You are on page 1of 6

Incidental and Task-Related Affect: A Re-Inquiry and Extension of the Inuence of Affect on Choice

NITIKA GARG J. JEFFREY INMAN VIKAS MITTAL*


We examine the interplay between incidental affect and task-related affect in the context of consumer choice. Specically, we examine the differential impact of two discrete negative affective statesanger and sadnessvis-a-vis a neutral affective ` state. We replicate Luces (1998) nding that people are more likely to rely on a status quo option when they have to make emotionally difcult trade-offs. However, incidental affect moderates this effect such that angry individuals are more inuenced by task-related affect, while sadness is less inuenced by it. These ndings support our thesis that consumers experiencing different negative emotions display differential reliance on avoidance choice strategies such as choosing the status quo.

esearch over the past 25 yr. has demonstrated the pervasive inuence of incidental affect on phenomena such as information acquisition (Lee and Sternthal 1999), encoding (Bower, Gilligan, and Monteiro 1981), recall and recognition (Isen et al. 1978), risk assessments (Johnson and Tversky 1983; Keller, Lipkus, and Rimer 2002), and attitudes (Cohen and Areni 1991). Most research has examined the impact of general positive and negative affective states on decision processes. However, recent research has demonstrated the importance of examining specic emotions rather than global feelings (DeSteno et al. 2000; Lerner and Keltner 2000). In this research, we examine the interplay between two sources of affect, task-related affect and incidental affect, in the context of consumer choice. In a decision-making context, task-related emotions are related to the decision difculty, such as when the consumer has to make an emotionally difcult trade-off (Luce, Bettman, and Payne 2001). One of the ways that people cope with this negative affect is by adopting avoidance strategies such as choosing a status quo option (Luce 1998) or relying on dominant alternatives

*Nitika Garg is assistant professor, School of Business, University of Mississippi, University, MS 38677 (nitika@olemiss.edu). J. Jeffrey Inman is the Albert Wesley Frey Professor of Marketing (jinman@katz.pitt.edu) and Vikas Mittal is associate professor (vmittal@katz.pitt.edu), both at the Joseph M. Katz Graduate School of Business, University of Pittsburgh, Pittsburgh PA 15260. Please address all correspondence to the rst author. The authors appreciate the positive and constructive feedback of the editor, the associate editor, and the reviewers.

(Garbarino and Edell 1997). Thus, task-related affect originates from factors that are embedded in the decision task itself. Incidental affect, by contrast, arises from factors external to the decision task (e.g., a consumer may be shopping after being cut off in the parking lot and be in an angry state). Such incidental affect can inuence the decision process and subsequent judgments by inuencing information processing (Lerner and Keltner 2000; Schwarz and Clore 1983; Wegener and Petty 1994). For example, Tiedens and Linton (2001) suggest that whereas anger leads to heuristic processing, sadness leads to relatively more systematic processing. Our research makes two important contributions. First, it shows that the two sources of affect interact to inuence consumers choice of a status quo option. It claries that the source of affect matters and shows what happens when consumers experiencing a particular incidental emotion face a decision where they need to make emotionally difcult trade-offs. Specically, we not only replicate Luces (1998) ndings, which suggest that in the absence of incidental affect, task-related affect leads to increased preference for the status quo option, but also show that this effect changes in the presence of incidental affect. Second, by examining two affective states of the same valenceanger and sadnesswe are able to argue that various incidental states moderate the inuence of task-related affect differently, even when these incidental affect states are of the same valence. Our results show that task-related affects inuence is exacerbated when individuals are ex154
2005 by JOURNAL OF CONSUMER RESEARCH, Inc. Vol. 32 June 2005 All rights reserved. 0093-5301/2005/3201-0014$10.00

INFLUENCE OF AFFECT ON CHOICE

155

periencing anger and is weakened when they are experiencing sadness. By moving beyond valence and examining the role of affective states that have similar valence but which induce different decision processes, we demonstrate that similarly valenced emotions should not be treated interchangeably. Exploring the relative inuence of anger and sadness is important as they are perhaps the two most common negative emotions associated with consumption experiences. From a theoretical perspective, anger and sadness are very interesting, because even though both have a negative valence, they arise from different appraisal tendencies that lead to differential impact on a subjects degree of certainty, control, and responsibility (Smith and Ellsworth 1985). In the following sections, we use relevant literature to develop hypotheses and describe the results of the study. We then discuss the implications of the ndings, avenues for future research, and the limitations of the current work.

H1: A main effect of trade-off difculty will emerge for individuals in a neutral incidental affective state: they will be less likely to choose the status quo option under low trade-off difculty than under high trade-off difculty.

Incidental Affect and Information Processing


Early research considers affect valence as the primary dimension and is therefore more relevant for comparing general positive and negative affect (Schwarz and Clore 1983; Wegener and Petty 1994). Recently, researchers in social cognition (DeSteno et al. 2000; Lerner et al. 2003; Lerner and Keltner 2000, 2001; Raghunathan and Pham 1999; Tiedens and Linton 2001) have established differences not only between negative and positive affect but also between different discrete emotions that have the same valence (e.g., anger vs. sadness, pride vs. happiness). Lerner and Keltner (2000, 477) introduced the appraisaltendency approach, which hypothesizes that each emotion activates a cognitive predisposition to appraise future events in line with the central appraisal dimension that triggered the emotion. Six cognitive dimensions (i.e., certainty, pleasantness, attentional activity, control, anticipated effort, and responsibility) were found best to dene the appraisal patterns of different underlying emotions (Smith and Ellsworth 1985). Linking specic incidental affect and information processing more closely, Tiedens and Linton (2001) focus on the certainty dimension and posit that the certainty associated with emotions inuences peoples certainty in subsequent judgments. They argue that an emotion associated with greater relative certainty (regardless of the valence) promotes heuristic processing, whereas one associated with the feeling of uncertainty promotes systematic processing. For example, emotions associated with certainty (vs. uncertainty) result in greater reliance on message source expertise, stereotypes, and less attention to argument quality. Anger and sadness are found to differ primarily on three appraisal dimensions (Smith and Ellsworth 1985): certainty, control, and responsibility. Certainty refers to the extent to which a person understands and is certain about what is happening in the situation. Control refers to the extent to which a person believes that a human agency is in control of the situation. Responsibility refers to the degree to which one feels responsible for what is happening in the situation. Even though anger and sadness differ on these three dimensions, we focus on the certainty dimension. We argue that in a consumer decision setting, an emotion associated with a higher degree of certainty such as anger (vs. sadness) will differentially inuence the decision process. Compared with sadness, anger is associated with a higher degree of certainty. This suggests that angry individuals should process information more heuristically, whereas sad individuals should process it more systematically (Tiedens and Linton 2001). How might this difference in information processing across these two affective states interact with task-related affect to differentially inuence consumers

HYPOTHESIS DEVELOPMENT Task-Related Affect and Choice


Task-related affect is affect that is generated because of the characteristics of the activity or undertaking. In a typical study it is manipulated by increasing or decreasing the difculty entailed in arriving at a judgment or choice. For example, Garbarino and Edell (1997) nd that when a task requires excessive cognitive effort on the decision makers part, it generates negative affect, and this reduces the probability of a subject choosing the alternative that is difcult to evaluate. Further, this leads participants to choose a dominant alternative if one is available. Luce (1998) reports that in the face of trade-off difculty in a decision context, participants experience negative emotions and cope with them by choosing an avoidant response such as prolonging a search, maintaining the status quo, or choosing a dominant alternative. This mitigates the negative affect generated by task complexity. Luce, Payne, and Bettman (1999) hypothesize that consumers are motivated by a desire to cope with emotional trade-off difculty, which results in predictable shifts in choice patterns. They argue that increases in the degree of emotional trade-off difculty associated with a choice task lead to increases in the likelihood that the status quo option or one that is best on the more-difcult-to-trade-off attribute is chosen. In summary, task-related affect has been shown to inuence decisions signicantly by motivating decision makers to fall back on the default option that requires little cognitive effort or on easy-to-justify options (e.g., status quo, dominant alternative). In this study we examine choice patterns among consumers in angry, sad, or neutral affective states when they face decisions low or high in emotional tradeoff difculty. Our rst hypothesis is a replication of Luces (1998) nding that consumers in an emotionally difcult decision tend to adopt avoidant strategies such as choosing a status quo option:

156

JOURNAL OF CONSUMER RESEARCH

choice of the status quo? We seek to answer this question in our research.

STUDY
The objectives of this study are twofold. First, we examine how incidental affect moderates the impact of task-related affect on choice (hypotheses 13). Second, we analyze the differences among anger and sadness on the appraisal dimensions (i.e., certainty, responsibility, and control) to ascertain that these specic discrete states were indeed being manipulated.

The Inuence of the Two Sources of Affect on Choice


Under high trade-off difculty conditions, participants in the anger condition face two forces. The rst is the increase in their certainty because of the incidental affect, and the second is the negative affect generated by the task. The rst should make participants in the anger condition process the information more heuristically (Tiedens and Linton 2001), while the second should make them want to escape the negative affect-generating situation. The choice of a status quo option can be a simplifying heuristic, and it also enables these participants to move away from the task-related affect. Combined with their sense of increased certainty (Lerner and Keltner 2001), this should make them more condent in their choice and, in turn, result in their being more comfortable with their heuristic choice of the status quo when pressurized by the task-related affect. This implies that when compared with individuals in the neutral condition, angry individuals in the high trade-off difculty condition should be more likely to choose the status quo option. Under low trade-off difculty condition, by contrast, individuals in the anger condition are under less pressure to maintain the status quo as the trade-offs are simple. This implies that angry individuals would be less likely to choose the status quo in the low (vs. high) trade-off difculty condition. Hence, H2: An interaction between incidental affect and taskrelated emotional trade-off difculty will emerge when comparing angry versus neutral individuals: angry individuals will be more affected by the task-related emotional trade-off difculty as compared with those in the neutral group. Bodenhausen, Sheppard, and Kramer (1994) found that angry individuals relied more on the source cue (heuristic) in a persuasive message than did sad individuals. Tiedens and Linton (2001) explain these results by linking differences in certainty across these two similarly valenced emotions to information processing. Sad individuals experience relatively lower levels of certainty as compared with angry individuals (Smith and Ellsworth 1985), implying that they will process the information relatively more carefully and systematically (Tiedens and Linton 2001). This tendency to process systematically should lead sad individuals to be less susceptible to the inuence of task-related affect: H3: An interaction between incidental affect and taskrelated emotional trade-off difculty will emerge when comparing sad versus neutral individuals: sad individuals will be less affected by the taskrelated emotional trade-off difculty as compared with those in the neutral group.

Participants and Experimental Design


In exchange for extra credit, 178 undergraduates at a large mid-Atlantic university took part in the study. A 3 (incidental affect: anger, sadness, and neutral) # 2 (emotional trade-off difculty level: high, low) between-subjects design was used. The dependent variable was the product choice that participants made (status quo versus others).

Incidental Affect Induction


Incidental affect was manipulated using a methodology rened by Lerner and Keltner (2001). Different affective states were induced by asking the participants to relate a personal incident that had evoked the target emotion (anger, sadness, or neutral). Participants were asked to describe in detail the one thing that makes (or made) them most angry or sad. They were encouraged to describe the incident in such a way that another person reading the description might experience the same emotion. Participants in the neutral condition were asked to rst describe three to ve things that they did that day and then describe in detail what their typical day is like and the activities they undertake on a routine day. A pretest with 61 participants suggests that the manipulation was effective. Once the participants had written up the incident, they were asked to indicate how they were feeling on a seven-point scale. The scale was anchored by (1) denitely do not feel the emotion and (7) denitely feel the emotion. These measures included three items for sadness (sad, gloomy, and down) and four items for anger (angry, annoyed, frustrated, and irritated). For angry participants (n p 21), the anger score was higher than the sadness score (Manger p 4.56 vs. Msad p 2.62; t 20 p 6.37, p ! 0.01), whereas for participants in the sadness condition (n p 19), the sadness score was higher than the anger score (Manger p 3.71 vs. Msad p 4.84; t18 p 3.89, p ! 0.01). By contrast, participants in the neutral group showed no difference between their anger and sadness scores (Manger p 2.39 vs. Msad p 2.25; t 20 p 0.79, NS). Participants in the anger condition reported a signicantly higher score for anger than did participants in the sadness condition (t 38 p 1.96, p ! 0.05). Similarly, participants in the sadness condition reported a greater score for sadness than did those in the anger condition (t 38 p 5.47, p ! 0.01).

INFLUENCE OF AFFECT ON CHOICE

157
FIGURE 1 CHOICE PROPORTIONS OF THE STATUS QUO OPTION

Trade-Off Difculty Stimuli


To manipulate trade-off difculty, we adopted Luces (1998) procedure and stimuli. That is, we inuenced the low and high emotional trade-off difculty through the attributes and their levels to give us two sets of cars, one for each condition. Car A was the status quo option (an avoidant response). On the basis of pretests, Luce found that occupant survival and pollution caused were high trade-off difculty attributes, whereas routine handling and sound system were the two low trade-off difculty attributes, but of equal average importance. The high and low trade-off attributes were altered across the high and low trade-off difculty conditions, respectively, to yield different available options to the participants.

Procedure
The experiment consisted of two parts, the affect induction part and the choice task. To dissociate these tasks, the former was presented to the participants as a life-events survey while the latter was called a marketing survey. The incidental affect manipulation was administered rst. After the participants described the relevant incident, they completed the emotional manipulation checks pertaining to the three relevant appraisal-tendency dimensions. We used the complete list of eight items from Smith and Ellsworth (1985) to capture the three dimensions. Once the participants completed the affect induction, they undertook the choice task. Before they attempted the choice task, participants were introduced to the four attributes that they had to consider in making their choice along with the denitions and the best and worst levels of these attributes. After this they completed an affect manipulation check as well as the importance ratings of the four attributes. The affect measures were adapted from Luce (1998). Luce used the Positive and Negative Affect Schedule (PANAS) scale (Watson, Clark, and Tellegen 1988) and added some more negative adjectives, such as worried and troubled, for capturing the postchoice emotion. We retained measures from the PANAS scale and added adjectives suited to represent anger and sadness emotions such as down, angry, gloomy, and sad. praisal dimensions successfully differentiate between the different discrete emotions.

Postchoice Affect Manipulations Checks. After making their choice, participants were asked to report how they would feel if they had to make the same choice for real. We elicited their response on 11 negative emotion adjectives and six positive emotion adjectives, based largely on the PANAS scale. Cronbachs alpha for the 11 negative affect adjectives was a neg p 0.86, so we created a composite measure of the average negative affect for each subject. Luce (1998) predicts that when participants in the high trade-off condition are offered the option of choosing an avoidant response, the negative affect generated from the task should dissipate. In keeping with this reasoning, the average negative affect was not signicantly different across the low (n p 91) and high (n p 84) trade-off difculty conditions (Mlo p 1.94 vs. Mhi p 1.84). Importance Ratings Manipulation Checks. The importance ratings for the four attributes across the six cells were not signicantly different. This is consistent with what Luce (1998) found in her sample. That is, although the attributes across high and low trade-off difculty conditions were similar in importance, the associated emotional difculty in trading off on these attributes was different.1 Choice Proportions. Figure 1 shows the proportion of participants choosing the status quo option in each cell. The data were analyzed using a 3 # 2 logistic regression. As expected, we found a signicant main effect of trade-off difculty (x 2 (1) p 14.49, p ! .01). The proportion of participants choosing the status quo is signicantly higher under high (n p 86) trade-off difculty than under low (n p 92) trade-off difculty (Mhi p .57 vs. Mlo p .30). The main effect of incidental affect is not signicant (x 2 (2) p 2.34,
1 The 24 means and standard deviations for the importance ratings across the six cells are available from the authors.

Results
Emotional Appraisal-Tendency Measures. This analysis revealed the expected differences across the two emotions, anger (n p 55) and sadness (n p 63). Certainty was signicantly higher for angry participants than for sad participants (anger p 7.16 vs. sadness p 6.12, F(1, 116) p 6.97, p ! .01), as were perceptions of individual control (anger p 5.96 vs. sadness p 4.35, F(1, 116) p 11.95, p ! .01). By contrast, self-responsibility attribution for the situation was higher for participants in the sad condition than for those in the anger condition (anger p 2.64 vs. sadness p 4.91, F(1, 116) p 32.09, p ! .01). Thus, the ap-

158

JOURNAL OF CONSUMER RESEARCH

p p .31). Collectively, hypotheses 2 and 3 predict an interaction between incidental affect and task-related affect. Consistent with our thesis, the interaction between incidental affect and emotional trade-off difculty is signicant (x 2 (2) p 10.05, p ! .01). Hypothesis 1 predicts that under high (vs. low) trade-off difculty, individuals in the neutral group would choose the status quo more often, replicating Luces (1998) results. To test this hypothesis, we conducted a one-tailed z-test of proportions between the low (n p 31) and high (n p 29) tradeoff difculty conditions for the neutral group. The difference between the two conditions is signicant (Mhi p .52 vs. Mlo p .28; z p 1.94, p ! .05). These results replicate Luces (1998) and support hypothesis 1. Hypothesis 2 predicts that in comparison with neutral participants, angry individuals will be more affected by the task-related emotional trade-off difculty and that this will result in a signicant interaction between the two sources of affect for angry and neutral groups. In the anger condition, participants reacted strongly to the task-related affect, and the proportion for status quo choice under high (n p 24) trade-off difculty is greater than under low (n p 31) tradeoff difculty (Mhi p .71 vs. Mlo p .13). A one-tailed z-test of proportions for the anger group shows that the difference between high and low trade-off difculty conditions is signicant for the group (z p 4.38, p ! .01). Importantly, a 2 # 2 logistic regression comparing the angry and neutral groups reveals not only a main effect of trade-off difculty (x 2 (1) p 18.52, p ! .01) but also a signicant interaction between trade-off difculty and incidental affect (x 2 (1) p 3.95, p ! .05). This suggests that angry individuals were more affected by emotional trade-off difculty than individuals in the neutral condition. Thus, hypothesis 2 is supported. Hypothesis 3 predicts that sad (vs. neutral) individuals will be less affected by the emotional trade-off difculty and that they will adopt avoidant strategies to the same extent irrespective of the level of the trade-off difculty. As can be seen from gure 1, the proportion of participants choosing the status quo is similar for participants in the sad group across the two conditions (Mhi p .52 vs. Mlo p .50; Ns p 31 and 32, respectively), whereas participants in the neutral condition chose the status quo to a greater extent under high trade-off difculty than under low trade-off difculty (Mhi p .52 vs. Mlo p .28). We ran a 2 # 2 logistic regression comparing the sad and neutral groups along with z-tests to test for differences in proportions across conditions. While the interaction between task-related affect and incidental affect for these two groups fails to reach statistical signicance (x 2 (1) p 1.68, p p .19), a z-test (one-tailed) between sad and neutral conditions in the low trade-off difculty condition reveals that the difference between the sad versus neutral individuals is signicant (Msad-lo p .50 vs. Mneutral-lo p .28; z p 1.80, p ! .05). Also, as predicted, a ztest of proportions between the low and high trade-off difculty conditions for sad individuals found that the differ-

ence was not signicant (z p .13, p p .90). This is consistent with hypothesis 3.

DISCUSSION
Our study examines the interplay between incidental affect (i.e., irrelevant affect) and task-related affect (i.e., relevant affect). It shows that incidental affect moderates the inuence of task-related affect on consumer choice. Thus, the incidental affect that participants bring to the decision situation moderates the process and outcome of such decisions. We also show that anger and sadness have differential moderating inuence even though they are similarly valenced. This suggests the need to enhance the simple valence perspective and to examine different discrete emotions individually. We found that participants in the neutral group increased their reliance on the status quo under the high trade-off difculty scenario, successfully replicating Luces (1998) ndings. We also found that participants in the sad and anger groups behaved differently in terms of the choices they made and that this resulted in a signicant interaction between the two sources of affect for these groups. This validates our premise that incidental affect exerts a signicant impact on participants assessment of the task at hand and leads to a differential choice pattern. Closer examination of the sadness group highlights another interesting nding. The choice proportion for this group exhibits an almost 50-50 split between cars A and B. Luce (1998) designed the stimuli such that most participants would select cars A or B based on the attributes and the attribute values. Thus, it seems that participants in this group are examining the stimuli relatively more closely and making an informed choice, exhibiting little status quo bias. Interestingly, the differences in the proportion of participants choosing the status quo across the three affective states narrowed in the high trade-off difculty condition. We speculate that this is because of various reasons depending on the different affective states (e.g., angry individuals might choose status quo as a heuristic choice, whereas sad individuals choose it because of its attributes). However, a reasonable explanation for this might be that the inuence of incidental affect is stronger in the low tradeoff difculty condition, but the presence of a salient, second source of affect (task-related) tends to bring the groups closer in the high trade-off difculty condition. Our ndings offer implications for consumer welfare. Consumers are differentially inuenced by discrete emotions that are seemingly irrelevant to the decision task. This brings up the question of whether the consumer is making the best choice in the circumstances. This needs to be studied further to enable consumers to adopt better choice strategies that increase their welfare. For practitioners, a clear implication is that this knowledge can be leveraged by associating a relevant affect with the product. This might be accomplished by inducing a certain emotion through product advertisements (e.g., humor, fear appeals, or sadness) and creating strong associations between the ad and the product at the

INFLUENCE OF AFFECT ON CHOICE

159 Isen, Alice M., Thomas E. Shalker, Margaret Clark, and Lynn Karp (1978), Affect, Accessibility of Material in Memory and Behavior: A Cognitive Loop? Journal of Personality and Social Psychology, 36 (January), 112. Johnson, Eric J. and Amos Tversky (1983), Affect, Generalization and the Perception of Risk, Journal of Personality and Social Psychology, 45 (July), 2031. Keller, Punam A., Isaac M. Lipkus, and Barbara K. Rimer (2002), Depressive Realism and Health Risk Accuracy: The Negative Consequences of Positive Mood, Journal of Consumer Research, 29 (June), 5769. Lee, Angela Y. and Brian Sternthal (1999), The Effects of Positive Mood on Memory, Journal of Consumer Research, 26 (September), 11527. Lerner, Jennifer S., Roxana M. Gonzalez, Deborah A. Small, and Baruch Fischhoff (2003), Effects of Fear and Anger on Perceived Risks of Terrorism: A National Field Experiment, Psychological Science, 14 (March), 14450. Lerner, Jennifer S. and Dacher Keltner (2000), Beyond Valence: Toward a Model of Emotion-Specic Inuences on Judgment and Choice, Cognition and Emotion, 14 (July), 47393. (2001), Fear, Anger and Risk, Journal of Personality and Social Psychology, 81 (July), 14659. Luce, Mary F. (1998), Choosing to Avoid: Coping with Negatively Emotion-Laden Consumer Decisions, Journal of Consumer Research, 24 (March), 40933. Luce, Mary F., James R. Bettman, and John W. Payne (2001), Emotional Decisions: Trade-Off Difculty and Coping in Consumer Choice, Monographs of the Journal of Consumer Research, No. 1, Chicago: University of Chicago Press. Luce, Mary F., John W. Payne, and James R. Bettman (1999), Emotional Trade-Off Difculty and Choice, Journal of Marketing Research, 36 (May), 14359. Raghunathan, Rajagopal and Michel T. Pham (1999), All Negative Moods Are Not Equal: Motivational Inuences of Anxiety and Sadness on Decision Making, Organizational Behavior and Human Decision Processes, 79 (July), 5677. Schwarz, Norbert and Gerald L. Clore (1983), Mood, Misattribution, and Judgments of Well-Being: Informative and Directive Functions of Affective States, Journal of Personality and Social Psychology, 45 (September), 51323. Smith, Craig A. and Phoebe C. Ellsworth (1985), Patterns of Cognitive Appraisal in Emotion, Journal of Personality and Social Psychology, 48 (April), 81338. Tiedens, Larissa Z. and Susan Linton (2001), Judgment under Emotional Uncertainty: The Effects of Specic Emotions on Information Processing, Journal of Personality and Social Psychology, 81 (December), 97388. Watson, David, Lee A. Clark, and Auke Tellegen (1988), Development and Validation of Brief Measures of Positive and Negative Affect: The PANAS Scales, Journal of Personality and Social Psychology, 54 (June), 106370. Wegener, Duane T. and Richard E. Petty (1994), Mood Management across Affective States: The Hedonic Contingency Hypothesis, Journal of Personality and Social Psychology, 66 (June), 103448.

point of purchase. This would perhaps lead to consumers reexperiencing the relevant emotion.

Limitations and Future Research


Our analysis reveals that differences in choices existed across incidental affect conditions. Although the appraisal dimensions differentiated the two affective states, they did not mediate the choice pattern. The reason for this could be that we are looking at the interaction of incidental affect with task-related affect and that their combined effect is reected in the choice pattern. Our case may not be comparable with past research that has found these dimensions to mediate the effect of incidental affect on their dependent variables, such as risk assessment, because this research examined incidental affect in isolation (e.g., Lerner and Keltner 2001). This underlying process needs to be explored further in future research. We showed that incidental affect is a key moderator for the inuence of task-related affect on choice. One of the next steps in this research is to identify other factors that interact with incidental affect. This might lead to the discovery of a common dimension that links these different factors (such as effort) and also helps us understand more about the effect of incidental affect. This research is a step toward understanding the specic individual and situational factors responsible for the inuence of incidental versus task-related affect on decision making. [Dawn Iacobucci served as editor and Stephen Nowlis served as associate editor for this article.]

REFERENCES
Bodenhausen, Galen V., Lori A. Sheppard, and Geoffrey P. Kramer (1994), Negative Affect and Social Judgment: The Differential Impact of Anger and Sadness, European Journal of Social Psychology, 24 (JanuaryFebruary), 4562. Bower, Gordon H., Stephen G. Gilligan, and Kenneth P. Monteiro (1981), Selectivity of Learning Caused by Affective States, Journal of Experimental Psychology: General, 110 (December), 45173. Cohen, Joel B. and Charles S. Areni (1991), Affect and Consumer Behavior, in Handbook of Consumer Behavior, ed. Thomas S. Robertson and Harold H. Kassarjian, Englewood Cliffs, NJ: Prentice-Hall, 188240. DeSteno, David, Richard E. Petty, Duane T. Wegener, and Derek D. Rucker (2000), Beyond Valence in the Perception of Likelihood: The Role of Emotion Specicity, Journal of Personality and Social Psychology, 78 (March), 397416. Garbarino, Ellen C. and Julie A. Edell (1997), Cognitive Effort, Affect and Choice, Journal of Consumer Research, 24 (September), 14758.

You might also like