consumer research study sanitary napkins

elaine lee

ma candidate in advertising

the university of texas at austin

fall 2006

table of contents

executive summary

introduction

methodology

analysis

conclusion

summary

appendix a

appendix b

home

executive summary

An online questionnaire is administered to test the effects of print ads on the respondents’ opinions about the brand.  This is a quantitative study evaluating responses from 18+ year old females.  Three brands of sanitary napkin are tested in this study, Always, Kotex, and Stayfree.

Initially, respondents favor the brand Always over Kotex and Stayfree.  There is not a huge difference between the preference for Kotex and Stayfree, but there is a bigger difference between Always and the other two brands.

After exposure to the print ad for Always, Kotex, and Stayfree, respondents still preferred the brand Always over Kotex and Stayfree.  The print ads do not have a great effect on the brand preference.  The print ad for Always did lower the brand preference for its brand, but the lower score is still above the preference for Kotex and Stayfree.  Always is the preferred brand before and after exposure to the print ads.

home

introduction

This online questionnaire generated 122 responses within a matter of 4 days with the help of friends and family.  The usage of Likert items is prevalent in this questionnaire.  A print ad for Always, Kotex, and Stayfree are presented to the respondents after they gave an initial preference score for these brands.  Then, they are presented with Likert items regarding the brand.  Respondents proceed to answer questions regarding the three ads.

 

Navigate through the table of contents, executive summary, methodology, analysis, conclusion, summary, appendix A, and appendix B for the results of this study.  Appendix A provides the sample questionnaire with the frequencies, means, and standard deviations reported under each question.  Appendix B provides the actual questionnaire administered.

 

home

methodology

questionnaire

section 1

Section 1 consists of an open question asking to list the brand of sanitary napkins last purchased.  Then the respondents divide up ten points among the three brands of sanitary napkins (Always, Kotex, Stayfree) they are most likely to purchase.  These questions are used to find a base score for the respondents’ brand preference.  Next, the respondents examined the three print ads for the three brands of sanitary napkins.

section 2

Section 2 begins by asking the respondents to divide up ten points among the three brands of sanitary napkins after examining the ad.  It continues by asking whether or not the print ads affected their impression of the brands.  The respondents can respond with an open answer.  These questions find the difference in brand preference pre-ad exposure and post-ad exposure.

section 3

Section 3 studies the consumers’ brand impressions through the utilization of ten Likert items regarding each brand of sanitary napkins.  The same questions are consistently asked for each brand.  There are both positive brand impression questions (ex. Always sanitary napkins are comfortable.) as well as negative brand impression questions (ex. Always sanitary napkins do NOT fit in my pockets.).  These questions help determine the driving factors behind each consumer’s brand preference.

section 4

Section 4 studies the consumers’ impression of the print ads for each brand.  A list of ten adjectives is presented.  Respondents check all applicable adjectives for each brand (ex. Creative, boring, hard to believe, easy to understand).  Next, a list of ten statements is presented and the respondents check all applicable statements regarding the print advertisements for each brand (ex. This ad is annoying.).  The results from this section can yield a connection or lack there of between the print ads and the brand preference.

section 5

Section 5 gathers demographic information of the 122 respondents.  Age, level of education, products used during menstrual cycle, and additional comments are collected.  This information helps break down respondents into different groups that can help determine if demographics have an effect on brand preferences.

design

This is a quantitative, cross-sectional study.  The questionnaire focuses on the pre and post ad exposure brand impressions to determine whether or not the print ads for each band has an effect on the respondents regarding their preference for a certain brand of sanitary napkin.  The usage of a questionnaire helps the respondents provide their honest answers due to the fact that they are not connected to the survey they answered.

sampling

The sample of 122 respondents derives from a network of females starting from friends to strangers throughout the United States.  With the convenience of forums and online-networking sites such as MySpace, data is quickly gathered within a few days.  Because this is a female product, only females who are 18+ are used in this study.

sample message to generate responses

i know you still have your period

you'll be interested in my consumer brand preference study i'm doing.

and you'll want to take this survey because it will prevent me from having left ventricular hypertrophy.

i need 60 surveys filled by tomorrow. i have 40 so far. and i really mean it when i say that every single one of you who will take the survey helps me more than you think.

thanks girls, and guys...who can get other girls to take this survey.

(for FEMALES 18+)

http://ciadvertising.org/sa/fall_06/adv380j/elaine/elaine.html

data collection

Quantitative data is collected online using ColdFusion to quickly determine statistical scores.  Respondents have the opportunity to provide qualitative answers where appropriate, leading to further insight not questioned through the survey.  The fact that it is an online questionnaire accessed through a website, technical difficulties tend to occur with connection and website development.

home

analysis

basic statistics

The means, brand index score, and ad index scores are all calculated using the ten Likert items for each brand to analyze the relationship between each brand and pre-post ad exposure.

Correlated t-test.

The means of the three brand index scores are calculated using the scores of the ten Likert items for each brand.

Sample size = 122

Always has a brand index score of 35.1 while Kotex and Stayfree have approximately similar scores of 30.8 and 30.1 respectively.   The respondents favor the brand Always over Kotex and Stayfree.  Kotex and Stayfree have about the same audience appeal in this sample of 122 respondents.

*p ≤ 0.15

The mean of each brand’s index score is compared to another, resulting in a t-ratio indicating that Always is preferred above Kotex and Stayfree.  Because the three t-ratios are significant, in 85 or more samples out of 100 samples drawn from the same population as this sample of 122 respondents, it will be expected that the brand index scores for the pairs Always and Kotex, Always and Stayfree, and Kotex and Stayfree will be about what they are here as they will be in the total population.

 

Between groups t-test.

The between groups t-test measures the movement in brand index score and ad index score after the respondents examine the print ad for Always sanitary napkins.

Brand Index Score Change for Always Sanitary Napkins

*p ≤ 0.15

Ad Index Score Change for Always Sanitary Napkins

*p ≤ 0.15

After exposure to the Always sanitary napkin print advertisement, 23 respondents report a higher brand index score and ad index score from their initial score while 46 respondents report a lower brand index score and ad index score from their initial score.  The 46 respondents who moved down after seeing the print ad have a higher brand preference for Always than the 23 respondents who moved up, with the mean of 36.8 and 32.7 respectively.  The means for the ad index scores, 8.6 for up movers and 7.5 for down movers, show that the 23 respondents who moved up like the print ad more than the 46 respondents who moved down.  The down movers, even though they like the print ad less, still prefer the brand Always more so than the up movers prefer the brand. The up movers like the print ad, but prefer the brand less than the down movers prefer the brand.

The t-ratios are significant, meaning that in 85 or more samples out of 100 samples drawn from the same population as this sample of 122 respondents, it will be expected that the means for the brand index score and the ad index score of up and down movers will be what they are here as they will be in the total population.

 

Chi-Squared for Always Sanitary Napkins.

The chi-squared test finds whether or not there is a significant relationship between up, same, and down movers of the pre-exposure/post-exposure test for the Always print ad and being above and below the median brand index score of 36.

Median = 36

Thirty respondents who moved down after seeing the print ad for Always are above the median brand index score of 36.  The print ad for Always slightly, negatively affects the 30 respondents’ preference for the brand, but the respondents show brand loyalty by reporting above the median brand index scores for Always. 

 

Chi-Squared

*p ≤ 0.15

The chi-squared value of 17.17 is significant with 2 degrees of freedom.  In 85 or more samples out of 100 samples drawn from the same population as this sample of 122 respondents, it will be expected that the relationship between up, same, and down movers of pre-exposure/post-exposure to the Always print ad and being above and below the median brand index score of 36 will be what it is here as it will be in the total population.

 

Move Frequencies.

This frequency test calculates the number of respondents who moves up, stays the same, and moves down in brand index score for each brand after viewing the print ads for each brand.

Sample size = 122

The majority of the 122 respondents reports the same brand index score before and after viewing the print ads, with the highest number of respondents, 67, reporting the same brand index score for Stayfree after seeing the Stayfree print ad.  Kotex’s print ad positively influences 48 respondents to report a higher brand index score for the Kotex brand, while Always’ print ad negatively influences 46 respondents to report a lower brand index score for the Always brand.

 

Comparison between Always and Stayfree.

In this sample size of 122 respondents, 87 respondents report higher brand index scores for Always than for Stayfree.

 

Correlation Coefficient between Kotex Brand Index Score and Stayfree Brand Index Score.

Kotex brand index score and Stayfree brand index score are compared to find a relationship between the two brands.

Sample size = 122

*p ≤ 0.15

There is a positive and moderate direct correlation between Kotex’s brand index score and Stayfree’s brand index score.  As brand index score for Kotex goes up, the brand index score for Stayfree also goes up.  Respondents interested in Kotex are also somewhat interested in Stayfree.  The coefficient is significant, meaning that in 85 or more samples out of 100 samples drawn from the same population as this sample of 122 respondents, it will be expected that the correlation score for Kotex and Stayfree will be about what it is here as it will be in the total population.

 

Correlation comparing Always and Kotex sanitary napkins for Tampon Users.

The brand index scores for Always and Kotex sanitary napkins are compared among the respondents who use tampons instead of sanitary napkins.

*p ≤ 0.15

The significant correlation value of -0.18 indicates a negative and weak correlation between Always and Kotex brand index scores for sanitary napkins among tampon users.  As the brand index score for Always move up, the brand index score for Kotex moves down.  Even though this correlation is weak, it is significant.  In 85 or more samples out of 100 samples drawn from the same population as this sample of 122 respondents, it will be expected that the correlation score between Always and Kotex brand index scores among tampon users will be about what it is here as it will be in the total population.

 

 

analysis

home

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

regression analysis

A multiple linear regression analysis is conducted for each brand of sanitary napkins to examine the relationship between pre-post ad exposure change score (dependent variable) and the ten Likert brand attribute items (independent variables).

 

Sample size = 122

*p ≤ 0.15

Always

The multiple correlation coefficient (R) for Always sanitary napkin is 0.5.  There is a moderate connection between the change score of pre-post exposure to the ad and the ten Likert brand attribute items.  The coefficient of multiple determination (R²) for Always is 23.8%, with a standard error of the estimate of 2.4.  The low variance indicates a low relationship between the pre-post ad exposure change score to the ten Likert brand attribute items.  The brand attributes do not explain the change score.  The standard error of the estimate of 2.4 is the average distance respondents report away from the regression equation.  The F-ratio of 3.46 is significant, meaning in 85 or more samples out of 100 samples drawn from the same population as this sample of 122 respondents, it will be expected that the multiple correlation coefficient (R) for Always will be about what it is here.

Kotex

The multiple correlation coefficient (R) for Kotex sanitary napkin is 0.3.  There is a weak connection between the change score of pre-post exposure to the ad and the ten Likert brand attribute items.  The coefficient of multiple determination (R²) for Kotex is 7.8%, with a standard error of the estimate of 2.5.  There is a low relationship between the pre-post ad exposure change score to the ten Likert brand attribute items.  The brand attributes do not explain the change score.  The standard error of the estimate of 2.5 is the average distance respondents report away from the regression equation.  Because the F-ratio of 0.94 is not significant, the multiple correlation coefficient (R) for Kotex cannot be generalized to the population.

Stayfree

The multiple correlation coefficient (R) for Stayfree sanitary napkin is 0.4.  There is a weak connection between the change score of pre-post exposure to the ad and the ten Likert brand attribute items.  The coefficient of multiple determination (R²) for Stayfree is 15.4%, with a standard error of the estimate of 1.9.  There is a low relationship between the pre-post ad exposure change score to the ten Likert brand attribute items.  The brand attributes do not explain the change score.  The standard error of the estimate of 1.9 is the average distance respondents report away from the regression equation.  The F-ratio of 2.02 is significant, meaning in 85 or more samples out of 100 samples drawn from the same population as this sample of 122 respondents, it will be expected that the multiple correlation coefficient (R) for Stayfree will be about what it is here.

 

*p ≤ 0.15

b = unstandardized regression coefficient

Beta = Standardized regression coefficient

The Beta values indicate the relative importance of the Likert items (independent variables) on the pre-post ad exposure change score (dependent variable).  Every Beta value equal or greater than the absolute value of the highest Beta value of each brand divided by two is considered important and marked in red.

Always

Always sanitary napkin has negative correlations between the variables Good Brand, Not Fit in Pocket, and Brand of Choice and the pre-post ad exposure change score.  As the values for these three variables increase, the score for pre-post ad exposure decrease.  The more the respondents believe Always is a good brand, does not fit in their pockets, and is their brand of choice, the less they like the ad.  For the rest of the positively correlated variables, as the values for the variables increase, the scores for pre-post ad exposure increase.  The important variables for Always, with Beta values equal or greater than 0.15, are Good Brand, Not Fit in Pocket, Brand of Choice, and Appealing Package.  These four variables have significant t-ratios, meaning in 85 or more samples out of 100 samples drawn from the same population as this sample of 122 respondents, it will be expected that the b values (unstandardized regression coefficients) for these variables for Always will be about what it is here.

The regression equation for Always sanitary napkin is:

Pre-post ad exposure change score of Always = 2.2 + (-0.7) (good brand) + (-0.1) (comfortable) + (-0.1) (against odor) + (-0.5) (not fit in pocket) + (-0.1) (prevent leaks) + 0.3 (not dry) + 0.3 (expensive) + (-0.5) (brand of choice) + 0.0 (high quality) + 0.5 (appealing package)

Kotex

Kotex sanitary napkin has negative correlations between the variables Good Brand, Not Fit in Pocket, Expensive, and Brand of Choice and the pre-post ad exposure change score.  As the values for these four variables increase, the score for pre-post ad exposure decrease.  The more the respondents believe Kotex is a good brand, does not fit in their pockets, is expensive, and is their brand of choice, the less they like the ad.  For the rest of the positively correlated variables, as the values for the variables increase, the scores for pre-post ad exposure increase.  The important variables for Kotex, with Beta values equal or greater than 0.1, are Good Brand, Against Odor, Not Fit in Pocket, Prevent Leaks, Not dry, Expensive, Brand of Choice, High Quality, and Appealing Package.  Out of these nine variables only Good Brand, Brand of Choice, and Appealing Package have significant t-ratios, meaning in 85 or more samples out of 100 samples drawn from the same population as this sample of 122 respondents, it will be expected that the b values (unstandardized regression coefficients) for these variables for Kotex will be about what it is here.

Stayfree

Stayfree sanitary napkin has negative correlations between the variables Not Dry, Brand of Choice, and High Quality and the pre-post ad exposure change score.  As the values for these three variables increase, the score for pre-post ad exposure decrease.  The more the respondents believe Stayfree does not keep them dry, is their brand of choice, and is high in quality, the less they like the ad.  For the rest of the positively correlated variables, as the values for the variables increase, the scores for pre-post ad exposure increase.  The important variables for Stayfree, with Beta values equal or greater than 0.1, are Against Odor, Not Dry, Expensive, Brand of Choice, High Quality, and Appealing Package.  Of these six variables only Not Dry, Brand of Choice, and Appealing Package have significant t-ratios, meaning in 85 or more samples out of 100 samples drawn from the same population as this sample of 122 respondents, it will be expected that the b values (unstandardized regression coefficients) of these variables for Stayfree will be about what it is here.

 

 

analysis

home

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

discriminant analysis

This discriminant analysis measures how well the Likert Items (independent variables) classify the respondents into Up Movers and Down Movers (dependent variables).  The brand Stayfree for sanitary napkins is used in this discriminant analysis.

 

Stayfree Discriminant Analysis

n = sample size

SD = Standard Deviation

In this discriminant analysis, 31 respondents (Up Movers) increase their preference score for Stayfree after viewing Stayfree’s print ad while 24 respondents (Down Movers) decrease their preference score.  On average, the Up Movers have a slightly lower mean score than the Down Movers, though the similar mean scores of both groups do not yield a good discriminant analysis.  Every Standardized Discriminant Function Coefficient equal or greater than the absolute value of the highest Standardized Discriminant Function Coefficient divided by two is considered important and marked in red (0.9/2= 0.45).  The Likert Items Not Dry and Comfortable are relatively important in predicting to the respondents’ changed score pre-post ad exposure.

 

Wilks’ Lambda

*p ≤ 0.15

 

Group Centroids

The Wilks’ Lambda tests the significance of the group centroids (the average discriminant scores by group, also known as Z-scores).  The group centroid for Up Movers is -0.5 and for Down Movers is 0.6.  In this discriminant analysis, the Wilks’ Lambda value of 0.79 with an insignificant Chi-squared value of 11.53 indicates the group centroids cannot be projected to the population.  In 85 or more samples out of 100 samples drawn from the same population as this sample, it will be expected that the group centroids will not be about what it is here.

 

Classification Matrix

74.5% of original grouped cases correctly classified

*p ≤ 0.15

Classification matrix charts Actual Movers against Predicted Movers.  There is a total of 29 Actual and Predicted Up Movers and a total of 12 Actual and Predicted Down Movers.  A total of 41 respondents out of 55 respondents in this analysis accurately moves in the predicted direction, resulting in a moderately high classification accuracy of 74.5%.  This analysis is considered relatively good according to the classification accuracy.  The significant t-ratio of 2.74 indicates that in 85 or more samples out of 100 samples drawn from the same population as this sample, it will be expected that the classification accuracy will be about what it is here.  The classification accuracy in this discriminant analysis can be projected to the population.

 

Conclusion:

This discriminant analysis is not very good nor is it strong because the mean scores between the Up and Down Movers are similar and the group centroids are not significant, but the percentage of classification accuracy is moderately strong and can be projected to the population.

 

 

analysis

home

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

anova

ANOVA (analysis of variance) is used to find the relationship between one or more categorically measured independent variable and one and only one continuously measured dependent variable.

A two-way ANOVA is used for this research; therefore there are two independent variables.

 

Descriptive Statistics

Independent Variables:  Age 21 and Under VS Age Over 21

Up Movers, Same Movers, Down Movers

Dependent Variable:  High Quality of Stayfree sanitary napkins