Consumer Preference Analysis |
CONCLUSIONS
These are substantive interpretations of statistical findings:
Basic Statistics:
Paired t-test: Dove is the highest brand index score of 32.7, meaning it is the most preferred body wash brand among the three brands. Olay is the second most preferred followed by Softsoap. In 85 or more samples out of every 100 samples drawn from the same population as this sample of 79 respondents, it would be expected that about the same result for Dove/Softsoap, Softsoap/Olay and Olay/Dove can be projected to the population.
Between Groups t-tests: The means for the brand index score and ad index score for Olay body wash are not statistically significant. In 85 or more samples out of 100 samples drawn from the same population as this sample of 79 respondents, it would be expected that the mean brand index score and ad index score for Olay would not be about the same as this sample. Overall, there are more respondents whose opinions about Olay body wash have improved than respondents whose opinions have deteriorated after viewing the ads.
Chi-Squared significance test for Olay: Of the 32 respondents who were more favorable toward Olay after viewing the advertisement, just over half were above the median of 30.0. There were 19 respondents who had no change towards Olay body wash, 60% of them were below the median. 71 respondents were less favorable after seeing the advertisement, but 48% were above the median. The Chi-squared show that the sample % cannot be projected to the population because the Chi-squared value of 1.28 is not statistically significant.
Pre to Post Exposure Change Scores by brands: The most favorable pre-to-post exposure change score was the Olay print ad. 36 respondents moved up in their impressions of Olay, while Dove print ad had 35 respondents lower their impressions. This shows that Olay had the most effective print advertisement since more respondents were more favorable towards Olay post exposure.
Brand Index Score Frequency Count for Olay vs. Dove: The result of the frequency test indicates that out of the 79 respondents, just 26 people (32.9%), ranked Olay higher than Dove.
Simple Correlation Coefficient for Olay vs. Dove: There is a weak positive correlation of coefficient between the Olay and Dove Brand Index Scores. In 85 or more out of every 100 samples drawn from the same population as this sample, we cannot project this result to the population.
According to the brand index score, respondents are likely to feel differently about each brand. The results indicate that when a respondent has a favorable perception towards Olay, it does not necessarily mean they have the same perception towards Dove.
Simple Correlation Coefficient for Olay vs. Dove (Female): There is no positive correlation of .0 between the Olay and Dove brand index scores among female respondents. In 85 or more samples out of 100 samples drawn from the same population as this sample of 52 respondents, we cannot project this result to the population. According to the brand index score, respondents are likely to feel differently about each brand. The results indicate that when a respondent has a favorable perception towards Olay, it does not necessarily mean she will also have the same perception towards Dove.
Regression Analysis:
In Dove's case, R was found to be .4 which indicates a positive and direct relationship between respondents’ feeling toward Dove and impression of its ad, however the correlation is weak in strength. The R values were slightly lower for both Olay (.3) and Softsoap (.3), suggesting a positive and even weaker correlation between respondents’ feeling for the brand and their respective ads.
The brand attributes that explain most of the variance accounted for “Olay makes skin glow,” “I don’t think Olay’s image fits my style,” and “I would not buy Olay body wash;” all with statistically significant impact. This means that in 85 or more samples out of every 100 samples drawn from the same population as this sample, it would be expected that the correlation for these attributes would be about what they are in this sample. As for Softsoap, three of the ten brand attributes: “Softsoap makes skin feel softer,” “I don’t think Softsoap’s image fits my style,” and “I would recommend Softsoap body wash to other people” are significant independent variables that explain the variance in pre-to-post ad exposure change scores. In Dove case, the significant brand attributes are “Dove is a good body wash,” “I prefer Dove over other body wash brands,” and “Dove makes me feel cleaner than other body washes.”
Discriminant Analysis:
Three variables are shown to be important in relation to the movement of post-exposure attitude toward the Dove brand. These variables are "good body wash", "prefer this brand ", and "make skin glow."The distance between up-mover and same-down mover groups' centroids was found to be .7, reflecting a moderate amount of discrimination.
The Wilks' Lambda is .90 and, based on the Chi-squared score of 4.98, it is significant. The result indicate that in 85 or more samples out of every 100 samples drawn from the same population as this sample, we would expect to find the same results. The difference between the Group Centroids’ scores for the up movers and down movers can be projected to the larger population.
67.9% of the total 79 respondents in this survey are correctly classified as up-movers or same-down-movers basing on the ten brand-related Likert items. Besides, in 85 or more samples out of every 100 samples drawn from the same population as this sample, it can be said that the classification accuracy will be about what it is for this sample.
ANOVA/MANOVA:
The majority of down-movers whose attitude toward Dove body wash decrease after viewing ad were female. Based on the critical F-ratio, there were no statistically significant relationships found and these relationships cannot be projected to the larger population.
Only the relationship between the “attitude change after seeing the Dove ad” and the mean score of the Likert item “Dove is a good body wash” as statistically significant. In 85 or more samples out of every 100 samples drawn from the same population as this sample, it is expected that the relationship between “Dove ad post-exposure attitude change” and the “mean score of Dove is a good body wash” will be about what it is for this sample.
Based on the Wilks’ Lambda to test for the significance of pre-post ad exposure change score, gender, and the relationship/interaction between change score and gender, we cannot project the mean scores of the same magnitude among pre-post ad exposure change score groups (up/same/down) in this sample to the entire population. We can also not project the mean scores based on gender in this sample to the entire population. Lastly, we cannot project the mean scores of the same magnitude among pre-post ad exposure change score/gender in this sample to the entire population.
Factor Analysis:
Olay
Only two factors were found to be significant and accounted for a total of 58.0% of Likert item variance for Olay. The Varimax Rotated Matrix was selected for Olay because it contained only one ambiguous row. In this matrix, Olay loaded highly on Factor I, representing a high correlation between Olay and Factor I. Since the Likert item of “good body wash” is evaluative in nature as an attitude item, all the other significant items loading on the same factor as the item of “good body wash” are considered related to it, and thus are selected to calculate the “brand attitude score.” In this case, the item of “good body wash” loaded on Factor I, and the other five important Likert items that loaded on Factor I were “prefer this brand,” “makes skin moisturizing,” “first brand to think of,” “Would not buy,” and “would recommend.”
Softsoap
62.8% of Likert item variance for Softsoap can be explained by the first three factors. The Varimax Rotated Matrix was used because it had one ambiguous row, and the loadings were slightly different in this row comparing to the same loadings in the ambiguous row in Factor Matrix. Similar to Olay brand, Dove also loaded highly on Factor I, representing a high correlation between Softsoap and Factor I. The item of “good body wash” loaded on Factor I, and the other five important Likert items loading on Factor I were completely the same as Olay brand. These items in Factor I established a relationship independent from the other factors that affect participants' attitudes toward Softsoap.
Dove
Only two factors were found to be significant and accounted for a total of 61.7% of Likert item variance for Dove. The Varimax Rotated Matrix was selected for Dove because it contained only one ambiguous row. In this matrix, Dove loaded slightly more on Factor I, representing a lighter correlation between Dove and Factor I comparing to Olay and Softsoap. The item of “goodbody wash” loaded on Factor I, and the other four important Likert items that loaded on Factor I were “prefer this brand,” “makes skin moisturizing”, “first brand to think of,” and “makes skin fell cleaner.”
Paired t-test for Brand Attitude Score
The mean of brand attitude scores showed that Olay and Softsoap had slightly higher scores (3.4) than Dove, meaning that respondents in this study liked Olay and Softsoap more than Dove. Besides, the t-ratios revealed that the pairs of Olay versus Softsoap, and Softsoap versus Dove were significant. That is, in 85 or more samples out of every 100 samples drawn from the same population as this sample, it can be expected that the comparison between mean brand attitude score for the pairs of Olay versus Softsoap, and Softsoap versus Dove will be about what they are for this sample.
Cluster Analysis:
The F-ratio for Two-cluster Group indicate that all the mean scores are significant meaning that in 85 or more samples out of every 100 samples drawn from the same population as this sample of 79 people, it would be expected that the mean scores would be the same magnitude as in this sample.
The F-ratio for Three-cluster Group indicate that all the mean scores are significant meaning that in 85 or more samples out of every 100 samples drawn from the same population as this sample of 79 people, it would be expected that the mean scores would be the same magnitude as in this sample.
The chi-squared test shows that both the percentages in the Two-cluster Group and Three-cluster Group are not significant meaning. So it is not expected to see the same distribution when 85 or more samples out of every 100 samples are drawn from the same population as this sample. Therefore, we cannot project the results for this sample to the population.