Individual differences in the social facilitation effect: A review and meta-analysis

Abstract

The present study reviews the social facilitation literature and identifies two overarching responses to social presence: negative-apprehensive and positive-self-assured. These responses correspond to two general orientations toward the social environment described in current personality literature: a negative orientation, comprised of neuroticism and low self-esteem; and a positive orientation, comprised of extraversion and high self-esteem. A meta-analysis reveals that social presence is associated with performance impairment for negatively oriented individuals, and with performance improvement for positively oriented individuals. Additional analysis reveals that personality is a more substantial moderator of the effect of social presence than is task complexity. The results of this study open a new avenue in the research on social facilitation, encouraging a closer inspection of the meaning that various individuals attribute to social presence.

Introduction

At one time or another, we have all found ourselves in situations in which we have been the focus of others’ attention. Occasionally, those others were individuals unfamiliar to us. Sometimes we were observed performing a task, whether it was giving a talk, taking a test, or playing ball. Nearly everyone can recall such an experience, yet the experience is unique to the individual who recalls it. How we react to being the focus of social attention seems to be closely related to the meaning that we assign to the situation. In turn, this meaning is largely the result of our prior experience as well as of our genetic heritage, or, in short, our personality. This premise lies at the foundation of the present review, which looks at the way personality moderates the response to social presence.

Despite being one of the early questions asked in the field of social psychology, research on the social facilitation effect (i.e., the effect of social presence) has not dealt much with the question of personality moderation. The present review shows that considering personality could considerably improve our understanding of this effect. The review begins by describing the major research directions and explanations of social facilitation. A broad look at this literature shows that the response to social presence follows one of two general orientations: positive-self-assured and negative-apprehensive. The review then explores the role of personality in social facilitation. Personality is suggested as an enduring basis for a relatively positive or negative orientation toward social presence. The results of an original meta-analysis support this claim by showing that positively oriented individuals experience performance improvement in social presence, whereas negatively oriented individuals experience performance impairment. The moderating effects of personality are discussed vis-à-vis the traditional focus in the social facilitation literature on task complexity as a moderating variable. The review concludes with suggestions for future research.

Triplett’s (1898) observation of cyclists riding faster when in competition compared to when alone stimulated the creation of a new research interest in social psychology, which focuses on the effects of social presence on behavior. The term social facilitation was coined by Allport (1924) to describe “an increase in response merely from the sight or sound of others making the same movement” (p. 262). Although early studies of social facilitation focused on a co-action paradigm, later studies introduced a passive observer paradigm (e.g., Dashiell, 1935) in which an individual is passively observed by one or few individuals. Current approaches to social facilitation refer to both paradigms. Specifically, social facilitation is said to occur when there is an increase or decrease in behavior by an animal¹ in the presence of another animal that does not otherwise interact with the first animal (Guerin, 1993).²

Zajonc’s (1965) seminal review was a major milestone in the research on social facilitation, paving the way for future research by proposing an interaction-based paradigm. According to Zajonc, social presence improves performance of simple or well-learned tasks, and impairs performance of complex or ill-learned tasks. Utilizing the Hull-spence drive model (Spence, 1956), Zajonc’s Drive theory asserted that while in the mere presence of others, an organism automatically and unconditionally experiences generalized drive (arousal), which interacts with its habit strength to yield a greater tendency toward issuing a habitual (dominant) response. On simple tasks, this response is usually the correct one, leading to performance improvement, but on complex tasks, it is usually the incorrect one, leading to performance impairment.

Zajonc’s contribution was twofold: first, by introducing task complexity³ as a moderating variable, he managed to account for previously seemingly inconsistent results and to establish what later became a consensual paradigm in social facilitation research. Subsequent theories regarded simple performance improvement and complex performance impairment as the social facilitation effect that needs to be explained. The second aspect of Zajonc’s contribution was his proposed explanation, which, as described above, was based on an automatic arousal increase. Later theories, however, have considered other explanations differing mainly in their suggestions for mediating mechanisms and the role they ascribe for arousal. A short consideration of this theoretical dispute follows (for more detailed reviews see Aiello and Douthitt, 2001, Geen, 1989, Geen, 1995, Geen and Gange, 1977, Guerin, 1993).

Four major mediating mechanisms of the social facilitation effect have thus far been proposed: uncertainty, evaluation apprehension, self-awareness, and distraction.

Two theories suggest that the social facilitation effect evolves from a sense of uncertainty that the organism experiences in social settings. The first to suggest the uncertainty mechanism was Zajonc (1980) in a modification of his Drive theory. This idea was further developed by Guerin, 1983, Guerin, 1993, Guerin and Innes, 1982 in their Monitoring theory. Both the Drive theory and the Monitoring theory claim that organisms are predisposed to monitor and be prepared to react to the ever-changing demands induced by social presence. Social situations pose various threats to the organism, necessitating it to maintain a high level of alertness in order to quickly react in case of attack. This high level of readiness causes the aforementioned increase in arousal, which in turn is claimed to be responsible for the social facilitation effect. Based on this approach, one can expect stronger effects when the performing organism feels threatened, unable to monitor the observer, or is unfamiliar with the observer.

In response to Zajonc’s original Drive theory, Cottrell, 1968, Cottrell, 1972 asserted in his Evaluation Apprehension theory that the drive in social situations is a learned one, originating from one’s past experience with social encounters. An individual’s past experiences produce a set of (either positive or negative) unconscious expectations regarding future encounters, forming the basis of the drive. According to Cottrell, a social facilitation effect will be evident only if the performing organism expects to be evaluated by others (similar ideas were presented by Bond, 1982 and Berger et al., 1981, emphasizing self-presentation concerns and various behavioral changes stemming from a tendency to conform in social presence). In support of this theory, several studies showed that social facilitation occurs only when the observers were able to evaluate the actor’s performance (e.g., Cottrell et al., 1968, Henchy and Glass, 1968, Paulus and Murdoch, 1971). Some researchers (e.g., Weiss & Miller, 1971) have suggested that the social facilitation effect occurs only under the expectation of receiving negative feedback, yet evidence has shown otherwise, demonstrating that positive feedback expectations produce social facilitation effects as well (e.g., Blascovich et al., 1999, Good, 1973).

Two theories argue that social presence makes one self-aware. Duval and Wicklund, 1972, Wicklund and Duval, 1971 Objective Self-Awareness theory proposes that while in social presence, one inspects oneself from an objective observer’s perspective. In so doing, the difference between the current self and an ideal self becomes salient and produces an aversive state, which motivates behavior. When the task is simple, the added energy improves performance; in contrast, when the task is complex, the excessive effort leads to impaired performance.

Similar ideas were introduced by Carver and Scheier, 1981a, Carver and Scheier, 1981b in their Control theory. This theory compares human nature to a control system, in which a process of matching to standards governs behavior. Social presence catalyzes a process of testing one’s behavior against a (private or public) standard, which in turn stimulates a negative feedback loop to close the gap. This affect-free mechanism contributes to enhanced performance of simple tasks with clear and achievable standards; however, the difficult and sometimes contradicting standards of complex tasks create cognitive withdrawal that negatively affects performance.

The Distraction-Conflict theory (Baron, 1986, Baron et al., 1978, Groff et al., 1983, Sanders and Baron, 1975) proposes that social presence distracts and creates an attentional conflict for the individual engaged in a task. Social presence summons the actor’s attention for various reasons, such as social comparison or monitoring for threats, and this distraction conflicts with the actor’s wish to accomplish the task. The unavoidable result is an attentional conflict that creates either increased drive or cognitive overload (depending on the version of the theory; see Baron, 1986); both supposedly facilitate performance of easy tasks (due to their low demands), but impair the performance of complex tasks.

All of the above theories have supporting evidence, yet all have marked weaknesses. Attempts to pinpoint a single exclusively accurate theory have proven unsuccessful (Guerin, 1993). In my opinion, such attempts are misguided because the existing theories are not mutually exclusive. The picture that emerges from the aforementioned theories is of an organism that, once aware that it is being observed, reacts in a way that is either compatible with its phylogenetic heritage (i.e., increased alertness to enhance survival), its ontogenetic experience (i.e., learned drive), its current evaluation of its abilities (resulting from a process of self-reflection), or its momentary sense of disturbance (i.e., reaction to being distracted). Considering the different bases for these reactions, the possibility of the co-occurrence of one or more of them seems a plausible alternative (see also Geen, 1989, Guerin, 1993, Paulus, 1983, Sanders, 1981, for other integrative ideas). A multi-level reaction to social presence seems reasonable when taking into consideration the ambiguity entailed in having a non-interacting stranger observe you. In such a situation, myriad interpretations and reactions are likely, and the theories emphasize what might reflect varying levels of control, awareness, and intentionality in the individual’s reaction to it.

For our purpose, however, more important is what cuts across these levels, and these are two overarching orientations toward social presence. The first is a negative orientation, which is reflected in explanations that emphasize threat, apprehension, or distraction. This orientation predominates in social facilitation theorizing (Borden, 1980). Other explanations emphasize a positive orientation, expressed by high levels of self-assurance and enthusiasm. Such a reaction can result from the potential of being prized by others after successfully accomplishing a task, from the boost of motivation in pursuing an ideal self that emerges while being observed, or from any other positive connotation that one has with regard to social situations.

With the bulk of social facilitation theories emphasizing anxiety-related reactions to social presence, it is no surprise that little research attention has been paid to the exploration of other reactions. In the instances when this issue was addressed, it was in the context of manipulating the relations between the actor and the observers (e.g., Cox, 1966, Cox, 1968, Geen, 1977, Henchy and Glass, 1968, Seta and Seta, 1995), or by manipulating the actor’s prior success with the task (e.g., Blascovich et al., 1999, Robinson-Staveley and Cooper, 1990, Sanna, 1992). These manipulations attempted to induce either anxiety or self-assurance in the actor, creating temporary states of either negative or positive activation. However, the effect of such manipulations is not always straightforward. For example, prior success with a task seems a good source for self-assurance when asked to re-perform the task before an audience. However, under some conditions (e.g., when the audience is aware of the actor’s prior success) such manipulation may act as a double-edged sword, actually increasing anxiety and eventually debilitating performance (Seta & Seta, 1995). Therefore, although there is a theoretical basis for a distinction between positive and negative activation in situations of social presence, experimental manipulations do not necessarily capture it, and when they do, they necessitate the manipulation of additional factors (e.g., familiarity with observers, prior experience with the task) that may compromise the researcher’s ability to manipulate in isolation the effect of the mere presence of others.

One way to overcome this limitation is by looking at personality traits. To the extent that there are positive and negative orientations toward social presence, one can reasonably expect that the relevant traits will moderate the reaction to social presence with no need for additional (potentially confounding) manipulations. To further explore this idea, the following sections focus on the role of personality in the social facilitation effect.

Recognition of the importance of individual differences in moderating the effect of social presence dates back to Triplett’s pioneering work. Triplett (1898) attributed the variance in his findings to differences in control among his subjects. Several years later, Allport (1924) also reported on individual differences in susceptibility to the influence of social presence, this time with regard to mental functions (thought and association). According to Allport (1920/1967), some individuals are inclined to be distracted and to experience performance impairment in group sessions. Later on, Hollingsworth (1935) concluded that audience effects are extremely variable across performers, and Dashiell (1935) marked individual differences as a source of complication in the understanding of this effect. Finally, writing from a theoretical perspective, Spence (1956) mentioned individual differences as a source of variance in drive and response.

However, following Zajonc (1965) most of the social facilitation theories ignored individual differences, and emphasized task complexity as the central moderating variable. This emphasis had some less-than-constructive consequences. The task-based interaction is, by its nature, nearly irrefutable: there are no objective criteria for determining the difficulty of a task, so virtually any outcome is theoretically accounted for. In addition, many sorts of tasks that require varying and even opposing skills (e.g., error detection vs. creativity) could be classified as equally complex. That is, without a theoretical basis for task classification, we are left with a very general and vague categorization of simple versus complex tasks. In addition, the emphasis on task characteristics shifted the theoretical debate to the last link in the causal chain—task performance. All theories aimed at explaining this behavioral effect while actually disagreeing on the mediating process(s) involved (e.g., uncertainty, distraction, etc.). Because most theories provided successful posteriori explanations for the bulk of results, they became undifferentiated and nearly impossible to refute (Aiello and Douthitt, 2001, Guerin, 1993).

The utility of the simple/complex dichotomy is also challenged by the results of a meta-analysis (Bond & Titus, 1983). Two of its most substantial conclusions were: (1) social presence accounts, on average, for only a small amount of variance (between 0.3 and 3%) in both simple and complex task performance; and (2) various studies typically yield inconsistent and sometimes contradictory results. On some effects (e.g., complex task performance quantity), over 40% of the results opposed the overall mean. This pattern of results points to the existence of moderating variables beyond the contribution of task complexity, once again indicating that there is more to social facilitation than the facilitation of simple performance and the impairment of complex performance.

An additional, broader reason for the neglect of individual differences in the social facilitation effect resides with the difference between situationism and dispositionism (Baumeister, 1999, Jones, 1998). These two disciplines of psychology have a long history of rivalry and mutual disregard (Cronbach, 1957), and from a historical standpoint, it is interesting to note that the growing body of research in social facilitation (starting in 1965) was concurrent with the beginning of the fierce “person–situation” debate in psychology (Mischel, 1968). At least in the early years of the debate, personality was attacked for its conceptual existence and empirical contribution. Yet even as the debate waned, and both sides were proved partially correct, the methodological insights gained along the way help in clarifying why personality would still find it difficult to be proven useful under the current social facilitation paradigm: experiments are “strong situations” (Snyder & Ickes, 1985), and the conventional social facilitation experiment is an “impact study” (Aronson, Wilson, & Brewer, 1998), usually relying on a between-subjects experimental design and not on repeated measures or longitudinal designs (for example, in Bond & Titus’s, 1983 meta-analysis, four out of every five studies was based on a between-subjects design). Narrow response options, brief experiment duration, detailed behavioral instructions—all typical of social facilitation experimentation—contribute to constraining the effect of personality traits and to enhancing effects associated with experimental manipulation (Buss, 1989).

From the above discussion, it appears that a blend of historical, methodological, and theoretical causes contributed to the relatively scant attention paid to personality in the study of the social facilitation effect. It is estimated that only 5–7% (approximately 30 studies) of the social facilitation studies thus far conducted have measured individual differences. Only two reviews have addressed this issue, the more recent published over 25 years ago (Geen, 1980, Paivio, 1965), and both offered only a qualitative analysis of the then-available literature. The present review, therefore, offers the first quantitative analysis on individual differences in the social facilitation effect.

Three traits dominate the existing literature on individual differences in the social facilitation effect: self-esteem, trait anxiety, and extraversion. Recent studies show that these traits are strongly associated with the two general orientations described above, the positive and the negative.

A considerable quantity of research reports on the relationship between the “Big Two” dimensions of personality—extraversion and neuroticism (in which anxiety is the most critical component; e.g., Eysenck & Eysenck, 1985)—and positive and negative affective states, accordingly (Abe and Izard, 1999, Carver et al., 2000, Chang, 1997, Costa and McCrea, 1980, Rusting, 1998, Rusting and Larsen, 1997, Tellegen, 1985, Watson et al., 1999, Wilson and Gullone, 1999). Compared to introverts, extraverts are predisposed to experience higher levels of positive affect in general (i.e., across conditions; e.g., Lucas & Baird, 2004), and in response to emotional inductions (e.g., Rusting & Larsen, 1997). Extraversion has little effect, however, on the emotional response along the negative affect scale (Rusting & Larsen, 1997).

Neuroticism, on the other hand, is strongly related to negative affect in general (Costa & McCrea, 1980), and in response to emotional inductions (e.g., Rusting & Larsen, 1997). Neuroticism has little effect, however, on positive activation (Rusting & Larsen, 1997). In addition, when faced with ambiguous events, extraversion and neuroticism promote trait-congruent interpretations. That is, extraversion promotes a positive interpretation, whereas neuroticism promotes a negative interpretation (e.g., Gomez et al., 2002, Rusting, 1999, Zelenski and Larsen, 2002), the one not exclusive of the other (Uziel, 2006).

Extraversion and neuroticism are considered independent of each other (e.g., Eysenck & Eysenck, 1985), yet both are related to self-esteem (e.g., Judge et al., 2002, Swickert et al., 2004). Robins, Tracy, Trzesniewski, Potter, and Gosling (2001) reported on a correlation of 0.38 between extraversion and self-esteem, and a correlation of −0.50 between neuroticism and self-esteem in an online survey with more than 320,000 respondents. Their results are consistent with nine other studies that reported an average correlation of 0.40 between extraversion and self-esteem, and −0.61 between neuroticism and self-esteem (cf. Robins et al., 2001).

As might be expected from the aforementioned correlations, in terms of associations with other variables, the self-esteem relationships resemble in part those of extraversion and in part those of neuroticism. For example, self-esteem appears to be associated with both positive and negative affect (Cheng and Furnham, 2003, Pelham and Swarm, 1989, Watson and Clark, 1984), and was found to mediate the effects of both extraversion and neuroticism on happiness and depression (Cheng & Furnham, 2003).

In addition, high self-esteem individuals, like extraverted individuals, seem to be especially responsive to positive stimuli (Wood, Heimpel, & Michela, 2003). In contrast, low self-esteem individuals, like neurotic individuals, focus on the negative aspects of situations (Smith and Petty, 1995, Wood et al., 2003). In light of the differential reactions by low and high self-esteem individuals, researchers have suggested that a bi-dimensional model, differentiating between positive and negative self-esteem, would better fit the existing data (e.g., Cheng and Furnham, 2003, Owens, 1993).

At least part of the association of self-esteem with extraversion and neuroticism has been attributed to social relations and perceived social acceptance (Leary & MacDonald, 2003). According to the Sociometer theory of Leary and his colleagues (e.g., Leary, 1999, Leary and Baumeister, 2000, Leary and MacDonald, 2003, Leary et al., 1995), self-esteem operates as a gauge of one’s interpersonal appeal and success. Extraverts tend to be dominant, optimistic, and assertive in their social interactions—behaviors that lead to social approval and to a sense of being socially accepted. This sense is in turn reflected in high levels of self-esteem. In contrast, neuroticism is associated with anxiety, depression, and bashfulness, which are not socially desired, and which lead to social rejection and to a reduced sense of social inclusion, which in turn is reflected in low self-esteem.

According to the Sociometer theory, trait self-esteem serves as a gauge especially “in the absence of explicit cues connoting inclusion or exclusion” (Leary et al., 1995, p. 527), that is, in conditions that prevent individuals from ascertaining how they are perceived and evaluated. In such conditions, “people with very low trait self-esteem may perceive others as rejecting most of the time, whereas those with higher self-esteem generally feel they are being accepted” (Leary et al., 1995, p. 527). These perceptions guide individuals’ overall approach (apprehensive vs. self-assured) and behavior in social conditions.

Taken together, the above discussion points to two basic tendencies that have affective and social roots. The first represents a tendency toward self-assurance and enthusiasm in general and toward the social environment, and is reflected in the traits of extraversion and high self-esteem. The other represents a tendency toward anxiety and apprehension in general and toward the social environment, and is reflected in the traits of neuroticism and low self-esteem.

The idea of two general predispositions related to social presence was suggested by Paivio (1965) in his research on children’s reaction to audience. Paivio differentiated between a predisposition to anxiety before observers (“self-consciousness”) and an attention-seeking tendency (“exhibitionism”), representing largely orthogonal avoidance and approach tendencies toward social performance (Levin, Baldwin, Gallwey, & Paivio, 1960). According to Paivio, such tendencies result from child-rearing practices and the development of conditioned (negative or positive) responses to audience by the child. In a series of experiments (summarized in Paivio, 1965), Paivio and his collaborators demonstrated the interaction between audience presence and personality by showing generally positive effects of exhibitionism and negative effects of self-consciousness on verbal behavior in the presence of audience.

Paivio’s work did not focus on social facilitation as it is currently defined (for example, Paivio’s experiments did not compare performance alone to performance under social presence). Still, in many respects, his work is a conceptual antecedence of the present study in proposing a dual personality-based approach to social presence. One noticeable difference in this regard between the present work and that of Paivio is that the latter’s research treated the approach tendency as a need or motivation that may or may not be fulfilled (depending on other factors, particularly the level of avoidance tendency), whereas in the present study, positive orientation is conceptualized as a characteristic quality of the individual.

Social presence is both an ambiguous situation and a significant one. It is ambiguous because one does not know for certain the consequences of such presence—both positive and negative consequences are possible. It is significant because as social creatures, each outcome, either positive or negative, has a potentially substantial effect on our future.

Personality predisposes us to adopt differing approaches to these situations. High self-esteem and extraverted individuals are predisposed to positive orientation, whereas low self-esteem and neurotic individuals are predisposed to negative orientation. The existing social facilitation literature has demonstrated that it is the individual’s approach—expressed in the level of comfort that she/he experiences in social presence (Geen, 1977); in the assessment of the impression that she/he makes (Bond, 1982); in his/her level of self-assurance (Robinson-Staveley and Cooper, 1990, Sanna, 1992, Sanna and Shotland, 1990); and in his/her sense of challenge and excitement (as opposed to threat and apprehension; Blascovich et al., 1999, Levin et al., 1960)— that determines the positive or negative effect that social presence will have on him/her (and in turn the perceived difficulty of the task being performed).

The present study proposes that positive and negative orientations promote an opposite cascade of affective (challenge, enthusiasm vs. anxiety, depression), motivational (approach vs. withdraw), and cognitive (active generation of ideas vs. rumination and distraction) processes (cf. Carver et al., 2000, Fiedler, 2001, Higgins, 1998, Isen, 1987, Mueller, 1992). These processes in turn cause positive orientation to be associated more often with performance improvement, and negative orientation to be associated more often with performance impairment in conditions of social presence compared to solitary performance.

The above studies suggest that social orientation will have a greater effect on performance than will the effect associated with the “objective” classification of the task as simple or complex (e.g., Bond, 1982, Geen, 1977). Still, positive orientation may be associated with an even greater improvement on simple tasks, because the more likely felt success during the performance of these tasks may further enhance this orientation. In contrast, negative orientation may be associated with an even poorer performance on complex tasks, because the more likely experienced difficulties during the performance of these tasks may further (and quicker) enhance this orientation (see Mueller, 1992, for additional theoretical mechanisms that predict performance impairment among negatively oriented individuals especially on complex tasks). To explore these propositions, the interaction between orientation and task complexity will also be included in the analysis.

The next section quantitatively explores the existing literature in order to check the above hypotheses. Efforts were made to trace all published studies in order to present a comprehensive overview of the field.