Conclusions
Basic Statistics
Paired Sample T-tests
The highest Brand Index Score with a mean of 31.1 belongs to Skyy, while the lowest score was for Ketel One with a mean of 29.0. Out of the 59 respondents, Skyy was determined to be the more favorable brand. The standard deviation scores were moderate, demonstrating that for each brand there was a slight deviation from the mean. Paired sample t-tests showed that the results for Skyy and Ketel One and Skyy and Level were significant and could be projected to the larger population.
Between-Groups t-tests
There were 26 respondents who felt more favorable towards Skyy and 20 who felt more negatively toward the brand after viewing the print advertisement. However, only the Ad Index Score for Skyy was significant, while the Brand Index Score was not.
Chi-Squared Significance Test
After viewing the ad for Skyy, 13 felt more positively towards the brand, five felt the same and 11 felt more negatively towards the brand. Of the respondents who liked the brand more, 23.6% were above the median. Those who reported no change in brand likeability were 9.1% above the median and those who liked the brand less were 20.0% above the median. However, there is no significance found in this set of data, so the results cannot be projected to the larger population.
Pre-Post Exposure Change Score Frequencies
The advertisements for Skyy and Level had the greatest positive influence, while Ketel One had the least. For Skyy and Level, 26 people had positive change scores after viewing their advertisements. After viewing the advertisement for Ketel One, only 15 people had positive change scores.
Brand Index Score Comparison
When comparing the Brand Index Scores for Skyy and Ketel One, over half of the respondents (59.3%) favored Skyy over Ketel One.
Simple Correlation Coefficients
There is a low positive correlation for the relationship between Skyy and Ketel One. However, this only concentrated on those respondents whose first liquor of choice is vodka. This correlation, although positive, is also not significant.
Regression Analysis
Skyy
The multiple regression analysis for Skyy shows a coefficient of multiple determination (R²) of 9.7%. This indicates a weak relationship. Only 9.7% of how much respondents liked the Skyy ad is explained by how much they like the brand. The coefficient of multiple correlations is 0.3 and therefore the data has a low correlation. The data is not statistically significant. None of the Likert items is significant in this data.
Ketel One
The multiple regression analysis for Ketel One shows a coefficient of multiple determination (R²) of 27.9%. This is higher than Skyy but still indicates a weak relationship. Only 27.9% of how much respondents liked the Ketel One ad is explained by how much they like the brand. The coefficient of multiple correlations is 0.5 and therefore the data has a moderate correlation. The data is statistically significant so it can be projected to the larger population. In 85 or more samples, out of 100, drawn from the same population as this sample, it would be expected that the correlation would be about the same. In addition, two of the 10 Likert items are significant in this analysis. They are “Ketel One is an upscale vodka” and “Ketel One is a strong vodka.”
Level
The multiple regression analysis for Level shows a coefficient of multiple determination (R²) of 21.5%. This indicates a weak relationship. Only 21.5% of how much respondents liked the Level ad is explained by how much they like the brand. The coefficient of multiple correlations is 0.5 and therefore the data has a moderate correlation. The data is not statistically significant so it cannot be projected to the larger population. In addition, two of the 10 Likert items are significant in this analysis. They are “I like the taste of Level” and “I would order Level at a bar.”
Discriminant Analysis
When comparing the means and standard deviations for all ten Likert items, differentiated by group membership, the mean scores for Up-Movers and Down-Movers are similar to each other. This indicates that there is no significant correlation between how respondents perceived Skyy and whether or not they liked the ad for the brand. In other words, we cannot accurately predict the Up-Movers and Down-Movers based on the Likert item responses. Also, the Wilks’ Lambda and Chi-Squared values are not significant and, therefore, the results cannot be projected to a larger population. Because the Wilks’ Lambda value is not significant, then the Group Centroids are also not significant. Discriminant function coefficient analyses were conducted to determine the relative importance of variables in explaining differences between groups. Three attributes were found to be important. They are "I cannot tell the difference between Skyy and other brands," "Skyy is the same as other brands," and "I would give Skyy to my friends." These variables best account for the differences in pre-post exposure change scores. Overall, 65.2% of the cases were predicted correctly. This indicates a moderate to high rate of accuracy.
ANOVA/MANOVA
ANOVA
There is no significant difference between the mean scores for male or females. Also, based on the F-ratios, there were no statistically significant relationships and therefore these relationships cannot be projected to the larger population.
MANOVA
Again, there is no significant difference between the mean scores for male or females, as well as between the up, same and down movers. Also, based on the F-ratios, there were no statistically significant relationships and therefore these relationships cannot be projected to the larger population.
Factor Analysis
Skyy
Three of the ten factors have Eigenvalues greater than or equal to one and are therefore important in this analysis because they explain the variance of at least one Likert item. Together, these three factors account for 69.7% of the variance, leaving 30.3% unexplained. Seven of the 10 attributes are loaded on Factor I. Factor I is used in this analysis because the evaluative statement, "Skyy is a good vodka.” This indicates that these attributes are related to each other in some way independent of the other variables.
Ketel One
Three of the ten factors have Eigenvalues greater than or equal to one and are therefore important in this analysis because they explain the variance of at least one Likert item. Together, these three factors account for 72.8% of the variance, leaving 27.2% unexplained. Seven of the 10 attributes are loaded on Factor I. This indicates that these attributes are related to each other in some way independent of the other variables.
Level
Three of the ten factors have Eigenvalues greater than or equal to one and are therefore important in this analysis because they explain the variance of at least one Likert item. Together, these three factors account for 76.5% of the variance, leaving 23.5% unexplained. Nine of the 10 attributes are loaded on Factor I. This indicates that these attributes are related to each other in some way independent of the other variables.
Cluster Analysis
When compared to Cluster 1, Cluster 2 exhibits a higher mean score for every brand attribute. This reveals that Cluster 2 rated, and liked, the brand (Skyy) better on all evaluative levels. Each of the Likert items were found to be significant, except for “Skyy vodka is no different from other vodka brands” and “Skyy vodka is the same as other brands of vodka.” Because these results are statistically significant, they can be used to describe a larger population. In 85 (or more) of 100 samples drawn from the same population as this sample, we would expect to find the same results. When divided into three groups, the three clusters all had similar mean scores. Again, each of the Likert items were found to be significant, except for “Skyy vodka is no different from other vodka brands” and “Skyy vodka is the same as other brands of vodka.”
Back to Table of Contents