Testing of Hypothesis

Example 1: A sample of 200 blubs made by a company give a life time mean of
1540 hours with a standard deviation of 42 hours. Is it likely that that the sample has
been drawn from a population with a mean lifetime of 1500 hours? You may use 5
per cent level of significance.

Solution:
In the above example, the size is large (n=200), sample mean (X ) = 1540 hours and
the standard deviation (s) is equal to 42 hours.

The null and alternate hypothesis can be written as:

H0 : µ = 1500 hrs
H1 : µ ≠ 1500 hrs

It is a two-tailed test with level of significance (α) to be equal to 0.05. Since n is large
(n>30), though population standard deviation σ is unknown, one can use Z-test. The
test statistics are given by:
X−μ
Z= σ
x
Where μ = Value of µ under the assumption that the null hypothesis is true

= Estimated Standard Error of mean

Here μ = 1500, = = = = 2.97

√ √ √
(Note that is estimated value of σ)

= = = = 13.47
. .
√

The value of α = 0.05 and since it is a two-tailed test, the critical value Z is given by -Z
α/2 and Z α/2 which could be obtained from the standard normal table.

1|Page
Since the computed value of Z=13.47 lies in the rejection region, the null hypothesis
is rejected. Therefore, it can be concluded that the average life of the bulb is
significantly different from 1500 hrs.

Alternative Approach to the test of Hypothesis:

There is an alternative approach called probability approach or simply p value

approach to test hypothesis. Under this approach, the researcher does not have to
refer to Z table to determine the critical value.

Referring to the above example the p value can be calculated as follows:

p = P ( Z > 13.47 ) + P ( Z < -13.47 )

We know that the problem is a two-sided test and Z has a systematic distribution,
therefore,
p = 2P ( Z > 13.47 ) = 2 x 0 = 0

Now the distribution rule is:

Reject H0 if p ≤ α
Accept H0 if p > α

In this example, α = 0.05 and p value is less than α, so the null hypothesis is rejected.
Therefore, it may be noted that the same conclusion is arrived at and there is no
need to look at the critical value of Z as given in the statistical table. Please note that
the p value provided there is for the two-sided test. In case the problem is of one-
sided test, the reported p value is divided by 2 to obtain the desired p value for the
problem and then compared with alpha (α), the level of significance so as to either

2|Page
accept or reject the null hypothesis. This is possible since Z-distribution is a
symmetrical distribution.

Example 2: Past records indicate that a golfer has averaged 82 on a certain course.
With a new set of clubs, he averages 7 over five rounds with a standard deviation of
2.65. Can we conclude that 0.025 level of significance, the new club has an adverse
effect on the performance?

Solution:
H0 : µ = 82
H1 : µ < 82

X = 7.9, n = 5, s = 2.65, α = 0.025. As the population standard deviation is

unknown and the sample size is small (n<30), a t test would be appropriate. The test
statistic is given by:
. – . – .
t n-1 = = = = = - 0.25
. .
√

σ 2.65
= = = 1.185
x √ √5
The critical value of t at 0.025 level of significance with four degrees of freedom is
given by -tα = -2.776 as the sample t value of -0.25 lies in the acceptance region, the
null hypothesis is accepted.

3|Page
Therefore, there is no adverse effect on the perfroamnce due to a change in the club
and the performance can beattributed to chance.

Example 3: A study is carried out to examine whether the mean hourly wages of the
unskilled workers in the two cities – Ambala Cantt and Lucknow are the same. The
random sample of hourly earnings in both the cities is taken and the results are
presented in the Table below.

City Sample Mean Standard Deviation of Sample Size

Hourly Earnings Sample

Ambala Cantt Rs. 8.98 (X1) 0.40 (s1) 200 (n1)

Lucknow Rs. 9.10 ( 2) 0.60 (s2) 175 (n2)

Using a 5 per cent level of significance, test the hypothesis of no difference in

the average wages of unskilled workers in the two cities.

Solution: We use subscripts 1 and 2 for Ambala Cantt and Lucknow respectively.

H0 : µ1 = µ2 → µ1 - µ2 = 0
H1 : µ1 ≠ µ2 → µ1 - µ2 = 0

The following survey data is given

X1 = 8.98, 2 = 9.10, s1 = 0.40, s2 = 0.60, n1 = 200, n2 = 175, α = 0.05

Since both n1, n2 are greater than 30 and the sample standard deviations are given, a
Z test would be appropriate.

The test statistic is given by:

− −( )
=

As σ1 and σ2 are unknown, their estimates would be used.

S1 = σ1, S2 = σ2

( . ) ( . )
+ = + = √0.0028 = 0.0053

4|Page
 ( . . )
= = -2.83
.

As the problem is of a two-tailed test, the critical values of Z at 5 per cent level of
significance are given by -Z α/2= - 1.96 and and Z α/2 = 1.96. The sample value of Z
= -2.83 lies in the rejection region as shown in the figure

Therefore, the null hypothesis is rejected and it may be concluded that there is a
difference in the average wages of unskilled workers in the two cities. Let us rework
the same problem using p value approach. As it is known that the problem is if a
two-tailed tezt, the p value is given by:

p = P (Z > 2.83) + P (Z < -2.83)

= 2P (Z < -2.83)
= 2 X (0.05 – 0.4977)
= 2 x 0.0023
= 0.0046

As the value of p is less than α (0.05), the null hypothesis is rejected. Similarly, the
problems on one-tailed tests can be solved.

5|Page
Case

Perception of People about Ban on Plastic Bags in Delhi

Plastic bags play an integral role in our daily life. Be it carrying groceries from
the local kirana store or the storing of household articles in a poly-bag. We never
actually run out of plastic bags. The omnipresence of this utility object brought to the
fore an impending problem that needed to be resolved. The problem associated with
using plastic bags is that they are not biodegradable and in fact take close to 60 years
to decompose. Apart from that, they are also the cause of various other problems
such as clogging of drain pipes and death of cattles that accidentally chew on plastic
bags.

This prompted the Delhi government to finally take notice and introduce a
blanket ban on plastic bags in 2009. The storages and sale of plastic bag in all places,
including shops, is banned. The penalty for violating the ban, could be a fine of Rs
1,00,000 or five years, imprisonment or both.

The officials empowered to enforce the ban are the staff of the health and
environment department. Food and supply officers and sub divisional magistrates
are also empowered to enforce the ban.

The Delhi pollution control committee (DPCC) has been assigned the task of
implementation. It has formed a special inspection team for the purpose. The team
would visit manufacturing units and retail shops, and would initiate punishment for
the violators. The scope of this ban has been widened by including four star hotels
under its purview.

The imposition of this widespread ban has prompted researchers to analyses

the impact and effectiveness of the ban from the perspective of both the consumer
and the vendor. They first checked whether the consumers and vendors are aware of
the ban or not. Along with that they analysed the preference, choices and willingness
of the consumers and vendors from diverse backgrounds to switch to eco-friendly
alternatives so as to ascertain the effectiveness of the ban on plastic bags.

A survey was conducted in Delhi to understand the perception of consumers

about the plastic bag ban. The statements related to the respondents’ perceptions
are listed below:

6|Page
 What are your views about plastic bag ban? (Tick one for each answer)

Parameters Strongly Moderately Neither Moderately Strongly

Agree Agree (2) Agree Nor Disagree Disagree
(1) Disagree (4) (5)
(3)
Plastic bag is a must
when buying
groceries/vegetables.
(X12a)
Plastic bag is harmful for
the environment. (X12b)
I do not wish to quit
using plastic bags.(X12c)
I try to avoid plastic
bags as much as I can.
(X12d)
Plastic bag ban is not
enforced properly.
(X12e)
Plastic bag is not a
useful substitute for
plastic bag. (X12f)
A sample of 44 respondents was chosen randomly. The data is presented in the Table
below and is also available in SPSS/EXCEL file.

7|Page
 Table

Selected data on perception and demographic profile of consumers regarding ban on

plastic bags

Resp No. X12a X12b X12c X12d X12e X12f Age Gender
1 1 2 3 4 2 2 2 1
2 2 1 3 4 1 5 2 1
3 4 1 4 3 2 4 2 1
4 1 1 3 3 2 4 2 2
5 3 1 5 4 1 5 2 2
6 2 1 3 3 1 2 2 1
7 3 1 4 2 2 4 2 1
8 1 5 5 5 3 1 3 1
9 3 2 3 3 2 2 2 2
10 2 1 5 2 2 4 2 1
11 5 1 1 1 1 2 1 1
12 5 1 2 2 1 2 2 1
13 2 1 3 2 1 2 2 2
14 3 1 2 2 1 2 2 1
15 2 1 5 2 2 4 2 2
16 2 1 4 4 1 5 2 1
17 2 3 3 3 4 1 2 2
18 2 2 4 2 2 3 3 1
19 2 1 4 4 1 5 2 1
20 1 2 2 3 3 2 3 1
21 5 1 3 2 2 2 3 1
22 5 1 4 1 1 5 2 1
23 2 2 4 2 2 2 2 2
24 3 1 3 2 2 2 2 1
25 4 1 5 1 1 2 1 2
26 2 1 3 2 1 2 2 1
27 2 4 5 2 4 5 2 1
28 2 1 2 5 2 5 2 1
29 2 1 4 2 1 2 2 2
30 1 1 4 3 2 4 2 1
31 1 1 2 5 2 4 2 1
32 5 1 5 3 1 5 2 2
33 5 3 2 4 4 2 2 1
34 5 1 5 2 2 2 2 1
8|Page
 35 4 1 2 1 3 1 2 1
36 3 1 5 3 1 2 2 2
37 2 1 4 2 1 5 2 1
38 4 1 3 2 1 4 2 1
39 5 1 5 2 2 4 2 1
40 5 1 5 3 4 4 2 1
41 2 2 2 5 1 4 2 1
42 2 1 3 2 1 3 2 2
43 2 1 4 4 1 4 2 1
44 2 2 2 4 2 3 2 2

 The variable age is coded as:

1 = Below 18 years

2 = 18 to 30 years

3 = 31 to 50 years

 The variable gender is coded as:

1 = Male
2 = Female

Questions:

1. By using a one-sample t-test, identify the parameters of the plastic bags

ban on which the consumers have a favorable opinion.

(Hint: Test the null hypothesis: µ = 3 against an appropriate alternative

hypothesis.)

2. Using a two-sample independent t-test, examine whether the views of the

male and the female respondents are the same.

3. Divide all the respondents into two groups by taking respondents aged 30
and below as the younger respondents and those who are 31 and above as
older respondents. Now statistically examine whether the views on the ban
on plastic bags are different for the younger and the older respondents.

4. Write a summary of your finding.

9|Page
Solution:
The Delhi government introduced a ban on plastic bags in 2009. This was
necessitated because these bags cause clogging of drains and deaths of cattle that
accidently chew these bags. Moreover, they are not biodegradable and take close
to 60 years to decompose. There is a penalty for storage and sale of plastic bags in
all places, including shops, which could be a fine of Rs.1,00,000 or five years’
imprisonment or both. A survey was conducted on a sample of 44 people to
understand the perception of consumers about the ban on plastic bags. Six
statements were evaluated on a five-point Likert scale, where 1 = strongly agree, 2 =
agree, 3 = neutral, 4 = disagree, 5 = strongly disagree.

The mean score, standard deviation and standard error of their mean were
computed and presented in Table 1.

Table 1 - One-Sample Statistics

Std. Error
N Mean Std. Deviation Mean
[Plastic Bag is must
when buying 44 2.7955 1.37383 .20711
groceries/vegetables ]
[Plastic Bag is harmful
44 1.4091 .87120 .13134
for environment]
[I do not wish to quit
using Plastic Bags] 44 3.5000 1.15134 .17357
[I try to avoid Plastic
44 2.7727 1.13841 .17162
Bags as much as I can]
[Plastic Bag Ban is not
enforced properly] 44 1.7955 .92960 .14014
[Paper Bag is not a
useful substitute for 44 3.1591 1.32846 .20027
plastic bag ]

Answer – 1
The null hypothesis to be tested in all cases is H0 : µ = 3. Looking at the mean score
of six statements in Table 1, the alternative hypothesis for statements 1, 3 and 6
would be H1 : µ > 3 whereas for the remaining statements 2, 4 and 5, the alternative
hypothesis would be H1 : µ < 3. The results of test of significance are reported in
Table 2.

10 | P a g e
 Table 2 - One-Sample Test

Test Value = 3
95% Confidence
Interval of the
Mean Difference
t df Sig. (2-tailed) Difference Lower Upper
[Plastic Bag is must
when buying -.988 43 .329 -.20455 -.6222 .2131
groceries/vegetables ]
[Plastic Bag is harmful
for environment] -12.113 43 .000 -1.59091 -1.8558 -1.3260
[I do not wish to quit
using Plastic Bags] 2.881 43 .006 .50000 .1500 .8500
[I try to avoid Plastic
Bags as much as I can] -1.324 43 .192 -.22727 -.5734 .1188
[Plastic Bag Ban is not
enforced properly] -8.595 43 .000 -1.20455 -1.4872 -.9219
[Paper Bag is not a
useful substitute for .794 43 .431 .15909 -.2448 .5630
plastic bag ]

A 5% level of significance is assumed. One could use t test with n – 1 = 43 degrees of

freedom or a Z test because for a large sample both would give the same results
approximately. The table value of one-tailed t at 5% level of significance is 1.645.
Since absolute value of computed t in case of hypotheses corresponding statements
2, 3 and 5 is greater than 1.645, these hypotheses are rejected.

The implications of rejections of hypotheses corresponding to statements 2 and 5

are –

 That the consumers are in favourable agreement that plastic bags are harmful
for environment and that ban is not enforced properly. The results are found
to be statistically significant.

The hypothesis corresponding to statement 3 is also rejected since the absolute

value of computed t is greater than the tabulated value. The rejection of the
statement means that they are in favour of quitting the use of plastic bags.

11 | P a g e
Answer – 2

The hypotheses to be tested corresponding to each of the six statements are:

H0 : µm = µf
H1 : µm ≠ µf
An independent two sample test is used for the purpose. The results are presented
in Tables 3 and 4.

Table 3 - Group Statistics

Std. Error
Gender N Mean Std. Deviation Mean
[Plastic Bag is must Male 31 2.9032 1.49119 .26783
when buying
groceries/vegetables ] Female 13 2.5385 1.05003 .29123
[Plastic Bag is harmful Male 31 1.4194 .95827 .17211
for environment] Female
13 1.3846 .65044 .18040
[I do not wish to quit Male 31 3.3548 1.17042 .21021
using Plastic Bags] Female 13 3.8462 1.06819 .29626
[I try to avoid Plastic Male 31 2.8387 1.24088 .22287
Bags as much as I can] Female 13 2.6154 .86972 .24122
[Plastic Bag Ban is not Male 31 1.8710 .95715 .17191
enforced properly] Female 13 1.6154 .86972 .24122
[Paper Bag is not a Male 31 3.2903 1.34644 .24183
useful substitute for Female 13 2.8462 1.28103 .35529
plastic bag ]

12 | P a g e
 Table 4 - Independent Samples Test

Levene's Test for

Equality of Variances t-test for Equality of Means
95% Confidence
Interval of the
Mean Std. Error Difference
F Sig. t df Sig. (2-tailed) Difference Difference Lower Upper
[Plastic Bag is must Equal variances
5.327 .026 .800 42 .428 .36476 .45586 -.55520 1.28473
when buying assumed
groceries/vegetables ] Equal variances
.922 31.787 .364 .36476 .39566 -.44137 1.17090
not assumed
[Plastic Bag is harmful Equal variances
for environment] assumed .331 .568 .119 42 .906 .03474 .29122 -.55297 .62245
Equal variances
.139 32.888 .890 .03474 .24933 -.47260 .54207
not assumed
[I do not wish to quit Equal variances
.192 .663 -1.302 42 .200 -.49132 .37740 -1.25293 .27030
using Plastic Bags] assumed
Equal variances
not assumed -1.352 24.628 .189 -.49132 .36326 -1.24005 .25742
[I try to avoid Plastic Equal variances
Bags as much as I can] assumed 3.195 .081 .589 42 .559 .22333 .37905 -.54163 .98828
Equal variances
.680 31.926 .501 .22333 .32841 -.44569 .89234
not assumed
[Plastic Bag Ban is not Equal variances
enforced properly] assumed .123 .727 .829 42 .412 .25558 .30829 -.36657 .87773
Equal variances
not assumed .863 24.733 .397 .25558 .29621 -.35480 .86596
[Paper Bag is not a Equal variances
.841 .364 1.012 42 .317 .44417 .43883 -.44143 1.32977
useful substitute for assumed
plastic bag ] Equal variances
1.033 23.663 .312 .44417 .42978 -.44353 1.33187
not assumed

In Table 3 are presented the results of descriptive statistics for male and female
respondents corresponding to six statements. The mean score for males and
females are generally close to each other except for statements 1 and 6. However,
to examine the whether the views are statistically same the results of t test under
the assumption of equal and unequal variances are presented in Table 4.

Again, the level of significance is assumed to be 5%. To test the hypothesis, we

would compare p values (in Table 4) with level of significance (α). If p ≤ α, null
hypothesis would be rejected. Examining the results in Table 4, we accept the null
hypotheses corresponding to all six statements. This means that the views about
plastic bag bans are independent of gender.

Answer – 3

The 44 respondents are divided into two groups based on age. Those 30 and below
are grouped as younger respondents and those more than 30 are termed as older
respondents. There are 40 respondents in younger group and only 4 in the older
13 | P a g e
group. The instructor should point out the limitation of same to the students. The
hypothesis to be tested corresponding to six statements is
H0 : µy = µo (views of younger and older respondents is same.)
H1 : µy ≠ µo
Table 5 - Group Statistics

Std. Error
age_redefined N Mean Std. Deviation Mean
[Plastic Bag is must Younger respondent 40 2.8500 1.33109 .21046
when buying
groceries/vegetables ]Older Respondent 4 2.2500 1.89297 .94648
[Plastic Bag is harmful
Younger respondent 40 1.3000 .68687 .10860
for environment] Older Respondent
4 2.5000 1.73205 .86603

[I do not wish to quit Younger respondent 40 3.5000 1.15470 .18257

using Plastic Bags] Older Respondent 4 3.5000 1.29099 .64550
[I try to avoid Plastic Younger respondent 40 2.7500 1.12660 .17813
Bags as much as I can] Older Respondent 4 3.0000 1.41421 .70711
[Plastic Bag Ban is not Younger respondent 40 1.7250 .93336 .14758
enforced properly] Older Respondent 4 2.5000 .57735 .28868
[Paper Bag is not a Younger respondent 40 3.2750 1.32021 .20874
useful substitute for Older Respondent 4 2.0000 .81650 .40825
plastic bag ]

Table 6 - Independent Samples Test

Levene's Test for

Equality of Variances t-test for Equality of Means
95% Confidence
Interval of the
Mean Std. Error Difference
F Sig. t df Sig. (2-tailed) Difference Difference Lower Upper
[Plastic Bag is must Equal variances
when buying assumed .424 .519 .830 42 .411 .60000 .72307 -.85921 2.05921
groceries/vegetables ] Equal variances
not assumed .619 3.303 .576 .60000 .96960 -2.33142 3.53142
[Plastic Bag is harmful Equal variances
for environment] assumed 7.588 .009 -2.833 42 .007 -1.20000 .42356 -2.05478 -.34522
Equal variances
-1.375 3.095 .260 -1.20000 .87281 -3.93003 1.53003
not assumed
[I do not wish to quit Equal variances
.000 1.000 .000 42 1.000 .00000 .61091 -1.23287 1.23287
using Plastic Bags] assumed
Equal variances
not assumed .000 3.497 1.000 .00000 .67082 -1.97293 1.97293
[I try to avoid Plastic Equal variances
Bags as much as I can] assumed .025 .876 -.415 42 .680 -.25000 .60282 -1.46654 .96654
Equal variances
not assumed -.343 3.392 .752 -.25000 .72920 -2.42612 1.92612
[Plastic Bag Ban is not Equal variances
.597 .444 -1.620 42 .113 -.77500 .47854 -1.74074 .19074
enforced properly] assumed
Equal variances
-2.390 4.748 .065 -.77500 .32421 -1.62189 .07189
not assumed
[Paper Bag is not a Equal variances
useful substitute for assumed 7.508 .009 1.884 42 .067 1.27500 .67688 -.09100 2.64100
plastic bag ] Equal variances
not assumed 2.781 4.749 .041 1.27500 .45852 .07733 2.47267

14 | P a g e
Table 5 presents the results of descriptive statistic. Ask students to interpret it. The
results test of significance for the hypotheses corresponding to each of the size
statements are given in Table 6. Again a 5% level of significance is assumed. The p
value is greater than level of significance (α) for all cases except the equal variance
case of hypothesis corresponding to statement 2 and unequal variance case for
hypothesis corresponding to statement 6. (The instructor at this stage should
emphasize the importance of testing the assumption of equality of variance before
carrying out this test.)

We could infer from the analysis that there is no difference in the views of younger
and older respondents on ban of plastic bags.

Answer – 4
The results could be summarized as given below:
 Views on ban on plastic bags are indifferent across gender and age group.
 The results indicate that respondents are of the view that plastic bags are
harmful for the environment, the ban on plastic bags is not enforced properly
and people want to stop using plastic bags.

15 | P a g e
 Correlation and Regression Analysis
Case

MRP Biscuit Company Pvt. Ltd.

The Indian biscuit industry has a turnover of around Rs 3000 crore. India is the
second largest manufacture of biscuit, after USA. The industry employs almost
3.5lakh people directly and 30lakh people indirectly. The biscuit industry can be
segmented into the organized and unorganised sectors. There are about 150 small
and medium sector is in the ratio of 55 to 45 per cent. Exports of biscuits have been
generally to the tune of 10 per cent of annual production. The industry is showing an
annual growth rate of about 14 to 16 per cent since 2003. The per capita
consumption of biscuits in India is only 1.8 kg per annum as compared to 2.5 kg to
5.5 kg in the South East Asian Countries, European countries and USA. The biscuits
could be broadly classified into various categories such as Glucose, Marie, Sweet,
Salty, Cream and Milk.

MRP biscuit company started its operations in Ambala city, Haryana, ion 2001. The
company was growing at an annual rate of 20 per cent, which was above the
industry average. However, for the last three years, the growth has been only to the
tune of 5 to 6 per cent. This very factor has been of a main concern to the top
management of the company. Mr P.K. Malhotra, the Senior Vice president,
Marketing, had a meeting of the senior marketing team and was wondering why
their company, which has been doing so well, has showed down in the last few
years. During the discussion it was suggested by one of the senior managers to
identify the factors which influence the preference for biscuits. It was argued that
once these are known, it will help the company to concentrate on those factors
accordingly. Therefore, the company decided to get a study done from a research
agency to identify the various factors that influence the preference for biscuits. A
sample of 40 individuals was chosen randomly from Ambala. The data was collected
on variables like preservation, quality, taste, nutrition value and preference on a 7-
point scale with the higher number indicating a more positive rating. The data is
presented in the Table below

16 | P a g e
 Table
Data on preference for biscuits
S. No. Preference Nutrition Value Taste Preservation Quality
1 7 5 6 5
2 6 4 6 6
3 5 5 7 4
4 6 6 7 5
5 4 3 2 4
6 2 2 1 2
7 3 3 2 3
8 6 5 6 5
9 7 7 7 6
10 5 6 5 4
11 4 4 3 2
12 3 6 2 3
13 1 1 2 1
14 2 2 3 1
15 4 5 4 3
16 4 4 5 4
17 3 2 1 3
18 6 7 5 4
19 6 5 5 6
20 7 6 4 5
21 5 5 4 4
22 3 2 3 4
23 2 2 1 1
24 5 5 4 4
25 6 5 6 4
26 7 6 5 7
27 2 1 1 2
28 4 2 1 2
29 6 4 5 5
30 7 6 5 5
31 6 3 6 5
32 6 3 6 5
33 2 1 1 2
34 3 2 1 1
35 4 3 2 2
36 6 5 7 6

17 | P a g e
 37 7 6 7 6
38 7 5 6 7
39 4 3 2 3
40 5 3 4 3

(i) Run a multiple regression exploring the preference for the brand of biscuits in
terms of the nutrition value, taste and preservation quality.
(ii) Interpret the partial regression coefficients.
(iii) Examine the significance of the partial regression coefficient using a 5 per cent
level of significance.
(iv) As a marketing manager of the biscuit company, on what attributes will you
concentrate more so as to improve the marketability of the brand?

18 | P a g e
Solution:
MRP Biscuit Company Pvt. Ltd.
The company started its operations in Ambala city in the year 2001. The company
was growing at an annual rate of 20%, which was above the industry average.
However, for the last three years, the growth has slowed down and the present rate
of growth is five to six per cent. The management is concerned about the reasons
for this decline. They are interested in determining the variables which influence the
preference for biscuits. They got a study done where preservation quality, taste and
nutrition values were considered as independent variables and preference rating
was taken as dependent variable. All these variables were measured on a seven-
point scale with a higher number indicating a more positive rating. We ran
regression with preference rating as dependent variable and nutrition value, taste
and preservation quality as independent variables. The results are presented in the
tables below:
Table - 1 : Model Summary

Adjusted Std. Error of

Model R R Square R Square the Estimate
1 .928a .860 .849 .69921
a. Predictors: (Constant), Preservation Quality, Nutrition
Value, Taste

b
Table - 2 : ANOVA

Sum of
Model Squares df Mean Square F Sig.
1 Regression a
108.375 3 36.125 73.891 .000
Residual 17.600 36 .489
Total 125.975 39
a. Predictors: (Constant), Preservation Quality, Nutrition Value, Taste
b. Dependent Variable: Preference

Table - 3 : Coefficients a

Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Sig.
1 (Constant) .733 .301 2.436 .020
Nutrition Value .295 .103 .284 2.865 .007
Taste .170 .103 .198 1.655 .107
Preservation Quality .548 .118 .522 4.660 .000
a. Dependent Variable: Preference

19 | P a g e
1. The estimated multiple regression model is
Pref = 0.733 + 0.295 Nutrition Value + 0.170 Taste + 0.548 Preservation
quality
Sig (p) 0.020 0.007 0.107 0.000
2
R = 0.86

The regression equation makes use of unstandardized regression coefficients as

reported in the first column of Table 3.
2. The results indicate that all the independent variables namely nutrition value,
taste and preservation quality have a positive impact on the preference for the
biscuit. This is because the signs of all the regression coefficients are positive.
3. The R2 value for the model is 0.860 which indicates that 86% of the variations in
the preference are explained by nutrition value, taste and preservation quality.
The significance of R2 is tested with the help of F statistic which works out to be
73.891 and is significant if we assume the significance level to be 5%. (Please
refer Table 2.) This is because the p value for F equals zero which is less than the
level of significance.
4. The significance of the individual coefficients can be tested using t-statistic. The
computed value of the t-statistic is presented in Table 3. However we will have
to refer to the table value of t to know which variable is significant. The table
value of t at 5% level of significance with 36 degrees of freedom is 1.96. Since
the computed t value is greater than table values of t for the coefficient of
presentation quality and nutrition value, these variables are statistically
significant. An easy way out is to look at the sig. (p value) as given in Table 3. Let
us assume that the level of significance is α = 0.05. Now if p value is less than α
for a particular coefficient means the variable attached to that coefficient is
significant. Therefore, we find that nutrition value and preservation quality are
significant variable in influencing preference. This was also found using t
statistic.

20 | P a g e
 5. Further, the relative importance of the independent variables in influencing
preference is determined by examining the standardized coefficients (called β).
They are reported in Table 3. The standardized coefficients are obtained by
running the regression of standardized values of dependent variable on the
standardized values of the independent variables. The standardized values of a
variable are obtained by subtracting from the variable its mean value and
dividing by its standard deviation. Higher the value of absolute standardized
coefficient, higher is the importance of that variable in influencing preference.
Therefore we note that preservation quality is the most important variable
followed by nutrition value in influencing the preference for biscuits.

Therefore, it is suggested that the company should work on the preservation quality
the most in improving the preference for its brand. It should also consider on
nutrition value which is the second important variable. This doesn’t mean that the
taste should be ignored. Taste has a positive impact on preference although it is not
highly significant. However, if we increase the level of significance α to 15% it will
become significant in influencing preference.

21 | P a g e
 Analysis of Variance Techniques
Case
Paid Kids’ Care Unit in a Mall

In the past few years, a large number of malls have sprouted in the Indian metros.
Malls are not only meant for shopping but are also combined with multiplexes and
provide other indoor modes of recreation. In this context, it has become a place to
hang out for most of the younger population.

Many young parents go to malls, usually with their children in tow. While it can be a
terrific family outing. Sometimes a break from the children while shopping can also
be a pleasant experience. A kid’s care center in a mall can give parents a fantastic
place to drop off their children while shopping or while exploring the mall for other
modes of entertainment or recreation.

Such facilities are already available in European markets. A study was conducted to
examine whether Indians need such a facility.

The unit of analysis for the study was young parents having kids in the age group1 to
6 years. The visit to a mall was considered to be the most appropriate method to find
the target population. A sample of 30 respondents was selected while they were
visiting malls. A questionnaire was administered to the respondents. A few questions
that were asked of the respondents were:

 If you are provide with a paid kids’ care facility in a mall, for the kids aged 1-6
years, would you be interested in availing of the facility? (Y)
(a) Very interested - (5)
(b) Interested - (4)
(C) indifferent - (3)
(d) Not interested - (2)
(e) Not at all interested - (1)

 According to you what should be the change on hourly basis, for a kids care
center in a mall? (X₁)
(a) Rs 100 – Rs 150 - (1)
(b) Rs 151 – Rs 200 - (2)
(C) Rs 201 – Rs 250 - (3)
(d) Rs 251 and above - (4)

22 | P a g e
  Yours Sex (X₂)
(a) Male - (1)
(b) Female - (2)

 Your education (X₃)

(a) Undergraduate - (1)
(b) Graduate - (2)
(c) Post graduate - (3)

 Your monthly household income (X₄)

(a) Less than or equal to Rs 15000 - (1)
(b) Rs 15001 – Rs 30000 - (2)
(c) Rs 30001 – Rs 45000 - (3)
(d) Rs 45001 and above - (4)

 Are both you and your spouse working (X₅)

(a) Both - (1)
(b) One - (2)

 You belong to (X₆)

(a) Nuclear family - (1)
(b) Joint family - (2)

The data on the variable Y is in the interval scale, whereas the data on the remaining
variable - X₁, X₂ up to X₆ - is normal scale. The coding for X variable is shown within
parenthesis. The value taken by the interval scale variable Y is shown within the
brackets. The entire data is reproduced below in Table.

23 | P a g e
 TABLE: Data for select variables

S. No. Y X₁ X₂ X₃ X₄ X₅ X₆
1 4.00 1.00 2.00 2.00 3.00 1.00 1.00
2 3.00 1.00 1.00 3.00 3.00 1.00 1.00
3 2.00 1.00 2.00 3.00 3.00 2.00 1.00
4 4.00 1.00 2.00 3.00 3.00 1.00 1.00
5 5.00 1.00 2.00 2.00 4.00 2.00 1.00
6 3.00 1.00 2.00 2.00 3.00 2.00 1.00
7 5.00 1.00 1.00 2.00 4.00 2.00 2.00
8 2.00 1.00 2.00 3.00 4.00 2.00 2.00
9 2.00 1.00 1.00 3.00 4.00 2.00 2.00
10 3.00 1.00 1.00 3.00 3.00 2.00 1.00
11 5.00 1.00 2.00 2.00 4.00 2.00 1.00
12 4.00 1.00 1.00 3.00 4.00 1.00 1.00
13 5.00 1.00 1.00 2.00 4.00 2.00 2.00
14 5.00 1.00 1.00 2.00 3.00 2.00 2.00
15 4.00 2.00 1.00 2.00 3.00 2.00 2.00
16 5.00 2.00 2.00 3.00 4.00 2.00 2.00
17 2.00 3.00 2.00 3.00 4.00 1.00 2.00
18 2.00 1.00 1.00 2.00 3.00 1.00 2.00
19 3.00 1.00 1.00 3.00 4.00 2.00 1.00
20 4.00 1.00 2.00 3.00 3.00 1.00 1.00
21 5.00 1.00 1.00 3.00 4.00 1.00 2.00
22 5.00 1.00 1.00 1.00 3.00 1.00 1.00
23 4.00 2.00 2.00 1.00 3.00 1.00 1.00
24 4.00 3.00 2.00 3.00 4.00 1.00 1.00
25 5.00 1.00 1.00 2.00 4.00 2.00 2.00
26 5.00 2.00 2.00 2.00 4.00 2.00 2.00
27 5.00 2.00 2.00 2.00 4.00 1.00 2.00
28 3.00 1.00 1.00 2.00 4.00 2.00 2.00
29 4.00 1.00 1.00 2.00 4.00 2.00 2.00
30 5.00 2.00 2.00 2.00 4.00 2.00 2.00

24 | P a g e
Question:

1. Treat X1, X2, X6 as independent variables. Run a one-way analysis of variance

using the independent variables X1, X3 and X4 with interest in Kids’ Cadre
Centre (Y) as a dependent variables. If the results are significant, carryout an
appropriate t-test for a further analysis

2. Conduct an appropriate test to examine whether there is a difference in the

interest in the Kid’s Care Centre because of gender (X2), spouse working (X5)
and type of family (X6). Interpret the result

3. Divide the interest in the Kids’Care Centre into two groups – low interest with
a score of 1 to 3 and high interest with a score to 4 -5. Cross-tabulate it with
the gender (X2), spouse working (X5) and type of family (X6). Interpret the
results

4. Write a management summary of the findings.

25 | P a g e
Solution

Paid Kids’ Care Unit in a Mall

1) The interest in the kids’ care facility in a mall is taken as a dependent variable
(Y) whereas each of X1, X3, X4 variables are independent variables. Each of
the independent variable is non-metric whereas the dependent variable is
metric.

The results of one-way ANOVA with Y as dependent variable and (X1) the
amount of money that would be charged on an hourly basis as independent
variables are given in Table 1 & 2.

The hypothesis to be tested is:

H0 : The interest in kids’ care facility is independent of their desire to pay
the amount of charges on hourly basis.
H1 : The interest in kids’ care facility is dependent on their desire to pay the
amount of charges on hourly basis.

TABLE – 1
Descriptives

Interest in facility
95% Confidence Interval for
Mean
N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum
Rs 100-Rs 150 22 3.77 1.152 .246 3.26 4.28 2 5
Rs 151-Rs 200 6 4.67 .516 .211 4.12 5.21 4 5
Rs 201-Rs 250 2 3.00 1.414 1.000 -9.71 15.71 2 4
Total 30 3.90 1.125 .205 3.48 4.32 2 5

TABLE – 2
ANOVA

Interest in facility
Sum of
Squares df Mean Square F Sig.
Between Groups 5.503 2 2.752 2.381 .112
Within Groups 31.197 27 1.155
Total 36.700 29

26 | P a g e
 In Table 2, computed = 2.381 while the table value of = 3.35 at 5%
level of significance. Since computed F is less than tabulated F, accept H0.
Therefore interest in Kids’ care facility is independent of their desire to pay
the amount of charges on hourly basis.

The results of one-way ANOVA with Y as dependent variable and (X3)

education level as independent variable is given in Tables 3 & 4. The
following hypothesis is to be tested.

H0 : Education has no influence on the desire to avail kids’ care facility.

H1 : Education influences the desire to avail kids’ care facility.

TABLE – 3
Descriptives
Interest in facility
95% Confidence Interval for
Mean
N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum
Less than graduate 2 4.50 .707 .500 -1.85 10.85 4 5
Graduate 15 4.33 .976 .252 3.79 4.87 2 5
Post Graduate and above 13 3.31 1.109 .308 2.64 3.98 2 5
Total 30 3.90 1.125 .205 3.48 4.32 2 5

TABLE - 4
ANOVA

Interest in facility
Sum of
Squares df Mean Square F Sig.
Between Groups 8.097 2 4.049 3.822 .035
Within Groups 28.603 27 1.059
Total 36.700 29

In Table 4, computed = 3.822 while table value of = 3.35 at 5%

level of significance. Since computed is greater than tabulated ,
reject H0. Therefore interest in kids care facility is dependent on their
education level.

27 | P a g e
 In order to test, which education level respondents have more interest in
Kids’ care facility and which have the least, t test is used to examine various
set of hypotheses as stated below:

I H0 : μless than graduate = μ graduate (Interest in the facility is same for

respondent having education less than
graduate and graduate.)
H1 : μless than graduate > μ graduate
II H0 : μless than graduate = μ post graduate (Interest in the facility is same for
respondents having education less than
graduate and above post graduate.)
H1 : μless than graduate > μ post graduate
III H0 : μ graduate = μ post graduate (Interest in the facility is same for
graduate and post graduate.)
H1 : μ graduate > μpost graduate

Table – 5
Group Statistics

Std. Error
Education N Mean Std. Deviation Mean
Interest in facility Less than graduate 2 4.50 .707 .500
Graduate 15 4.33 .976 .252

Table – 6
Independent Samples Test

Levene's Test for

Equality of Variances t-test for Equality of Means
95% Confidence
Interval of the
Mean Std. Error Difference
F Sig. t df Sig. (2-tailed) Difference Difference Lower Upper
Interest in facility Equal variances
.638 .437 .231 15 .821 .167 .723 -1.374 1.708
assumed
Equal variances
.298 1.565 .801 .167 .560 -3.012 3.346
not assumed

To test hypothesis I, Tables 5 & 6 are used. The p value for one-tailed test for
the cases of equal variance and unequal variances is 0.82/2 = 0.41 and
0.80/2 = 0.4 respectively. Assuming the level of significance to be 5%, p
values are greater than the level of significance thereby implying that null
hypothesis cannot be rejected. Therefore there is no difference in the
interest in the facility for respondents with education less than graduation
and that of graduation.

28 | P a g e
 To test hypothesis II Tables 7 & 8 are used.

Table – 7

Group Statistics

Std. Error
Education N Mean Std. Deviation Mean
Interest in facility Less than graduate 2 4.50 .707 .500
Post Graduate and above 13 3.31 1.109 .308

Table – 8
Independent Samples Test

Levene's Test for

Equality of Variances t-test for Equality of Means
95% Confidence
Interval of the
Mean Std. Error Difference
F Sig. t df Sig. (2-tailed) Difference Difference Lower Upper
Interest in facility Equal variances
1.443 .251 1.448 13 .171 1.192 .823 -.586 2.971
assumed
Equal variances
not assumed 2.031 1.878 .188 1.192 .587 -1.497 3.882

The p values for one tailed tests are 0.171/2 = 0.085 and 0.188/2 = 0.094 for
the cases of equal and unequal variances respectively. Again the level of
significance is assumed to be 5%, these p values are greater than the level of
significance and therefore, we do not have enough evidence to reject the
null hypothesis. Therefore, there is no difference in the interest for centers
for respondents with qualifications less than graduate and those with post
graduate and above.

To test the hypothesis III, Tables 9 and 10 are used.

Table – 9
Group Statistics

Std. Error
Education N Mean Std. Deviation Mean
Interest in facility Graduate 15 4.33 .976 .252
Post Graduate and above 13 3.31 1.109 .308

29 | P a g e
 Table – 10
Independent Samples Test

Levene's Test for

Equality of Variances t-test for Equality of Means
95% Confidence
Interval of the
Mean Std. Error Difference
F Sig. t df Sig. (2-tailed) Difference Difference Lower Upper
Interest in facility Equal variances
assumed .569 .457 2.603 26 .015 1.026 .394 .216 1.835
Equal variances
2.579 24.173 .016 1.026 .398 .205 1.846
not assumed

For a one-tailed test the p values are given by 0.015/2 = 0.0075 and 0.016/2
= 0.008 for the cases of equal and unequal variances. Assuming a 5% level of
significance, the p values are less than the level of significance thereby
rejecting the null hypothesis. Therefore, the interest in the kids’ care facility
is more for respondents who are graduate than those having qualification of
post graduate and above.

The results of ANOVA table with Y as dependent variable and (X4) monthly
household income as independent variable are given in Tables 11 & 12.

Table – 11
Descriptives

Interest in facility
95% Confidence Interval for
Mean
N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum
Rs 30001-Rs 45000 12 3.58 .996 .288 2.95 4.22 2 5
> Rs 45000 18 4.11 1.183 .279 3.52 4.70 2 5
Total 30 3.90 1.125 .205 3.48 4.32 2 5

Table – 12
ANOVA

Interest in facility
Sum of
Squares df Mean Square F Sig.
Between Groups 2.006 1 2.006 1.619 .214
Within Groups 34.694 28 1.239
Total 36.700 29

Table 11 indicates that respondents are distributed in two income class i.e.
Rs.30,001 – Rs.45,000 and above Rs. 45,000.

The ANOVA results are given in Table 12. The hypothesis to be tested is
30 | P a g e
 H0 : Income level has no effect on the respondents’ interest in kids' care fac
H1 : Income level influences the respondent’s interest in Kids care facility.

In Table 12, computed = 1.619 while table value of = 4.20 at 5%

level of significance. Since computed F is less than tabulated F accept H0.
Therefore, income level has no effect on respondents’ interest in kids’ care
facility.

2) A two independent same t test will be used to test the following set of
hypotheses.

I H0 : μM = μF (Interest in kid’s care facility is same for males and females)

H1 : μM ≠ μF

II H0 : μB = μ0 (Interest in kid’s care facility is same whether both or one

spouse is working.)
H1 : μB ≠ μ 0

III H0 : μN = μJ (Interest in kid’s care facility is same for respondents

belonging to nuclear or joint family.)
H1 : μM ≠ μF

The results for the above hypotheses are presented in Table 13, 14 & 15.
Table – 13
Independent Samples Test

Levene's Test for

Equality of Variances t-test for Equality of Means
95% Confidence
Interval of the
Mean Std. Error Difference
F Sig. t df Sig. (2-tailed) Difference Difference Lower Upper
Interest in facility Equal variances
assumed .076 .785 -.160 28 .874 -.067 .418 -.923 .789
Equal variances
not assumed -.160 27.970 .874 -.067 .418 -.923 .789

Table – 14

31 | P a g e
 Independent Samples Test

Levene's Test for

Equality of Variances t-test for Equality of Means
95% Confidence
Interval of the
Mean Std. Error Difference
F Sig. t df Sig. (2-tailed) Difference Difference Lower Upper
Interest in facility Equal variances
assumed 2.183 .151 -.261 28 .796 -.111 .426 -.984 .762
Equal variances
not assumed -.270 26.217 .790 -.111 .412 -.958 .736

Table – 15
Independent Samples Test

Levene's Test for

Equality of Variances t-test for Equality of Means
95% Confidence
Interval of the
Mean Std. Error Difference
F Sig. t df Sig. (2-tailed) Difference Difference Lower Upper
Interest in facility Equal variances
4.041 .054 -.514 28 .611 -.214 .417 -1.069 .640
assumed
Equal variances
-.527 26.470 .602 -.214 .406 -1.049 .621
not assumed

It is seen that t statistic is insignificant in all the above cases as the p values
in these tables are above 0.05.

Therefore, the interest in the Kids’ care centre is independent of gender,

whether both or one spouse is working and whether respondents belong to
nuclear or joint families.

3. We are asked to divide the sample into two groups. One group is for those
who have given a score of 1 to 3 on the question on interest in the kids’ care
facility. This group we will name it as the one which is not interested in the
facility. The second group is for those who have given a score of either 4 or 5
to this question. This group will be of interested respondents and we would
name it the one who are interested in this facility.

We have to cross-tabulate the above created two groups with gender (X2),
spouse working (X5) and type of family (X6) separately. For the purpose of

32 | P a g e
interpretation our dependent variable will be the one indicating the interest
in the facility and the other three would be treated separately as
independent variables. As explained in Chapter 11 for the purpose of
interpretation, the percentages are to be computed in the direction of
independent variables.

The cross- tables are prepared and percentages computed in the directions
of independent variable with the help of SPSS software. Please remember
that interest in the facility has been divided into two groups – one showing
interest in the facility and the other one not interested in the facility. The
resulting cross-tables are presented below.

Table 16 gives the cross-tabulation between gender and the interest. It is

seen that out of 15 male respondents, 60% of them are interested in the
facility. There are 15 females in the sample, of which 73.3% are interested in
the facility. Therefore, it is seen that females are keener to go for this facility
as compared to the males. One probable reason could be that females have
to take care of the kids when they go out for shopping in a mall.

Table – 16
Interest Redefined * Sex Crosstabulation

Sex
Male Female Total
Interest Not interested in the Count 6 4 10
Redefined facility % within Sex 40.0% 26.7% 33.3%
Interested in the facility Count 9 11 20
% within Sex 60.0% 73.3% 66.7%
Total Count 15 15 30
% within Sex 100.0% 100.0% 100.0%

Table 17 presents the cross-tabulation between interest in the facility and

whether both spouse are working or one is working. The results indicate

33 | P a g e
that 61% of those respondents where one spouse is working is interested in
the facility, whereas 75% of those where both are working are interested in
availing of this facility. Therefore, the interest in the facility is more for those
couples where both of them are employed. This is but natural because going
to the mall is not only for shopping but also for outing and they would want
children to be taken care of by some reliable organization in the mall itself.

Table – 17
Interest Redefined * Both Working Crosstabulation

Both Working
Both one Total
Interest Not interested in the Count 3 7 10
Redefined facility % within Both Working 25.0% 38.9% 33.3%
Interested in the facility Count 9 11 20
% within Both Working 75.0% 61.1% 66.7%
Total Count 12 18 30
% within Both Working 100.0% 100.0% 100.0%

Table 18 below presents the cross-tables between interest in the facility and
family type. Here it is seen that of the respondents from nuclear families,
64.3% are interested in the facility, whereas among those with joint family
68.8% are interested in the facility. There does not seem to be much
difference in the two family structures but nevertheless both of them have
shown a high interest in the scheme.

Table – 18
Interest Redefined * Family Crosstabulation

Family
Nuclear
Family Joint Family Total
Interest Not interested in the Count 5 5 10
Redefined facility % within Family 35.7% 31.3% 33.3%
Interested in the facility Count 9 11 20
% within Family 64.3% 68.8% 66.7%
Total Count 14 16 30
% within Family 100.0% 100.0% 100.0%

34 | P a g e
4. Management Summary: The results indicate that the interest in kids’ care
facility is the same for gender, both or single spouse working and living in
single or joint family as indicated by t test. However, the cross-tabulation
results show that it is more for female; both spouse working and joint
families. Please note that no test of significance is conducted for cross-
tabulation, otherwise perhaps the results might have been in the same
direction. The statistical test on this is conducted in the chapter on non-
parametric tests. One thing which clearly comes from cross-table is that
about 67% of respondents are interested in kids’ care facility. Further, the
interest is independent of amount of charges, income of respondents.
However, education level has some effect. As the sample size for the study
is small, it may not be appropriate to make generalizations.

35 | P a g e
Case
Kumar Soft Drink Bottling Company

Kumar Soft Drink Company came into operation in 1984 and was operating in the
NCR of Delhi and in the states of Punjab and Haryana. The turnover of the company
was Rs 1.5crore in 2010 and it was growing at the rate of 10 per cent per annum. The
chairman of the company, Mr. Kumar, wanted to examine whether the flavour of the
soft drink and the price level had any impact upon the sales. He wanted this because
the results could have implications for changing the product mix if required. Three
types of flavours were considered, namely, pineapple, mango and orange. Further,
three level of prices were taken into consideration- Rs 10, Rs 12 and Rs 14. An
experiment was conducted by randomly choosing a sample of 18 stores where the
flavour of the soft drink and the price level were varied. The experiment period was
one month. The result of the experiment is shown in Table.

Table
Data for Select Variables
Store No. Sales (in Rs. Lakh) Flavour Price
1 5.5 1 1
2 4.2 2 1
3 3.7 3 1
4 3.6 1 2
5 2.9 2 2
6 2.5 3 2
7 2 1 3
8 1.9 2 3
9 2.8 3 3
10 5.6 1 1
11 4.3 2 1
12 5.4 3 1
13 4 1 2
14 3.8 2 2
15 3.2 3 2
16 2.6 1 3
17 2.8 2 3
18 2 3 3

36 | P a g e
Coding of Flavour: Pineapple = 1

Mango = 2

Orange = 3

Coding for Price: Rs. 10 /- = 1

Rs. 12 /- = 2

Rs. 13 /- = 3

Questions:

1. Is there any impact of the flavour or the price level independently upon the
sales? Conduct the test using a 5 per cent level of significance
2. Examine if there is any combined effect of the flavour and the price level
(interaction effect) on sales

37 | P a g e
Solution
Kumar Soft Drinks Bottling Co.

This is the case of factorial design. The chairman of the company is interested in
studying the effect of flavour and price on sales. He is also interested in examining
the interaction effect of these variables on sales. The dependent variable, sales, is in
ratio scale measurement, whereas the independent variables are flavor (pineapple,
mango and orange) and price (Rs.10, Rs.12 and Rs.14) in nominal scale format.
(Table 1)

TABLE – 1 : INDEPENDENT VARIABLES

Value Label N
Flavour 1 Pineapple 6
2 Mango 6
3 Orange 6
Price 1 Rs.10/- 6
2 Rs.12/- 6
3 Rs.14/- 6

The three hypotheses to be tested are:

I Flavour

H0 : Average sale for all flavours is same.

H1 : Averages sale for all flavours is not same.

II Price

H0 : Average sale for all prices is same.

H1 : Average sale for all prices is not same.

38 | P a g e
III Interaction

H0 : The average sale for all interactions is same.

H1 : The average sale for all interactions is not same.

The ANOVA results are reported in Table 2 below.

TABLE – 2: ANOVA RESULTS

Type II Sum of
Source Df Mean Square F Sig.
Squares
Corrected 20.748 8 2.593 7.554 0.003
Model
Intercept 219.102 1 219.102 638.162 0.000
Flavour 1.408 2 0.704 2.050 0.185
Price 17.981 2 8.991 26.186 0.000
Flavour * Price 1.359 4 0.340 0.989 0.461
Error 3.090 9 0.343
Total 242.940 18
Corrected Total 23.838 17

The computed F statistic corresponding to flavour, price and interaction are

= 2.05, = 26.186, = 0.969 respectively. Assuming α = 0.05, The table
values are are = 4.26, = 3.63. We find that only in case of price, computed F
is greater than tabulated F. Therefore, null hypothesis is rejected in case of price
only. Therefore price is the only significant variable influencing sales. The effect of
flavour and interactions is insignificant.

The results indicate that the sales of soft drink are price sensitive. The flavour and its
interaction with the price have no effect on sales. Therefore, the company should
concentrate on its pricing strategy. An exercise on the effect of varying price levels
on sales should also be carried out.

39 | P a g e