Special Session Summary How Do I Prefer Thee? Let the Task Decide For Me: Examining Task Effects and Consumer Preference Formation

Ziv Carmon, Duke University
Stephen M. Nowlis, Arizona State University
[ to cite ]:
Ziv Carmon and Stephen M. Nowlis (1997) ,"Special Session Summary How Do I Prefer Thee? Let the Task Decide For Me: Examining Task Effects and Consumer Preference Formation", in NA - Advances in Consumer Research Volume 24, eds. Merrie Brucks and Deborah J. MacInnis, Provo, UT : Association for Consumer Research, Pages: 72.

Advances in Consumer Research Volume 24, 1997      Page 72



Ziv Carmon, Duke University

Stephen M. Nowlis, Arizona State University

Previous research has shown that the task, or the manner in which consumer preferences are elicited, can significantly affect these preferences (e.g., Payne, Bettman, and Johnson 1992). Examples of task effects that have been documented over the years include effects of time pressure, of the response mode, and of scaling. Different theoretical perspectives have attempted to explain these results, with varying degrees of success. The major objective of this session was to improve our understanding of why, how and when task characteristics influence consumer preferences by examining some of the central task effects. To accomplish this, each of the papers in the session focused on a different aspect of task effects. One focused on the way in which preferences are measured, another on the perceived goal of elicitation tasks, and the third on the effect of justification and anticipated evaluations. Each aimed to improve our understanding of how and why these factors affect the resulting preferences.

The first paper, by Alan Cooke and Barbara Mellers, examined how the response mode and the context jointly affect elicited preferences. Respondents evaluated apartments described by monthly rent and distance to campus. Both response tasks and contexts varied, and preferences for the same pairs of apartments reversed in different contexts (holding the task constant) and in different tasks (holding the context constant). The authors found support for their suggestion that context appears to influence attribute perceptions, whereas the response task appears to impact the weights of the attributes. The paper provided a better understanding of the differential impact of the task and the context, which are typically studied separately.

The second paper, by Greg Fischer, Ziv Carmon, Dan Ariely, and Gal Zauberman investigated the effect of the perceived goal of the response task on differences in the prefeences that it elicits. Focusing on the Prominence Effect (greater weighting of the prominent attribute in choice versus in matching), they compared preferences elicited by several variants of choice and of matching tasks. They found strong support for their hypothesis that equivalent tasks that have different underlying task-goals result in significantly different preferences. In fact, when they directly manipulated the perceived goal of two tasks the preferences were significantly altered. They concluded with a brief review of other task effects that can be better understood in light of respondents’ underlying task-goals.

The third paper, by Itamar Simonson and Stephen M. Nowlis, looked at how aspects of the social situation can impact consumer preferences. Specifically, social interactions following a purchase decision have two key components: the buyer explains the choice made to some audience (e.g., superior, spouse) and the audience forms an evaluation of the decision. This paper proposed that, although these task components have often been studied simultaneously (often referred to as accountability), they have independent and often conflicting effects on buyer behavior. Specifically, a need to provide reasons was hypothesized to promote behavior that is designed to appear rational and unique, whereas anticipation of evaluations by others encourages conformity to norms. Consistent with the analysis, the results of the studies demonstrated a wide range of choice phenomena for which providing reasons and being evaluated by others lead to systematically different purchase behaviors. The studies also provided insights into the mechanisms underlying these effects.

The discussion leader, Donald Lehmann, suggested that research on task effects was initially mostly concerned with documenting their existence. At the current stage the focus has turned more to a deeper understanding of when and why these effects occur. He suggested that future research investigate how substantial these effects are (e.g., in terms of market inefficiencies). Another direction would be to study how to control for these effects either in terms of "inoculating" decision makers or countering these biases with the use of decision aids.


Payne, J. W., J. R. Bettman, and E. J. Johnson (1992), "Behavioral Decision Research: A Constructive Processing Perspective," Annual Review of Psychology 43, 87-131.