Structure of the questionnaire

Throughout the survey, respondents were permitted to move only forward, not backward. Certain responses were required and the respondent could not advance without answering them.

Introduction

The survey opened with a brief introduction, assuring respondents that the survey was for educational purposes, that it asked no questions about personal medical history and that all responses would be kept strictly confidential.

Pre-exposure (Steps 1 and 2)

In Step 1, the respondent was asked to identify, without prompting, the last brand of over-the-counter allergy medication, over-the-counter liquid antacid and over-the-counter pain relief medication she purchased. If she had never purchased one of these products or had forgotten which brand she purchased last, she was asked to name the first brand that came to mind.

In Step 2, the respondent was asked to divide ten points on a constant-sum scale between three brands in each of these categories: over-the-counter allergy medication, over-the-counter liquid antacid and over-the-counter pain relief medication. Only the over-the-counter pain relief medication scores were used in this analysis. The respondent was asked to divide these 10 points immediately before and immediately after viewing the three ads under consideration. The difference between the score for each brand before and after exposure is referred to in this study as the "change score."

Advertising exposure (Step 3)

In Step 3, the respondent was shown one full-page, four-colour ad for each of the three brands (Advil, Bayer and Tylenol) for fifteen seconds. The exposure was timed and the respondent could not return to the ad after viewing it. The ads can be seen by clicking the links below:

Post-exposure (Steps 4 and 5)

After viewing the ads, the respondent was asked to rate her likelihood of purchasing each of the brands using the same constant-sum scale as in Step 2. In Step 5, the respondent was asked if her impression of any of the brands changed after viewing the ads and, if so, which one.

Brand attitudes (Steps 6, 7 and 8)

For each of the three brands, the respondent was presented with ten Likert items asking her if she strongly agreed, agreed, neither agreed nor disagreed, disagreed or strongly disagreed with ten statements about the brand. Although some questions were phrased negatively, scores were coded so that higher scores (with 5 being the highest and 1 being the lowest) indicated more positive perceptions of the brand.

Advertising attitudes (Steps 9 and 10)

The next two steps asked questions about the advertisements the respondent was shown earlier. In Step 9, the respondent was asked to check all the boxes beside the words that she felt applied to the advertising for each brand. Similarly, in Step 10, the respondent was asked to check all the boxes beside the phrases she felt applied to the advertising for each brand. The Advertising Index Score was calculated by counting the number of boxes that were checked representing positive attributes.

Demographics and lifestyle (Step 11)

The respondent was asked basic demographic questions including her sex, age, household income and level of education. She was also asked three questions to determine how and how often she used over-the-counter pain relief medication.

Design

The study was designed to measure brand perceptions and the impact of advertising using a copy-testing method that took measurements before and after exposure to the ads. Respondents were asked to score three brands immediately before and after viewing ads for those brands. Other attitudes were measured after exposure to the ad. Due to time constraints and limited resources, this study was conducted without a control group with which to compare responses.

Sampling

Due to time and resource constraints, this study used a convenience sample made up mostly of friends and colleagues. They were contacted using the following e-mail:

Yes, you too can help me fail my research methods class! All you need to do is ignore this e-mail.

If, on the other hand, you don't want me to fail my research methods class, please take ten minutes to complete my online survey at http://tinyurl.com/34novt This completely confidential survey is short and easy to complete. But if I don't receive a certain number of responses, the consequences for my grade will be dire.

Besides, four out of five doctors recommend taking my survey for fast, effective relief of a burning need to express your opinion.

I really appreciate your help. But you can help me still further: please ask a few of your friends and colleagues to complete this survey too! All you need to do is send them to http://tinyurl.com/34novt I need as many responses as I can get.

Questions? Comments? E-mail cameronfleming@mail.utexas.edu.

Best,

Cameron

Respondents were in turn asked to forward the link to the survey to their own friends and colleagues, creating a snowball sample. Sixty responses were considered sufficient for analysis, and 127 responses were collected, of which 126 were complete enough for analysis.

Online data collection

The data for this survey were collected using the online questionnaire in Appendix B. Respondents accessed the questionnaire using the link contained in an e-mail sent to friends and colleagues. The responses were collected using Cold Fusion in a Microsoft Access database, which was imported into SPSS for analysis.