اثرات سببی برونگرایی بر روی عاطفه مثبت و اثر روان رنجوری بر روی عاطفه منفی: حالت دستکاری برونگرایی و حالت روان رنجوری در یک رویکرد تجربی

Although the relationships between extraversion and positive affect and between neuroticism and negative affect are powerful, the patterns of causality accounting for them are unknown. We employed an experimental methodology to manipulate state extraversion and state neuroticism to determine their causal status. In Experiment 1, state extraversion was manipulated by instructing participants to act extraverted and introverted during two different discussions with two other participants. Participants reported more positive affect when instructed to act extraverted than when instructed to act introverted, and this finding was supported by the ratings of observers. In Experiment 2, state neuroticism was manipulated by instructing participants to act neurotic and emotionally stable during two different discussions. Participants reported more negative affect when instructed to act neurotic than when instructed to act stable. Thus, at least some part of the extraversion-positive affect and neuroticism-negative affect relationships may be accounted for by the causal influences of extraverted behavior and neurotic behavior, respectively.

The purpose of this study is to determine whether extraverted behavior causes positive affect and whether neurotic behavior causes negative affect. Demonstrating these causal relationships would be important for at least three reasons. First, overwhelming evidence has established that more extraverted individuals experience higher levels of positive affect than less extraverted individuals do, and that individuals higher in neuroticism experience higher levels of negative affect than individuals lower in neuroticism do. However, none of this evidence has established whether extraversion causes positive affect or neuroticism causes negative affect (Costa and McCrae, 1980, Diener et al., 2003 and Eid et al., 2003). Without evidence for such causal impacts, the important potential implications of these relationships for interventions to improve well-being cannot be realized. In contrast, showing that extraverted behavior causes positive affect and that neurotic behavior causes negative affect may open the door to the development of relevant interventions. Specifically, acting as though one is extraverted or not neurotic may be sufficient for increasing psychological well-being, by both increasing positive emotional states and decreasing negative emotional states. Second, the methodology employed will be to manipulate by instruction participants’ current level of extraverted and neurotic behavior and to test for effects on positive and negative affect. Successful manipulation of extraverted behavior and neurotic behavior may open up new directions of research by demonstrating the possibility of manipulating states (current levels of extraversion and neuroticism) in order to test theories about traits. A third reason this research is important is that it evaluates the meaningfulness of states. Demonstrating consequences of states would suggest that deviation from one’s baseline way of acting is not purely error but rather is meaningful and potent. Such a finding may encourage more theory and research directed at understanding processes of within-person variation in states (Fleeson & Jolley, in press). 1.1. Evidence of extraversion-positive affect and neuroticism-negative affect relationships Considerable research documents the relationships between extraversion and positive affect and between neuroticism and negative affect. Extraverts are happier than introverts, and individuals higher in neuroticism are more unhappy than those lower in neuroticism (e.g., Charles et al., 2001, Costa and McCrae, 1980, David et al., 1997, DeNeve and Cooper, 1998, Diener and Lucas, 1999, Diener et al., 1999, Diener et al., 2003, Fossum and Barrett, 2000, Gross et al., 1998, Larsen and Ketelaar, 1989, Larsen and Ketelaar, 1991, McCrae and Costa, 1991, Rusting, 1999, Watson and Clark, 1992 and Wilson and Gullone, 1999). The extraversion-positive affect and neuroticism-negative affect relationships are quite strong. Extraversion typically accounts for about 15% and in one study accounted for 34% of the variance in positive affect (Costa and McCrae, 1980, David et al., 1997, Gross et al., 1998, Larsen and Ketelaar, 1989, Larsen and Ketelaar, 1991, Lucas and Baird, 2004, McCrae and Costa, 1991, Rusting, 1999 and Watson and Clark, 1992). These same studies found that neuroticism typically accounts for about 25% and in one study accounted for 36% of the variance in negative affect. Evidence for the robustness of the extraversion-positive affect and neuroticism-negative affect relationships comes from replication across multiple methodologies. The most common methodology is one-time self-report of both general trait behavior and the general experience of affect (e.g., Lucas et al., 2000 and Watson and Clark, 1992). The relationships also exist when self-reports of affect are provided daily (e.g., David et al., 1997 and Suls et al., 1998). Extraversion has been shown to predict amount of positive affect experienced when watching a comedy film, and neuroticism has been shown to predict amount of negative affect experienced when watching a sad or disturbing film (Gross et al., 1998 and Larsen and Ketelaar, 1991). Extraversion and neuroticism also predict mood-related cognitive variables, such as free recall of positive and negative words (Rusting and Larsen, 1998 and Sandvik et al., 1993). Both the extraversion-positive affect and neuroticism-negative affect relationships have been found in ratings by spouses (e.g., Costa & McCrae, 1988), to exist for children, adolescents, adults (Lucas and Diener, 2000 and Wilson and Gullone, 1999), and chimpanzees (Weiss, King, & Enns, 2002), and across countries and cultures (Allik and Realo, 1997, Hamid and Cheng, 1996, Headey and Wearing, 1989, Lucas et al., 2000 and Steel and Ones, 2002). One of the reasons this relationship is important is because it demonstrates the impressive power that internal personality factors may have in driving psychological well-being. Although both internal and external factors are related to affect, the internal factors of extraversion and neuroticism can account for more variance in affect than can some, but not all, external factors. For example, income (e.g., George et al., 1985, Lykken and Tellegen, 1996 and Pinquart and Soerensen, 2000), objectively rated health (e.g., Brief et al., 1993 and Larson, 1978), marriage (e.g., Argyle, 1999 and Haring-Hidore et al., 1985), age (e.g., Diener and Suh, 1998 and Mroczek, 2001), and gender (e.g., Haring et al., 1984 and Inglehart, 1990) each typically accounts for less than 5% of the variance in affect when zero-order relationships are considered. Although it is likely that methodological similarities between measures of extraversion/neuroticism and affect (both typically on rating scales) may inflate these relationships as compared to the relationships between affect and other variables (e.g., income), it appears that extraversion and neuroticism are the best predictors of positive and negative affect, respectively. 1.2. How to establish causality? The possibility that psychological well-being is strongly influenced by internal personality factors suggests that well-being may be enhanced through personality change. However, such application can be realized only if changes in extraversion and in neuroticism indeed do produce changes in positive affect and negative affect, respectively. Specifically, if extraverted behavior and neurotic behavior cause positive and negative affect, respectively, then individuals may be able to act extraverted and emotionally stable to achieve a higher quality of life. Research has not yet established whether extraversion and neuroticism cause emotion primarily because extraversion and neuroticism are relatively stable traits and thus are essentially impossible to manipulate. Without manipulation, it is difficult to know whether the affect differences came first. An alternative strategy is to employ a cross-lagged longitudinal design so that individuals can be statistically equated on starting affect; any later differences in affect that are associated with starting levels of traits must either be caused by those traits or by third variables. However, no research to date has tested this. The closest study showed correlations between initial trait levels and affect levels ten years later, but starting differences in affect were not controlled (Costa & McCrae, 1980). The density-distributions approach to traits (Fleeson & Jolley, in press) suggests a third option. Specifically, manipulating extraversion and neuroticism states may be a way to test the causality of traits in a way that gains the benefits of experimental methodology and the ability to conclude causality. The density-distributions approach starts with the notion of a personality state (Fleeson, 2001). A personality state describes the individual’s personality at the moment, and uses the same content and dimensions as personality traits. For example, the amount an individual is acting extraverted at the moment, on a seven-point scale, is his or her current extraversion state. Note that this means the state is very closely matched to the trait, both in the items and the scales used to assess and define it. The density-distributions approach proposes that an individual’s trait fruitfully may be conceived of as his or her distribution of such states across a suitable period of time (e.g., a few weeks). Because such distributions are typically very wide—the typical individual routinely and regularly acts at most levels of most traits—referring to an individual’s trait with just one level of the trait may not accurately reflect his or her behavior. Rather, it is the entire distribution of states that should be used to describe the individual’s trait. States may be isomorphic with traits in at least some regards (Fleeson, in press, Fleeson et al., 2002 and Pytlik-Zillig, 2001). That is, at least some of the properties of traits may be carried by the corresponding state. For example, at least some of the properties of trait extraversion are contained in the act of state extraversion. State extraversion will influence positive affect only if the property responsible for trait extraversion’s association to positive affect varies with state extraversion (except in the unlikely event that state extraversion also causes positive affect but for a different reason than is underlying the association of trait extraversion to positive affect). Similarly, state neuroticism will influence negative affect only if the property responsible for trait neuroticism’s relationship to negative affect varies with state neuroticism. Some properties that are possibly responsible for the extraversion to positive affect relationship do not vary with state extraversion. For example, size of a particular brain area, resting frontal lobe activity, basic values, strength of the behavioral activation system, activity preferences, and optimal cortical arousal are properties generally presumed not to vary with state extraversion (Davidson, 1992, Elliot and Thrash, 2002, Eysenck, 1990, Lucas and Baird, 2004 and Schmidtke and Heller, 2004). If such a property is responsible for extraverts’ higher positive affect, then manipulating state extraversion cannot influence positive affect. Thus, this strategy can be successful only if the properties responsible for the extraversion to positive affect and neuroticism to negative affect associations vary within a person, vary with acting extraverted or acting neurotic, and are powerful in that variation. In fact, Fleeson et al. (2002) showed that variations in state extraversion were strongly and robustly associated with variations in positive affect. Thus, it is likely that the critical property of trait extraversion that is associated with positive affect does vary within a person and can be manipulated to investigate whether extraversion causes positive affect. Participants can be instructed to act extraverted or to act introverted and resulting differences in positive affect can be assessed; resulting differences can be attributed to the instructed state extraversion. To the extent that this works and that it can be used with other traits and in other situations, it may open many experimental options for personality psychologists. In previous research following this logic, Fleeson et al. (2002) instructed participants to act extraverted or to act introverted during a discussion. Because this was the first test of the theory it was not clear whether participants could act extraverted or introverted on demand, so a warm-up was added in which participants described a recent time in which they had acted extraverted (or introverted). Supporting this logic, self-reports of positive affect after the discussion differed significantly across the two instructional conditions. The present research is intended as a more direct test of this logic in two ways. First, it tests whether this finding for extraversion and positive affect can be obtained with a confound removed, with improved measures, and in neutral observer ratings of positive affect. This finding requires an independent demonstration before it can generally be accepted that this innovative approach to traits and to personality research is viable. More importantly, the warm-up in the previous research inadvertently introduced a confound that undermines the ability to conclude causality. Specifically, by writing in detail about a recent occasion in which they acted extraverted or introverted, participants may have recalled the affect they experienced on that occasion, and this affect may have carried through to the discussion. Thus, to conclude that the behavior and not the recall caused the subsequent affect requires instructions without such a warm-up. Finally, one participant, who receives no instructions and is informed of no instructions nor manipulation, will also be present in the discussion and will rate the other participants’ positive affect. This participant’s ratings are included to reduce any demand characteristic interpretations of the data, although this possibility may not be ruled out entirely. Second, the current research tests whether the same logic holds for a second trait and for a second trait-affect association, namely the association between neuroticism and negative affect. It is important to determine what factors may contribute to individual differences in negative affect, and neuroticism is the factor that has been found to account for the most variance in negative affect (e.g., Watson & Clark, 1992). If instructing individuals to act neurotic and stable at different times is found to influence the experience of negative affect, the hypothesis that neurotic behavior causes negative affect would be supported. 1.3. Hypotheses The hypotheses of the present study are: (a) the manipulation of extraverted behavior will influence positive affect, such that individuals will experience more positive affect when instructed to act extraverted than when instructed to act introverted; (b) the manipulation of neurotic behavior will influence negative affect, such that individuals will experience more negative affect when instructed to act neurotic than when instructed to act emotionally stable; (c) these effects will be the same whether individuals are dispositionally extraverted or introverted, and whether they are dispositionally higher or lower on neuroticism (i.e., the effects of state and trait will be additive).

A great deal of personality research concerns the relationships between extraversion and positive affect and between neuroticism and negative affect, but it has not yet been clear whether traits are potent in this relationship. The present research consisted of two experiments manipulating state extraversion and state neuroticism, and was the first to show that that these states cause positive and negative affect, respectively. Although we believe the generalization to the associated traits causing affect is warranted, it must be clear that it is a generalization because the traits themselves were not manipulated. To make such generalizations clearer, and to clarify the basic nature of traits, more research needs to be directed at understanding the relationships between traits and states. What the current experiments have done is demonstrate a causal link between the behavior that manifests such fundamental dimensions of personality as extraversion and neuroticism and the positive and negative affect that people experience as a consequence of exhibiting that behavior.