Cluster Analysis

The data were subjected to a cluster analysis to determine if respondents could be grouped into clusters based on their responses to the ten brand attitude Likert items for Axe. The clusters were then analyzed to determine if the clusters were differentiated on the basis of whether respondents took showers or baths.

Axe…	Two-means cluster analysis
Axe…	Cluster 1	Cluster 2
*Member count*	34	24
... is a good brand of body wash.	2.9 (0.7)	4 (0.4)
... is preferred over other body washes.	2.5 (0.7)	3.7 (0.8)
... gets me cleaner than other body washes	2.6 (0.7)	3.2 (0.7)
... smells nicer than other body washes.	2.7 (0.9)	4.1 (0.7)
...isn’t recommended	2.7 (0.9)	3.5 (1.1)
... is a good value.	2.9 (0.5)	3.2 (1.1)
...is healthy.	2.7 (0.7)	3.5(0.7)
... makes me feel good.	2.6 (0.8)	3.3 (1.0)
...is bad for my skin.	3.1 (0.5)	3.9 (0.7)
...is a brand I would purchase	2.3 (0.8)	3.9 (0.6)
Mean (Standard deviation)

In the two-means cluster analysis, the mean Likert score is higher in cluster 2 than cluster 1 for all attributes.
Table 1. Summarized ANOVA Table

Axe	F-ratio
... is a good brand of body wash.	51.69*
... is preferred over other body washes.	33.34*
... gets me cleaner than other body washes	10.75*
... smells nicer than other body washes.	61.70*
...isn’t recommended	9.36*
... is a good value.	2.96*
...is healthy.	14.41*
... makes me feel good.	9.20*
...is bad for my skin.	32.72*
...is a brand I would purchase	77.88*

*p < .15

	Bath	Shower
Cluster 1 Row % Column % Total %	6 17.6% 60.0% 10.3%	28 82.4% 58.3% 48.3%
Cluster 2 Row % Column % Total %	4 16.7% 40.0% 6.9%	20 83.3% 41.7% 34.5%
	Chi Square = .01	* p ≤ 0.15

Cluster 2 has a higher perception of Axe on all of the brand Likert items The members of cluster I and II are equally likely to contain people who prefer showers over baths. The percentages are in line with the rest of the population.

In 85 samples out of 100 samples drawn from the same population as this sample, the cross-tabulation displayed above would be about what it is here. The results of this test cannot be projected to the population.