The normality of the data was tested through the Kolmogorov–
Smirnov and Shapiro–Wilk tests. The p-values for all variables
and tests were extremely small, indicating that the variables are
not normally distributed (See Table II) and, therefore,
parametric statistical tests were excluded to test the hypotheses.
The non-parametric Kruskal–Wallis one way analysis of variance
by ranks test was, therefore, employed to enable differences in
perceptions between the different advertising images to be
explored. The means and standard deviations of the brand image
dimensions are shown in Table III.
As can be seen from Table IV, in relation to
non-product-related brand attributes, no significant differences
were found in perceptions of brand personality or brand user
imagery across the different advertising images (p 0.05). Thus,
in relation to RQ1a and RQ1b, we can conclude that the size of
the model used in the advertising image does not significantly
impact perceptions of the brand users’ imagery, or of the brand
personality. Brand price perceptions (RQ1c) displayed some
differences. Pairwise comparisons (Table V) of advertising
images, found that image A (underweight model), and B (slender
model) were perceived to represent a higher price point than
image C (normal size) (p 0.016 and 0.061, respectively). No
significant differences were found with respect to other image
pairings.
With respect to RQ2 (Table IV), differences were found in
perceptions relating to the quality and designer positioning of the
brand. Pairwise comparisons (see Table V for mean ranks of
significant variables) suggest that advertisement C (normal size)
was found to have a lower brand quality perception than
advertisement A (underweight model) (p 0.059), although this
difference was only of marginal significance. In addition,
advertisement A (underweight model) was significantly more
likely to be believed to symbolise a designer brand in comparison
with advertisement C (normal size) (p 0.049). No other
pairings showed significant differences. In relation to the
statement that the brand represented a status symbol, no
significant differences were found across the different advertising
images.
To explore research questions RQ3a-RQ3d, the analysis was
rerun separating the data into five age groups, broadly speaking
Ta