|
|
Analysis:
After collection of the data, many different analyses were performed using SPSS software. The results of the ten analyses can be found on this page. Click the links below to see the various analyses: Paired t-tests
Correlated t-tests were conducted on all three skin lotion brands (Aveeno, Keri, & Vaseline), comparing the mean brand index scores. Brand index scores were derived from the summation of input given by subjects in response to Likert items.
*p <= .15 Due to the level of significance found in only the Aveeno vs. Keri brands and the Keri vs. Vaseline brands, it can be concluded that in 85 or more samples out of every 100 taken from the same population as this one, their brand index scores can be projected to the population. We are unable to determine whether or not 85 out of 100 samples drawn from the same population of this sample would result in the same magnitude of difference in mean scores when comparing Aveeno and Vaseline. Therefore, we cannot project the results on the population. From this analysis, we can infer according to the above charts that subjects had the most positive feelings about Aveeno, followed closely by Vaseline. Subjects had the least possible feelings toward Keri. The t-ratio between Keri and Aveeno, and Keri and Vaseline is the same.
The following table shows those respondents whose perceptions of Aveeno moved up or down after viewing the advertisements. Analysis shows only respondents who moved up or down.
*p <= .15 Due to lack of statistical significance, in 85 or more samples out of every 100 drawn from the same population, we cannot expect the data to reflect the results shown above. The following table shows those respondents whose perceptions of Aveeno moved up or down after viewing the advertisements. Analysis shows only those respondents who moved up or down.
*p <= .15 In 85 or more samples out of 100 drawn from the same population of this sample, we would expect respondents who moved up or down in their perception of the advertisement to have the same mean scores as those above. From this analysis, we can infer that more people will have a positive view of the brand after viewing the ad than a negative view, creating a positive relation to the brand after viewing. More people like the brand after viewing the ad.
The table below represents the correspondence between the movement of the respondents’ Brand Index Score for Aveeno and the median Brand Index Score for Aveeno. The median Brand Index Score for Aveeno is 35. The sample size is 61.
Chi-Square = .876* *p <= .15 0 cells have expected count less than 5. The minimum expected count is 6.39. Due to the level of significance in 85 or more samples of 100 samples taken from the same population as this one, we would be able to project a cross tab distribution as this sample for Aveeno. The median brand index score for Aveeno is 35. From this sample, there were 14 people or 23% of the population whose pre and post test measurements did not change and whose brand index scores were below the median. There were also 14 people or 23% of the population whose pre and posttest measurements did not change and whose brand index scores were above the median. Those who favored the Aveeno brand more after viewing the print ad had 11 people, or 18% of the population, who rated the brand above the median and 9, or 14.8% of the population, who still rated below the median. Those who favored the brand less after viewing still had 5 (8.2%) who rated the brand above the median and 8 (13.1%) below. From this analysis, we can infer that people who had a negative perception of the brand before viewing the ad would dislike the brand even more after viewing it. Likewise, those people with a positive perception of the brand before the ad had a stronger affinity to the brand after viewing it.
Frequency tests were conducted on all three brands to show pre-to-post exposure on the constant-sum scale.
Aveeno had the most individuals moving up in favorability (22 individuals or 34.0% of the population) in pre-to-post exposure and the fewest moving down (14 individuals or 22% of the population. Vaseline had the least moving up (13 individuals or 20% of the population), the most individuals staying the same (29 or 45.0%) and the most moving down (22 individuals or 34%). From this analysis, we can infer that for the majority of the sample, the ads had no effect on brand perception, as they did not move up or down. However, we can also infer that the Aveeno ad was more effective, since more people moved up for it than either of the other two. Frequency tests were then conducted to compare Aveeno’s and Keri’s brand index scores.
The results of the sample indicate that 27 individuals (42.2%) rated Aveeno higher than Keri Lotion and 37 individuals (57.8%) rated Aveeno less than or equal to Keri. From this analysis, we can infer that the Aveeno ad had a larger positive impact on the sample than the Keri ad did. Therefore, the Aveeno ad was generally more effective.
Simple Correlation Coefficient: A correlation test was conducted between Aveeno’s and Keri’s brand index scores to determine if any relationship exists.
Sample size = 64 *p <= .15 The correlation between brand index scores for Aveeno and Keri in this sample is quite small. It is expected that in 85 or more samples out of every 100 drawn from the same population as this sample, the correlation between brand preferences of Aveeno and Keri would be about the same. From this analysis, we can see that in fact, a relationship does exist between the Aveeno and Keri brands in this sample. Although it is small, the relationship is still significant to have and affect on the two competing brands.
Modified Simple Correlation Coefficient: A separate correlation test was conducted between Aveeno’s and Keri’s brand index scores to determine if any relationship exists. This time; however, only respondents who had at least some college were selected from the sample for this test.
Sample size = 60 *p <= .15 The correlation between brand index scores for Aveeno and Keri in this sample of respondents with at least some college is still small. It is expected that in 85 or more samples out of every 100 drawn from the same population as this sample, the correlation between brand preferences of Aveeno and Keri would be about the same. Despite removing those respondents who only had a high school education, we can still conclude that a relationship does exist between the Aveeno and Keri brands in this sample. Although it is small, the relationship is still significant to have and affect on the two competing brands.
This section contains a regression analysis of copy-testing research on print ads for three popular brands of moisturizing lotion (Aveeno, Keri, and Vaseline), where the dependent variable is the pre-post ad exposure change score and the independent variables are the responses for the ten Likert items regarding specific brand attributes. An on-line questionnaire was constructed and sent via e-mail, returning a total of 64 completed surveys. Data was imported into SPSS for the following analysis:
*p <= .15
*p <= .15; b= unstandardized coefficient; Beta = Standardized coefficient
Aveeno:The multiple regression analysis for Aveeno reveals a low coefficient of multiple determinants (R 2), 14.2%, indicating that there is little to no relationship between how respondents rated the Aveeno brand on the ten Likert items regarding brand attributes and their favorability of the Aveeno ad as indicated by their pre-post ad exposure change score. In other words, the Likert items account for only 14.2% of the variance in the pre-post ad exposure change scores. The standard error of estimate (se) is high at 2.7, indicating respondents were an average of 2.7 units away from the regression line. Given the statistically significant F ratio of .88, in 85 or more samples of every 100 samples drawn from the same population as this sample, we would expect to find a multiple regression coefficient of the same magnitude. The unstandardized coefficients (b) suggest the relationship between each Aveeno brand attribute (Likert item) and how respondents favored the Aveeno print ad. The following illustrates how to interpret the unstandardized coefficients (b): The more respondents think Aveeno is a good brand of moisturizing lotion, the more they like the ad. The more respondents think Aveeno is too oily, the more they like the ad. The more respondents think Aveeno will keep their skip smooth and soft, the less they like the ad. The more respondents think Aveeno will protect their skin, the less they like the ad. The more respondents think Aveeno is too expensive, the more they like the ad. The more respondents think Aveeno will replenish their skin, the more they like the ad. The more respondents think Aveeno is a trusted brand, the less they like the ad. The more respondents think Aveeno is not a hip brand, the less they like the ad. The more respondents think Aveeno helps fight dryness in skin, the more they like the ad. The more respondents think Aveeno does not contain as much moisture as other brands, the more they like the ad.
The multiple regression equation is written as follows: Pre-Post Ad Exposure Change Score for Aveeno = .3 + .2 (good) + 0.8 (oily) – 0.1 (smooth) – 0.5 (protect) + .4 (expensive) + 0.3 (replenish skin) – 1.2 (trust) - 0.2 (hip) + 0.3 (fights dryness) + 0.4 (moisture) Only two of the ten brand attributes can be considered important, meaning they best explain the variance in pre-post ad exposure change scores relative to the other brand attributes (Is too oily, and Is a trusted brand). In 85 or more of 100 samples drawn from the same population as this sample, we would expect all the Aveeno brand attributes except for “is a trusted brand”, as well as the constant term (a), to have the same magnitude of impact.
Keri:The multiple regression analysis for Keri reveals a low coefficient of multiple determinants (R 2), 21%, indicating that there is little to no relationship between how respondents rated the Keri brand on the ten Likert items regarding brand attributes and their favorability of the Keri ad as indicated by their pre-post ad exposure change score. In other words, the Likert items account for only 21% of the variance in the pre-post ad exposure change scores. The standard error of estimate (se) is high at 2.3, indicating respondents were an average of 2.3 units away from the regression line. Given the statistically significant F ratio of 1.39, in 85 or more samples of every 100 samples drawn from the same population as this sample we would expect to find a multiple regression coefficient of the same magnitude. Six of the ten Keri brand attributes can be considered important: is a good moisturizing lotion; will keep my skin smooth and soft; will protect my skin; will replenish my skin; is not a hip brand; and does not contain as much moisture as other brands. In 85 or more of 100 samples drawn from the same population as this sample, we would expect all of the Keri brand attributes except will protect my skin, as well as the constant term (a), to have the same magnitude of impact.
Vaseline:The multiple regression analysis for Vaseline reveals a low coefficient of multiple determinants (R 2), 13.2%, indicating that there is little to no relationship between how respondents rated the Vaseline brand on the ten Likert items regarding brand attributes and their favorability of the Vaseline ad as indicated by their pre-post ad exposure change score. In other words, the Likert items account for only 13.2% of the variance in the pre-post ad exposure change scores. The standard error of estimate (se) is high at 3.0, indicating respondents were an average of 3 units away from the regression line. Given the statistically significant F ratio of .81, in 85 or more samples of every 100 samples drawn from the same population as this sample, we would expect to find a multiple regression coefficient of the same magnitude. Only two of the ten Vaseline brand attributes can be considered important: will keep my skin smooth and soft; and is a trusted brand. In 85 or more of 100 samples drawn from the same population as this sample, we would expect the all Vaseline brand attributes but “will keep my skin smooth and soft”, as well as the constant term (a), to have the same magnitude of impact.
Brand Management Implications: Aveeno, Keri, and VaselineBrand managers for all three brands (Aveeno, Keri, and Vaseline) should reevaluate the ability of the brand attributes (Likert items) to explain variance in respondents’ pre-post ad exposure change scores, since very little variance can be explained by these brand attributes and the standard error of estimate is high for all three brands. In order to maximize the usefulness of regression analysis in practical application, it is important for all brands to find a more suitable regression equation through further research and analysis.
This report contains a multiple discriminant analysis for Aveeno Moisturizing lotion consisting of two groups, those who had positive pre-post ad exposure change scores and those who had negative pre-post ad exposure change scores. Those respondents who indicated no change in their pre-post ad exposure brand evaluations were not included in this analysis. The categorical dependent variable in this analysis is group membership and the independent variables are the brand attributes as indicated by the Likert items in the questionnaire. This analysis was conducted to determine whether or not we can accurately predict, with a better than chance accuracy, the “up-movers” (those with positive change scores) from the “down-movers” (those with negative change scores) according to their brand attribute (Likert item) responses. An on-line questionnaire was constructed and sent via e-mail, returning a total of 64 completed surveys, 50 of which will be used in this analysis. Data was imported into SPSS for the following analysis.
*indicates coefficients equal to at least half the value of the largest coefficient
Aveeno Brand Attributes: “Up-Movers” vs. “Down-Movers”Comparison of the mean brand attribute (Likert item) scores does not reveal any dramatic differences between up-movers and down-movers. The largest differences between mean scores for up-mover and down-mover groups were on items 7 (“is not a hip brand”) and 8 (“helps fight dryness”). The magnitude of change for these items is not great (0.2), but these items revealed the largest differences between groups’ mean scores. Six of the ten brand attribute items can be considered important, meaning they best account for the differences in pre-post exposure change scores between the two groups, up-movers and down-movers: is a good brand (1), is too oily (2), will protect my skin (4), is a trusted brand (7), is not a hip brand (8), and Helps fight dryness (9). The unstandardized discriminant coefficients suggest the relationship between each brand attribute and how much respondents liked the ad and can be interpreted as follows: The more respondents think Aveeno is a good brand of moisturizing lotion, the more they like the ad. The more respondents think Aveeno has too oily, the less they like the ad. The more respondents think Aveeno helps keep skin soft, the more they like the ad. The more respondents think Aveeno helps protect skin, the more they like the ad. The more respondents think Aveeno is too expensive, the more they like the ad. The more respondents think Aveeno helps replenish skin, the less they like the ad. The more respondents think Aveeno is a trusted brand, the less they like the ad. The more respondents think Aveeno is not a hip brand, the more they like the ad. The more respondents think Aveeno fights dryness in skin, the less they like the ad. The more respondents think Aveeno does not contain as much moisture as other brands, the more they like the ad.
Functions at Group Centroids
* p <= .15 Group CentroidsThe group centroids (mean of the discriminant function scores) for the up-movers and down-movers are far apart and distinguishable from each other. Both the Wilks’ Lambda and Chi-Square values are significant at the 0.15 level, meaning that in 85 or more out of 100 samples drawn from the same population as this one, we would expect to find group centroid values of the same magnitude. Classification Matrix
66.0& of original grouped cases correctly classified. t = (.66 - .5) / √ {[(.66*.34)/50] + [(.5*.5)/50]} = 1.35* * p <= .15 The discriminant function correctly predicted group membership to up-movers or down-movers for 33 of these 50 respondents, resulting in an accuracy rate of 66%. Based on the observed t-ratio, this accuracy result can be projected to the entire population. In other words, in 85 or more out of 100 samples drawn from the same population as this one, we would expect to find an accuracy rate of the same magnitude. ImplicationsThis is a valuable discriminant analysis since all of the findings were significant at the 0.15 significant level. In other words, the discriminate analysis results based on this sample of 50 respondents can be projected to the entire population. If we know the attitudes someone in this population has toward Aveeno (as defined by the Likert items on the questionnaire), we can predict with 66% accuracy whether or not they will like the ad (as indicated by a positive pre-post ad exposure change score). Aveeno brand managers could use the discriminant function as a valuable tool for determining the favorability, applicability, and appropriateness of the ad tested.
Using the Likert item “Aveeno is too expensive” as a dependent variable, the two-way ANOVA test will try to determine the relationship between two independent variables: 1) sex and 2) likeability of the Aveeno print ad (using up/same/down constant sum scale). When using a factorial design, we can draw conclusions regarding effects of the independent variables separately, as well as the combined effects of the independent variables. The MANOVA test will then take all ten Likert items into account to try to determine the relationship between the sexes and how well they liked the Aveeno print ad. ANOVA Descriptive Statistics Dependant Variable: “Aveeno is too expensive.”
ANOVA Inferential Statistics
From the descriptive data, it can be noted that the mean scores of respondents who moved up, down, or stayed the same pre-to-post exposure to the Aveeno ad are moderately similar, and identical in value for those who stayed the same. The F-ratio was found to be statistically insignificant for the effects of sex, change score, and the interaction of sex and the Aveeno Change Score. This means that in 85 or more samples out of 100 drawn from the same population, we cannot project these magnitudes of variance on the population.
MANOVA Descriptive Statistics
MANOVA Inferential Statistics
*p <= .15 In this multi-variate factorial analysis of variance, we used all brand Likert items as the dependent variable, with the respondent’s sex and Aveeno Change Score as the independent variables. The mean scores of the ten brand attributes, change score direction, and sex of respondents are relatively similar in value. Females who did not like the ad according to their constant sum change scores comprised the smallest sample, yet had the highest mean scores for most likert items, indicating they reacted strongly in a negative magnitude to the towards the ad. The Wilks’ Lamda was found to be statistically insignificant for the main effect of Change Score, and Sex interaction with Change Score, meaning that in 85 or more samples out of 100 drawn from the same population we would not expect to get the same results found in this study. However, it was found to be significant for Sex alone, which means in 85 or more samples out of 100 drawn from the same population, we would expect to get the same results. Therefore, these results of this study can only partially be projected to the population. In general, this data cannot be used to determine whether or not respondents liked the advertisement based on their responses to the ten Likert items.
A factor analysis was performed for each test brand to place related brand attributes into independent groups (or factors) for the calculation of a brand attitude score. The variables used were all 10 Likert items. Three paired t-tests were then performed to find potential statistical significance for the differences of means between brand attitude scores. The results were obtained via an online questionnaire. Sixty-four people participated in the study. 10 Likert Items:
Communalities:
Total Variance Explained
Four of the 10 factors created for Aveeno have Eigenvalues greater than 1, indicating that they explain the variance of at least a single Likert item. These 4 factors explain 73.1% of the variance (26.9% is unexplained variance). Four of the 10 factors created for Keri have Eigenvalues greater than 1. These four factors explain 71.1% of the variance (28.9% is unexplained variance). Four of the 10 factors created for Vaseline have Eigenvalues greater than 1. These four factors explain 70.8% of the variance (29.2% is unexplained variance). For all three brands, a Varimax rotation was used to create a component matrix for these factors. For all three brands, the “good” factor (“Aveeno/Keri/Vaseline is a good moisturizing lotion”) loaded on factor one. For Aveeno, the brand attributes that loaded on factor one are “good”, “protect”, “trust”, and “dry.” For Keri, the brand attributes that loaded on factor one are “good”, “soft”, “protect”, and “dry.” For Vaseline, the brand attributes that loaded on factor one are “good”, “protect”, “replenish”, “trust”, and “dry.” The factors that loaded in factor one are somehow related to one another independent of other factors. Function and Rotated Matrices Factors: Aveeno
Function and Rotated Matrices Factors: Keri
Function and Rotated Matrices Factors: Vaseline
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||