Special Session Summary Beyond Valence: Negative Affect and Its Effects on Consumer Decision Making


Nitika Garg (2003) ,"Special Session Summary Beyond Valence: Negative Affect and Its Effects on Consumer Decision Making", in NA - Advances in Consumer Research Volume 30, eds. Punam Anand Keller and Dennis W. Rook, Valdosta, GA : Association for Consumer Research, Pages: 232-235.

Advances in Consumer Research Volume 30, 2003     Pages 232-235



Nitika Garg, University of Pittsburgh


Anger, sadness, anxiety, fearBwhat do they have in common? All these are emotions that possess a negative valence. Does that mean that their effect on variables such as risk assessment, choice, and perception of future events is the same? This is the question that this session addressed. All the three papers presented in the session highlighted the impact of different discrete negative emotions (e.g., anger, fear, sadness) on a variety of decision-related dependent variables.

Researchers in the areas of social cognition and marketing have been examining incidental or context-based emotion and its effect on different aspects of consumer decision making. However, it is only recently that researchers have begun to explore the impact of different types of discrete emotions that have the same valence. The interesting finding across these studies was that same valence emotions differentially impact consumers’ decision making. The papers in the session combined the insights from both marketing as well as the social cognition literature to present a broader spectrum of research that is being currently undertaken to draw new insights about consumer behavior.

Some researchers have looked at negative affect and its various components (Pham and Raghunathan, 1999; Pham et al, 2001; Keltner, Ellsworth, and Edwards, 1993; Lerner and Keltner, 2000; Lerner and Keltner, 2001; Tiedens and Linton, 2001). The research presented in this session built on this literature and combined it with learnings from other research areas such as decision complexity (Luce, 1998; Bettman, Luce and Payne, 1998; and Luce, Payne and Bettman, 1999) and affect and cognition (Zajonc, 1980; Ortony, Clore, and Collins, 1988). By integrating different research areas, the studies drew on a richer knowledge base and thereby increased our understanding of consumer decision making. Specifically, the objective of the session was to present a range of studies that draw on different theoretical bases and methodological aspects to highlight various facets of negative affect and thus present a more holistic view.

In the first paper, Michel Tuan Pham (Columbia University), Rajagopal Raghunathan (University of Texas at Austin) and Tamar Avnet (Columbia University), empirically examined the interface between affect and cognition. The authors investigated whether affect and cognition tap into separate systems in relation to the environment or instead simply represent different inputs operating within a single system. The authors suggested that one should look not at the phenomenology of "feeling" versus "thinking" or the correlation between feelings and thoughts, but at the functional properties of the two types of judgment inputs. The results of two studies were reported, along with theoretical and practical implications.

The second paper examined the interactive effect of task-related negative affect and incidental negative affect on consumer choice. The paper was co-authored by Nitika Garg, Jeff Inman, and Vikas Mittal (University of Pittsburgh). Specifically, in this research the authors explored the impact of different incidental negative emotions namely, anger and sadness, on consumer choice under high and low emotional trade-off difficulty. The authors argued that these negative emotions impact subject’s degree of certainty and control causing subjects to display differential reliance on various avoidance choice strategies such that angry versus sad subjects exhibit greater reliance on status-quo under high emotional trade-off difficulty. This is because subjects in the anger condition process the information more heuristically and thus, rely on the status-quo as a means to avoid the task-related negative affect whereas sad subjects process the information more systematically and exhibit a different choice pattern.

The last paper drew on growing literature that shows how emotions and risk perceptions affect one another. Co-authored by Jennifer S. Lerner, Roxana M. Gonzalez, Deborah A. Small, and Baruch Fischhoff (Carnegie Mellon University), the paper examined how the recent 9/11 terrorist attacks evoked negative emotions like anger and fear in people and its effect on perceptions of risk in a nationally representative field survey. The authors also explored the role of media portrayals of these news events that could evoke these emotions and consequently affect the risk perceptions.

This session brought together several leading scholars (e.g. Fischhoff, Pham, Lerner, Raghunathan) who have been very active in the field of negative affect and decision making. Further, the proposed session offered a variety of implications of negative affect useful for managers to public policy professional and highlighted its impact on people and their behavior in general. By employing a diverse set of methodologies (e.g., experiments, surveys) and settings (field, lab), the papers in the session also enabled an assessment of the findings’ generalizability.



Michel Tuan Pham, Columbia University

Rajagopal Raghunathan, University of Texas at Austin

Tamar Avnet, Columbia University

The interface between affect and cognitionCand relatedly the relation between emotion and reasonChas always intrigued social scientists. That affect and cognition represent different mental phenomena is well accepted. From a phenomenological standpoint, people, laypersons and scientists alike, experience no difficulty differentiating between the two kinds of phenomena. Empirically, research on the structure of attitudes has consistently identified two separable components: (a) an affective component, sometimes called hedonic or experiential, and (b) a cognitive component, sometimes called utilitarian or instrumental. Yet, that affect and cognition can be separated either conceptually or empirically does not necessarily mean that they represent separate systems, as Zajonc (1980) once argued in his well-known essay. For instance, research on emotional appraisal (e.g., Roseman 1991) and on the cognitive representation of emotion (e.g., Ortony, Clore, and Collins 1988) suggests that affect and cognition often interact with one another. The question of whether affect and cognition tap into separate systems of relation to the environment or instead simply represent different inputs operating within a single system remains open.

We suggest that to examine this important but very difficult question, one should look not at the phenomenology of "feeling" versus "thinking" or the correlation between feelings and thoughts, but at the functional properties of the two types of judgment inputs. In other words, regardless of how they are experienced and regardless of whether they are correlated, does "feeling" versus "thinking" make a difference in how people relate to their environment? We report two studies that tentatively suggest that it does.

The first study elaborates on Raghunathan and Pham (1999, OBHDP)’s recent finding that states of anxiety versus sadness provoke marked differences in people’s decisions. In particular, Raghunathan and Pham showed that, when given a choice between a high-risk/high-reward option and a low-risk/low-reward option, anxious individuals tend to prefer the low-risk/low-reward option, whereas sad individuals tend to prefer the high-risk/high-reward option. According to the authors, this is because people often make decisions by asking themselves "how do I feel about" the options (e.g., Pham 1998; Schwarz and Clore 1996). Feelings of anxiety signal that risks and uncertainty are relatively more important, whereas feelings of sadness signal that potential rewards are more important.

An important limitation of the Raghunathan and Pham (1999) studies is that they manipulated anxiety and sadness by having subjects empathize with anxiety-producing and sadness-producing scenarios. It is therefore possible that their results were not driven by feelings of anxiety and sadness per se, but by anxiety and sadness-related cognitions primed by the scenarios. In order to test explanation it is important to examine the effects of feelings of anxiety and sadness independent of anxiety and sadness-related cognitions. In this new study, 161 subjects were randomly assigned to the conditions of a 3 (scenario) x 2 (frame) design. As in Raghunathan and Pham, subjects were first exposed to a detailed scenario, then asked to make a choice between two options: one involving greater rewards and greater risks and one involving lower rewards and lowers risks. The first factor varied whether the scenario was anxiety-producing (learning about the possibility of having cancer), sadness-producing (witnessing a mother’s death), or affect-neutral (experiencing a series of innocuous events). The second factor manipulated the frame of mind subjects adopted while reading the scenarios. In the "hot" frame condition, subjects were asked to empathize with the described situation (as in Raghunathan and Pham 1999). In the "cold" frame condition, subjects were asked to analyze the described situation. It was anticipated that, compared with subjects with the "empathize" frame, subjects with the "analyze" frame would be less likely to actually experience feelings of anxiety or sadness (e.g., Strack et al. 1985). Therefore, while al subjects were exposed to the same cognitive content (which was either anxiety-related, sadness-related, or affect-neutral) only those in the hot (empathize) frame condition were expected to experience genuine feelings. Results from an extended pretest among 86 subjects supported these assumptions. Specifically, they show that subjects exposed to the anxiety or sadness scenarios in the hot-frame condition did experience genuine feelings of anxiety or sadness. Those exposed to the same scenarios in the cold frame condition did not experience these feelings significantly. Results also show that the hot/cold framing manipulations did not influence the type of thoughts subjects had. In other words the framing manipulation was effective in blocking the feelings in the cold-frame conditions without blocking subjects’ scenario-related thoughts.

The results of the main study show that, as predicted, the Raghunathan and Pham findings held in the hot frame condition but not in the cold frame condition. In the hot frame condition, where feelings were presumably indeed experienced, anxious subjects were again more likely to prefer the low-risk/low-reward option than sad subjects who were more likely to prefer the high-risk/high-reward option (neutral affect subjects had intermediate preferences). However, in the cold frame condition, there were no differences: subjects exposed to the anxiety scenario had similar preferences as those exposed to the sadness scenario and those exposed to the neutral affect scenario. These findings suggest that a necessary condition for the Raghunathan and Pham result is the actual experience of feelings of anxiety and sadness. The mere accessibility of anxiety and sadness-related cognitions does not suffice. More generally, this is consistent with the hypothesis of a functional independence between feeling and thinking.

The second study elaborates on recent results by Pham et al (2001, JCR). They recently showed that, contrary to common wisdom, under certain conditions different people agree more on their feelings toward stimuli than they do on their reason-based (cold evaluative) assessments of the same stimuli. An important characteristic of the Pham et al.’s studies is that subjects reported their feelings and reason-based assessments online (i.e., as they were exposed to the stimuli). This new study examines what happens when subjects are asked to make memory-based reports of their feelings or reason-based assessments (using their recollection of the stimuli). The study was based on the popular reality-TV show "SurvivorCAfrica." This show presents a unique opportunity to study the dynamics of judgment. First, previously unknown individuals (the cast members) become celebrities within a span of several weeks. Second, each cast member’s personae is almost uniquely determined by his or her portrayal on the show (i.e., the available information is heavily controlled). Finally, one cast member is voted off the show every week, effectively disappearing from exposure. The study was conducted in the 12th week of the show. Subjects were asked to make either feeling-based or reason-based evaluations of the 16 cast members. As in Pham et al., the analyses focused on how much subjects agree among themselves. Agreement coefficients were calculated (a) for the first eight members to be voted off the show (who had "disappeared" from the show for at least four weeks) and (b) for the last eight members (who were still on the show or had disappeared for less than four weeks). Interestingly, agreement coefficients for reason-based assessments were equally high for the first eight and last eight members of the show. That is, subjects agreement in their reason-based assessments where independent of whether the target has been on the show for a long (last eight) or short (first eight) period of time and whether the target was relatively accessible in memory (last eight) or relatively inaccessible (first eight). In contrast, agreement coefficients of feeling-based judgments were much lower for the first eight members voted off the show than for the last eight. This finding can have two interpretations. The first, less interesting interpretation is that feelings generate high interpersonal agreement only when people have a high amount of information about (here exposure to) the target. Although plausible, this explanation seems unlikely because in the Pham et al. studies subjects had relatively limited exposure. The second and more intriguing possibility is that feelings generate high interpersonal agreement only when the target is very accessible (here for the last eight members), that is, again when feelings are genuinely experienced. Should the second interpretation be confirmed in subsequent studies, this would be further evidence of a functional independence of feeling versus thinking. As in the other study, the experience of feelings seems to influence a person’s mode of relation to the environment in a way that is functionally distinct from the colder representation of the same feelings.

That genuine feelings may have very different functional properties from colder representations of these feelings has important implications for research on affect. This finding would suggest that measurement-based studies of feelings (e.g., questionnaire-based correlational studies) may not really tap into the true affective system.


Ortony, Andrew, Gerald L. Clore, and Allan Collins (1988), The Cognitive Structure of Emotions. Cambridge: Cambridge University Press.

Pham, Michel Tuan (1998), "Representativeness, Relevance, and the Use of Feelings in Decision Making," Journal of Consumer Research, Vol. 25 (September), 144-159.

Pham, Michel Tuan, Joel B. Cohen, John Pracejus, and G. David Hughes (2001), "Affect Monitoring and the Primacy of Feelings in Judgment," Journal of Consumer Research, Vol. 28 (September), 167-188.

Raghunathan, Rajagopal and Michel Tuan Pham (1999), "All Negative Moods are Not Equal: Motivational Influences of Anxiety and Sadness in Decision Making," Organizational Behavior and Human Decision Processes, Vol. 71 (July), 56-77.

Roseman, Ira J. (1991), "Appraisal Determinants of Descrete Emotions," Cognition and Emotion, 5 (3), 161-200.

Schwarz, Norbert and Gerald L. Clore "Feelings and Phenomenal Experiences," in Social Psychology: Handbook of Basic Principles, eds. E. Tory Higgins and Arie W. Kruglanski, New York: The Gilford Press, 433-465.

Strack, Fritz, Norbert Schwarz, and Elisabeth Gschneidinger (1985), "Happiness and Reminiscing: The Role of Time Perspective, Affect, and Mode of Thinking," Journal of Personality and Social Psychology, 49 (6), 1460-1469.

Zajonc, Robert B. (1980), "Feeling and Thinking, Preferences Need No Inferences," American Psychologist, 35 (2), 151-175.



Nitika Garg, University of Pittsburgh

J. Jeffrey Inman, University of Pittsburgh

Vikas Mittal, University of Pittsburgh

Negative affect in decision making has captured the interest of researchers for quite some time in both social cognition and marketing. There are two distinct streams of research. One, which focuses on task-related negative emotion arising due to the decision based complexity and the consequent trad-off difficulty and the other, which looks at the more general, incidental negative affect and its effect on various dimensions, like risk taking and future appraisal.

Task related emotions result from whatever is currently engaging one’s attention. In the decision making context, task related emotions are related to the decision difficulty where the consumer has to make an emotionally difficult trade-off (e.g., Luce, Bettman and Payne, Monographs of JCR, 2001). Luce (1998) argues that trade-off difficulties are a major source of task related emotion and decision difficulty. Luce predicts that consumers caught in a decision involving emotion laden trade-offs experience negative emotions and choose avoidant responses such as prolonging search, maintaining the status-quo and choosing a dominating alternative.

Incidental affect, also known as context-based affect, is affect that a consumer may imbue from her environment in isolation to the complexity of the decision on hand. Recently researchers have begun to look beyond simple emotion valence (Desteno et al., 2000; Lerner and Keltner, 2000, 2001; Tiedens and Linton 2001; Raghunathan and Pham, 1999). Regarding incidental affect, these researchers have found differences not only between negative and positive affect but also between the different discrete emotions having the same valence (e.g., anger/sadness, pride/happiness). Lerner and Keltner (2000, 2001) have developed the 'appraisal-tendency’ approach, which hypothesizes that 'each emotion activates a predisposition to appraise future events in line with the central appraisal dimension that triggers the emotion’. They find that fearful people made pessimistic risk assessments and future events judgments whereas angry people make optimistic assessments in the same scenarios.

The above discussion provides evidence that these two sources of affect have been studied in the literature. However, to the best of our knowledge, these two types of emotions have been studied in isolation. No one has attempted to bring the two together and examine their interaction. The objective of this research then, is to examine the interactive effect of task-related negative affect and context-based negative affect on consumer choice. In the process, the research also aims to replicate existing results for each type of affect separately.

In this research we examine the differential impact of different incidental negative emotionsBanger and sadnessBon consumer choice in the context of high versus low emotional trade-off difficulty. We hypothesize that consumers experiencing different negative emotions will display differential reliance on the avoidance choice strategies such as choosing status-quo, asymmetrically dominated option or delaying the decision, when considering an emotionally difficult trade-off. Our reasoning for the above hypothesis is that anger versus sadness differentially impacts subject’s degree of certainty and control. Specifically, previous research has shown anger to have a higher degree of certainty associated with it as it has a target as compared to sadness or fear (Smith and Ellsworth, 1985). Further, anger is associated with greater degree of control compared to the other two affects. Thus, we expect that in a consumer decision setting, higher degree of certainty and control will lead to higher degree of optimism and more heuristic processing of available information. Consequently, angry subjects should exhibit greater reliance on status-quo under high emotional trade-off difficulty compared to sad subjects who process the information relatively more systematically, leading to a different choice pattern.

In two studies, we test the above hypotheses and show how these different discrete negative emotions influence subjects’ decision making process and consequently their choice under low and high decision complexity situation. Study 1 is a 3 (emotions: angry, sad, neutral) x 2 (emotional trade-off difficulty: high, low) between subjects design. We expect that neutral mood subjects will replicate Luce’s (1998) results. Study 2 builds on study 1 and examines the underlying process across the conditions in greater detail. The implications of these findings to academicians as well as managers are then discussed.


DeSteno, D., Petty, R.E., Wegner, D.T. & Rucker, D.D.(2000). Beyond Valence in the perception of likelihood: The role of emotion specificity, Journal of Personality & Social Psychology, 78(3), 397-416.

Lerner, J.S. & Keltner, D.(2000). Beyond Valence: Toward a model of emotion-specific influences on judgment and choice, Cognition and Emotion, 14(4), 473-493.

Lerner, J.S. & Keltner, D.(2001). Fear, Anger and Risk, Journal of Personality & Social Psychology, 81(1), 146-159.

Luce, M.F.(1998). Choosing to avoid: Coping with negatively emotion-laden consumer decisions, Journal of Consumer Research, 24(March), 409-433.

Luce, M.F., Bettman, J.R. & Payne, J.W.(2001). Emotional decisions: Trade-off difficulty and coping in consumer choice, Monographs of the Journal of Consumer Research, Number 1.

Raghunathan, R. & Pham, M.T.(1999). All negative moods are not equal: Motivational influences of anxiety and sadness on decision making, Organizational Behavior & Human Decision Processes, 79(1), 56-77.

Smith, C.A. & Ellsworth, P.C.(1985). Patterns of cognitive appraisal in emotion, Journal of Personality & Social Psychology, 48, 813-838.

Tiedens, L.Z. & Linton, S.(2001). Judgment under emotional uncertainty: The effects of specific emotions on information processing, Journal of Personality & Social Psychology, 81(6), 973-988.



Jennifer S. Lerner, Carnegie Mellon University

Roxana M. Gonzalez, Carnegie Mellon University

Deborah A. Small, Carnegie Mellon University

Baruch Fischhoff, Carnegie Mellon University

The recent terrorist attacks have intensely affected many individuals and institutions. Equity markets sold off, consumer spending declined, air travel dropped precipitously, and public opinion toward government shifted dramatically. These responses reflected both intense deliberations and intense emotions. The attacksBand prospect of sustained conflict with a diffuse, unfamiliar enemyBcreated anger, fear, and sadness.

A growing literature (for reviews, see Loewenstein and Lerner, in press; Loewenstein, Weber, Hsee, and Welch, 2001; Mellers, Schwartz, and Ritov, 1999; Schwarz and Clore, 1996) shows how emotions and risk perceptions affect one another. Its theories can both illuminate current events and be tested by them. Early research found that positive emotions triggered relatively optimistic risk assessments, while negative emotions evoked more pessimistic ones (Johnson and Tversky, 1983). According to recent theories, however, the distinct cognitive appraisals that evoke some specific negative emotions can differentially color perceptions of future events (Keltner, Ellsworth, and Edwards, 1993; Lerner and Keltner, 2000; Lerner and Keltner, 2001; Tiedens and Linton, 2001). For example, although fear and anger are both negative emotions, only fear arises from (Smith and Ellsworth, 1985) and evokes strong appraisals of uncertainty and situational control (Lerner and Keltner, 2001), two central determinants of risk judgments (Slovic, 1987). Anger, by contrast, is associated with appraisals of certainty and individual control (Lerner and Keltner, 2001; Smith and Ellswoth, 1985).

Based on appraisal-tendency theory (Lerner and Keltner, 2000; Lerner and Keltner, 2001), we predicted that anger and fear would have opposite effects on perceptions of risk. Moreover, we predicted that media portrayals of news events could evoke these emotions and, thereby, influence risk perception. In a nationally representative sample of Americans (N=973, ages 13-88) that matched Census figures on major demographic indicators, fear increased risk estimates and plans for precautionary measures; anger did the opposite. These patterns emerged with both experimentally induced emotions and naturally occurring ones. Males’ risk estimates were less pessimistic than females; differences in emotion explained 80% of the gender difference. Emotions also predicted diverging public policy preferences. Discussion focuses on theoretical, methodological, and policy implications.


Johnson, E. J., & Tversky, A. (1983). Affect, generalization, and the perception of risk. Journal of Personality and Social Psychology, 45, 20-31.

Keltner, D., Ellsworth, P. C., & Edwards, K. (1993). Beyond simple pessimism: Effects of sadness and anger on social perception. Journal of Personality and Social Psychology, 64, 740-752.

Lerner, J. S., & Keltner, D. (2000). Beyond valence: Toward a model of emotion-specific influences on judgment and choice. Cognition and Emotion, 14(4), 473-493.

Lerner, J. S., & Keltner, D. (2001). Fear, anger, and risk. Journal of Personality and Social Psychology, 81(1), 146-159.

Loewenstein, G., & Lerner, J. S. (in press). The role of affect in decision making. In R. Davidson, K. Scherer, & H. Goldsmith (Eds.), Handbook of Affective Science. New York: Oxford University Press.

Loewenstein, G. F., Weber, E. U., Hsee, C. K., & Welch, E. (2001). Risk as feelings. Psychological Bulletin, 127, 267-286.

Mellers, B., Schwartz, A., & Ritov, I. (1999). Emotion-based choice. Journal of Experimental Psychology: General, 128, 332-345.

Schwarz, N., & Clore, G. (1996). Feelings and phenomenal experiences. In E. T. Higgins & A. W. Kruglanski (Eds.), Social Psychology: Handbook of Basic Principles (pp. 453-465). New York: The Guilford Press.

Slovic, P. (1987). Perception of risk. Science, 236, 280-285.

Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in emotion. Journal of Personality and Social Psychology, 48, 813-838.

Tiedens, L. Z., & Linton, S. (2001). Judgment under emotional certainty and uncertainty: The effects of specific emotions on information processing. Journal of Personality and Social Psychology, 81(6), 973-988.



Nitika Garg, University of Pittsburgh


NA - Advances in Consumer Research Volume 30 | 2003

Share Proceeding

Featured papers

See More


Inequality and Market (In)efficiency

Serena Hagerty, Harvard Business School, USA
Michael Norton, Harvard Business School, USA

Read More


M8. Nostalgia Increases Healthy Attitudes and Behaviors

Jannine Lasaleta, Yeshiva University
Carolina O. C. Werle, Grenoble Ecole de Management
Amanda Pruski Yamim, Grenoble Ecole de Management

Read More


Approach and Loss Aversion: Consumer Responses to Approaching and Receding Stimuli in Advertising

Lana Mulier, Ghent University, Belgium
Iris Vermeir, Ghent University, Belgium
Hendrik Slabbinck, Ghent University, Belgium

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.