Attraction and Compromise Effects in Choice: Moderating Influences and Differential Loss Aversion to Quality and Nonquality Attributes

Timothy B. Heath, University of Pittsburgh
Subimal Chatterjee, University of Pittsburgh
[ to cite ]:
Timothy B. Heath and Subimal Chatterjee (1993) ,"Attraction and Compromise Effects in Choice: Moderating Influences and Differential Loss Aversion to Quality and Nonquality Attributes", in NA - Advances in Consumer Research Volume 20, eds. Leigh McAlister and Michael L. Rothschild, Provo, UT : Association for Consumer Research, Pages: 475.

Advances in Consumer Research Volume 20, 1993      Page 475


Timothy B. Heath, University of Pittsburgh

Subimal Chatterjee, University of Pittsburgh

Two studies in this session assessed the importance of the type of comparisons made when making decisions. Sen and Johnson found that experimentally created reference brands were chosen more often than non-reference brands. More importantly, referent designation caused systematic variations in search patterns. Comparisons involving a particular brand were more common when it was the referent, especially when losses were involved. Simonson, Nowlis, and Lemon manipulated type of potential comparisons in two different ways across three experiments. In one case, comparisons were manipulated by having some consumers engage in pairwise comparisons ({A,B}, {B,C}, and {A,C}) prior to choosing from the set of all three brands ({A,B,C}). Pairwise comparisons increased the market share of the least expensive brand. Similar effects were found from variations in the ways in which alternatives were displayed.

Kahn, Huber, and Holbrook presented a number of experiments in this session demonstrating an unexpected effect of time pressure. Consumers were initially given well-known brand names within a product category and asked to indicate their preferences. They were later given a choice involving these brands where (1) information on price and quality was present, and (2) their favorite brand was highest on both attributes. It was expected that consumers would simplify time-pressured decisions by choosing the brand that they had previously indicated as their favorite. However, the opposite pattern held. Time pressure decreased the tendency to choose the previously favored brand. The results stimulated a vigorous discussion of potential mechanisms, although no clear-cut explanation emerged.

The session also included Chatterjee and Heath's meta-analysis of research reporting that asymmetrically dominated entrants can increase the share of existing brands (the attraction effect; Huber, Payne, and Puto 1982). Such entrants were found to consistently increase the share of higher quality brands, but to rarely increase the share of lower quality brands. Attraction to lower quality brands occurred only when, relative to the entrant, the target brand was superior on the quality dimension and at the same level on the non-quality dimension.

The studies reported here illustrate the importance of anchoring and adjustment processes in multiattribute decisions and are therefore consistent with theories of referent-dependent choice (Tversky and Kahneman 1991). They further support the principle of loss aversion in general, and more specifically the growing body of evidence indicating that loss aversion to quality dimensions is more severe than loss aversion to nonquality dimensions (e.g., Hardie, Johnson, and Fader 1991). Differential loss aversion despite initial choices of lower quality alternatives suggests the possibility that relative attribute weights shift following choice.

Finally, the effects of paired-alternative presentations and display formats support theories of extremeness aversion (Simonson and Tversky 1992). They suggest (1) that using paired-comparisons in conjoint measurement may inflate estimates of price's importance relative to estimates based on rankings or ratings of the entire choice set, and (2) ways through which marketers can influence consumer decisions.


Hardie, Bruce G. S., Eric J. Johnson, and Peter S. Fader (1991), "Modeling Loss Aversion and Reference Dependence Effects on Brand Choice," Working Paper #91-025, Wharton Marketing Department, University of Pennsylvania, Philadelphia, PA 19104-6371.

Huber, Joel, John W. Payne, and Christopher Puto (1982), "Adding Asymmetrically Dominated Alternatives: Violations of Regularity and the Similarity Hypothesis," Journal of Consumer Research, 9 (June), 90-98.

Simonson, Itamar and Amos Tversky (1992), "Choice in Context: Tradeoff Contrast and Extremeness Aversion," Journal of Marketing Research, 29 (August), 281-295.

Tversky, Amos and Daniel Kahneman (1991), "Loss Aversion in Riskless Choice: A Reference Dependent Model," The Quarterly Journal of Economics, 106 (November), 1039-1062.