Consumer Preference Analysis
of three brands of anti-aging foundations
Conclusions
Basic Statistics
In 85 or more samples from every 100 people drawn from the same population of th is sample of the 60 people, it is expected to find the same mean scores for Brand A and Brand B as found in this sample (0.12 < 0.15). The same can be expected for the pair for Brand A and Brand C (0.02 < 0.15) as well. However, in 85 or more samples from every 100 people drawn from the same population of the 60 people, we would not expect to find the same mean scores for Brand B and Brand C as found in this sample (0.16 > 0.15).
While brand preference toward Clinique between up-movers and down-movers were almost the same, ad preference were significantly higher in those who showed increased favorability. Therefore, we could conclude that respondents liked the ad for Clinique but that did not necessarily lead to brand favorability.
In 85 or more samples from every 100 people drawn from the same population of th is sample of the 6 0 people, it is expected to find the same mean Ad Index S cores for up-movers and down-movers as found in this sample (0. 01 < 0.15). However, we cannot expect that for Brand Index Scores for these groups.
From the chi-square significance test, we can conclude that the ad for Clinique should be improved since the salience of dislike in Clinique among those who had negative attitudes toward the brand before are more likely to grow stronger, while it did not have any effect on those with above-median brand preferences and improved attitudes toward the brand.
The correlation between Lancome and Clinique is very weak ( correlation coefficient = 0.02, significance = 0.44). Therefore, we cannot say that favorability toward Lancome and Clinique have any effect on each other.
Lancome and Clinique showed decreased favorability by about 40%. Thus, Shiseido's advertisement seems to have had more positive influence on the viewers toward the product than the others.
Shiseido showed more women not changing their attitudes toward the product and less negative outcomes. Therefore, we can conclude that the ad for Shiseido did not necessarily contribute to more positive attitudes of female respondents, but had the most influence preventing favorability diminish.
Regression Analysis
The regression analysis showed that there is very little relationship between change in brand favorability and ad favorability. Clinique had the closest relationship between all factors in question, meaning that these factors did relate to how they favored the advertisement shown.
The multiple regression analysis for Lancome shows a low coefficient of multiple determinants (R Square) of 2.6%, indicating that there is very little relationship between how respondents rated the Lancome brand on the ten Likert items and their favorability of the Lancome ad derived from their score changes before and after exposure to the ads. The F-ration (0.13) indicates that in 85 or more samples of every 100 samples drawn from the same population as this sample of 60 people, we would not expect to find the same multiple regression coefficient for Lancome.
Shiseido shows an even lower coefficient of multiple determinants (R Square) than Lancome with 1.1%, indicating that there is very little relationship between brand favorability and ad favorability. The F-ration (0.05) indicates that in 85 or more samples of every 100 samples drawn from the same population as this sample, we would not expect to find the same multiple regression coefficient for Shiseido. The only attribute that respondents thought important of Shiseido was brand credibility. Unstandardized coefficients (b) suggest that the more respondents trust Shiseido, the more they favored the ad.
Clinique shows a coefficient of multiple determinants (R Square) of 24.4%, showing some relationship between brand and ad favorability. in 85 or more samples of every 100 samples drawn from the same population as this sample, we would expect to find the same multiple regression coefficient for Clinique.
Discriminant Analysis
From the standardized discriminant coefficients, respondents considered the attributes that ¡°Clinique is a good anti-aging foundation¡± and ¡°would prefer Clinique over other brands¡± important. The unstandardized discriminant coefficients suggest the relationship between the brand attributes and ad favorability as follows:
• The more the respondents considered Clinique a good anti-aging foundation, the more they liked the ad.
• The more the respondents preferred Clinique over other brands, the less they liked the ad.
The Chi-square value of 8.21 indicates that there is little relationship between those with changed favorabilities and brand attributes. In 85 or more samples of every 100 samples drawn from the same population as this sample of 60 people, we would not expect to find the relationship and Chi-square score for Clinique (0.30 ¡Ã 0.15).
The group centroid for those with increased favorability was -0.6 and that for those with decreased favorability was 0.4. However, we cannot project the same results of this sample to the population since the Wilks' Lambda score (0.78) is not statistically significant. In other words, in 85 or more samples of every 100 samples drawn from the same population as this sample of 60 people, we would not expect to find the same group centroids for Clinique (0.30¡Ã0.15).
The outcome of this study was not successful which is made clear by the low percentage of respondents that were accurately classified. This is a mere 60 percent of the respondents which are not much better than the 50 percent that could be achieved by random selection. The scores show that the percentage of persons studied was not significant enough to make the data collected successful.
ANOVA/MANOVA
The sample shows obvious differences in favorability toward the brand by age, indicating that older women were more skeptical that the product would make them look younger. The difference of preference by age was more prominent especially for those who had changed favorability toward the brand after exposure to the ad.
We can conclude that in 85 or more samples of every 100 samples drawn from the same population as this sample, we would expect to find the same mean scores by age. However, the results for change scores and their relationship to age is not statistically significant; thus, in 85 or more samples of every 100 samples drawn from the same population as this sample, we would not expect to find the same means for neither change scores nor their relationship with age.
The relationship between the change in brand preference and age was statistically significant. Therefore, in 85 or more samples of every 100 samples drawn from the same population as this sample, we would expect to find the same mean in change scores by age. However, we cannot project the same results to the population for each independent variables; change of favorability and age.
When factoring in both age and the movement upwards and downwards in favor of Clinique, there is a positive interaction between the two making it an important factor. This relays the idea that when considering whether or not Clinique is a good brand, respondents' age played a part in the way they favored the ad and moved up or down in this favor.
Factor Analysis
The Factor Analysis revealed those Likert items that were evaluative in nature and related to the most unambiguous Likert item of each group: the ¡°good¡± item. These items were then taken to compute an attitude score for each respective test brand. These attitude scores did reveal that in this sample, Lancome is the brand liked the most followed by Shiseido and Clinique. Lancome is favored more than the other brands in this study and in the population as is the advertising for this brand.
Based on the results from a paired t-test to observe the significance of the difference between the three mean attitude scores, we can conclude that that in 85 or more samples of every 100 samples drawn from the same population as this sample, we would expect to find the same difference between mean scores for pairs Lancome-Shiseido and Lancome-Clinique . However, in 85 or more samples of every 100 samples drawn from the same population as this sample, we would not expect to find the same mean score difference s for pair Shiseido-Clinique.