Behaviors that are linked to emotions range from the very simple, such as defensive reflex actions, to the complex, such as sequences of action tendencies. Emotions likely evolved to produce adaptive actions, such as to approach desired objects or to withdraw from dangerous objects, as well as to support more flexible action tendencies associated with survival. Researchers may take advantage of these behavioral outputs to estimate emotions.
Behavior action tendencies. One behavioral manifestation of emotion concerns the action tendencies that become more or less likely during emotion. Tasks that inquire about various actions or intentions may be linked to emotion states. One task is to ask participants how much they would like to engage in various behaviors, such as talk with a good friend, engage in some exercise, or have a pleasant meal. Teasdale and colleagues (Teasdale, Taylor, & Fogarty, 1980) reported that this task is sensitive to depressed mood or sadness, which has the action tendency of social withdrawal. When sad, people often lose interest in activities that formerly gave them pleasure. Sadness is also thought to be associated with depressed psychomotor function. Writing speed, for example, is negatively correlated with sadness and depression (see Velten, 1968, who used this task as one criterion measure in the validation study of the mood induction that bears his name). Other psychomotor tasks that have been used in emotion research include letter cancellation and smooth pursuit motor tracking tasks. Performance speed is most sensitive to sadness or depressed emotional states. Pleasant emotions, however, do not appear to increase psychomotor speed.
A variety of other behavioral tasks have been shown to be sensitive to affective states (Mayer, 1986; Mayer & Bremer, 1985; Mayer, Mamberg, & Volanth, 1988). One category of emotion-sensitive tasks consists of judgment tasks. One assessment strategy is to have participants make probability estimates of the likelihood of various good and bad events. For example, participants may be asked the probability of being killed in an airplane crash, dying in a car accident, or contracting cancer in their lifetime. It has been shown that persons in generalized unpleasant emotional states overestimate the probability of such bad events (Johnson & Tversky, 1983). The converse—increased probability estimates of good events while in positive emotional states—also appears true (Zelenski & Larsen, 2002). Results also apply to specific emotions; for example, fearful people make pessimistic judgments of future events (Lerner & Keltner, 2000). General appraisals of events also show emotion-specific patterns (Herrald & Tomaka, 2002; Siemer, 2001).
Another emotion-sensitive behavioral task is to ask participants to generate associations to positive, neutral, and negative stimuli. For example, have participants write down as many words as come to mind in 60 seconds when they hear each of the following stimulus words: happy, disappointed, generous, destroy, peace, or pain. Mayer and Bremer (1985) showed that performance on this task correlated with naturally occurring mood. Seidlitz and Diener (1993) used a variation wherein participants recalled as many happy experiences from their own life as they could in a given time period. Participants higher on trait-positive affect recalled more pleasant experiences, in the same time period, than participants lower on trait happiness. Teasdale and colleagues (Teasdale & Fogarty, 1979; Teasdale & Russell, 1983) have also demonstrated that emotion inductions influence recall of pleasant and unpleasant events in predictable (i.e., hedonically consistent) ways.
Another behavioral strategy for assessing emotion involves various information processing parameters. Reaction times in lexical decision tasks, for example, have been shown to be sensitive to affective states (Challis & Krane, 1988). For example, the participant's task might be to judge whether a string of letters represents a word or a nonword. On each trial the letters represent either: a nonword, an emotion word (e.g., anger), or a neutral word (e.g., house). Participants in positive affective states are quicker and sometimes more accurate at judging positive words compared to participants in neutral states, and vice versa for unpleasant moods (Niedenthal & Setterlund, 1994).
A related assessment task is to present participants with incomplete word stems and ask them to add letters to complete the word. Word stems are selected so that they can be completed as an emotion term or as a neutral term. For example, ANG_ _ could be completed as ANGER or as ANGLE or ANGEL or as ANGLO; JO_ could be completed as JOY or as JOB (e.g., Rusting & Larsen, 1998). A related technique is the use of homophones (words that sound alike but have different meanings). With this technique, the subject hears the word (die or dye, for example) and is asked to write that word. Participants in an unpleasant mood are more likely to write or complete the word stems in a manner congruent with their mood (Halberstadt, Niedenthal, & Kushner, 1995).
Behaviors that are enhanced or disrupted by emotion. So far we have discussed how certain behaviors directly follow from emotional states, and how specific emotion-related tasks may be influenced by the emotional state of the participant. However, other categories of behaviors, such as the defensive reflex or perception or attentional control, may be enhanced or disrupted by emotion, and thus might be used as an indicator or measure of emotion (Compton, 2000). One such emotion-sensitive task relies on a very simple behavior—the startle reflex. The startle reflex involves a rapid shutting of the eyes (blink), pulling the chin down, and a rapid inhalation. The startle reflex is easy to elicit through the application of a sudden and loud acoustic stimulus. Startle potentiation refers to an increase in the startle response (measured as a faster or stronger eye blink) when the person is startled while they are in an unpleasant emotional state (Vrana, Spence, & Lang, 1988). The researcher most responsible for developing this technique in humans is Peter Lang (e.g., Lang, Bradley, & Cuth-bert, 1990). Lang and colleagues, as well as others (Skolnick & Davidson, 2002), have demonstrated startle potentiation for unpleasant emotions compared to neutral states. The converse—slower and weaker startle when in positive emotional states—is rarely found.
A final behavioral paradigm with potential for measurement concerns the effects emotion has on cognitive parameters such as attention (Buodo, Sarlo, & Palomba, 2002). One effect is the automatic vigilance effect (Cothran, Zelenski, Prizmic, & Larsen, 2003; Pratto & John, 1991), which refers to the "grabbing" of attention by aversive or threatening information. The so-called emotional Stroop paradigm is one example of the automatic vigilance effect, where, in naming the colors of various words, people are generally slower to name the color if the word is threat-related. Presumably, threatening stimuli are processed more carefully, especially if one is already in an aversive emotional state, resulting in slowing on the primary, nonemotional task (color naming). Another behavioral paradigm, where cognitive parameters are influenced by both the emotional state of the participant, as well as the emotional content of the stimuli, concerns emotional priming (Wentura & Rothermund, 2003), as well as other irrelevant feature tasks, such as the Affective Simon task (De Houwer, 2003), where the participant in supposed to ignore the emotional content of a stimulus while responding to some other relevant feature.
Was this article helpful?