Practice Questions

Chapter No.: 9

Chapter Title: Contingency Tables and

The Chi-Square Test

Class: BSc

Subject: Statistics-1

Course Code: 04a

Compiled By: Sir Waseem Mustafa

Cell No.: 0334 8111881

E-Mail: waseem.mustafa.ibd@gmail.com
 Page |1

Important Points:
9.1) Introduction:
In Chapter 8 we focused on testing the value of a particular population parameter of interest,
such as a mean, μ, or proportion, 𝝅. Here we shall look with testing for association between
two categorical variables

9.2) Association:
We often wish to answer questions such as ‘Do vegetarians tend to support Party X?’ or ‘Do
Party X supporters tend to be vegetarians?’ If either question has the affirmative answer, we say
that there is association between the two categories (or ‘attributes’ or ‘factors’). We will need
frequency data in order to answer such questions, but in this course we will not be concerned
with the strength of any such association, but rather just testing for its existence, and with
commenting on its nature.

9.3) Tests for association:

This type of test, tests the null hypothesis that two factors (or attributes) are not associated,
against the alternative hypothesis that they are associated.
For example; suppose that we are sampling people, and that one factor of interest is hair colour
(blonde, brown, black, etc.) while another factor of interest is eye colour (blue, brown, green,
etc.).
In this example, we wish to test whether or not these factors are associated. Hence,
𝐇𝟎 : No association between hair colour and eye colour.
𝐇𝟏 : There is association between hair colour and eye colour.
Note: When conducting tests for association, between the categorical variables, the null
hypothesis can be expressed as ‘No Association’.

Compiled by: Sir Waseem Mustafa (0334 8111881)

9.4) Contingency Tables:

A contingency table (also referred to as cross tabulation or cross tab) is a type of table in a
matrix format that displays the (multivariate) frequency distribution of the variables.
For example; suppose, in three areas of a city a record has been kept of the numbers of
burglaries, robberies and car thefts that take place in a year. If the total number of offences
was 150, then the result can be displayed by the following contingency table:

The cell frequencies are known as observed frequencies and show how the data are spread
across the different combinations of factors.
 Page |2

9.5) General Procedure for Testing Hypothesis from Contingency Tables:

The procedure for testing a hypothesis about for categorical variables involves the following
steps:
Step 1: Define the Hypothesis: State your problem and formulate an appropriate null
hypothesis 𝐇𝟎 with an alternative hypothesis 𝐇𝟏 , which is to be accepted when 𝐇𝟎 is
rejected.
Note: When conducting tests for association, between the categorical variables, the null
hypothesis can be expressed as ‘No Association’.
Step 2: Expected Frequencies: We proceed by computing a corresponding set of expected
frequencies, the expected frequency 𝑬𝒊𝒋 is obtained corresponding to the cell in row i and
column j by the following formula:
(𝑹𝒐𝒘 𝒊 𝒕𝒐𝒕𝒂𝒍) × (𝑪𝒐𝒍𝒖𝒎𝒏 𝒋 𝒕𝒐𝒕𝒂𝒍)
𝑬𝒊𝒋 =
𝑮𝒓𝒂𝒏𝒅 𝑻𝒐𝒕𝒂𝒍
Therefore, the completed expected frequency table for the offences data is

Step 3: Test statistic: The test statistic used for tests for association is as follows:
𝒓 𝒄
𝟐
(𝑶𝒊𝒋 − 𝑬𝒊𝒋 )𝟐
𝑿𝒄 = ∑ ∑
𝑬𝒊𝒋
𝒊=𝟏 𝒋=𝟏

Compiled by: Sir Waseem Mustafa (0334 8111881)

Step 4: Decide significance level and find the value of 𝑿𝒗 𝟐 : A chi-square value is only
defined over positive values and it is dependent on the degrees of freedom
𝒗 = (𝒓 − 𝟏)(𝒄 − 𝟏), where r is no of rows and c is no of columns and for significance level
we use same decision tree which we have studied us in chapter-8. It is also presented here.
Note: In this case, we calculate value of 𝑿𝒗 𝟐 from (table −8) (Will be given in past papers).

(Topic continues on next page)

 Page |3

Step 5: Determine the critical region: 𝑿𝟐 tests of association are always upper tailed tests.
Which means, we have to check 𝑿𝒄 𝟐 > 𝑿𝒗 𝟐
Step 6: Choose hypothesis: Now we decide whether or not to reject 𝐇𝟎 (our working
hypothesis which, up to now, is assumed to be true). If the test statistic value lies in the critical
region then we will reject 𝐇𝟎 , if not we do not reject 𝐇𝟎 . A rejected 𝐇𝟎 means the test is
statistically significant at the specific significance level used.

Step 7: Draw Conclusions: It is always important to draw conclusions in the context of the
variables of the original problem, we can divide it into two types which are as follows:
a) value of the test statistic falls in the critical region and conclude that 𝐇𝟏 is true.
b) Accept the null hypothesis 𝐇𝟎 , otherwise.

Compiled by: Sir Waseem Mustafa (0334 8111881)

Returning to the crime example, now Test for an association between area and crime at 0.01
Solution: 𝐇𝟎 = There is no association between area and crime
𝐇𝟏 = There is association between area and crime
The observed frequencies, the expected frequencies and the test statistic contributor values are
given in the following table:

(Question continues on next page)

 Page |4

This gives a test statistic value of

Since r = c = 3, we have 𝒗 = (r − 1) (c − 1) = (3 − 1) (3 − 1) = 4 degrees of freedom.

For 𝜶 = 𝟎. 𝟎𝟏, from the table-8 we obtain an upper-tail critical value of 13.277.
So in this case the test is highly significant at the 1% level since 23.13 > 13.277. There is
definitely association between the factors.
9.6) Goodness-of-fit Tests: This is a special case of Chi- Square in which we only deal with
one row or one column. In this case, expected frequencies can be computed by
𝟏
𝑬𝒊 = 𝒏 × , i = 1, 2, 3, , k
𝒌
𝟏
Where n denotes the sample size and is the uniform (equal, same) probability for each
𝒌
characteristic.
Having computed expected frequencies, we will use the same test statistic.
The degrees of freedom, 𝒗, in this case will be
𝒗=k−1

Compiled by: Sir Waseem Mustafa (0334 8111881)

Example: A confectionery company is trying out different wrappers for a chocolate bar: its
original A, and two new ones B and C. It puts the bars out in a supermarket and looks to see
how many of each wrapper type have been sold in the first hour. Here are the results:

Is there a difference between wrapper types in the choices made?

(Solution of this example is on next page)

 Page |5

Solution: 𝐇𝟎 = There is no difference in preference for the wrapper types

𝐇𝟏 = There is such a difference.
As there is only one row, so each expected frequency will be equal and it will be

Applying our test statistic formula, we obtain:

Compiled by: Sir Waseem Mustafa (0334 8111881)

The degrees of freedom will be 𝒗 = k − 1 = 3 − 1 = 2.

At the 5% significance level, the upper-tail critical value is 5.99, so since 2.364 < 5.99 we do
not reject 𝐇𝟎 .

If we now consider the 10% level the critical value is 4.61, so again we do not reject 𝐇𝟎 . So
it looks as if there are no preferences for a particular wrapper type on the choices so far.
 Page |6

Exercise:
Q 1) A survey has been made of levels of satisfaction with housing by people living in different
types of accommodation. Levels of satisfaction are:
high, medium, low, very dissatisfied
and housing types are:
public housing apartment, public housing house, private apartment, private detached house,
private semi-detached house, miscellaneous (includes boat, caravan etc!)
Give:
a) the null and alternative hypotheses
b) the degrees of freedom
c) the 5% and 1% critical values for 𝑿𝟐 .
Q 2) In a survey made in order to decide where to locate a factory, samples from five towns
were examined to see the numbers of skilled and unskilled workers. The data were as follows:

Does the population proportion of skilled workers vary with the area?

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 3) Look at the following table taken from a study of gender differences in perception. One
of the tests was of scores in verbal reasoning.

Do these figures show a difference in verbal reasoning by gender (using chi-square)?

Q 4) The following figures give live births by season in town X:

Is there any evidence that births vary over the year?

 Page |7

Questions from Past Papers:

Q 5) (Paper-2003, Zone-A, Section-B, Question-4 (a))
A survey was commissioned recently to examine differences of opinion on various government
initiated measures in different areas of England. A stratified random sample was used with either
60 or 90 respondents in each of the four major regions. The responses, on a five-point scale
ranging from very satisfied to very dissatisfied, for one question are shown below:
Very Satisfied No Dissatisfied Very Total
Satisfied opinion Dissatisfied
South East 17 24 25 12 12 90
South 20 22 20 13 15 90
West
North 6 7 10 18 19 60
East
North 9 8 12 15 16 60
West
Total 52 61 67 58 62 300
Is there an association between response and region?

Q 6) (Paper-2003, Zone-B, Section-B, Question-4 (a))

A survey was conducted to investigate the relationship between frequency of newspaper
readership and reader’s educational background. The following table shows the results of this
survey:

Carry out a suitable test at at least two levels of significance and report on your findings.

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 7) (Paper-2004, Zone-A & B, Section-A, Question-1 (c))

 Page |8

Q 8) (Paper-2004, Zone-A, Section-B, Question-4 (a))

Q 9) (Paper-2004, Zone-B, Section-B, Question-4 (a))

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 10) (Paper-2005, Zone-A, Section-B, Question-3 (a))

240 students were asked to taste 3 of cola in a random order and were not told their identity.
The contingency table below shows the counts of preferred colas classified by the subject
specialism of the students.

(Question continues on next page)

 Page |9

Q 11) (Paper-2005, Zone-B, Section-B, Question-3 (a))

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 12) (Paper-2006, Zone-A & B, Section-B, Question-4 (a))

Q 13) (Paper-2007, Zone-A & B, Section-A, Question-1 (i))

Give an example of a 2 x 2 contingency table in which there is:
i) No association.
ii) Strong association.
 P a g e | 10

Q 14) (Paper-2007, Zone-A & B, Section-B, Question-3 (a))

i) In a large country each district is permitted to have its own policy on the death penalty.
Some districts choose to have it; others choose not to.
The table below shows the relationship between having the death penalty (No, Yes) and the
crime rate (Low, High) for a sample of 200 districts.

Calculate the value of chi-squared for the table and say what you would conclude.
ii) If you look separately at the relationship for poor districts and rich districts you get
the following two tables:

For the poor districts the value of chi-squared is 1.76. For the rich districts the value of chi-
squared is 0.023. What would you conclude for each table, and overall?

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 15) (Paper-2008, Zone-A, Section-B, Question-2 (a))

The table below shows the relationship between gender and party identification in a US state.

Test for association between gender and party affiliation at two appropriate levels and comment
on your results.

Q 16) (Paper-2008, Zone-B, Section-B, Question-2 (a))

An analyst of the retail trade uses as analytical tools the concepts of ‘Footfall’ (the daily number
of customers per unit sales area of a shop) and ‘Ticket Price’ (the average sale price of an item
in the shop’s offer).
Shops are classified as offering Low, Medium or High price items and, during any sales period,
as having Low, Medium or High footfall.
During the January Sales the analyst studies a sample of shops and obtains the following
frequency data for the nine possible combined classifications:
(Question continues on next page)
 P a g e | 11

Conduct a suitable test for association between Ticket classification and Footfall level, and
report on your findings.

Q 17) (Paper-2009, Zone-A, Section-A, Question-1 (e), (iii))

State whether the following are possible or not and give a brief explanation. (Note that no marks
will be awarded for a simple possible/not possible reply).
A negative chi-squared value shows that there is a big association between the variables tested.

Q 18) (Paper-2009, Zone-A & B, Section-B, Question-3 (b))

The table below cross classifies Education by Religious Beliefs and is based on a sample of
2726 randomly selected adult respondents in the USA.

i) Use an appropriate test to see whether there is an association between Education and Religious
Belief according to these data. Give the null and alternative hypotheses and test at two
appropriate levels.
ii) What are your conclusions?
iii) Compare the proportion of those with less than high school education who have liberal
religious beliefs with that for those with whose highest degree was of bachelor or graduate level.
Does this agree with your earlier findings?

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 19) (Paper-2009, Zone-B, Section-A, Question-1 (e), (iii))
State whether the following are possible or not and give a brief explanation. (Note that no
marks will be awarded for a simple possible/not possible reply).
A negative chi-squared value shows that there is a little association between the variables
tested.
Q 20) (Paper-2010, Zone-A & B, Section-B, Question-3 (b))
A researcher into the use of computational aids conducts a survey in which 195 students, as a
homework assignment, compute the standard deviation of a set of data. The students are asked
what computational aids, if any, they used. The researcher is particularly interested as to whether
male and female students differ in this respect. The following table shows the results of this
survey:
(Question continues on next page)
 P a g e | 12

i) Carry out an overall test for association between gender and method of calculation at two
levels. Give the null and alternative hypotheses and comment on your results.
ii) The researcher is interested in whether there are any gender differences in preferred method
of computation. Discuss any potential gender differences which appeared in the test for
association.

Q 21) (Paper-2011, Zone-A, Section-A, Question-1 (b), (iii))

State whether the following are possible or not and give a brief explanation. (Note that no
marks will be awarded for a simple possible/not possible answer.)
A chi-squared value can be positive.

Q 22) (Paper-2011, Zone-A & B, Section-B, Question-3 (a))

The ministry of education is considering funding pre-school education. Before making their
recommendations, administrators take a random sample of 100 students from various areas to
compare the performance of students in algebra between those who attended pre-school and
those who did not. The results are summarised in the table below:

i) Test for association between performance in algebra and pre-school attendance at two
appropriate significance levels. State the null and alternative hypotheses clearly.
ii) Comment on your results describing potential associations in detail. Discuss any potential
differences in the algebra marks between students who did and students who did not attend pre-
school.

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 23) (Paper-2012, Zone-A, Section-B, Question-2 (a))

A survey was conducted in order to examine potential differences of opinion regarding a new
taxation policy in 3 major cities of England (London, Birmingham, Manchester). The responses
were measured on a binary scale (in favour, against) and are summarised in the table below:

i) Test for an association between a person’s opinion on the new taxation policy and the city of
residence at two appropriate significance levels. State the null and alternative hypotheses
clearly.
ii) Comment on your results describing potential associations in detail. Discuss the potential
differences in rates in favour of the new taxation policy across different cities.
 P a g e | 13

Q 24) (Paper-2012, Zone-B, Section-B, Question-2 (a))

An experiment was conducted in order to determine whether contacting people by phone or by
letter before sending them a survey will increase the response rate. Specifically, one group of
people received a letter before getting the survey; one group received a phone call before
receiving the survey; and one group did not receive any information before the survey arrived.
For this study, a response was defined as returning the survey within 2 weeks.

i) Test for an association between the method of contact prior to the survey and response at two
appropriate significance levels. State the null and alternative hypotheses clearly.
ii) Comment on your results describing potential associations in detail. Discuss the potential
differences in response rates for different methods of contact.

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 25) (Paper-2013, Zone-A, Section-B, Question-2 (a))

A social survey in the United States asked subjects, `Would you say that homeopathy is very
scientific, sort of scientific, or not at all scientific?' The table below cross-classifies their
responses with their highest level of education.

i) Based on the data in the table, and without doing a significance test, how would you describe
the relationship between education and opinion on whether or not homeopathy is scientific?
ii) Calculate the 𝑋 2 statistic and use it to test for independence, using a 1% significance level.
What do you conclude?

Q 26) (Paper-2013, Zone-B, Section-B, Question-2 (a))

The 2006 General Social Survey in the United States asked subjects, `Would you say that
astrology is very scientific, sort of scientific, or not at all scientific?' The table below cross-
classifies their responses with their highest level of education.

(Question continues on next page)

 P a g e | 14

i) Based on the data in the table, and without doing a significance test, how would you describe
the relationship between education and opinion on whether or not astrology is scientific?
ii) Calculate the 𝑋 2 statistic and use it to test for independence, using a 1% significance level.
What do you conclude?

Q 27) (Paper-2014, Zone-A, Section-B, Question-2 (a))

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 28) (Paper-2014, Zone-B, Section-B, Question-2 (a))

Q 29) (Paper-2015, Zone-A, Section-A, Question-1 (h), (v))

State whether the following are true or false and give a brief explanation. (Note that no marks
will be awarded for a simple true/false answer.)
In a chi-squared test of association, the larger the test statistic value, the larger the corresponding
p-value.
Q 30) (Paper-2015, Zone-A, Section-B, Question-3 (a))
A survey was conducted to investigate the relationship between the frequency of newspaper
readership and readers' educational background. The following table shows the results of this
survey:
(Question continues on next page)
 P a g e | 15

i) Based on the data in the table, and without conducting a significance test, would you say
there is an association between the frequency of newspaper readership and reader's
educational background?
ii) Calculate the 𝑋 2 statistic and use it to test for independence, using two appropriate
significance levels. What do you conclude?

Q 31) (Paper-2015, Zone-B, Section-B, Question-3 (a))

Questionnaires were mailed to 300 households, in three different areas of a city, to assess the
level of local sporting facilities. The collected data are shown in the table below.

i) Based on the data in the table, and without conducting a significance test, would you say
there is an association between areas and level of local sporting facilities?
ii) Calculate the 𝑋 2 statistic and use it to test for independence, using two appropriate
significance levels. What do you conclude?

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 32) (Paper-2016, Zone-A, Section-B, Question-2 (a))

A factory uses four different machines to manufacture a particular type of machine component.
A random sample of 400 components is selected from the output of the factory. Each component
in the sample is inspected to determine whether or not it is faulty. The machine that produced
the component is also recorded. The results are as follows:

i) Based on the data in the table, and without conducting any significance test, would you
say there is an association between the machine number and the component being faulty?
ii) Calculate the 𝑋 2 statistic and use it to test for independence, using a 5% significance
level. What do you conclude?
 P a g e | 16

Q 33) (Paper-2016, Zone-B, Section-B, Question-2 (a))

A sample consisting of 400 randomly selected students was classified in terms of personality
type (introvert or extrovert) and in terms of their favourite colour (red, yellow, green or blue).
Their responses are summarised in the table below:

i) Based on the data in the table, and without conducting any significance test, would you
say there is an association between the student's type of personality and colour
preference?
ii) Calculate the 𝑋 2 statistic and use it to test for independence, using a 5% significance
level. What do you conclude?

Q 34) (Paper-2017, Zone-A, Section-B, Question-4 (a))

A sample consisting of 100 randomly-selected adults in the USA was classified in terms of their
political affiliation (Democrat or Republican) and opinion on a tax reform bill (in favour,
indifferent or opposed). The data are summarised in the table below.

i) Based on the data in the table, and without conducting any significance test, would you
say there is an association between the political affiliation and opinion on the tax reform
bill?
ii) Calculate the 𝑋 2 statistic and use it to test for independence of political affiliation and
opinion on the tax reform bill. What do you conclude?

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 35) (Paper-2017, Zone-B, Section-A, Question-1 (f), (vi))

State whether the following are true or false and give a brief explanation (no marks will be
awarded for a simple true/false answer).
A contingency table is used to display two measurable variables.

Q 36) (Paper-2017, Zone-B, Section-B, Question-4 (a))

A sample consisting of 100 randomly-selected students in a UK university was classified in
terms of a student's origin (either UK/EU or overseas) and in terms of their satisfaction with
university life (satisfied, indifferent or dissatisfied). The data are summarised in the table below.

(Question continues on next page)

 P a g e | 17

i) Based on the data in the table, and without conducting any significance test, would you
say there is an association between the student's origin and satisfaction with university
life?
ii) Calculate the 𝑋 2 statistic and use it to test for independence of student's origin and
satisfaction with university life. What do you conclude?

Q 37) (Paper-2018, Zone-A, Section-B, Question-2 (a))

An experiment was conducted to examine whether age, in particular being over 30 or not, has
any effect on preferences for a digital or an analogue watch. Specifically, 129 randomly-selected
people were asked what watch they prefer and their responses are summarised in the table
below:

i) Based on the data in the table, and without conducting any significance test, would you
say there is an association between age and watch preference? Provide a brief justification
for your answer.
ii) Calculate the 𝑋 2 statistic for the hypothesis of independence between age and watch
preference, and test that hypothesis. What do you conclude?

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 38) (Paper-2018, Zone-B, Section-B, Question-2 (a))

A survey was conducted in order to examine whether the final grade of students taking a class
is associated with their attendance of a revision session a few days before the examination. The
data, consisting of students' final grades and revision session attendance, are summarised in the
table below.

i) Based on the data in the table, and without conducting any significance test, would you
say there is an association between final grade and attending revision? Provide a brief
justification for your answer.
ii) Calculate the 𝑋 2 statistic for the hypothesis of independence between final grade and
attending revision, and test that hypothesis. What do you conclude?
 P a g e | 18

Answers:
Q 1) (a)
H0 = There is no association between the kind of accommodation people have and their level of
satisfaction with it.
H1 = There is such a difference.
(b) Degrees of freedom = 15
(c) 5% value = 25.00 and the 1% value = 30.58

Q 2) The proportion of skilled workers does not appear to vary with area.

Q 3) There is no difference between gender and verbal reasoning.

Q 4) There is very strong evidence that births vary over the year.

Q 5) Do it by yourself, Q 6) Do it by yourself.

Q 7) Do it by yourself., Q 8) Do it by yourself.

Q 9) Do it by yourself.

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 10) The results demonstrate a strong association between preference and ‘subject specialism.

Q 11) The results demonstrate an association, though not a strong one, between attitude and
type of household.

Q 12) There is a strong connection between attitude and social class.

The interesting figures here is; differences between the upper social class and the others in their
attitudes.

Q 13)
 P a g e | 19

Q 14) (i) There is a strong connection between crime rate and the existence of the death penalty.
(ii) There is no connection between crime rate and the existence of the death penalty in both
districts.

Q 15) There is some association between gender and party identification but it is not very
marked.

Q 16) There is some association between footfall level and ticket price, but it is not a strong
one.

Q 17) This is clearly not possible as the formula for chi-squared is, in fact, a square (and hence
has a positive value).

Q 18) (i) H0 : There is no association between education and religious beliefs.

H1 : There is association between education and religious beliefs.
(ii) There is strong connection education and religious beliefs.
(iii) High school education and strong religious beliefs = 178/424 = 42.0%
Bachelor/graduate education and strong beliefs = 138/642 =21.5%
This shows that those with degrees are half as likely as others to have strong beliefs and
agrees with the strong significance of our chi-squared test.

Q 19) This is clearly not possible as the formula for chi-squared is, in fact, a square (and hence
has a positive value).

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 20) (i) There is some evidence of association between gender and method of computation,
but the evidence is not terribly strong.
(ii) Looking at individual ‘observed’ and ‘expected’ values, we can see that there is no
difference between men and women in their using no aids.
Slightly fewer women than men than might have been expected use a computer. But the big
difference is that men are much less likely than women to use a statistical function on a
calculator than expected, while women are less likely to use a basic calculator compared with
men.

Q 21) The chi-squared value is always positive, so yes.

Q 22) (i) We conclude that there is some evidence of an association between pre-school
attendance and algebra marks.
(ii) There are a number of statements that can be drawn from the previous results. By checking
differences between expected and observed numbers we can extract various arguments that aid
in the interpretation of the results. For example, we may say things like:

(Answer continues on next page)

 P a g e | 20

 Main sources of association: pre-school v. below grade and at grade.

 Students who attended pre-school are more likely to obtain grade algebra marks than
students who did not.
 Pre-school attendance reduces the chances of a below grade level algebra mark.

Q 23) (i) We conclude that the association between opinion on the new taxation policy and city
is weakly significant.
(ii) A number of statements can be drawn from the previous results. By checking differences
between expected and observed numbers we can extract various arguments that aid in the
interpretation of the results. For example, we may say things like
 Main sources of association: London v. Other cities.
 People in London appear to be neutral regarding the new taxation policy whereas people
in Birmingham and Manchester appear to be slightly against it.
 There does not seem to be a difference of opinion among the people of Birmingham and
Manchester.

Q 24) (i) We conclude that the association between response and method of contact is highly
significant.
(ii) A number of statements can be drawn from the previous results. By checking differences
between expected and observed numbers we can extract various arguments that aid in the
interpretation of the results. For example, we may say things like
 Main sources of association: no contact v. any type of contact.
 Contact prior to the survey increases the response rate.
 Contact by phone results in higher response rates than contact by letter.

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 25) (i) Using the percentages we see that the higher someone's education, the smaller the
belief that homeopathy is very scientific and the higher the belief that it is not at all scientific.
For example, 79% of those who attended college or higher education responded that
homeopathy is not at all scientific, whereas the corresponding proportion for those with less
than high school education is 48%.
(ii) We conclude that the association between views on homeopathy and educational level is
highly significant.

Q 26) (i) Using the percentages we see that the higher someone's education, the smaller the
belief that astrology is very scientific and the higher the belief that it is not at all scientific.
For example, 79% of those who attended college or higher education responded that astrology
is not at all scientific, whereas the corresponding proportion for those with less than high school
education is 48%.
(ii) We conclude that the association between views on astrology and educational level is highly
significant.
 P a g e | 21

Q 27) (i) Looking at the percentages, we see some differences between males and females. More
specifically, 39% of females shop online frequently versus 26% of males. Moreover, the
percentage of males who never shop online is 27% versus 18% for females. Hence, there may
be an association between gender and tendency to shop online, although this needs to be
investigated further.
(ii) There is some evidence of an association between gender and tendency to shop online.

Q 28) (i) Looking at the percentages, we see that distributions for the answers regarding buying
organic products are similar in rural and urban areas. More specifically, the percentage of people
who answered `Yes' is quite close in these two cases (17% and 21%, respectively). The
percentages of those who replied `No' were not too far either (38% vs. 33%). Hence, there does
not seem to be a strong association between place of residence and buying organic products,
although this needs to be investigated further.
(ii) There is no evidence to support an association between place of residence and buying
organic products.

Q 29) False. The larger the test statistic value, the smaller the p-value.

Q 30) i) There are some differences in the distributions within readership levels. More
specifically, graduates appear more frequent readers than low readership compared to those with
lower educational attainment than A-levels (46% vs. 19% and 14% vs. 49%, respectively). For
those with A-levels, most are of a moderate readership type (52%). Hence, there seems to be an
association between readership levels and reader's educational background, although this needs
to be investigated further.
ii) There is strong evidence of an association between readership level and educational
background.

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 31) i) There are some differences in the distributions within areas. More specifically, very
good sporting facilities appear more frequently than poor sporting facilities in Areas 1 and 3
(44% vs. 30% and 45% vs. 28%, respectively). In Area 2, however, poor sporting facilities are
more common than very good ones (45% vs. 29%). There seems to be an association between
area and level of sporting facilities, although this needs to be investigated further.
ii) There is some evidence to support an association between area and level of sporting facilities.

Q 32) i) There are some differences in the proportions of faulty components for each machine.
More specifically, 2% of the components from Machine 2 are faulty, whereas the corresponding
proportion for Machine 3 is 11%, and for Machine 4 is 14%. Hence, there seems to be an
association between machine number and the component being faulty, although this needs to be
investigated further.
ii) There is evidence of an association between machine number and the component being
faulty.
 P a g e | 22

Q 33) i) There are some differences in rates of introvert students for each colour preference.
More specifically, 21% of the students who prefer the green colour are introvert, whereas the
corresponding proportion for students who prefer red is 32%, and for students preferring blue
is 46%. Hence, there seems to be an association between personality type and colour preference,
although this needs to be investigated further.
ii) There is evidence of an association between personality type and colour preference.

Q 34) i) There are some differences in the opinion on the tax reform bill between Democrats
and Republicans. More specifically, only 40% of those in favour are Democrats, whereas more
than 50% of those opposed are Democrats. Hence there seems to be an association between
political affiliation and opinion on the tax reform bill, although this needs to be investigated
further.
ii) There is moderate evidence of an association between political affiliation and opinion on the
tax reform bill.

Q 35) False. A contingency table is used to display two categorical variables. An alternative
justification is that a scatter diagram is used to display two measurable variables.

Q 36) i) There are some differences in the rates of satisfaction between UK/EU and overseas
students. More specifically, two thirds of the satisfied students were overseas students, whereas
only half of the dissatisfied students were overseas students. Hence there seems to be an
association between a student's origin and satisfaction with university life, although this needs
to be investigated further.
ii) There is moderate evidence of an association between a student's origin and satisfaction with
university life.

Compiled by: Sir Waseem Mustafa (0334 8111881)

Q 37) i) There are some differences between younger and older people regarding watch
preference. More specifically, 16% of younger people prefer an analogue watch compared to
48% for people over 30. Hence there seems to be an association between age and watch
preference, although this needs to be investigated further.
ii) There is strong evidence of an association between age and watch preference.

Q 38) i) There are some differences in the final grades between students who did and did not
attend the revision session. More specifically, 56% of those who got a final grade A attended
the revision session, whereas only 40% of those who got a final grade C attended it. Hence there
seems to be an association between final grade and attending revision although this needs to be
investigated further.
ii) There is weak evidence of an association between attending revision and final grade.

The End, Good Luck