Asymmetric strength of preference in the domains of gains and losses

In comparative judgments that follow binary choices, judgments of "How much better is a preferred option?" and "How much worse is a less preferred option?" may differ in their magnitudes. I call such differences "valence effects," and analyzed cognitive processes that underlie positive valence effects ("Better" exceeding "Worse") and negative valence effects ("Worse" exceeding "Better"). My analyses used a "focus shift model." The focus shift model postulates that choice options are represented as sets of desirable and undesirable features. Difference judgments are reached by assigning subjective weights onto such features and integrating weighted feature contributions. Positive and negative valence effects reflect the differences in subjective weighting depending on valence of judgments. Experiments 1 and 2 showed that systematic positive valence effects were observed in the domain of gains, and negative valence effects in the domain of losses. Estimates of subjective weights showed that valence effects occurred when subjects heavily weighted desirable features in preferred options and undesirable features in less preferred options. These patterns of subjective weighting for positive and negative valence effects were consistent with the focus shift model. Experiments 3 through 5 used the directions of valence effects as a diagnostic tool to assess decision makers' interpretation of choice tasks. Previous studies (Tversky & Kahneman, 1986) showed that, in binary choices between certain and uncertain options, preferential orders switched from certain options in the domain of gains to uncertain options in the domain of losses. I found that preferences for certain options were associated with positive valence effects, whereas preferences for uncertain options were associated with negative valence effects. Experiments 4 and 5 examined conditions under which such reversals in preferential ordering ceased to occur. I found that such reversals ceased to occur when decision makers maintained consistent task interpretations as pertaining to gains or to losses. Possibilities for extending the current findings to understanding related economic and psychological phenomena are discussed.