Affect-as-Information Definition

How do we know whether or not we approve of some action or like some person? According to the affect-as-information hypothesis, our feelings provide such information. Just as our smiles and frowns provide information about our reactions to others, our positive and negative feelings provide such information to our-selves. Like many psychological processes, emotional appraisals are generally unconscious. Hence, having evaluative information available from affective feelings can be highly useful.

Affect-as-InformationAffective reactions are forms of evaluation, and experiencing one’s own affective reactions provides information that something good or bad has been encountered. Such information can be compelling, because it may involve not only thoughts but also feelings, bodily reactions, and even action. Thus, specific emotions, like embarrassment, involve distinctive thoughts, feelings, and expressions, whereas general moods are less differentiated. Affective states can be thought of as having two components—affective valence, which provides information about how good or bad something is, and affective arousal, which signals its importance or urgency. Most research focuses on valence, but recent studies also examine arousal. They find that assessing events as important causes a release of adrenaline, which results in its consolidation into long-term memory. Thus, people remember well the events of September 11, 2001, but perhaps not so well those of September 10, 2001. For victims of highly traumatic events, such arousal-powered memories can become stressful and even disabling.


Psychologists have traditionally argued that attitudes and evaluations depend on people’s beliefs about what they are judging. In the early 1970s, social psychologist Charles Gouaux examined how variation in feelings (from mood-inducing films) and beliefs (about another person’s political opinions) influenced liking. Gouaux found that the affective feelings of one person influenced attraction or dislike of another over and above the influence of the cognitive beliefs about that person.

But even after many demonstrations that affect influences attitude, the assumption persisted that such evaluative judgments must reflect evaluative beliefs. Positive or negative feelings were assumed to activate positive or negative beliefs about the person, which in turn influenced judgment. In contrast, the affect-as-information view said that evaluative judgments are often made simply by asking oneself, “How do I feel about it?”

As part of a study of this process, people were telephoned and asked questions about their life satisfaction. They were called on early spring days that were either warm and sunny or cold and rainy. People reported more positive moods and greater life satisfaction on sunny than on rainy days. The explanation was that the weather influenced satisfaction ratings, because people misattributed their feelings about the weather as feelings about their “life as a whole.” To test this explanation, experimenters said they were calling from another city, so that they could ask some respondents, “How’s the weather down there?” When respondents’ attention was directed to the weather, the mood influences on life satisfaction disappeared. Asking about the weather did not influence the feelings themselves, but it did influence their apparent meaning. The experiment established that affect could influence evaluative judgment directly by conveying information about value.

Since emotions are rapid reactions to current mental and perceptual content, people generally know what their emotions are about. But the causes of moods and depressed feelings are often unclear. Without a salient cause, feelings become promiscuous, attaching themselves to whatever comes to mind. As a result, the affect from moods can influence judgments, and enduring feelings of depression and anxiety can create a discouraging and threatening world.

These considerations suggest that many influences of affect depend on the attributions that people make for their feelings, rather than on the feelings themselves. To study this process, experiments often encourage misattributions of feelings from their true source to a different object. Efforts to get people to misattribute their feelings are also common in everyday life. For example, advertisers often pair products with exciting or suggestive images to foster misattribution of that excitement to the product being marketed.

Despite the fact that experiments and advertising are sometimes designed to fool people, social psychologists generally view affect as adaptive and functional, in contrast to traditional views of affect as a source of irrationality and bias. Emotion does sometimes conflict with rational choice, but affect is also essential to good judgment. Studies of neurological damage show that the inability to use affective reactions to guide judgments and decisions is costly. Similarly, research on emotional intelligence suggests that being able to extract information from one’s own and others’ affective reactions is highly beneficial.

Decision Making

Psychologists now believe that the process of decision making takes place largely unconsciously. As a result, deciding explicitly often involves entertaining alternatives until one is visited by a feeling that one has decided. When ordering food from a menu or selecting a video to watch, one may look until something feels right. Thus, decisions are hard when none of the alternatives feels right or when more than one alternative elicits such feelings. Making important decisions in the absence of an experience of rightness may therefore be stressful. For men and women considering marriage, for example, saying yes without feeling anything would surely be anxiety provoking.

A well-known model and actress recently described her devastation when, after realizing her lifelong dream of having a baby, she felt nothing as she held her new daughter. Feelings of attachment, intimacy, and nurturance are so basic to birth and motherhood that the woman concluded from their absence that she was profoundly unworthy. She even considered suicide, but fortunately, treatment for postpartum depression allowed the appropriate feelings to arise. Only then could she say confidently that she loved her daughter or herself.


The affect-as-information hypothesis assumes that people’s feelings inform them about what they like, want, and value. When a belief that one values something is not validated by embodied affective reactions, the person is faced with an epistemic problem. Such disparities between affective beliefs and embodied affect have been studied in the laboratory. Investigators have developed simple procedures for activating happy or sad thoughts and also for eliciting feelings, facial expressions, and actions characteristic of happiness and sadness. They find that when people’s cognitions and affect do not agree, their ability to remember presented material suffers, as does the speed with which they can make simple choices. From the standpoint of cognitive efficiency, when thinking sad thoughts, it is apparently better to feel sad than to feel happy. Just as people’s beliefs about the world are subject to validation by what they see and hear, so too do evaluative beliefs appear to require validation by one’s own feelings, expressions, and actions.


Affect guides not only judgments and decisions but also attention and styles of thinking. During task performance, affect may be experienced as information about the task or about how one is doing, rather than as information about how much one likes something. Such task information leads to adjustments in cognitive processing or cognitive tuning. Research suggests that positive affect promotes global, interpretative processing and negative affect leads to local, perceptual processing. Thus, whether one focuses on the forest or the trees and whether one uses one’s own mental associations or not appear to be controlled by affect. Since many of the phenomena that have defined cognitive psychology involve reliance on such cognitive responses, it turns out that many of them are not observed in sad moods. Research shows that such textbook phenomena as categorization, stereotyping, persuasion, impression formation, false memory, heuristic reasoning, and others are all more apparent in happy moods than in sad moods. Ultimately, whether it is better to be happy or sad when engaged in cognitive tasks depends on the nature of the task. Positive affect may promote creativity and performance on constructive cognitive tasks, but it may promote error on some detailed tasks such as solving logical syllogisms. These effects too have been found to depend on the attributions that participants make for their affect.


According to the affect-as-information view, people are informed by their affect, even though they produce it themselves. Moreover, rather than being fixed and reflex-like, affective influences can often be altered by simple cognitive manipulations. Thus, the information value of the affect, rather than the affect itself, is often the critical factor in its influence. This view can also be generalized to nonaffective feelings. For example, the information from bodily feelings of pain depends on attributions about its source (e.g., where it hurts). Likewise, cognitive feelings of the ease of recalling something influence whether it seems true. Moreover, the impact of these feelings also depends on attributions about their source.


  1. Gasper, K., & Clore, G. L. (2002). Attending to the big picture: Mood and global vs. local processing of visual information. Psychological .Science, 13, 34-40.
  2. Gouaux, C. (1971). Induced affective states and interpersonal attraction. Journal of Personality and Social Psychology, 20, 37-43.
  3. Martin, L. L., & Clore, G. L. (Eds.). (2001). Theories of mood and cognition: A user s handbook. Mahwah, NJ: Erlbaum.
  4. Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and judgments of well-being: Informative and directive functions of affective states. Journal of Personality and Social Psychology, 45, 513-523.