Guest Speaker:
Barbara E. Kahn
Professor of Marketing, The Wharton School, University of Pennsylvania
Topic:
Coping with Ambivalence: The Effect of Removing a "Fence Sitting" Option on Consumer Attitude and Preference Measurement
Chairperson:
Associate Professor of Marketing Seshan Ramaswami
School of Business, Singapore Management University
Venue:
Business Block, Level 2, Seminar Room 4
Singapore Management University
469 Bukit Timah Road, Singapore 259756
Date:
Friday, 5 July 2002, at 4.00pm
Reservation:
This seminar is free. Places are limited. Please confirm your attendance by Thursday, 4 July 2002, 12 noon with Ms. Lim Lih Yeng at lylim@smu.edu.sg or telephone: 6822-0197.
About the Seminar:
The concepts of attitude and preference remain among the most important in consumer and social psychology. Attitudes are typically based on consumers' evaluation of both positive and negative components about a stimulus, and as such are typically conceptualized and measured as lying along a bipolar continuum that ranges from unfavorable to favorable. For example, attitudes and preferences are measured in consumer and marketing research through self-reports in which respondents are asked to indicate their beliefs or feelings towards an object or class of objects on bipolar scales, such as Likert (or summated ratings) scales, semantic differential scales, or itemized ratings scales. One of the critical issues in survey and marketing research is whether or not to allow respondents the option of "sitting on the fence," i.e., not committing to a positive or negative position. In attitude measurement, allowing respondents to sit on the fence corresponds to the use of odd-point scales, where there is a middle response alternative (hereafter referred to as a "neutral position"). In contrast, the use of even-point scales, where a neutral point is not offered, is akin to forcing respondents to choose a position, or to "jump off the fence". The traditional idea suggests that the qualitative results between the two scales are unaffected since if the respondents are truly neutral, then they will randomly choose one or the other, so forcing them to choose should not bias the overall results.
In contrast to this traditional viewpoint, we identify the conditions under which the exclusion of a neutral position will shift the relative distribution of responses in a systematic manner. As long as the evoked reactions are (1) neither positive nor negative (i.e., close to indifference), or (2) either positive or negative, then using these bipolar scales with or without a neutral position may be appropriate and the traditional assumptions behind using odd or even scales may be acceptable. However, if respondents feel strongly conflicted between the positive and negative aspects of an object, then we hypothesize that the decision as whether or not to include a neutral option becomes more important. In such instances, the removal of the neutral position option in even-point scales may result in a predictable, systematic bias in the expression of attitudes. This conclusion results in the uncomfortable suggestion that attitude measurement can be systematically manipulated by changing the scale from odd to even.
We hypothesize that if consumers are experiencing ambivalence or feeling conflicted about making tradeoffs between the positive and negative components involved in attitude formation, they are likely to alleviate this discomfort by engaging in effort or conflict reducing heuristics if the "fence sitting" alternative is removed. In particular, if an attitude or preference evaluation requires making difficult tradeoffs, consumers forced to make the decision are likely to resort to a lexicographic rule and choose the option that is superior on the more important attribute. To understand this relationship more fully we present the findings of six studies, which compare the distribution of responses when there is a middle option in an odd-point scale to the distribution of opinion when the middle option is not present in an even-point scale. In these studies we examine factors that moderate the relationship, as well as provide evidence for the underlying process.
