Mood-Induced Processing Styles in Judgment

Other Judgment and Decision Tasks

These mood-induced differences in spontaneous processing style are likely to affect individuals’ performance on many judgment and decision tasks. For example, Luce, Bettman, and Payne (1997) observed that “decision processing under increasing negative emotion both becomes more extensive and proceeds more by focusing on one attribute at a time” (p. 384), consistent with the assumption that negative feelings foster a more detail-oriented processing style. Moreover, Hertel, Neuhof, Theuer, and Kerr (2000) observed pronounced mood effects on individuals’ decision behavior in a chicken game. Consistent with the present theorizing, their findings suggest that individuals in a happy mood are likely to heuristically imitate the behavior of other players, whereas individuals in a sad mood base their moves on a rational analysis of the structure of the game.

Note, however, that more extensive reasoning about a task does not always result in better solutions. As an example, consider the well-known phenomenon of anchoring effects in quantitative judgments (Tversky & Kahneman, 1974). In a typical study, participants are provided with an arbitrary number (e.g., 50, 5,000) and are asked to make a binary comparative judgment (e.g., “Is the Mississippi River longer or shorter than 5,000 miles?”). Next, they must provide an absolute estimate (e.g., “How long is the Mississippi River?”). The latter estimate is typically biased toward the anchor value. As Strack and Mussweiler (1997; Mussweiler & Strack, 1999) demonstrated, this bias is due to a process of positive hypothesis testing (e.g., “Can the Mississippi River really be that long/short?”) that renders hypothesis-consistent information selectively accessible. If so, this bias may be exaggerated the more individuals engage in elaborative hypothesis testing. Confirming this prediction, Bodenhausen, Gabriel, and Lineberger (1999) observed that sad individuals showed more pronounced anchoring effects than individuals in a neutral mood, presumably because their sad mood fostered more elaborative hypothesis testing, thus rendering more hypothesis-consistent information accessible.

As these examples illustrate, the differential processing styles elicited by being in a happy or sad mood can increase as well as decrease judgmental biases, depending on the nature of the task. Future research will benefit from testing the impact of affective states across a wider variety of tasks.