISE 2.5 (Page 109)
QUESTION 2

The specification for the pull strength of a wire that connects an integrated circuit to its frame is 10g or more. In a
sample of 85 units made with gold wire, 68 units met the specification, while in a sample of 120 units made
with aluminium wire, 105 units met specifications. Find a 98% confidence interval for the difference between
the population proportions of circuits made with aluminium and gold wires that not met the specification
(Beware of the trick). Give a comment on the parameter estimate.

SLIDE | 43
 EXERCISE 2.5 (Page 109)
SOLUTION

Information Gold Aluminium

Based on the question, we know that ,

Interpretation: We are 98% confident that the difference between population proportions of circuit made with
aluminium and gold wires that not met the specification lies within -0.1980 and 0.0480.

SLIDE | 44
 EXERCISE 2.5 (Page 109)
QUESTION 3

Two processes for manufacturing a certain microchip are being compared. A sample of 400 chips was selected
from a less expensive process, and 16% were found to be defective. A sample of 100 chips was selected from a
more expensive process, and 12% were found to be defective. Find a 90% confidence interval for the difference
between the population proportions of defective chips produced by the two processes. Explain your answer.

SLIDE | 45
 EXERCISE 2.5 (Page 109)
SOLUTION

Based on the question, we know that ,

Information Less Expensive More Expensive

Interpretation: We are 90% confident that the difference between the population proportions of mints defective
chips produced by the two processes lies within -0.0214 and 0.1014.

SLIDE | 46
 MIND EXPANDING EXERCISE (Page 122)
QUESTION ME.8 b) (MODIFIED)

A candy manufacturer selects mints at random from the production line and weighs them. For one week, the day
shift weighed 194 mints, and the night weighed 162 mints. The number of these mints that weighed at most 21
grams was 28 for the day shift and 11 for the night shift. Let and denote the proportions of mints that weigh at
most 21 grams for the day and night shifts, respectively. Find a one-sided 94% confidence interval that gives a
lower bound for .

SLIDE | 47
 MIND EXPANDING EXERCISE (Page 122)
SOLUTION

Information Day Night

Based on the question, we know that ,

Interpretation: We are 94% confident that the minimum difference between population proportions of the mints
weighed at most 21 grams between night and day shift greater than -0.1262.

SLIDE | 48
 THANK YOU
END OF CHAPTER 2 (PART 2)

SLIDE | 49