| Table of Contents | Analysis
The following tests were carried out on a sample of 60 respondents (unless otherwise noted) regarding three brands of skin moisturizers: Lubriderm, Oil of Olay, and Vaseline.
1. PARED t-TESTS Brand Index Scores
Paired t- tests
Sample size = 60 By looking at the Mean Brand Index Scores, it looks as though the respondents have an almost equal perception of Lubriderm and Vaseline, but the t-ratio between these two brands is not significant (p=.203,) and thus we cannot project this hypothesis to a larger population. Managerial implications of these t-tests are ambiguous. In 85 or more samples of 100 samples drawn from the same population as this sample, we would expect to find a more favorable brand perception of Oil of Olay than of Lubriderm. The implications of this test are that people might be more likely to buy Oil of Olay, and that perhaps Lubriderm should change its advertising to compete with that brand. In 85 or more samples of 100 samples drawn from the same population as this sample, we would expect to find a more favorable brand perception of Oil of Olay than of Vaseline. The implications of this test are that Vaseline's advertising may need to change in order to compete with Oil of Olay. 2. BETWEEN GROUPS t-TESTS Two between-groups t-tests were conducted for Lubriderm, comparing respondents who moved up to those who moved down pre-to-post exposure to the magazine advertisement. Brand Index Scores and Ad Index Scores were analyzed for this comparison. 35 people out of the total sample of 60 respondents were used for this analysis .
In 85 or more samples out of every 100 samples drawn from the same population as this sample of 60 people, we would probably not find about the same average scores for those whose scores increased and decreased for the brand index score, but we would probably find about those same average scores for the Ad Index Score. Most people's perception of Lubriderm improves after exposure to the magazine advertisement. 3. CHI-SQUARE SIGNIFICANCE TEST - LUBRIDERM A chi-squared significance test was conducted to find out if there is a significant relationship between a changed score on Lubriderm after viewing the advertisement and being above or below the median Brand Index Score for Lubriderm. 60 people were used in this analysis.
* p < .15 4. FREQUENCY TEST FOR CHANGED SCORES A frequency test was conducted for all three brands of skin moisturizer to analyze pre-to-post exposure changed scores for constant sum scale responses. 60 respondents were used in this analysis.
As expected, a larger percentage of respondents did not change their attitudes about any of the brands after seeing the ads, compared to those whose attitudes changed favorably or unfavorably. What is interesting is that even though Oil of Olay has the highest Brand Index Score, more people saw the ad and their attitudes became less favorable towards the ad, compared to those who saw it and increased their scores. This outcome might mean that people's favorable impression of the Oil of Olay brand has less to do with the company's advertising and more to do with something else. More research would be needed to explore the relationship between people's attitude toward that brand and its advertising. 5. FREQUENCY FOR BRAND INDEX SCORES A frequency test was conducted to compare the Brand Index Scores of Lubriderm and Oil of Olay. 60 respondents were used in this analysis .
The results obtained from this comparison of Brand Index scores indicate that 35 respondents (58.3% of the sample) rated Oil of Olay higher than Lubriderm. As a result of these measurements, it can be concluded for this sample that a more favorable brand perception exists for Oil of Olay than Lubriderm. This suggestion is verified, coincidentally, in the Ad Index change scores; there wasn't an increase in favorable perception of that brand after respondents saw the ad for that brand. The company might look at its other promotions to find out what exactly is working. 6. SIMPLE CORRELATION COEFFICIENT A simple correlation coefficient analysis was conducted, comparing the Brand Index Scores for Oil of Olay and Vaseline, to determine if any relationship exists between the two brands. 60 respondents were used in this analysis.
There is a moderately strong positive correlation between Oil of Olay and Vaseline's Brand Index Scores, indicating that the perception of Oil of Olay might be related to the perception of Vaseline for the purposes of this sample. In 85 or more of every 100 samples drawn from the same population as this sample of 60 respondents, the correlation between Brand Index Scores for Oil of Olay and Vaseline slightly suggests that it would remain the same as that of this sample of 60 respondents. 7. SIMPLE CORRELATION COEFFICIENT: WOMEN & LUBRIDERM PREFERENCE A simple correlation coefficient analysis was conducted between the Lubriderm Brand Index Scores and the Oil of Olay Brand Index Scores to determine if any relationship exists between the two brands within the female respondent set. The 55 female respondents were used in this analysis.
There is a moderately strong positive correlation between Oil of Olay and Vaseline's Brand Index Scores, in the female sample set, indicating that the perception of Oil of Olay could be related to the perception of Vaseline for the purposes of this sample. In 85 or more of every 100 samples drawn from the same population as this sample of 55 respondents, the correlation between Brand Index Scores for Oil of Olay and Vaseline slightly suggests that it would remain the same as that of this sample of 55 respondents.
Lubriderm. Oil of Olay, and Vaseline A linear regression test was conducted for each of the three brands of skin moisturizer, using brand attributes as independent variables and the pre-to-post change score as the dependent variable. The results were obtained through an online consumer preference survey of 60 respondents.
*p <0.15
Equations:
Explanation of data: In the case of Lubriderm; in 85 or more samples out of a 100 we would expect to find a moderately strong R (coefficient of multiple regressions) of .5 because F is 1.86, and is significant at a significance level of p <=.15. Thus, we can project the results to the population from which this sample was taken. The coefficient of multiple regression (R 2 ) indicates that we can expect that only 27.5% of the cases can be explained by the variables expressed in the Likert items for Lubriderm, implying that the independent variables tested are not enough to explain the brand index score change. The Standard Error of the Estimate represents the unexplained variance in a multiple regression analysis and signifies that the average distance the survey respondents were from the regression line was 2.3 units. The following illustrates the most important six indexes and shows how to interpret the unstandardized coefficients( b): The more respondents thought that Lubriderm is a good skin moisturizer, the more they liked the ad. The more respondents thought that Lubriderm smells good, the more they liked the ad. The more respondents thought that their skin absorbs Lubriderm well, the less they liked the ad. The more respondents thought Lubriderm has a pleasing texture, the less they liked the ad. The more respondents thought they fell healthy when they use Lubriderm, the less they liked the ad. The more respondents thought they like the way their hands feel using Lubriderm, the more they liked the ad. In the case of Oil of Olay; in 85 or more samples out of a 100 we would expect to find an R (coefficient of multiple regressions) of .5 since F is 1.6, and is significant at a significance level of p <=.15. Thus, we can project the results to the population from which this sample was taken. The coefficient of multiple regression (R 2 ) indicates that we can expect that only 24.6% of the cases can be explained by the variables expressed in the Likert items for Oil of Olay, meaning that the independent variables tested are not enough to explain the brand index score change. The Standard Error of the Estimate represents the unexplained variance in a multiple regression analysis and signifies that the average distance the survey respondents were from the regression line was 2.6 units The following illustrates the most important six indicies and shows how to interpret the unstandardized coefficients (b): The more respondents thought that Oil of Olay is a good skin moisturizer, the less they liked the ad. The more respondents thought they would use Oil of Olay, the more they liked the ad. The more respondents thought that Oil of Olay smells good, the more they liked the ad. The more respondents thought that they feel healthy when they use Oil of Olay, the more they liked the ad. The more respondents thought they feel healthy when they use Oil of Olay, the less they liked the ad. The more respondents thought they like the way their hands feel using Oil of Olay, the less they liked the ad. For Vaseline in 85 or more samples out of a 100 we would not expect to find an R (coefficient of multiple regressions) of .5 because our F of 1.43 is not significant at a significance level of p<=.15. In other words the value for R can not be projected to the population from which this sample was taken. The coefficient of multiple regression (R 2 ) indicates that we can expect that only 22.6% of the cases can be explained by the variables expressed in the Likert items for Vaseline, implying that the independent variables tested are not enough to explain the brand index score change. The Standard Error of the Estimate represents the unexplained variance in a multiple regression analysis and signifies that the average distance the survey respondents were from the regression line was 2.3 units. The following illustrates the most important five indices and shows how to interpret the unstandardized coefficients (b): The more respondents thought they would use Vaseline, the less they liked the ad. The more respondents thought Vaseline is the longest lasting, the more they liked the ad. The more respondents thought that Vaseline smells good, the more they liked the ad. The more respondents thought they feel healthy when they use Vaseline, the less they liked the ad. The more respondents thought they like the way their hands feel using Vaseline, the more they liked the ad. We can conclude from these results that there is not much of a connection between how respondents liked the moisturizer brands and how they received the ad. One issue to address is the fact that few of the brand characteristics questioned had anything to do with the lifestyle and self-image of the hypothetical brand user, and those characteristics might be important for skin moisturizers. Other questions could have been posed in order to obtain data that was more significant and conclusive of the relationship between each skin moisturizer's brand characteristics and how respondents favored each brand's advertisement.
This report analyzes two groups, respondents who had a positive pre-post ad exposure score (up-movers) and respondents who had a negative pre-post ad exposure score (down-movers) for Lubriderm. The "up" and "down" movers were used as the categorically dependent variables, and brand attributes (reflected in 10 Likert items) functioned as independent variables. This analysis was conducted to show if an accurate discrimination can be predicted to divide the "up" movers from the "down" movers, according to their Likert item responses with a "better than chance accuracy."
When looking at the mean scores of each Likert item and comparing these scores between those whose scores increased or decreased, the 6 notable items are "b, prefer it over other brands" (1.1,) "c, would use" (1.4, ) "f, my skin absorbs it well" (1.4,) "g, has a pleasing texture" (1.4,) "h, feel healthy" (1.2,) and "i, like the way it feels" (1.4.) It seems that overall, those whose general impression of the Lubriderm ad was positive (whose constant-sum scores went up,) also rated Lubriderm positively on the majority of the Likert items. Those whose constant-sum scores decreased did not have as the positive responses to the Likert items that the "up-movers" did.
*p < 0.15 The group centroids (average z scores for each of the two groups) are not significant at -.5 for the up-movers and .5 for the down-movers, due to the lack of significance found for Wilks' Lambda at .94 and chi-square at 3.32. Therefore, in 85 or more samples out of every 100 drawn from the same population as this sample of 49, we would not expect to find group centroids of the same magnitude.
88.1% of original grouped cases were correctly classified. t= .383/ Ö [(.883*.117)/49] + [(5*.5)/49] = 4.04 The classification matrix shows that 88.1% of the original grouped cases are predicted correctly. There were 27 out of 29 of the predicted increased scores to actually increase, and 18 out of 20 of those who's scores decreased that were correctly predicted to decrease. The observed t-ratio of 4.04 is significant. Since the t-ratio is significant, in 85 or more samples out of 100 samples drawn from the same population as these 49 people, we would expect to find accurate classification of the same magnitude.
ANOVA In this Analysis of Variance, we are comparing the results of two discretely measured independent variables (men vs. women) and one continuously measured dependent variable (Likert item: "Lubriderm smells good"). We then use this information and determine whether we can project the results to the entire population from which this sample was drawn. Brand attribute: "Lubriderm smells good."
*p < 0.15 The mean scores of the change score movement are relatively similar in value. However, the F-ratio is not significant for gender; the F-ratio is significant for the change score movement, and therefore in 85 or more samples out of every 100 drawn from the same population, we would expect to find the same results in the same sample as these 60 respondents. This cannot be said of the mean for gender because it is not significant. Men = 5 Women = 55
The mean scores illustrative of the relationships between the ten brand attributes, change score movement and gender) are relatively similar in value. The F-ratio for change score movement and gender was found to be significant. Thus, in 85 or more samples out of every 100 drawn from the same population, we would expect the mean scores for change scores among the ten brand attributes and gender to have the same magnitude as this sample of 60 people. The F-ratio for the interaction of both independent variables was not significant. Thus, in 85 or more samples out of 100 samples drawn from this same population, we would not expect to find mean scores of the same magnitude found in this study.
A factor analysis was performed for each skin moisturizer 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. A paired t-test was then performed to find potential statistical significance for the differences of means between brand attitude scores.
(Factor Matrix) (Verimax matrix non-existent; only one component extracted.)
For Lubriderm, it can be said that 55.5% of the variance can be explained by one factor (44.5% is unexplained or lost variance). This was the only factor out of ten that had an Eigenvalue greater than one. The Varimax Rotated Matrix was not used in this analysis because only one component was extracted. The good item (Lubriderm is a good skin mosturizer) loaded on factor 1. All other items loaded on factor 1. These items were also used in computing the attitude score of 3.2 for Lubriderm. Oil of Olay
(Factor Matrix) (Verimax Matrix)
For Oil of Olay, it can be said that 59.4% of the variance can be explained by two factors (40.6% Is unexplained or lost variance). These two were the only factors out of ten that had Eigenvalues greater than one. The Varimax Rotated Matrix tabulated but not used in this analysis as it had the greater amount of ambiguous rows. The good item (Oil of Olay is a good skin moisturizer) loaded on factor 1. All other items loaded on factor 1 except for "Oil of Olay is more of a necessity than a luxury." These items were also used in computing the attitude score of 3.4 for Oil of Olay. Vaseline
(Factor Matrix) (Verimax Matrix)
For Vaseline, it can be said that two factors explain 64.9% of the variance. (35.1% is unexplained or lost variance). These were the only two out of ten that had Eigenvalues greater than one. The Varimax Rotated Matrix was tabulated but not used in this analysis. The good item ("Vaseline is a good skim moisturizer" loaded on factor 1. All other items loaded on factor 1 except for "Vaseline smells good." These items were also used in computing the attitude score of 3.1 for Vaseline.
The significant t-ratios suggest that in 85 or more samples out of every 100 samples drawn from the same population as this sample of 60 people, respondents' attitudes would be slightly more favorable towards Oil of Olay, but because the attitude scores are so similar, the results are inconclusive.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||