The first question on the survey measured unaided awareness of brands in the respective producT&service category. Respondents were asked to list as many brands as they could, from which were recorded: top-of-mind awareness of the brands under investigation and total brand mentions. This question was administered separately from the rest of the survey to ensire unbiased responses. Aided brand awareness and degree of brand familiarity were then measured
The next set of questions elicited brand associations. Respondents were asked to list all descriptive words, thoughts, characteristics, symbols, or images tbat came to mind when the brand was mentioned. From this question, four variables were created: total associations, total positive associations, total neutral and total negative associations. Two trained researchers coded tbe responses. Using the procedvtre advocated by Perreault and Leigh (1989), interjudge reliability was calculated to be .90, with a lower confidence limit for reliability of .82.
Advertising awareness was measured next. Respondents were asked if they had ever seen any advertising for the respective brand. (It was decided not to put any time constraints on advertising exposure, since such a question would be difficult to answer and impossible to verify.) A follow-up question for subjects who reported advertising awareness asked for a description of what tbe advertising said or showed. Perceived quality was asked last in this sectioti.