The purpose of this research is to use copy testing as a means to examine brand preferences among three brands of cat food: Purina, Iams, and Whiskas. This study will determine the initial preferences for the three brands, analyze common attributes associated with each brand, find the effect the print advertising on brand preference, and make a comparison of which brand has the highest preference overall.
An on-line questionnaire was created and data were collected via email, asking to fill out the survey. The results completed by 80 respondents were collected in database format, and then imported into SPSS for a variety of statistical analyses. The descriptive statistics are included in Appendix A. The research design used a pre-to-post methodology rather than using a control group.
Purina achieved the most favorable pre-to-post exposure change scores, indicating the positive effects of the advertising on brand preference. Purina is a dominant player in this market and still the most preferred brand overall. The mean differences of Brand Index Scores between for the pairs of Purina and Iams and Purina and Whiskas can be projected to the larger population. This study indicates that there is a weak correlation for all three brands between how respondents rated the brand attributes and their liking toward the advertising. For more in depth analysis and further discussion, please continue through this report.
|
  
Among marketers, their challenge is to create memorable and effective communications and decide where to place them so that they have the greatest impact. And it all comes down to whether people like the ad or not (Clancy, 2001). Copy testing including both pre-testing and post-testing is a means of measuring how much people would like the ad and how effective the advertising message is. It can gauge the risk associated with producing new advertising by providing quantitative measures of consumer attitudes and perceptions on brand and advertising message.
This report contains copy-testing research for three brands of cat food, a low involvement consumer convenient good. The three brands of cat food are: Purina, Iams, and Whiskas. Their prices and targets (adult cat foods with non-specialty) are similar, and available at the same distribution channels (not like one available only at veterinary hospitals). Three print ads were selected from two women's magazines and one cat lovers' magazine: Martha Stewart Living, and Better Homes and Gardens, and Cat Fancy . An on-line questionnaire was constructed, distributed via email and gathered 80 completed responses. Once all surveys were returned, the data were imported into SPSS and analyzed using various statistical methods.
The purpose of this research is to examine brand preferences among three brands of cat food and the effect of print advertising upon brand preference. This research contains results from the questionnaire which will determine initial preferences for the three brands, analysis of common attributes associated with those brands, the effect of print advertising on brand preference, and which brand was liked the most. The results will be examined for statistical significance in order to determine if we can project these research results to the larger population which means in 85 or more samples out of 100 samples taken from the same population the results would be expected to be in as these 80 respondents.
This final report provides a comprehensive description of the methodologies used (structure of the questionnaire, the design, the sample, and the data collection), analysis of all statistical test results, conclusions drawn from the results, and a summary of the findings. This report also includes the actual questionnaire distributed (Appendix B) as well as a questionnaire that has the descriptive statistics (all responses) for each question (Appendix A).
Reference: Clancy, Kevin. (2001). Copy Testing Overview. Copernicus Marketing Consulting and Research, and MarketingPower Inc. American Marketing Association.
|
Structure of the Questionnaire
The questionnaire developed for this study was created for the purpose of analyzing three brands of wine: Louis M. Martini, Sequoia Grove, and Santa Rita. This survey was created using Dreamweaver and the layout was designed specifically for online data collection. The questionnaire was titled "Consumer Preference Survey" and did not mention any brand or product names on purpose not to give respondents any prejudice. The questionnaire consists of 8 main sections under which has many questions. A detailed breakdown of each section is as follows:
Title Page
A title page appears when a respondent clicks on the questionnaire's link. The title page is designed to motivate participants to take the survey and specifically states that all responses will be held strictly confidential and educational purpose only.
Part I: Usership of Relevant Brands to Test Brand and Pre-Ad Exposure Rating
Section 1 : The questions ask the respondents to type in the last brand of pet shampoo, cat treat, and cat food which are relevant to cat food. The purpose of this question is to find out which brands come to their minds at first to the respondent and which brands the respondent have a constant usage. At this point, the respondent does not know on which brands the survey will further ask; thus the goal is to minimize biases when collecting this data. For the purposes of the study, the responses on cat food will be the only brands that will be analyzed.
Section 2 : Respondents respond to question asking them to rate each brands presented on a 10-point constant sum scale. These questions ask the respondents to report their attitude or intention to buy pet shampoo, cat treat, and cat food according to how much they like each brand without seeing any ads. The values from this scale are used as pre ad exposure and later as change score. For the purposes of this study, the cat food responses will be analyzed.
Part II: Print Ad Exposure and Post-Ad Exposure Rating
Section 3 : The respondents are led to view three ads of each brand for a short time. Once they finish viewing the ads, they are advised not to go back to see the ads again whether they can fully recall or not. This survey is designed to be as realistic as normally people flip through a magazine.
Three print ads were selected from two women's magazines and one cat lovers' magazine: Martha Stewart Living, and Better Homes and Gardens , and Cat Fancy . All three ads are in the same layout and in color. Below are the ads shown to the respondents:
Section 4 : This question is the same as Section 2 except that respondents are exposed to three ads. The questions are designed to determine if there are any changes in ratings of three brands after viewing the ads. This "post viewing" question will allow comparisons of pre-viewed index scores and post-viewed index scores to create a change score for each brand.
Part III: Post-Ad Exposure Differences
Section 5 : This question has four parts, three of which are optional. The first part is to determine if viewing the ad makes the respondent feel differently toward any of the three brands. If they clicked ¡°No¡±, the respondent skipped ahead to Section 5. If there is any difference, the following questions asked which brand was affected, whether the impact was favorable, and typed in any other comments about the reasons behind.
Parts IV: Brand Attributes
Sections 6 : These questions use a Likert scale to measure attitudes, preferences, and subjective reactions. The questions reflect common attributes associated with cat food. The respondents correspond to only one scale for each attribute that best reflects their opinion: strongly agree, agree, neither agree nor disagree, disagree, or strongly Disagree.The Likert scale is used to give quantitative results of brand attributes; each answer has a numerical value from one to five. Negative and positive questions do not follow any pattern not to trick respondents.
Test Brand is a good brand of cat food.
Test Brand has a taste cats like.
Test Brand cat food is too expensive.
Test Brand is better than other cat food.
Test Brand offers a good source of nutrition.
Test Brand is a trustable brand.
I would not buy Test Brand cat food.
Test Brand prevents hair from shedding.
Test Brand is highly recommended by veterinarian.
Test Brand does not strengthen bones and muscle tone.
Part V: Attitudes toward Ads
Sections 7 : These questions ask the respondent to give a "Yes" or "No" response for 10 different impressions of the brand from the print ad. These are both negative and positive to receive general reactions to the print ads. If the words provided do not fully reflect of what respondents think, items could remain unchecked. What was said about the product in the "Test Brand" ad was:
Boring
Informative
Hard to believe
Memorable
Different/Unique
Important
Confusing
Creative
Common/Ordinary
Easy to Understand
Part VI: Effectiveness of Ads
Section 8 : These questions ask the respondent to give a "Yes" or "No" response for ten different statements of each print ad's effectiveness. The intention of these questions is to be more specific about execution of the actual print ads. This Section is tested without viewing the ads and is based on recall of the respondents. The following statements are questions on impressions of ads:
This ad was annoying.
This ad contained useful information.
This ad was believable.
This ad would be easy to remember.
This ad would catch my attention if it was in a magazine.
This ad made me think differently about cat food.
This ad would appeal to women more than men.
I would like to see this ad again.
This ad would be easy to forget.
This ad taught me something new about cat food.
Section VII: Demographic/Lifestyle Information
Section 9 : The final five questions collect information on basic demographic and lifestyle: the respondent's sex, age, education level, income level, and how many cats they own (for those who have cats are asked how many times they have changed cat food). When finishing answering all the questions, respondents are asked to click the "Submit" button by giving consents to give their answers.
To view the exact survey respondents distributed, please see Appendix B (Hyperlink) and if you would like to see the statistical results of how each respondents answered to each question, please refer to Appendix A .
Design
This study was conducted using survey methodology and follows the pseudo-experiment group format meaning there was no control group. Instead of having a control group, comparison is made between before and after advertising exposure using the same sample. The purpose of this study is to explore consumer preferences for three popular cat food brands and to see which brand was most preferred. The change score (post exposure scores less pre exposure ratings) is used to measure the effectiveness of adverting. The change score indicate whether respondents moved up, stayed the same, or moved down in their ratings of the brands after exposure to the print ads.
Sampling Description
Due to time and monetary constraints, this study opted for a non-random, convenience sample. It was decided that a sample size of at least 60 would be decent enough to perform statistical analysis. This sample size also allows for enough representation so that if statistical significance is found, projection can be made in 85 or more samples of 100 samples within the same population as these 60 were drawn we would find results of the same magnitude.
An email was sent out to personal and professional acquaintances of the researcher. This email contained the link to the survey asking individuals to complete the survey and to forward on to any other individuals who may be interested in participation. By clicking the submit button, all respondents gave consents to give their answers to the researcher. A total of 120 respondents submitted their answers; however, only 80 were completed the survey that this study has a sample size of 80.
A total of 175 respondents attempted to complete the survey. Of those, 136 surveys were completely and correctly completed. Thus, the sample size for this research is 136.
Data Collection
Data in this study were collected on an on-line basis from 80 respondents. All data in this study were collected electronically from September 5 to September 25 in 2006. The respondents received an email that contains link to the survey site and were asked to do the survey by clicking the link. By clicking the submit button, all responses were stored in a database file in the Center for Interactive Advertising. The database file and the online questionnaire were connected to each other through a Cold Fusion file, which contained a thank you message to respondents for their participation. The data were exported to SPSS for the statistical analysis. A total of 120 respondents attempted to complete the survey; among them, 80 respondents completed the survey making the sample size for this research is 80.
The online survey's web page was constructed to appear as if it were multiple pages; however, it is one continuous page. The detailed instructions guided respondents through to completion. The questionnaire included text fields to gather written comments, drop-down menus, and radio buttons so respondents could select appropriate answers in an efficient manner. When the questionnaire was created, each answer was given a field name and a coded numerical values (if applicable) so that the numbered data could be quantitatively used in a statistical program for analysis. However, the method to ask respondents to go back to answer the questions they missed during the survey was not used in this research for respondents not to quit in the middle of survey.
|
Basic Statistics (t-tests, Chi-squared, Frequencies)
Paired t-tests for Brand Index Score
Statistical results interpretation
Overall, 80 respondents generally have a more favorable perception of Purina than Iams or Whiskas. Two sets of brands (Purina-Iams and Purina-Whiskas) indicate significant difference in means of Brand Index Scores at an alpha level of 0.15. This means that in 85 or more samples drawn from the same population as this sample, it would be expected that the differences between Brand Index Score mean for these pairs. The result for Iams and Whiskas is not significant and cannot be projected to the population.
Between groups t-tests : Purina
Statistical results interpretation
After seeing the Purina ad, up movers had a slightly more favorable perception of the Purina brand; however, we cannot project this result to a larger population because the difference in means of Brand Index Score is not statistically significant. In 85 or more samples of every 100 samples drawn from the same population as this sample, it would be expected that down movers would have a slightly more favorable perception of the Purina ad and mean of Ad Index Score for Purina for up and down movers would be about what they are in the population.
Chi-Squared Significance test for Purina
Statistical results interpretation
19 respondents who moved up after pre-post testing were below median of Brand Index Score, representing approximately 55.9% of all up movers. 13 respondents stayed the same and were above the median of Brand Index Score which represents 59.1% of all respondents who remained the same. It means that more people who liked the ad were not shared with the people who liked the brand; furthermore, we cannot project these results to a larger population because these statistical results are not statistically significant.
Pre-Post Ad Exposure Change Score Frequencies
Statistical results interpretation
Purina had the most respondents with improved brand perception and the least respondents with lowered brand perceptions after exposure to the ad. Iams had the same percentage of the respondents who moved down and stayed the same. Whiskas after exposure to the ad, had respondents who stayed the same outnumbering those who moved down. Purina had the highest Brand Index Score and the Purina ad does the most effective job of improving general consumer brand perceptions in comparison to the other brands. It is notable that Iams had the second highest Brand Index Score; however, after exposure to the ad, Iams had the most respondents who moved down among three brands.
Brand Index Score Comparison: Purina and Whiskas
Statistical results interpretation
The results indicate that majority of respondents - 51 out of 80 total respondents accounting for 63.8% - had a more favorable brand perception of Purina than Whiskas. 29 respondents, approximately 36.3% rated Purina as lower or equal favorable brand to Whiskas.
Simple Correlation Coefficient: Iams and Whiskas
Statistical results interpretation
There is a weak positive correlation between Brand Index Scores of Iams and Whiskas. Since correlation r is not statistically significant, we cannot project this result to a larger population.
Simple Correlation Coefficient: Cat Owners Sub Sample
Statistical results interpretation
There is a weak positive correlation between the Iams and Whiskas Brand Index Scores in the cat owner sub sample. This means that among cat owners those who show favorable perception on Iams brand are more likely to show the same perception on Whiskas brand. In 85 or more out of every 100 samples drawn from the same population as this one, it would be expected to find a weak positive correlation between Iams brand perceptions and Whiskas brand perceptions among cat owners.
Regression Analysis
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.
Purina
The multiple regression analysis for Purina shows that there is a weak relationship between how respondents rated the Purina brand on the ten Likert items and their liking toward the Purina ad by the result of their pre-post ad exposure change score. In other words, only 24.9% of the variance of the pre-post ad exposure change scores can be explained by Likert items. As F-ratio is statistically significant, in 85 or more samples out of every 100 drawn from the same population as this sample, it would be expected that the coefficient of multiple determination would be about the same as it is in the sample.
The unstandardized coefficients (b) indicate the relationship between each Purina brand attributes and their liking toward the Purina ad. The more respondents think Purina is a good brand of cat food, the more they like the ad. The more respondents think they would not buy Purina cat food, the less they like the ad. In 85 or more samples out of every 100 samples drawn from the same population as this sample of 80, it would be expected to find about the same result of the constant term, and the following attributes of ¡°Purina is a good brand of cat food¡±, ¡°Purina has a taste cats like¡±, ¡°Purina cat food is too expensive¡±, ¡°Purina offers a good source of nutrition¡±, ¡°Purina is a trustable brand¡±, and ¡°Purina is highly recommended by veterinarians¡± as we found in this sample. Those items are considered to be the most important variables in explaining the ad exposure change scores relative to other brand attributes.
Iams
There is a weak relationship between how respondents rated the Iams brand based on the ten Likert questions and their liking toward the Iams ad by the result of their pre-post ad exposure change score. In other words, only 18.9% of the variance of the pre-post ad exposure change scores can be explained by Likert items. The standard error of estimate is high at 1.4, which means on average, each respondent is 1.4 units away from the regression line. In 85 or more samples out of every 100 drawn from the same population as this sample, it would be expected that the coefficient of multiple determination would be about the same as it is in the sample.
Six of the ten brand attributes can be considered important, indicating they best explain the variance in pre-post ad exposure change scores relative to the other brand attributes: ¡°Iams is a good brand of cat food¡±, ¡°Iams is better than other cat food¡±, ¡°Iams offers a good source of nutrition¡±, ¡°I would not buy Iams cat food¡±, ¡°Iams is highly recommended by veterinarians¡± and ¡°Iams does not strengthen bones and muscle tone¡±. In 85 or more samples out of every 100 samples drawn from the same population as this sample, it would be expected to find about the same result (not the constant term) of the following attributes of ¡°I would not buy Iams cat food¡± and ¡°Iams is highly recommended by veterinarians¡± as we found in this sample.
Whiskas
There is a weak relationship between how respondents rated the Whiskas brand on the ten Likert questions and their liking toward the Whiskas ad by their pre-post ad exposure change score. In other words, only 16.9% of the variance of the pre-post ad exposure change scores can be explained by Likert items. The standard error of estimate is high at 1.4, which means on average, each respondent is 1.4 units away from the regression line. In 85 or more samples out of every 100 drawn from the same population as this sample, we would not expect to find the coefficient of multiple determinations to be the same as in this sample.
Four of the ten brand attributes can be considered important, indicating they best explain the variance in pre-post ad exposure change scores relative to the other brand attributes: ¡°Whiskas offers a good source of nutrition¡±, ¡°Whiskas is a trustable brand¡±, ¡°Whiskas prevents hair from shedding¡±, and ¡°Whiskas does not strengthen bones and muscle tone¡±. In 85 or more samples out of every 100 samples drawn from the same population as this sample, it would be expected to find about the same result as well as the constant term, of the following attributes of ¡°Whiskas prevents hair from shedding¡± and ¡°Whiskas does not strengthen bones and muscle tone¡± as we found in this sample.
Discrimant Analysis
The categorical dependent variable is group membership and the continuous independent variables are the responses for the ten Likert items regarding specific brand attributes.
When comparing the mean brand attribute (Likert item) scores between up and down movers, these scores do not reveal any dramatic differences for each item between two groups . Some differ more greatly than others. The largest difference between mean scores for up and down mover groups was on item ¡°Whiskas offers a good source of nutrition¡±.
Five of the ten brand attributes can be considered important, indicating they best account for the differences in pre-post exposure change scores between the two groups, up movers and down movers: ¡°Whiskas offers a good source of nutrition¡±, ¡°Whiskas is a trustable brand¡±, ¡°Whiskas prevents hair from shedding¡±, ¡°Whiskas is highly recommended by veterinarians¡± and ¡°Whiskas does not strengthen bones and muscle tone¡±. The unstandardized discriminant coefficients indicate that the relationship between each Likert item to how much the respondents liked the Whiskas ad.
The group centroids for the up and down movers are not significant at 0.5 for the up movers and -0.6 for down movers from insignificant Wilks' Lambda of 0.77. Therefore, in 85 or more samples out of every 100 drawn from the same population as this sample of 45, we would not expect to find group centroids of the same magnitude. The Chi-Squared is also not significant and shows a lack of a relationship between the up and down movers and the brand attributes that we cannot project this result to the population.
The discriminant function correctly predicted group membership to up movers or down movers for 27 of 45 respondents, resulting in an accuracy rate of 60.0%. Based on the observed t-ratio (through calculation=0.68), this accuracy result cannot be projected to a larger population since it is not greater or equal to critical t-ratio of 1.04.
ANOVA/MANOVA Analysis
In the ANOVA, one of the Likert items is the dependent variable and changes in brand scores after the ad exposure are two independent variables. For the MAVONA, the independent variables remain the same while all the Likert items are chosen as dependent variables for the same test.
Two-way Factorial ANOVA
There is little difference across the mean scores on the Likert item of ¡°Whiskas is a good brand of cat food¡± between those who have a cat and those who do not have a cat. There was no difference among the pre-post ad exposure change score groups. The F-Ratios are not significant for the change scores, cat ownership or the interaction of the two, in 85 or more of 100 samples drawn from the same population as this sample, we would not expect to find mean scores of the same magnitude among pre to post ad exposure change score groups, cat owners, and the interaction between the two variables, meaning we cannot project the same mean scores among change score groups, cat owner groups and among the interaction between the two in this sample to the larger population.
Multivariate Analysis of Variance (MANOVA)
Mean scores between those who have a cat and those who do not have a cat across all Likert items for all change conditions do not show notable differences except for one item: ¡°Whiskas has a taste cats like¡±.
The F-Ratios for change score, cat ownership, and the interaction of these two variables are not significant; none of the mean scores are not statistically significant, meaning in 85 or more of 100 samples drawn from the same population as this sample, we cannot expect to find mean scores of the same magnitude among pre to post ad exposure change score groups, cat owner groups and pre to post ad exposure change score/cat ownership groups.
There are significant relationships between up, same and down movers and ¡° Whiskas cat food is too expensive¡±, ¡°I would not buy Whiskas cat food¡± and ¡°Whiskas does not strengthen bones and muscle tone¡±. There is also a significant relationship found between cat ownership and ¡°Whiskas cat food is too expensive¡±. For the interaction with those two variables, there is a significant relationship with one item ¡°Whiskas has a taste cats like¡±. In each of the stated relationships above, in 85 or more samples out of every 100 drawn from the same population, we would expect to find results of the same magnitude as this sample, the magnitude of the variance for these results can be projected to the larger population.
Factor Analysis
This report contains factor analysis of each test brand (Purina, Iams, and Whiskas), where brand attribute (Likert items) variables are analyzed for relationships between attributes and new composite factors are created to identify related characteristics of the variables for three brands of cat food.
For Purina, three of the 10 factors have Eigenvalues greater than 1, indicating that they explain the variance of at least a single Likert item. It can be said that 64.7% of the variance can be explained by three factors while 35.3% is unexplained or lost variance. Also three factors for Iams have Eigenvalues greater than 1; these three factors explain 66.3% of the variance whereas 33.7% is unexplained variance. Three of the 10 factors for Whiskas have Eigenvalues greater than 1, and these three factors explain 70.3% of the variance while 29.7% is unexplained or lost variance.
For all three brands, a Varimax rotation was not used in this analysis because Factor Matrix had fewer or no problems at all than Varimax Rotated Matrix; the problem means ambiguous rows that have more than one factor greater than or equal to absolute value of 0.5, or none of the factors are above 0.5. For all three brands, the good factor (¡¦is a good brand of cat food) loaded on factor I. For the attitude scores, only values loaded on factor I were chosen to calculate. The factors that loaded in factor I are somehow related to one another independent of other factors.
The mean attitude scores for Purina is 3.7 which is the highest, Iams is 3.5 and Whiskas is 3.4; in this study, Purina is liked the best but not by much. Although Purina has higher brand perception than Iams and Whiskas, the magnitude of this difference is not big enough to be considered important. Based on the results from a paired t-test to test for the significance of the mean attitude scores, the following can be stated: in 85 or more samples out of every 100 samples drawn from the same population as this sample of 80 people, we would expect that the differences between mean of attitude score for the pairs of Purina and Iams and Purina and Whiskas would be about the same as they are in this sample. However, the difference between mean of attitude score of Iams and Whiskas cannot be projected to the population.
|
Paired t-tests
Correlated t-tests were conducted comparing all three Brand Index Scores to their corresponding Likert questions. Overall, respondents have the most positive perception of Purina cat food in relation to Iams and Whiskas. It is not clear, however, which brand more preferred between Iams and Whiskas. In 85 or more samples out of every 100 samples drawn from the same population as this sample, we would expect to find a more favorable brand perception of Purina than Iams or Whiskas, but we cannot project with confidence the brand favorability between Iams and Whiskas. In addition, the magnitude of this difference compared to two other brands is not big enough to be considered important. Purina has a slight higher Brand Index Score over both Iams and Whiskas; however, all three Brand Index Scores are so close (different by approximately only 1.0). As a brand manager of one long established and dominant Purina should note this perception among consumers and try to appeal Purina's advantages over other brands to widen the gap between close followers.
Between-Groups t-tests
A between-groups t-test was conducted in order to determine whether the Purina ad shown in the survey may have had an effect on respondents. After seeing the Purina ad, up movers had a slightly more favorable perception of the Purina brand; however, we cannot project this result to a larger population because the difference in means of Brand Index Score is not statistically significant. The results indicate that respondents who improved their perception on Purina brand after viewing the advertisements had higher brand preferences. However, up movers did not show the same preferences toward the ad itself. In 85 or more samples of every 100 samples drawn from the same population as this sample, it would be expected that down movers (people with a less favorable perception of Purina after seeing the Purina ad than before) would have a slightly more favorable perception of the Purina ad and mean of Ad Index Score for Purina for up and down movers would be about what they are in the population.
Chi-Squared
Chi-Squared significance test is to find out if there is a significant relationship between up, same and down movers after seeing Purina ad on Purina brand and being above or below the median Brand Index Score on Purina brand. More people who liked the ad were not shared with the people who liked the brand. This result can be interpreted consistently with between-groups t-tests that people who moved up (showed more brand preferences) showed lower AD Index Scores. However, we cannot project these results to a larger population because these statistical results are not statistically significant at a significance level of 0.5. In 85 or more out of every 100 samples drawn from the same population as this sample, we would not expect to find a cross tabulation distribution with the same proportions as this sample.
Frequency
The Purina brand gained the most respondents with improved brand perception and the least respondents with lowered brand perceptions after exposure to the print ad. Iams had the same number of the respondents who moved down and stayed the same. Whiskas after exposure to the ad, had respondents who stayed the same outnumbering those who moved down. It is notable that Iams had the second highest Brand Index Score; however, after exposure to the ad, Iams had the most respondents who moved down among three brands. As a brand manger of Iams and Whiskas, it is imperative to note that after exposure to ad, there are more down movers than up movers, indicating that respondents showed less favorability toward the brands after the ad. The Iams and Whiskas advertising had not done a good job of influencing brand perception although there are small differences of Brand Index Scores.
Brand Index Score Comparison
The results indicate that majority of respondents - 51 out of 80 total respondents accounting for 63.8% - had a more favorable brand perception of Purina than Whiskas. 29 respondents, approximately 36.3% rated Purina as lower or equal favorable brand to Whiskas. More respondents in this sample rated Purina higher than Whiskas.
Simple Correlation Coefficient
There is a weak positive correlation between Brand Index Scores of Iams and Whiskas. It means that people who have more favorable perception on Iams brand do not necessarily have the same favorable perception on Whiskas brand. Since correlation r is not statistically significant, we cannot project this result to a larger population, which means that in 85 or more samples out of every 100 samples drawn from the same population as this sample, the correlation for Iams and Whiskas on their Brand Index Scores would not be about the same as what they are in this sample.
Simple Correlation Coefficient (Sub Sample)
There is a weak positive correlation between the Iams and Whiskas Brand Index Scores in the cat owner sub sample (51 are cat owners). This means that among cat owners those who show favorable perception on Iams brand are more likely to show the same perception on Whiskas brand. In 85 or more out of every 100 samples drawn from the same population as this one, it would be expected to find a weak positive correlation between Iams brand perceptions and Whiskas brand perceptions among cat owners.
Regression Analysis
PURINA
The multiple regression analysis for Purina shows that there is a weak relationship between how respondents rated the Purina brand on the ten Likert items and their liking toward the Purina ad by the result of their pre-post ad exposure change score. In other words, the Likert items account for a low percentage of the variance in the pre-post ad exposure change scores; however, with statistical significance, we can project the results to the population.
The following attributes of ¡°Purina is a good brand of cat food¡±, ¡°Purina has a taste cats like¡±, ¡°Purina cat food is too expensive¡±, ¡°Purina offers a good source of nutrition¡±, ¡°Purina is a trustable brand¡±, and ¡°Purina is highly recommended by veterinarians¡± are considered to be the most important variables in explaining the ad exposure change scores relative to other brand attributes.
IAMS
There is a weak relationship between how respondents rated the Iams brand based on the ten Likert questions and their liking toward the Iams ad by the result of their pre-post ad exposure change score. As F-ratio of 1.61 is statistically significant, the result can be projected to the population. In 85 or more samples out of every 100 samples drawn from the same population as this sample, it would be expected to find about the same result (not the constant term) of the following attributes of ¡°I would not buy Iams cat food¡± and ¡°Iams is highly recommended by veterinarians¡± as we found in this sample.
WHISKAS
There is a weak relationship between how respondents rated the Whiskas brand on the Likert items and their liking toward the ad by their pre-post ad exposure change score. Due to statistical insignificant, the result cannot be projected to the population.
In 85 or more samples out of every 100 samples drawn from the same population as this sample, it would be expected to find about the same result as well as the constant term, of the following attributes of ¡°Whiskas prevents hair from shedding¡± and ¡°Whiskas does not strengthen bones and muscle tone¡± as we found in this sample.
Discriminant Analysis
Five of the ten brand attributes can be considered important, indicating they best account for the differences in pre-post exposure change scores between the two groups, up movers and down movers: ¡°Whiskas offers a good source of nutrition¡±, ¡°Whiskas is a trustable brand¡±, ¡°Whiskas prevents hair from shedding¡±, ¡°Whiskas is highly recommended by veterinarians¡± and ¡°Whiskas does not strengthen bones and muscle tone¡±.
The relationship between each Likert item and how much the respondents liked the Whiskas ad are as follows:
• The more respondents think ¡°Whiskas offers a good source of nutrition¡±, the less they like the ad.
• The more respondents think ¡°Whiskas is a trustable brand¡±, the more they liked the ad.
• The more respondents think ¡°Whiskas prevents hair from shedding¡±, the more they liked the ad.
• The more respondents think ¡°Whiskas is highly recommended by veterinarians¡±, the less they liked the ad.
• The more respondents think ¡°Whiskas does not strengthen bones and muscle tone¡±, the more they liked the ad.
The discriminant function correctly predicted group membership to up movers or down movers, resulting in an accuracy rate of 60.0%. The accuracy rate is not fairly high and the classification results are not significant. Wilks' Lambda and Chi-Squared are not also significant, which indicates that the group centroids are close and we cannot project the results to the population. It appears that these ten brand attributes do not discriminate between these two groups of up and down movers at a high level for Whiskas. Perceptions of the respondents on brand attributes of Whiskas have no or weak relationship between how much they like the ad.
ANOVA/MANOVA
ANOVA
These analyses test if there is any significant relationship between one brand attribute, to cat ownership and change scores of the brand after the ad exposure. There is little difference across the mean scores on the Likert item of ¡°Whiskas is a good brand of cat food¡± between those who have a cat and those who do not have a cat. There was no difference among the pre-post ad exposure change score groups.
Statistically, in 85 or more of 100 samples drawn from the same population as this sample, we would not expect to find mean scores of the same magnitude among pre to post ad exposure change score groups, cat owners and the interaction between the two variables.
MANOVA
The MANOVA test examines if there is any relationship between the ten brand attributes, whether the respondents liked the ad, and cat owners. The F-Ratios for change score, cat ownership, and the interaction of these two variables are not significant; none of the mean scores are not statistically significant, meaning the results cannot be projected to the larger population.
When analyzing each Likert item with each independent variable and the interaction between the two, there are some significant relationships to be stated. There are significant relationships between up, same and down movers and ¡° Whiskas cat food is too expensive¡±, ¡°I would not buy Whiskas cat food¡± and ¡°Whiskas does not strengthen bones and muscle tone¡±. There is also a significant relationship found between cat ownership and ¡°Whiskas cat food is too expensive¡±. For the interaction with those two variables, there is a significant relationship with one item ¡°Whiskas has a taste cats like¡±. In each of the stated relationships above, the results can be projected to the larger population.
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. In this study, Purina is liked the best but not by much. Although Purina has higher brand perception than Iams and Whiskas, the magnitude of this difference is not big enough to be considered important. Based on the results from a paired t-test to test for the significance of the mean attitude scores, the following can be stated: in 85 or more samples out of every 100 samples drawn from the same population as this sample of 80 people, we would expect that the differences between mean of attitude score for the pairs of Purina-Iams and Purina-Whiskas would be about the same as they are in this sample. However, the difference between mean of attitude score of Iams-Whiskas cannot be projected to the population; in other words, we cannot expect that we would find the same magnitude of difference between Iams and Whiskas mean attitude scores.
|
This study attempts to discover variables that might help explain different advertising effects and test the effectiveness of print advertising: increased, decreased, and same purchase intention after exposure to print advertisements for three brands of cat food. 80 people completed an online questionnaire. The collected data were analyzed using various statistical tests in SPSS that included: correlated t-test, between groups t-test, chi-squared significance test, frequency counting, simple correlation coefficient, regression analysis, discriminant analysis, ANOVA/MANOVA, and factor analysis. The results from the sample were statistically examined to determine whether or not they can be projected to the entire population with statistical confidence. Statistical significance throughout (indicating that in 85 or more samples drawn from the same population as this sample, the expected values would be the same as those reported here) was set at p < .15.
The three brands tested in this research were Purina, Iams, and Whiskas. Initially, Purina was the most favored brand, then Iams, and finally Whiskas. Purina had the highest brand preference mean (on the constant sum scale) pre-exposure and remained the highest post viewing (up movers were as twice as down movers). Respondents who improved their perception of the Purina brand after viewing the advertisements had more positive perceptions of the Purina brand than those who felt less positive after viewing the ads. The following attributes are most important in explaining the variance in pre-post ad exposure change scores: ¡°Purina is a good brand of cat food¡±, ¡°Purina has a taste cats like¡±, ¡°Purina cat food is too expensive¡±, ¡°Purina offers a good source of nutrition¡±, ¡°Purina is a trustable brand¡±, and ¡°Purina is highly recommended by veterinarians¡±, which the result can be projected to the population.
Iams achieved second favorable Brand Index Score; however in terms of pre-post change score, down movers are almost as twice as up movers, indicating the negative impact of the advertising on brand preference. There is also a weak relationship between how respondents rated the Iams brand based on the ten Likert questions and their liking toward the Iams ad by the result of their pre-post ad exposure change score. The following attributes of ¡°I would not buy Iams cat food¡± and ¡°Iams is highly recommended by veterinarians¡± are important in explaining the variance in pre-post ad exposure change scores, which is statistically significant at the level of 0.15.
The Whiskas brand showed the lowest Brand Index Scores and the largest number of people who showed the same brand preferences after ad exposure, indicating that the ad was not effective enough to change preferences. The results showed that there is a weak relationship between how respondents rated the Whiskas brand on the Likert items and their liking toward the ad by their pre-post ad exposure change score. The following attributes of ¡°Whiskas prevents hair from shedding¡± and ¡°Whiskas does not strengthen bones and muscle tone¡± are important in explaining the variance in pre-post ad exposure change scores. However, the result cannot be projected to the population.
This study indicates that there is little or no relationship between how respondents rated the brand attributes and how they responded to the brand advertising. Only Purina showed the highest Brand Index Score and the largest number of up movers; however, up movers showed lower Ad Index Scores. Given the statistical significance, these findings for the Purina and Iams can be projected to the entire population at the p¡Â0.15 level. In other words, in 85 or more of 100 samples drawn from the same population as this sample, we would expect to find the same results.
Limitations of this study lies on the relatively small sample size. A convenience sample was used instead of a preferred random sample and a control group was not available due to constraints of both time and resources. Also, randomization was not used in this questionnaire to correct or lessen order biases, indicating three cat food ads were not randomly presented each time ads were displayed or responses taken through repeat measures of a randomized order. Instead, both the advertisements and questions regarding them always appeared in the same order, with Purina first, Iams second followed by Whiskas.
|
Terminology
Ad Index Score : |
Summation of positively worded phrases or descriptors of Likert items that were identified by respondents as seeing the ad. |
Attitude Score: |
Summation of Likert items of which loaded into Factor 1 (looking for absolute values greater than 0.5 in each variable rows). |
Brand Index Score: |
Summed score of 10 Likert items, rated after viewing the ads for each brand. |
Change Score : |
Scores viewers rated after seeing the ad less scores viewers reported before seeing the ad. This indicates how much people change preferences on brand by the pre-to-post ad exposure. |
Communality : |
Summed squared factor loadings over all the factors. |
Eigenvalue: |
Sum of squared factor loadings for a given factor. When an Eigenvalue is divided by the number of variables in the factor analysis, the result of percent of variance accounted for by a factor. |
Factor Score: |
Equals to Factor. Loadings times Ratings on questions which represent underlying dimensions that summarize the original set of variables. |
Factor loadings: |
Relationship between the original variable and the factors; they explain the nature of a particular factor. |
Down Movers: |
Respondents whose ratings on brand decreased after viewing advertisement. |
Group Centroid: |
Also, called the Centroid of a Cluster. The average point of discriminant scores within each group. |
Likert Items: |
10 items (statements) describing about each brand are rated by respondents on a 5 point Likert scale (strongly agree/agree/neither agree nor disagree/disagree/strongly disagree) after exposure to the ad. |
Same: |
Respondents whose scores remained the same for brand after viewing advertisement. |
Statistical Significance |
Indicates that in 85 or more samples drawn from the same population as this sample, the values would be expected to be the same as those reported here. Rule of thumb; p < 0.15. |
Up Movers: |
Respondents whose ratings on brand increased after viewing advertisement. |
Rule of Reporting Statistical Results
Statistical Description
; Mean, Standard Deviation, Percentile, Variance, mode, etc.
- One place beyond the measurement
For example, 3.492; report as 3.5
Statistical Inference (How to infer things about the population from the sample drawn)
; Chi-Squared, T-Ratio, F-Ratio, etc.
- Two place beyond the measurement
For example, 3.492; report as 3.49
|
|
