1.
What can we conclude regarding degree of perceived credibility
between car make of franchises? (25 marks)
Since we are
comparing mean values of more than 2 samples ANOVA is the appropriate
test. Although more than one mean ANOVA
is a robust technique so do not need to test for equality of variance (unless
we have to carry out multiple comparison tests).
H0 =
There is no significant difference in the mean credibility value between the
five car make franchises
HA = At
least two car make franchises have significantly different mean credibility
values
α = 0.05
From the ANOVA
table Sig. = 0.01 which is < α and consequently reject the null hypothesis
and conclude that the mean credibility values of at least two of the car make
franchises are significantly different.

In order to
isolate the differences we need to carry out a multiple comparison test. To determine the appropriate test we first
need to test for equality of variance.
H0 =
There is no significant difference in the variances of credibility value
between the five car make franchises
HA = At
least two car make franchises have significantly different variances of
credibility values
α = 0.05
From the following
table Sig. = 0.136 which is > α and consequently retain the null hypothesis
and conclude that the variances are equal.

Given equality of
variances we perform Scheffe’s test. The
solution presented below indicates the existence of two groups:
![]() |
Group A = Ford,
Nissan, Citroen and Honda
Group B = Citroen,
Honda and Audi
This implies that
the location/position of Citroen and Honda is unclear/overlapping. However, looking at the Sig. at the bottom of
the table we observe that the one for Group B is considerably lower than the
corresponding for Group A indicating greater degree of heterogeneity. Consequently, we could allocated the two
overlapping makes to Group A and conclude that there is significantly greater
degree of perceived credibility in franchises of Audi cars when compared to
Ford, Nissan, Citroen and Honda.
Comments on the mean values and comparing them to the scale of 1-to-7
will gain extra marks.
Get Free Quote!
318 Experts Online