You are on page 1of 26

INTERPERSONAL RELATIONS AND GROUP PROCESSES

Social Loafing: A Meta-Analytic Review and Theoretical Integration
Steven 1. Karau and Kipling D. Williams
Social loafing is the tendency for individuals to expend less effort when working collectively than when working individually. A meta-analysis of78 studies demonstrates that social loafing is robust and generalizes across tasks and S populations. A large number of variables were found to moderate social loafing. Evaluation potential, expectations of co-worker performance, task meaningfulness, and culture had especially strong influence. These findings are interpreted in the light of a Collective Effort Model that integrates elements of expectancy-value, social identity, and self-validation theories.

Many of life's most important tasks can only be accomplished in groups, and many group tasks are collective tasks that require the pooling of individual members' inputs. Government task forces, sports teams, organizational committees, symphony orchestras, juries, and quality control teams provide but a few examples of groups that combine individual efforts to form a single product. Because collective work settings are so pervasive and indispensable, it is important to determine which factors motivate and demotivate individuals within these collective contexts. Intuition might lead to the conclusion that working with others should inspire individuals to maximize their potential and work especially hard. Research on social loafing, however, has revealed that individuals frequently exert less effort on collective tasks than on individual tasks. Formally, social loafing is the reduction in motivation and effort when individuals work collectively compared with when they work individually or coactively.When working collectively, individuals work in the real or imagined presence of others with whom they combine their inputs to form a single group product. When working coactively, individuals work in the real or imagined presence of others, but their inputs are not combined with the inputs of others. Determining the conditions under which individuals do or do not engage in social loafing is a problem of both theoretical and practical importance. At a practical level, the identification of moderating variables may

Parts of this article were presented at the 1992 Midwestern Psychological Association convention in Chicago. We thank Martin Bourgeois, Donal Carlston, Alice Eagly, Rebecca Henry, Janice Kelly, Norbert Kerr, Brian Mullen, Kristin Sommer, and several anonymous reviewers for their valuable comments on drafts of this article. We also thank Jeffrey Jackson and William Gabrenya for providing unpublished data. Correspondence should be addressed to Steven 1. Karau, Department of Psychology, Clemson University, Clemson, South Carolina 29634-1511 or to Kipling D. Williams, Department of Psychology, University of Toledo, Toledo, Ohio 43606-3390. Electronic mail may be sent to skarauetclemson.edu or kipiing@uoft02.utoledo.edu.

suggest means for devising interventions by which social loafing may be reduced or overcome in everyday groups and organizations. Latane, Williams, & Harkins (1979) even suggested that social loafing is a type of social disease, having "negative consequences for individuals, social institutions, and societies" (p. 831). Perhaps as a result of this characterization, social loafing research has been focused on identifying conditions under which the effect can be reduced or eliminated. No studies have been designed to determine what factors increase social loafing. This emphasis on eliminating the effect is understandable, given the potential applicability of social loafing research to real-world contexts. However, this emphasis on studying conditions in which the effect is not likely to occur may also result in an underestimation of the magnitude of social loafing across a wider range of situations. At a theoretical level, specifying which variables moderate social loafing is central to developing a fuller understanding of the dynamics underlying the performance and motivation of both individuals and groups. The social loafing literature offers a host of findings relevant to theorists interested in evaluation processes, the self, and group dynamics. Indeed, in a recent review of social motivation, Geen (1991) regarded social loafing as one of only three dominant phenomena addressing this issue in the 1980s. Despite the theoretical importance of this topic, surprisingly little attention has been devoted to systematically reviewing or integrating the social loafing research. Although a number of researchers have discussed social loafing or have presented theories of particular causes of social loafing (e.g., Harkins, 1987; Harkins & Szymanski, 1987; Jackson & Williams, 1985; Latane, 1981; Mullen, 1983; Paulus, 1983; Shepperd, 1993; Stroebe& Frey,1982), they have drawn theirconc1usions from only small portions of the available empirical studies, which were selected by unspecified criteria. Presently, the magnitude and consistency of social loafing across studies has never been estimated, and the extent to which particular variables moderate social loafing has remained unclear. Given the large empirical literature that is now available, a thorough integration and analysis of this research is long overdue.

Journal of Personality and Social Psychology. 1993. Vol. 65. No.4. 681-706 Copyright 1993 by the American Psychological Association. Inc. 0022-3514/93/$3.00

681

682

STEVEN J. KARAU AND KIPLING

D. WILLIAMS

The current review has three major purposes. First, we present a model that integrates expectancy theories with theories of group-level social comparison and social identity to account for the findings of studies examining individual effort in collective settings. This model, the Collective Effort Model (CEM), is then used to generate predictions for the meta-analysis. Second, we systematically review the available empirical findings. Meta-analytic methods (Cooper, 1989; Glass, McGaw, & Smith, 1981; Hedges & Olkin, 1985; Rosenthal, 1991) are used to estimate the overall magnitude of social loafing and to compare it with the magnitude of the effects of other variables on social behavior. This quantitative method also allows for an examination of the consistency of social loafing across studies and accounts for inconsistencies by identifying moderating variables that affect the tendency for individuals to engage in social loafing. The empirical status of such moderating variables is critical to a sophisticated understanding of the dynamics of motivation losses in groups. Third, we use the results of the meta-analysis to identify gaps and ambiguities still existing in the social loafing literature and discuss the potential of the CEM for generating novel predictions to guide future research toward answering these questions.

shouting with others. Latane et al. demonstrated that a substantial portion of the decreased performance of groups was attributable to reduced individual effort, distinct from coordination loss, and that audience size did not account for these results. They also coined the term social loafing for the demotivating effects of working in groups. Since 1974, nearly 80 studies on social loafing have been conducted in which individuals' coactive efforts were compared with individuals' collective efforts. These studies have used a wide variety oftasks, including physical tasks (e.g.,shouting, rope-pulling, and swimming), cognitive tasks (e.g.,generating ideas), evaluative tasks (e.g.,quality ratings of poems, editorials, and clinical therapists), and perceptual tasks (e.g., maze performance and vigilance tasks on a computer screen). Both laboratory experiments and field studies have been conducted using a range of subject populations varying in age, gender, and culture.

Theoretical Accounts for Social Loafing
Although several researchers have offered theories of social loafing, these viewpoints have typically been restricted to explaining only one of several possible causal mechanisms and generally do not attempt to include the wide range of variables that moderate social loafing. In contrast, social loafing research has adopted an orientation toward isolating conditions under which the effect can be reduced or eliminated. As a result, despite the tendency of researchers to manipulate a wide array of moderating variables across individual studies, there has not been a concerted attempt to integrate all of these moderators into a single theoretical model. Indeed, each moderating variable is frequently associated with a separate theory. Nevertheless, the theories that have been advanced do provide some insight into the nature and possible causes of social loafing. Therefore, before presenting the CEM, we first present an overview of viewpoints offered by prior researchers and discuss the need for an integrative model.

Brief History of Social Loafing Research and Paradigms
The first experiment to suggest a possible decrement in individual motivation as a result of working in a group was conducted over a hundred years ago by Ringelmann (cited in Kravitz & Martin, 1986). Male volunteers were asked to pull on a rope, tug-of-war fashion, as hard as they could in groups of varying sizes. The rope was connected to a strain gauge that measured the group's total effort. The results showed that as group size increased, group performance was increasingly lower than would be expected from the simple addition of individual performances. However, Steiner (1972) proposed two possible causes for this performance decrement: (a) reduced individual motivation or (b) coordination loss. Steiner favored the latter cause as most parsimonious, concluding that individuals may fail to synchronize their efforts in a maximally efficient manner (e.g., pulling while others are pausing), thus evidencing less productivity, but not necessarily less effort. To separate effort reduction from coordination loss, Ingham, Levinger, Graves, and Peckham (1974) had subjects perform a rope-pulling task both in actual groups and in pseudogroups in which blindfolded male students believed they were pulling with other group members but were actually pulling alone. Data for the pseudogroup trials showed that performance still decreased as perceived group size increased, suggesting that individuals exerted less effort when working in groups than when working individually. However, group size covaried inversely with audience size in Ingham et al.'s studies, leaving open the possibility that effort was facilitated by larger audiences in the individual trials. Latane et al. (1979) conceptually replicated these findings while holding audience size constant. They had subjects shout and clap as loudly as they could, both individually and with others. Blindfolded and wearing headphones that masked the noise, college men shouted both in actual groups and in pseudogroups in which they shouted alone but believed they were

Social Impact Theory
According to Latanes (1981) social impact theory, people can be viewed as either sources or targets of social impact. The amount of social impact experienced in a situation is thought to be a function of the strength, immediacy, and number of sources and targets present. In the collective condition of the typical social loafing experiment, the experimenter serves as a single source of social impact, whereas the group members usually serve as multiple targets of social impact. Social impact theory suggests that the experimenter's request to try as hard as possible on the task should be divided across targets, resulting in reduced effort as group size increases. The division of impact is predicted to follow an inverse power function, with a negative exponent having an absolute value less than I, thereby resulting in marginally decreasing impact as group size increases. In the coactive condition, however, in which inputs are not combined, each individual feels the full impact of the experimenter's request and works hard. The strength of social impact theory appears to be its ability to specify group size effects. The strength and immediacy factors, however, have been neglected

resulting in decreased performance en simple tasks (where the dominant response is likely to' be correct) and increased performance en novel. the present review adopts a less restrictive definition of social loafing consistent with the original formulation: the tendency to reduce one's effort when working collectively compared with coactively on the same task. 1987. individual inputs are combined into one group product. 1981). of course. however. Kerr & Bruun. therefore. Wrightsman. Mullen. Recent research by Harkins and colleagues (Harkins. Matching of Effort Jackson and Harkins (1985) proposed that people tend to' match their co-workers' efforts when working collectively According to' this position. individuals' inputs can only be evaluated in the coactive condition. 1989. They argued that the presence of other co-workers is drive reducing because these others serve as cotargets of an outside source of social impact. social loafing occurs because individuals expect others to' slack off in groups and. subjects performed better collectively than coactively Jackson and Williams argued that working collectively led to' reduced drive and effort. the discovery of new causes for the same effect. participants' expectations QfhQW hard their co-worker would work were manipulated such that participants expected their co-worker either to' try hard or not to' try hard on a shouting task. and the experimenter. Some researchers (Harkins. difficult tasks (where the dominant response is likely to' be in error). 1985. This research has also demonstrated that two requirements must be met fer evaluation by any source (the experimenter. Subjects completed simple and complex computer mazes either alene. in this study. and (b) there must be a standard (personal. and Latane (1981). Harkins & Szymanski. 1985. reduce their own efforts to' maintain equity. 1990) suggests that workers' perceptions of and motivations toward their task are highly influenced by co-workers' task assessments. In the high-effort condition. she said that she thought the experiment was boring. Individuals only loafed when identifiability and the coactive-collective variable were confounded. individuals can "hide in the crowd" (Davis. 1987. social. Thus. Collective tasks may also lead individuals to' feel "lest in the crowd" (Latane et aI. 1979) such that they cannot receive their fair share of the credit fer a geed group performance. Harkins. Kerr & Bruun. Social loafing was eliminated and participants matched their co-worker's anticipated effort. never identifiable. Citing studies in which people faced with a fearful situation tended to' prefer being in the company of others (e. The role of identifiability in social leafing is illustrated in a study by Williams. Research on job attitudes (see Zalesny & Ford. 1959. Jackson and Williams (1985) found SUPPQrt for this logic in an experiment that combined features of the social loafing and social facilitation paradigms. Kerr. are dispensable). 1983) have even defined social loafing as motivation less in groups caused by reduced identifiability or evaluation. Therefore. Individuals performed a shouting task coactively and collectively Evaluation potential was manipulated such that individual inputs were either always identifiable. In the collective condition.g. In the low-effort condition. 1960). by definition. Harkins & Szymanski. 1987. In their "free rider" paradigm. however. their partner. they reasoned that the presence of others is not necessarily drive inducing (as has been repeatedly demonstrated in social facilitation research). 1985). However. When working en collective tasks. coactively. en some tasks individuals may be unwilling to' exert effort if they feel their input will have little impact on the resulting group product. the confederates assessment of the worth of the task could have alDne accounted for Jackson and Harkins's Evaluation Potential Many interpretations of social loafing invoke the concept of evaluation potential (e.g. Schachter. and the theory itself has been criticized for not addressing the underlying psychological processes that it describes (e... tions. Williams et aI. it seems premature to define social loafing in these terms because (a) defining the phenomenon in terms of its causes prevents. 1969) and avoid taking the blame fer a peer group performance. the group succeeds and further effort is unnecessary. or objective) with which this output can be compared. 1983) have suggested another possible cause of social loafing: Individuals may exert less effort when working collectively because they feel that their inputs are net essential to' a high-quality group product (i. or oneself') to' be possible: (a) the participant's output must be known or identifiable. Thus. 1983). 1 They argue that social leafing occurs because. the confederate co-worker's statement of her intended effort was confounded with her evaluation of the experiment's worth. 1983. and (b) evidence suggests that there are other causes of social loafing. such as redundancy of contributions (Harkins & Petty. 1982) and dispensability of effort (Kerr & Bruun.. one's co-workers. 1987. 683 Arousal Reduction Jackson and Williams (1985) proposed a drive explanation to' accompany a social impact theory explanation of social loafing. but should be drive reducing when they serve as cotargets.. in most studies. Harkins & Jackson.g. Moreover..SOCIAL LOAFING by researchers.1989) has suggested that making individuals' collective inputs evaluable to' anyone (including oneself) may be enough to' eliminate social loafing in many situa- I Although evaluation is likely to play an important role in social loafing. Harkins.e. Harkins & Jackson.. she said that she thought the experiment was interesting and that she was going to' try hard. In Jackson and Harkins's (1985) experiment. the presence of others should only be drive inducing when those others serve as sources of impact. or collectively On simple tasks. subjects performed better coactively than collectively On complex tasks. Instead.. . this reduction in individual effort occurs even though each group member's contribution is made identifiable to' themselves. Dispensability of Effort Kerr and his colleagues (e. or identifiable only in the coactive condition.g. and that she was not going to' try hard. 1987. these researchers have found that individuals tend to' reduce their collective efforts when working on threshold tasks that use a disjunctive rule whereby if any one of the group members reaches a certain performance criterion.

g. collective performance is lower than coactive performance. Naylor. WILLIAMS results. several researchers have implicitly used expectancy-value logic to cast light on the findings of social dilemma and social loafing studies (e. Individual effort must relate to individual performance. The group's performance must then lead to a favorable group outcome. and Ilgen (1980) have also suggested that expectancy-value logic can be extended into a model that can account for organization-level motivational phenomena. it is in need of empirical support. Similarly. We believe that social loafing occurs because there is usually a stronger perceived contingency between individual effort and valued outcomes when working individually. The model suggests that individuals will be willing to exert effort on a collective task only to the degree that they expect their efforts to be instrumental in obtaining valued outcomes.. More important. 1965. Heckhausen.. although the self-attention hypothesis is provocative. leading individuals to disregard salient performance standards and to engage in less self-regulation. Only one study (Stevenson. a number offactors must be perceived as existing before individuals will be willing to exert high levels of effort.factors other than the individual's effort frequently determine performance. 1968. Pritchard. the task's importance to the individual and group. adapts individual-level expectancy-value models of effort to collective contexts to highlight the most likely threats to motivation and uses recent theories of self-evaluation in group contexts to clarify which outcomes are likely to be valued by individuals when working collectively. an effect referred to as social compensation. we quantitatively examine the adequacy of some of these viewpoints for explaining social loafing. According to this self-attention perspective. Mullen has found evidence consistent with the notion that self-attention may influence aspects of social behavior in various meta-analyses (see Mullen. The most significant limitation common to all of these viewpoints is that they tend to offer explanations and make predictions about conditions under which social loafing will occur within a limited domain. Thus. Taylor & Brown. 1986). the CEM. is unique in specifying additional contingencies between effort and outcomes that are distinctive to collective contexts. and the degree to which the outcome provides information relevant to the individual's self-evaluation. the CEM assumes that individuals behave hedonistically and try to maximize the expected utility of their actions. When possible in the current review. The resulting model provides a framework for clearly specifying the implications of any key attribute of a given collective setting (e. 1977. However. Shepperd. This combination allows for the clear specification of potential threats to collective motivation in a way that is not restricted to mere task-performance outcomes. 1988). Williams & Karau. self-evaluation information. Olson. although current theories of social loafing provide insights as to why the effect occurs at all. we present a unified theory that proposes to envelop most if not all of the findings in the existing social loafing literature and that offers specific predictions for future research. including the meaningfulness and intrinsic value of the task. Self-Attention Mullen (1983) proposed that self-attention underlies social loafing. We propose that expectancy-value models of effort (e. which must be related to a favorable individual outcome. and feelings of purpose or belonging in one's group. individuals will not work as hard when the available outcomes are not valued. the degree to which the individual is dispositionally predisposed to view collective outcomes as important. When working collectively. The relative value of these outcomes depends on a number of factors. The notion that expectancyvalue models can be applied to motivation in group contexts is not entirely new: For example. In Vroom's (1964) original model. individuals' motivational force is dependent on . Harkins (1987) offered an excellent analysis of the impact of evaluation on social loafing. Tesser. 1991. 1991). working on a collective task leads to a decrease in self-awareness.Kerr. KARAU AND KIPLING D. Moreover. Similarly. Thus. and the results did not support the self-attention interpretation. and valued outcomes are often divided among all of the group members. The key features of the CEM are shown in Figure I. however. several researchers have discussed the potential implications of self-evaluation for individual motivation (e. for a summary). Integrative Model of Individual Effort on Collective Tasks The integrated model of individual effort on collective tasks. Conclusions Prior theories advanced to explain social loafing all appear to have some limitations. and in using group-level social comparison and self-evaluation theories to specify which factors should influence individual motivation in collective contexts. The CEM. 1993.. recent research by Williams and Karau (1991) has demonstrated that individuals may actually increase their collective effort when they expect their co-workers to perform poorly on a meaningful task. Porter & Lawler. Like traditional expectancy-value models of effort. 1987. because participants are more attentive to task demands and performance standards when working alone.684 STEVEN 1. Therefore. 1983. but there is currently no evidence that reduced self-attention causes social loafing (see Jackson.g. then they will not be likely to view their efforts as useful and will not work as hard on the task.g. If the task performance setting disrupts any of these relationships in a way that is salient to individuals. even if these outcomes are directly related to individual effort. which must in turn have some impact on the group's performance. Vroom.g. and Kerr (1983) contributed an equally effective analysis of the role that dispensability plays.group size and nature of task) for the motivation of individuals in that setting.. 1990) has manipulated self-attention across coactive and collective conditions.Goethals & Darley. 1964) can be extended to collective contexts by specifying how working in a group influences individuals' perceptions of the relationship between their effort and their expected outcomes. 1988. For example. they do not provide a framework that can clearly specify which particular factors should moderate social loafing under different conditions. Relevant individual outcomes include things such as objective outcomes (such as pay).

or the degree to which the outcome is viewed as desirable. even if such high effort has little or no impact on tangible performance outcomes.E f. The CEM can be described as "cognitive" both because perceived rather than actual contingencies are hypothesized to influence behavior and because individuals are hypothesized to either consciously or subconsciously select a level of effort to exert on the task.-------. or other important outcomes. or the degree to which high-quality performance is perceived as instrumental in obtaining an outcome. For example. Lepper & Greene. an analysis of individual effort is by necessity more complex for collective tasks than for coactive tasks. However. exerting high levels of effort may lead to self-satisfaction. three factors: (a)expectancy. '" . the model is not necessarily deliberative because individuals are unlikely to systematically process all of the available information about the task performance situation. there are some valued outcomes that do not depend on performance. (b) the perceived relationship between group performance and group outcomes.. or the degree to which high levels of effort are expected to lead to high levels of performance. unless they are particularly motivated to process it because of situational constraints or individual differences.g. The CEM expands on this logic by specifying that instrumentality on collective tasks is determined by three factors: (a) the perceived relationship between individual performance and group performance. gender) • Evaluation Apprehension 5:!ij ". the model places particularemphasis on group-level outcomes that have implications for the individual's self-evaluation. however. collectivism. (b) instrumentality.Deci.. Effort I Individual Outcomes I .~~ c: <II &!l. Valued outcomes can consist of either objective outcomes such as payor subjective outcomes such as enjoyment. For this reason. The CEM suggests that individuals generally consider salient features of the task setting when attempting to maximize the expected utility of their actions. and feelings of self-worth. o~ ~> C:. Research has repeatedly demonstrated that individuals are quite concerned with maintaining a . the model emphasizes performance-related outcomes because many of the most important outcomes available in a performance context are contingent on actual performance. some situations may lead individuals to respond automatically to a preexisting effort script. although individual and group performance are strongly related to many of the valued outcomes in performance settings.. Because the CEM is oriented to explaining collective phenomena. .. 1975..SOCIAL LOAFING gJ 1ii Ci5 685 ~ I Expectancy I x 'go s: (J e. approval from the group. and (c) the perceived relationship between group outcomes and individual outcomes. The CEM also acknowledges that. satisfaction.1 Individual Effort I . the model suggests that working on a collective task introduces additional contingencies between individuals' efforts and their outcomes._ c: <II 8 1 Group Performance • • • • Self-evaluation Feelings of Belonging Intrinsic Rewards Extrinsic Rewards H i Group Outcomes • Group Evaluation • Group Cohesiveness • Extrinsic Rewards Figure 1. and (c) valence of the outcome. 1978). The Collective Effort Model (CEM). whereas other situations may lead individuals to strategically increase or decrease their collective effort.rI I Instrumentality I L----··----··------ Individual Performance Individual Outcomes -----1 i 1 i --~ [ Outcome Valence • Task Importance • Task & Reward Meaningfulness • Individual Differences (e.g. it is the individual's evaluation of the outcome rather than the outcome itself that determines its valence (e...l I Individual .. The CEM also expands on prior expectancy-value frameworks by specifying which outcomes are likely to be valued in collective settings. However. Therefore. when working on intrinsically meaningful tasks or when working with highly respected co-workers. Thus. culture. In the case of objective outcomes. feelings of group spirit and belonging.

Suls & Miller. 1977. Taken as a whole.. Eastern or Oriental culture is often depicted as group. Similarly. 1984) has demonstrated that men are expected to possess high levels of agentic qualities (such as being independent and assertive). 1987). social loafing should be reduced when individuals: (a) believe that their collective performances can be evaluated by the experimenter. because the model suggests that individuals' outcomes are frequently less reliant on their efforts when working collectively than coacti- vely. (f) work with respected others (high group valence. 1991. Carli. Thus. (d)are provided with a standard with which to compare their group's performance. Reis. (e)work on tasks that are either intrinsically interesting. & Bond. Breckler and Greenwald's analysis suggests that individuals can be motivated by appeals to any of the three motivational facets of the self (see also Crocker & Luhtanen. In a similar manner. 1988. one's co-workers. Caporael. (b) a public self that responds mainly to others' evaluations. Therefore. Vogel. whereas Western or American culture is often depicted as individualistically oriented (e.. 1986) proposes that individuals gain positive self-identity through the accomplishments of the groups to which they belong (see also Cialdini et al. Predictions for Meta-Analysis On the basis of the preceding logic.686 STEVEN 1. 1982). & van de Kragt. task situations that provide clear information relevant to self-evaluation.. we predicted that. Greenwald. the CEM provides a number of predictions for our meta-analysis. Recent gender research seems to suggest that women tend to be more group. 1989). as discussed earlier. (g) expect their co-workers to perform poorly. Therefore. 1970. should be more inherently motivating than situations that do not provide such information or that make such information ambiguous and less potent. Tesser. or others.g. Dawes. Although most social loafing research has been focused on the motivation produced by the experimenter's evaluation of performance (for prominent exceptions. Leary & Forsyth. themselves. 1976. subjects from Eastern cultures are more likely to view performance on collective tasks as important than are subjects from Western cultures. rather than redundant with the inputs of others. and concerned with others. meaningful to the individual.Hsu. They distinguished between (a) a private self that responds primarily to internal values.g. Regarding the final prediction. holding other factors constant. but that they also compare the groups to which they belong to other groups for the same reason.or collectively-oriented than men. Second. the recent theories of group level self-evaluation processes all suggest that collective contexts that provide a great deal of information relevant to one's self-validation should be more important to individuals than contexts that provide less information (or none). 1972. Breckler and Greenwald (1986) proposed that the self has distinct motivational facets that are influenced most strongly by three corresponding audiences. Group performance situations produce the potential for self-evaluation from a variety of relevant sources (Crocker & Luhtanen. 1987).. the CEM suggests that individuals will work harder on a collective task when they expect their efforts to be instrumental in obtaining valued outcomes. can only be fulfilled in a collective setting (e. whereas women tend to specialize in activity oriented toward meeting interpersonal demands within the group (Anderson & Blanchard. whereas women appear to be oriented toward interpersonal and cooperative concerns. Clarkson. First. 1988). or others. Wheeler. 1988.or socially oriented. and (c)a collective self that is sensitive primarily toward fulfilling one's role within the context of the goals of important reference groups. several recent analyses suggest that some social motivations. 1982. Eagly & Steffen. (966). several factors are likely to influence the degree to which an individual values collective outcomes relative to individual outcomes. we predicted that subjects in countries such as Japan. 1989).Abrams & Hogg. Broverrnan. although both sexes are likely to engage in social loafing. KARAU AND KIPLING D. Orbell. such as a need for belonging or a need for social communication. partners. Tajfel & Turner. reference groups. Thus. (b) work in smaller rather than larger groups. and China should loaf less than subjects in the United States and Canada.. These perspectives also suggest that self-evaluation information can come from a variety of sources and that information relevant to one's role in valued reference groups may be especially influential. With respect to the CEM. 1990. whereas women are expected to possess high levels of communal qualities (such as being friendly. Social identity theory and group-level versions of social comparison theory lead to similar conclusions. their co-workers. For example. important to one's reference group or to valued others. 1989. Collins. Social identity theory (e.g. Jackson and Williams (I985) .g. and (h) have a dispositional tendency to view favorable collective outcomes as valuable and important. & Rosenkrantz. Two especially prominent factors may be gender and culture. Finally. 1990). & Schmidt.Brewer. unselfish. Indeed. Goethals and Darley's (1987) group-level revision of social comparison theory suggests that individuals not only compare themselves with other individuals to obtain self-evaluation information.g. teammates.. Taken together. Similarly. friends. or high in personal involvement. 1990.we predicted that individuals would generally tend to engage in social loafing and that this effect would be reliable across studies. and respected coworkers) or in a situation that activates a salient group identity. women should tend to loaf less than men. 1982. Goethals and Darley further stated that individuals may seek such group-level comparison information both to obtain self-knowledge (information about the actual level of one's performance compared with others) and self-validation (information that level of one's performance is superior to that of others). whether from oneself. 1989). An extension of this logic to social loafing suggests that women are more likely than men to view performing well on collective tasks as important. see Bakan. research on gender stereotypes (e. 1984. (c)perceive that their contributions to the collective product are unique. Greenwald & Pratkanis. the degree to which the dominant culture from which subject populations were selected emphasizes individualistic versus collectivistic concerns may moderate the social loafing effect. these findings suggest that men appear to be oriented toward individualistic and competitive concerns. Triandis. research conducted in small-group settings has found that men tend to specialize in activity oriented toward task completion.Broverman. WILLIAMS favorable self-evaluation (e. Levine & Moreland. see Harkins & Szymanski. Taiwan.Brown. we predicted that.

(b) type of coactive-collective comparison (within-subjects or between-subjects).g. studying the effects of "sensory sound dynamics. Huddleston. contradictory results have emerged from studies examining gender. and group valence. These key words were searched in the following databases: Psychological Abstracts (PsycINFO). junior high or high school students. Similarly. difficult. 1974. as well as the reference lists of all located studies. if at all. fourth through sixth grade students. maze performance.. be combined with those of other subjects (e. coactive condition only. several researchers provided data from unpublished studies that had been presented at regional or national meetings. e.. Thus. or have produced conflicting evidence... & Ruder. Harkins & Szymanksi. Experiment 1. (c)number of subjects. 1981.g. Experiments 1and 2. Kerr & Bruun. because they intentionally varied the number of individuals present in the alone and group conditions. Some of our predictions mirror well-established effects in the social-loafing literature. the study had to compare the effort of individuals working either individually or coactively with the effort of individuals working collectively. mere acquaintances.g. Second. generating ideas. 1985) had subjects perform a task "alone" in the collective condition by leading these individuals to believe that their inputs would be combined with those of others who would participate at different sessions. were searched. and unclear). This criterion also excluded studies in which only evaluation potential was manipulated. e. thus holding group size constant across individual and collective conditions. researchers generally accept the conclusion that individuals tend to engage in social loafing when working collectively. (d) age of subjects. Variables Coded From Each Study The following general information was coded for each study: (a)date of publication. e. cognitive. Ingham et aI.g. college students. Experiment I. close friends or couples. and task uniqueness have only been examined in a handful of studies. e..g. e. In our meta-analysis. Educational Resources Information Center (ERIC). Doody. In such studies.. motivation loss. the following attributes of the task were coded: (a)complexity of task (simple or well-learned. motivation decrement.g. moderate. (e) expectations of co-worker performance (high. With respect to well-established findings. that is. perceptual. compensatory. meeting paper. and unclear or unspecified). e.. low. (b) task valence (high. e. (f) uniqueness of individual task contributions (unique. again using Steiner's typology (additive. Sociological Abstracts (SOCA). and (c) type of cover story used (effort or performance related or other. using all available comparisons. our meta-analytic methods allowed us to estimate the magnitude of social loafing and examine its consistency across studies. and (e) status of subjects (third grade students or younger. Experiment I). studies were excluded if the number of task performers (i. or unpublished document).g. unknown or unclear.g.. working collectively should lead to reduced drive. studies were excluded if they merely compared the effort of individuals across different group sizes without including a coactive control condition (e." improving writing in a college journalism department. the following predictors relevant to the CEM were also coded: (a)the conditions under which individual inputs could be evaluated.g. this criterion excluded most social facilitation studies.. strangers. . we addressed the adequacy of this viewpoint by classifying studies according to whether they used simple or complex tasks. Price. and mixed or unclear). Finally.. The degree to which these variables consistently influence social loafing across studies is not known. culture. or unclear or unspecified). group cohesiveness manipulation. we quantitatively examine the influence of all of these variables on social loafing.. or obtaining evaluations of clinical therapists). to provide a clearer understanding of what factors moderate social loafing in different situations. e. and mixed or unclear). quality rating of poems or editorials). Kerr & Bruun. Three key criteria were adopted for including studies in the sample. In addition. Ringelmann. Finally. evaluation potential (none.Collective performance was defined as a situation in which individuals' inputs were combined into one group score. free-rider. (d) opportunity for group evaluation (present or absent). Zaccaro. In addition.. coactors or co-workers) present was confounded across coactive and collective conditions (e. and only tentative conclusions about their effects can currently be drawn.. tasks. or teammates. resulting in decreased performance on simple tasks and increased performance on novel. For example. Weldon & Gargano. is not manipulated in such studies.g. complex or novel. Finally. (b) type of effort required by task (physical. brainstorming tasks in which each subject generates uses for a different Method Sample of Studies Computer-based information searches were conducted using the key words social loafing.g." Studies were also omitted if individual effort could not be separated from coordination loss (e. First. and (d) method of combining individual inputs. dissertation or master's thesis.SOCIAL LOAFING 687 proposed that the presence of others is drive reducing when other group members serve as cotargets of an outside source of social influence.e. (c)quantity or quality emphasis of task using Steiner's (1972) typology (maximizing.. vigilance. This criterion excluded studies that merely compared the average individual performance with the average group performance without measuring individual performances under collective conditions. and collective performance. task meaningfulness. effort cannot be separated from distraction and mere presence effects. or organizational employees).1983.clear possibility that subjects might be acquainted but not stated in report). low. However. low. or unknown or unclear). whereas other predictions have either been examined only sporadically. 1987..g. optimizing. Of course. According to drive models of anxiety.. and a worldwide business and management database (ABI/ INFORM). see introduction). the effects of group-level comparison standards. (b) publication form (journal article.. 1984): The individual-versus-group work context. coactive and collective conditions. group size. 1988. The following methodological characteristics were also coded: (a)setting of study (laboratory or field). (c)group valence (high. 1985. which is central to the social loafing construct. and a number of studies have found that social loafing can be reduced or eliminated when evaluation is provided in the collective condition.shouting or rope-pulling. we were able to assess quantitatively the impact of evaluation on social loafing and compare it with the influence of other variables affecting collective effort.g. without providing subjects with an expectation that their inputs would 2 Several studies (e. Dissertation Abstracts International (DISS). or evaluative. e. In the present review. expectations of co-worker performance. collective task.g. disjunctive. between-subjects comparisons). There were no significant differences between studies that had subjects work alone (across individual and collective conditions) and studies that had subjects work in the presence of others (across coactive and collective conditions).g. the reference lists of numerous review articles and chapters. e.

effect sizes based on self-report measures of effort were also calculated so that supplementary analyses could be conducted. 178 units were available for analysis. or personal involvement in such a way that subjects were provided with a clear rationale for viewing the task as especially important or trivial (e. or objective standard was available with which their outputs could be compared. The categories listed for some of these variables were simplified from an initially more detailed set of categories either because (a)some categories had few or no entries or (b) prediction of the effect sizes was not improved by the more detailed categories. the strategy was necessary to allow theory-relevant interactions in the studies to be represented and to test the CEM's predictions about moderating variables.. 1986. or the method used to combine individual inputs (10 studies). collegiate swimmers). kappa = ..e. Experiment 1) and Gabrenya (from Gabrenya. This partitioning was undertaken when (a) subjects of different status or from different cultures were used (7 studies).and low-valence tasks. 1989). evaluative tasks that use a compensatory method of combining inputs. KARAU AND KIPLING D. Latane. All variables were coded by Steven 1. whereas others used individuals or individuals across repeated factors as the unit of analysis. along with their characteristics and effect sizes. potentially redundant. pulling on a rope.. culture of subjects.g. (h) number of task performers present at session.49 for objective measures based on evaluative tasks.subjects high and low in need for cognition performing an idea-generation task-Petty. e. or both). Expectations of co-worker performance were coded as high (or low) on the basis of researchers' manipulations of co-worker ability or coworker effort. The gs were converted to ds by correcting them for bias.e. Baca. The analysis was based on the methods detailed by Hedges and Olkin (1985). engaged in social loafing).e. The effect sizes were calculated by Steven 1. Cacioppo. tasks that were scored as unclear or unspecified are likely to fall somewhere between high. expectations were coded as unclear or unspecified. type of effort measured. The variables for which the initial coding was more detailed are the following: status of subjects. Analysis of effect sizes. & Latane. 1981. Kipling D. we accounted for variability in the effect sizes by relating them to attributes of the studies using both categorical and continuous models." A homogeneity statistic. The median agreement was 100% (M agreement = 98%. separate effect sizes were computed for each dependent variable. China. Williams independently coded a subsample of 19 studies (24%). Under some circumstances.. 1985. the relevant study outcomes were then combined by averaging the ds. Disagreements were resolved by discussion. 1989). Karau with the aid of a computer program (Johnson. completely redundant.e. 1989). task valence was scored as unclear or unspecified. the difference between individual effort in the coactive and collective conditions.. task meaningfulness. or pumping air).. (d) evaluation potential varied (17 studies). To estimate reliability. Expectations of co-workerperformance were also coded as high if subjects were experts at the task (e. For simplicity of exposition. women. Participants' efforts were only coded as evaluable if (a) their output was identifiable either to the experimenter. was consistent across the studies). Harkins & Szymanski. or Taiwan or Western.. "Complexity of task" yielded the lowest agreement. their co-workers. the number of subjects in the coactive and collective conditions were used in all cases to represent sample size in these calculations in an attempt to avoid arbitrarily weighting some effect sizes dramatically higher than others despite comparable numbers of subjects.' Multiple effect sizes from single studies. and a negative sign indicates that subjects exerted more effort collectively than coactively.. Karau. brainstorming tasks in which all subjects generate uses for the same object. To test the CEM's predictions about moderating variables. cern to this meta-analysis. or themselves. (f) group valence differed (5 studies). For practical purposes. By this strategy.g. WILLIAMS object or vigilance tasks in which each subject watches for randomly occurring blips in different quadrants of a video terminal. number of individual inputs combined into the group product). studies were partitioned and separate effect sizes were computed within levels of an independent variable. social. and (b) a personal. When this occurred. and therefore can reasonably be regarded as somewhat moderate in valence. . (i) sex of subjects (men. Qo (analogous to a main effect in an (text continues on page 694) Computation and Analysis of Effect Sizes The effect size calculated is g. the original sample of78 studies obtained from 59 documents produced 166 units. and (j) culture of subjects (Eastern. 1987. as well as on the basis of subjects' selection on interpersonal trust scores (low trusters may expect others to engage in social loafing. shouting and clapping. Although objective effort measures are of central con- 3 The between-measures correlations that were used when implementing this formula were estimated from correlations (a) given in several studies in the sample and (b) estimated from data provided by Williams (from Williams & Williams.71).e.95). is presented in Table I. 1988. To obtain an overall estimate of the difference in individuals' coactive and collective effort reported in the available research. Japan. cf.g. Harkins. 83% (kappa = .688 STEVEN 1.71 for objective measures based on shouting and clapping tasks. Although this partitioning of studies created some non independence in the data. Harkins. Task valence was also coded as high (or low) if subjects were selected on a personality variable that would make them likely to view the task used as especially involving (or aversive). these units are referred to as studies in the remainder of this manuscript. divided by the pooled standard deviation (see Hedges & Olkin. & Ostrom. collegiate swimmers at a competitive meet-Williams. Williams& Karau. was also calculated to determine whether each set of ds shared a common effect size (i. Nida. the 166 units were further partitioned by sex of subject when data were separately reported for male and female subjects. 4 Because some studies used the group as the unit of analysis. and evaluation potential. The categorical models.Brickner. Several studies provided data for more than one relevant dependent variable. (983).. Williams & Karau. or (g) expectations of co-worker performance were manipulated (7 studies). e. United States or Canada). and .g. This treatment is consistent with widely accepted interpretations of evaluation potential (e. Task valence was coded as high (or low) if researchers manipulated task importance.. A study that manipulated one of these variables was partitioned only ifits reported findings were sufficient to allow the computation of effect sizes within the levels of the variable or to determine the direction or significance of findings within the levels. 1985). & Wang.g. In all other cases. the uniqueness of individual inputs to the collective product. (g) group size (i. i. (c)different tasks were used that varied either in their complexity. All such means were computed with each effect size weighted by the reciprocal of its variance. which are analogous to analyses of variance (ANOVAs). The resulting correlations were . For some analyses. & Kasmer. or were selected on the basis of their probable intrinsic interest (or disinterest) in the task (e. (b) group size varied (II studies).. When this latter partitioning was undertaken. and these separate effect sizes were then combined using Rosenthal and Rubin's (1986) suggested formula. 1991). .82 for subjective effort. i. A positive sign indicates that subjects exerted more effort coactively than collectively (i. method of combining individual inputs. a procedure that gives more weight to effect sizes that are more reliably estimated. 26 of which were intact studies and 140 of which were subdivided parts of studies. (991). A complete list of the studies included in the meta-analysis. In all other cases. Q. provide a between-classes effect.g. (e) task valence varied (6 studies).

---::::-::::-:::-::::-::::-::::::::::-::::-::::-::::-::::NNNNN-----N NNNNNNNN--- ------ --- -<.s '" s 01) U ...'<.-.-.-. --- NN -<.----.... -----N -<..::! .".-...----.--::::-::::-::.....-.._-... .. '<.. -<.§. ' '<t'<t"""'<t -o N '<t --.c:: f-< Q) NN -<. NN --... ._ -<.. '<.. s Q) ...-<. NNNN -........ ------ --------....M~ __ ______ .5 ~ r/lt:Q t:Q t:Q 'S.. -.("1"') N ._ '<..-.-..0. '<...... -N --.......-------....-. --NN <.-. Q) Q) e .-----NNNN NNNN NNNN ---.'<.0\0\0----M -... -<.§ ... ----..i: l':i ·8 ... f-< '" ~ -<.-......._ -NM-.-... __ --.....::! 00 88 88 00 00 00 00 NN NN NN NN N00 --...--f<).-....-..-.SOCIAL LOAFING NN f<)f<) -<.-....--...........-....-----.... NNNN --NNN("f") ------------ --.--NN --..--. -<. NN --.. --NN -<.... ~ ~ .. ----------... -....f......--..f. -..------. ~888 r-...:: 01) :...----. l':i ..-.. ..~ '6 l':i i.._ NMNM-~ 8888~8 ~.. ---<.. 000 000 '" . -.. --- NN '<...... .. ... --NN .:.. --......~e ..-.~$ ~~ 1111 NNN '" '" --. -<. -...... ------ --...----.----.-....._-.-. -<..-. --'<.._-.c:: f-< 0 . .....-<.. '<t'<t '<.--......-.-.-<...-...-.::! 01) .-----.0. -<..--..5 ..0.- ~~ .-._-.-.....f.-....------ NNNN -. --'<..-. Nt"""lNt"i OOMOOM -... ::.-....... NN '<.... NNNNNN NN (""iC"'''''j('''''')('fj('f')t''''''') rr')('r')("f")("f")N '<.-..'<......-.....~ l':i ] .E .."N ...... '" 0 0...... '<.... N ·u . --- NN .. 01) l':i 01) f-< '" OS 01) ·8 .. -----'<.. .-----N -<.-.-... ~ l':i 01) ......_-.-. -<..) N-N- ---..0.-.-). --NN NN --..--....:: 01) :.-.--- NN --..... NN -<..----. NN --.. --- ----N '1n <) B r/l ._r("f")("f")("fj~ -... <.....iV) 0000oo'<t'<t-..-.-...-..0\.-. -... NOOONoo 00000 NNNNN -. <) .. 2 '" c .--<. --- &> 01) >NN -<..-<. NN --. 000 N f<) """0 --- -e..:~-.-<.... 0 Q) . -.-- -_NNC"t_ -...--- -....--...._ ("f")r"'j("''''j('fjN -.. -N-N-..------------.......-..-...._ -.=~ u_ gjl..888888 0080 ___ 8888 ~~ 55N~1 00 -----------..N '" 0"'" 00 ONMN 000 O\N 0\ Vi Vir-OCr- ("I")NO'\-V ~ ~ <) I I I I I 0000 ON-- Vrf')M- 0000 ~O I l':i '<t .......".5 o~ ..-----888 o.-----::::::::::-::: --.---. ~~~!::!..-.f NNN '<t'<t'<t M "' .. 01) 0...-..-. ---- NNN --N -<.. NNNNNN C"'1("1")("f")("f")- '<.....~ l':i l':i.-. --'<.-. -N --'<.. --- -<. --.-.----. -<.--. '<t -. -----N -..('f") ("fj(·· ..---- NNN NNN '<t'<t'<t -<..-.-----------....------- -------------'<. NN NNNN -------- .-----------...-..-. N....:.....-.-00 '" 00 00 00 I OOMNr-- IE)lI"'l\OOo\N OOO~OO I ".-..... . 0 & . -... -..--.0 ~ Q) .f~..-. - '<t'<t'<t'<t'<t 88888 \.--'" . 0 ... ::::. 0 ...-----00 0000 NNN 888888 00 ggg8 888 .c:: r/l " 8 i=5 0....---- NN-<. NNNN NN-N NNNN -. --f<)....-...-...... ------ '<t'<t'<t'<t-N '<. --. --'<... Q) .---(""')('''i V NN <....------ --.-.. --......'<t '<t'<t '<t'<t 888888 '<t'<t'<t'<t'<t'<t -.. --NN -<...-..------. ---- NNNNN- -. ~<o::tM-.. NN NN NN NN f<)f<) -<.." NNNN . -...-<.....f<) '<. -<..-.-..... -.....-.-----.. --- 689 -e '6 ...-.. MM('f"'')(''''j NNNN -----.... '0 .c r/l 0 .---...---("f")("·H"'··H·~ (""'H""')('f")C"')("f")-....-.. -----<..fN ("f"'H"'f")("fj("f") --..'<. NN - NNNNNN NNNNNN M("f")("fj("f")("f")("'j NNNNNN MN('f')NNN -------..-.. 2 "§ c '" t:Q l':i 01) '" ..-.....I I I -------... 01) .. -. .-.:. 8g 00 NN '<t'<t '<t'<t 00 '<.-<. ~ Q) l':i OJ) ~ l':i . f<) -<. NN --......-.c:: . -. -...... l':i ~ ~ 2 l':i '" ~ <2 .. 00... -..-.... -...f.--------------------.... .....s 01) 01) .-. NN --. '" '" NNNNNN OOOO<-N '<t'<t'<t'<t . o ·8 . -<..: u ..."')(·')c'r)("'. NNNNNN '<.-........---.........:.-....fNO M~V~~V) -<..'<. --f<)... --- --.iii N ..-. ------ ("i'i("f")("f")t'f"'l -. -<.--00 00 88 ~~ NN NN 88 II I I NN --. U -o e 0. - 000000 -... i=5 ~ > Q) -. NN --..---....-----888 ~~§.-..... --- -e.....-.. NN --..-----..-----.888888 ------I ----I 00 8888 -. Mc.---- C"'·')f""iv"d" ---..c:: ~ .------------NNNN NNNNNN NNNNNN '<t'<t'<t'<t'<tf<) NNNNMN '<t'<t'<t'<t .. --- --.--------~ 00 ---.-::::::::::-::: --..-....-.----.<...f<) '<.. NN '<...". .-....----... NN '<.".... '<t --. NNNN NN--('I")N ~.------ --._-.....:. -<........ 00 ..: l':i .._-.-....on . -<... ----..::::-::::- --..---..._ -N-N- -<.-.-.._-. U 0.......---. H·<")t""')('fj("f") · -<.-. IE) Ir) IE) IE) -<.-..... NNNNN NNNN- -..-.. -r-Mr-OO M-Q'\VOO 00000 I I NNO O~ :. '<..-----.. f<) -<..i NNNN -.----. NN ...-...f..-.'<.-.....--.-.. NN --..-..-..-... ~ l':i 01) ..-.---- NNNNN NNNNN ("'H'~ NNNNN NNNNv -----.0.-~N NNNN .._-.. -----NNN NNN NNNN NNN~ NNNN NNNN -. -<. 00NM-c::t'\ r.. --- -... '<. NN '<. --NN N- '<..-...------ -----. ---<...:....-.. -<.-... --- -------... 00 --......--.. -.: ..

------------...------..--------.---..."""" --M~ 00 00 00 NN -<.. 00 00 --V) V) r')r') -<.."l • ("'fj ---..--- N~~~\Ci -<.. NN --..'<. Oll . ---- -----------------.---.--"""" <. --00 'l"'l" "<t'l" 00 00 NN 00 --.-.-.0.88888 -...--..-.---N---------------.-. --.--. --00 00 NNC"-I(""'.-----.---NNNN --------.-..o r.-:::- ----.-----------.'<.. NN ---<.:::-::::-:::-:::-:::::-. ..c: '5 o .--NN 00 00 NN "<t'l" 00 00 00 "<t'l" 00 00 NN 00 00 """" --.----.§ ..§ :::J (/J ec eo 00 :::J .------.) v: v: -<.. -<.0". -c.------Mr-INN r e.c: (/J .-<. --r') C') -.-..-...--- 00 ~r. -c. WILLIAMS ('.-.------NN NN--- --..r)00f-.--- 00 NN --.0.c:. NN NNNNN -. --NN NNN NN ---<.---------.)MM 00 IIII ---- M~ 00 00 --..--------..-----"""""" 000 000 000 000 NN -c.--88 00 NN -<.0.-.'1(.("') -("f")NNN NNNN --..-.0 "" I 00 00 N V) N 00 I NO rI -=0 00 rN- M Q'\ 0 NO I 0-=0-= -MM\C) r-- 000-=0 OOMf'-O-O 1.-----"""""" --.-.0 00000 00000 NNNN~ r·. ('4")("..----.--r...-----NNN --...-----~NN --. -r') --- --.....---r'lM 00000 00000 ~r""'''H'f''')\'''f''')M NNNN~ -"" NNNNN ("")("<'j( .--.----:::.690 STEVEN 1.-.-.-:::--:::-:::.... 00 00 r') r') 00 --.-- -..-----.-c. -<..-----. -<.---.s (/J Oll o OJ 0...-----. NN ---- -N --.--- MMt"I'iM ----..--------..IN --.-----..---.0>. --"" r') --. """" 00--00 --.jr-l -------00000 00000 NNNf")(""') '<. --00 00 ~~ ... KARAU AND KIPLING D.----..-OO V)t---oo- 00-=00 -MMOOM ::: ....------.r- ~00 Nt- t-.

.....OOl) Ol) .> 0..._....._.---.'" ---..<T on 00 00 V) V) o-..M NN NN NN MN NN ------.-.> ......------------. ....---..-------...-..... t> 'i3 ...:::.......-.'" . 0 t.'" -----..:: ... -. 0 MM NN . ("fj __ -.......-------.D ·c '" t:: OJ) Ol) u ... cev M <'l NNN MN <'l N -...-.---.-::::-::::-::::- ---('....---~~ 00 00 NN ~~~~~~ 000000 000000 Nr-lNNNN . ---..-----.-.....-----.---...-.---"'~ -.....---__ ------------.. '" "§ s ..M ("f)t"'I")("f')('fj NNNNNN -.._. t:: '" ~ '" 0 0...-:::::-:::::::-:::-::::-:::.'" '" ----------... --N-N- --'Vf'f")("f')N NN ------..........---....-.......--__ -. 8888 888--80 ------......~~ ~~ ~~~-...... 0) .-.----NNNNNN NNN NNN-N--N MM ("fj __ -.... N Ol) t:: o ._...._..---.. NNNNNN -.------~~ ~~ 00 00 00 NN NO V.---..._..---.-._..2 .'!l ..-..88 ....§ C/) OJ) "<: ...-N-<. .88 88 ---00 00 000 -----.......00 ~ ·8 '" '" 00 ..-.......-------............~ N-\O "":"":0 00 0000 0"": on o-. '" .. NN MM MM -N ------88 ---------....._.s::~~ ~o~s::c t::- 8c .---. "·c Ol) '" E "'.IM- ---. ::> E III "<: ...---...._.------...-.....------------888888 888 ---.------...... NNN 000 000 M '" v) v) v-.......0 -0 '0 0 ~ 0 NN -N ~ ~ ~ ...... .......------. V V) 00 00 00 I1 O'-DI- 00-0 000 I 000 00 I ~- ._ '-D\oN~~(""')f'I")\OO\O.:....------('''''l\'''''''i ---::.... ..-.. ---M Iv ~ ....----...: N N-'<. t:: Ol) E "5._. ("f')("f')('f")("1"') NNNN -. '" ..-. ::> ...... 0 00 .-...---.-. '" 8 ::> ~~~ ~ ~ ..2 ~ 8 8 8 ~o. ---...---..---.o'l5MI~r:: 88~88~ ---88 88 888888 __ ---.----... .._ . '" OOO\O\O\Ct.. ".....-----. '" -...... NNN NN NN NNNN NNNN MMM MM __ -...._ -<...---~~ ~~ ---. 8.-----.....S? '5 . ·0 \C5~ N\..~~ ------.---NN NN NN NN MN NN NN MM N <'l NN NN --I __ 0 1-... __ ---.-----.------------........-.2 E .. t:: t:: ....---~~ ~~~ NN NN ...........---...---.. .....: Ol)O .--------..' 0 t!.... NN MM __ ----N -...---. ...---- 000 000 ......__ -----._ 0000 0000 NNNN ..-._...._...---.00 00 ---.-.----------.--------N----------------..------...__M ---- 8-.....on 0 .-........ '" 8 ...... o- ~~\O-<...-------...........8888 ... __ 0 00 on"....... 15' C/) . '" ------------.._..-.. "~ ~ ~ t) 00 00 000000 I III I MN M V) NM M~ O-NOO<o::j'OO ~~-tr)oooo N I.---..: ...------__ -..-..._ ~~~~ 0000 888888 000000 NNNNNN .e: ---------.5._...---...._.> 8 ... .------... M"" ('f')("I"')('f')("'f'j __ '<.......---.-----------...- .---.---...:.. ---------NN NN NN NN NN MM <'l N MM M.......:.........~~ ...---...---..> ::> "'''' ~ai .....---....M N M0 ~~ ~~ -----.....-. v) rri~r<"iM vv v: -.-----. ::> o t:: t:: ... ::> 0 ° "S·S ...: f-< '" '" ._... ------------------'" '" ---------------NN -i-i 00 on V) -i-i NN NN NN -N NN ._ 00 88 88 00 888 00 00 00 000 NN NN NN NNN 00 NN 00 00 00 00 00 00 ---0000 00 ---....-ac..-------.. __ NNNN ___ C''I'''H'''''l(''fjf'l'')('f'')(''f') NNN-NN NN .. 1-0 MN -~ ~if}NN O--N (""')0'\0\0000 ~vV)~V)"¢ MV "":"":"":oo<--i I 0'" o....-.00 ---..::::...._ ---------.---------...> ::>::> sss 0l)0l)0l) -s E ~ e..---.-----......---.:...e: 0.---NNNNNN ('1")("1"')("1"')("'1"')("1"')("'1') -....__ -....5... ".... -..---..._. '" '" f-< ------88 ---88 ------------------------NN NN M '" -i-i NN 00 00 NN NN M.-..--.-...2 0...-..... ------.'-D -------I---- NNN-<.... ._... ....0l)0l) "s ..---. M M--<...---.......e: f-< NN MN ::"...000 88 8~ 000 v)o 000 v- ~.. 000000 -----.......... NNNNNN <'l M vv NNNNNN ("f'j("l")('f')("fj NNC"H'I"')NN ------..:::...-----NNN----NNNN NN N- :::. '" ---------------<--i-i ....q- v~v~~~ .-._... ~ _~>1Il1Il '" '" c...-. ~~~ .'" 'Ch"a'a .t:: ......---..---....-----------...---. NNNNNN NNN <--i-i <--i-i NN N("f')N~ -.....SOCIAL LOAFING 691 N N N --<.. t:: Ol) .-----.--.-.---.:......---.. -----.-...---..........t::t..-----. 0. '" t:: '" ~ -c ~ "~ . .---..88 -----..~ -"_____ __ __ ('<"')-("1"')0 NN MM 00 NNNNNN .s t:: "~ .--------..---NM NNN NNN NN NN----- 00 00 ::> NN NN MM ~~ ------. ° t:: ~ ~ t:: ::>. M '" '<....-:::.:: c:: "._ NNN '-D '-D o-. '" t::1S ".. ---- -----'" '" '" -----NNN NNN --M NMN NNN -<.-.n NC""H'''''')NNN '" ..---------------..:....OVN-VN N ('i"')("j ("'f") tv) N '" NN -MN_('f')N NN NNNNNN NN NNNNNN ---rf')('t')('f") NN NN NN-- -. ---- MN ~~~ MM ----------------<.-..---.

...----.....--.1V1-0' "<:tooOOO- ""':""':0000 c '5..-.. .. '<T'<T NN -------- -------. NN <') -------r') 000 000 -------- -...o rlMr"I")~ -.---. ~ fOJ -c..-:::::-:::-::::-:::-::::- 000 000 MM MM -------- -------- NN -------- -----------~~~ -----------e')MN M 0iI...... '<T'<T --N -----------.. 0 .........-N '<.. e')M NN -<... tJ ~ ------ NNN -------- 2 :t a N- NN N- -------..... MM -------<') -------NN NN ---.a-..___ --..2 C <ii .......ononO ---------000 NNN - "<:t'<T "<:t'<T"<:t "<:t'<T'<T"<:t 0000 0000 NN r--r-- -D -D -----... f- C -" >.-:::-:::......:f t"")MMMf""iM --....: E '" 'i:: <l) <1) <U 0OJ ~ fOJ u "0 OJ '" c 0 c 0[) 0[) c 0[) 0[) .. ~~~~~~ ----------. <U > <l) 0 >.. ----.. ---- -<......-..... N '<T '<T'<T on ------ 00 -D '<T NN ~c:: 00 00 00 00 00 00 -----... ...........l U 00 ON - <') -"<:to q ""':0""': - -D 0-- on 0 I I 0 0""': 0""': 000 I I N -D <n 0 -D-D N'<T N 00 <') NN- 0-.'<...------- :::--::::-:::::-:::::- ::::-:::-:::-:::...... C::'_ ---- -------<.-.'<.-----<.----- ("I"')MMt"'f')("'I"')M ---------NNNNNN -..~ 0 ..-.... -.-------<n ""':""':0 0000 <') NNNN <n<n0 00 00 00 00 '<T'<T"<:t'<T -..... ---- -------- NN '<..'<. N~N~N........... '<. >.-..:: . -----------NNN M '" ------- -------- ".........-<. '<.. -NN <') <') 000 NNN "<:tD-D NNN ---------- C::'_ -D 0-- 0 0 0 0 0 '<T "<:t 00 00 C::'_ -D 0-- NN -<..- 0 s00 0-- . . WILLIAMS NNN ""'" .Q'-C5-.. -------...........----._---.-.'<.)\0\0'-0\0\.. ---e')Me') C::'_ M ...... NN NN f- .M --------------------NN NN NN '<T'<T ------....D-C.:.._ -----.. -------- 000 000 C---:c5 00 r--- 000 .-::::-::::-:::-::::-:::NMNNN<"'-l --------.r:~ '<...--..0 -------- -------- -0--0 NN 00 NN ----- ----------------------qqq NNN NNNN NNM -<. NN NN '" '" M NNN NNN ------------------.. C::'_ C::'_ M NN NN M'" . ---q N --------------qq r') NN <') 00 00 00 00 00 NN --------------'" 000 000 000 000 ~ q NN -------II -------- NN NN -------- -------- ----------------------~~~ -----------~~~ -----------<') '" NNN M '" --------------qq 00 -D -D 00 00 NN MM --------------qq 00 00 00 NN MM 00 M --------------'n on --------------- NN N ____ M N---M0 -<. 000 q 0 .. ---000 000 " 0-.D " 'iil <U "0 OJ > <ii ... MMNN -- NNMt'f"l-'<..lffJ 0[) cIs cIs Li: 0 .. ..'<..... 00 B o: U <U a. .------ -... ---- NN NN -<...<:: <l) <l) '" .692 ----------------------qqq M <') '" STEVEN 1. ..-. NNNN -. ---'<. N '<..- .. KARAU AND KIPLING D.... '<....... __ 0000 0000 0000 0000 ---- NNNNNN 000000 000000 000000 ---000000 ::::. r.888888 '<T'<T'<T'<T"<:t'<T 000000 NNNNNN ----<..D -. -------- ----------....<:: ffJ :: 0 2 on c ~ C .. \(..... NN -------- ------------------..0...Ci ------------ 00 q 00N 00 MMM ------------ 00 -<.... '<.. NN- NN-- -. ----'<..-----__ --.... r. NNN -NM N N '<.... 0[) ~ .. c E '13 'OJ C (....S > <1) C :a <ii > <l) . 00 -D "<:to-- "<:t"<:t"<:t -N 00 NNN 000 '<T'<T"<:t 000 . ---- NN NN '<T'<T '<. -.---NN --M... -<.....---- ---. -.....1. -c. NN NN -<.......- --NNNN -.' r ('--c...-------'<T"<:t 00 00 NN -........ NN NN -<.. -------- ------ -------- --N -------------- -..------------<.....'<..§ ~ u.'-D 0""': ""':0 0""': 0""': 000 I I -r-- '<T <n on '-D 0-'<T <')r-N- - .... .. .......--...D " > 0[) ..... . ......c OJ OJ. '<T'<T --------------- MM -N -<........------::-::::-::::-::::NNNN ------- -----:::.. M N '<..---..CiN. NN NN r') .... '" C . NN NN N- NN -------- ----D -----'<T'<T 00 00 -------- -<.. '"c-..... ------. -<..---..<') <') r-- OONN NV)OON -V)Nr- 00\('"'. . -<.. NNNN NNNNNN NNNNNN NNN -----------'" '" ----------------------- rl r''I(' . -"<:t'<T '" <') 00 -----. ----'<. 0 . '5...... N---<. :a .. ..<:: -e') NN ....-----.. '<. '<T'<T"<:t '<T '<T'<T -<. ... NN Q~ NNN NNN --N -<. ..---. ---MN '<T'<T ---------------------- - . ---.. ---- --------------- -------------------<'1 ..-- <.. -- ---'<........ ---- r--r-00 00 on r--. O[) "0 -0 0 -5 <U 0 -------... . M ... NN --------------- ------- NN NN -------.. -<. 0 '~ E E u: C v: 'OJ (. NN N MN -------. N '<...-<.-:::-::::-:::-:::-:::-:::--. 0 00 0' ' -D N- o on 000 - . ---NM NN N"'- ------------<.. '<.......... M - -c...'" <') ------NN NN .....<:: :..--. . <U OJ 3 o 0- 00 00 00 00 NN -------- M 0 ~ <') C::'_ NN M -<..... -c.....----------- v ......'<.........D " . ---........ ---MN --------------- NNN NM ..0 MMMMMM s <U .. M '') NNNN ---....----...~ ~ u. ---- NN N- ---. c 0[) '5 .. .

rN NNN NNN NNN NN NC"lNN ::::-::::.-.----------...-..--------.i ""f'"l 00 00 NN -------N 0-000 NN .i-...._-._ --M("f') NN NNNN NNN NNN ....-.r...-.---...- ---..r.r.---...-..r.---.:::-::::.---NN -N NNNN NNN --------- ::::-:::: ..-- 88 NN 00 00 00 ---.-----.....-N------. ---..-.----..--....---.-------.... ---.---------.i 00 00 ----------.r 0000 0000 -----...------------. NN MM NNNN .-----. ("'''H''')(''I'')(''f'') -..o 00 00 -----------~~~ -..---........o..._ 0000 NNNN NNNN 0000 -----....-. ---....----f'"l -("'1")-(") -------------.. -----........~ OIl c .-----........-------------------------------- NN NN "'f'"l NN NN ------------------------NN NNN --N ------------- ------.i-...rO<n "":0 I r-~O'\O\ V")-t-N 0"":0"": I "2 '" <!) ·u <:: .. '" NN -----------000 -----------000 000 000 NNN f'"l 'C5..--------..r -N- 000 00<n N V)O~ __ -00 -\ON .......SOCIAL LOAFING 693 NNN ..r.

(b) were published as journal articles. regarding key theory-relevant predictors. mixed. numbers in table represent frequency of effort comparisons in each class. studies (a)typically used simple tasks. Finally.and between-subjects comparisons of individuals' effort across coactive and collective conditions. 1987. Summaries of continuous based on reports for which information was available on each variable. The continuous models. In addition. Hedges & Olkin. and (e)only occasionally included organizational or nonadult samples. studies (a)were generally laboratory experiments with some field studies. 1985). As shown by the central tendencies of the characteristics presented in Table 2. the last 10 variables in Table I show that studies generally (a)varied evalua- Table 2 Summary of Study Characteristics Variable and class Value 1985 121 Conditions None Coactive Coactive Unclear High Moderate Low Variable and class under which individual can be evaluated condition only and collective conditions Task valence 30 or no basis for determining Group valence 9 II 129 17 12 151 outputs 5 117 28 16 Value Median date of publication Publication form Journal article Meeting paper Dissertation or master's thesis Unpublished document Median number of subjects Median age of subjects (years) Status of subjects Third grade students or younger Fourth through sixth grade students Junior high or high school students College students Organizational employees Setting of study Laboratory Field Type of coactive-collective comparison Within-subjects Between-subjects Type of cover story used Effort or performance related Other Complexity of task Simple Complex Unknown or unclear Type of effort measured Physical Cognitive Perceptual Evaluative Quantity or quality emphasis of task Maximizing Optimizing Mixed or unclear Method of combining individual inputs Additive Compensatory Other. provide a test of the significance of the predictor as well as a test of model specification (Hedges & Olkin. additive tasks. (b) used both within. studies generally (a)were published recently. (c) used a moderate number of subjects. we conducted outlier analyses to determine whether the effect sizes were relatively homogeneous aside from the presence of a limited number of aberrant values (see Hedges. and (c)generally used maximizing. 1985). the frequencies were 40 for men. Finally. Results and Discussion Characteristics of Studies Before considering the findings of the social loafing literature. and (c)typically informed subjects that the purpose of the experiment was to examine effort or task performance. (b) tended to measure either physical or cognitive effort with some studies of perceptual or evaluative effort. For categorical variables. WILLIAMS ANOVA) and a test of the homogeneity of the effect sizes within each class. which are least squares linear regressions calculated with each effect size weighted by the reciprocal of its variance. a When studies were subdivided on sex of subjects. . we first present the characteristics of the studies from which our conclusions will be drawn.694 STEVEN J. 24 for women. Regarding task characteristics. KARAU AND KIPLING D. or unclear 23 14 8 48 20 6 5 6 131 5 143 6 142 24 69 97 115 51 135 7 24 70 53 29 14 127 30 9 High Moderate Low Unknown or unclear Opportunity for group evaluation Present Absent Expectations of co-worker performance High Moderate or no basis for determining Low Uniqueness of individual contributions to collective outcome Unique Potentially redundant Completely redundant Median group size Median number of task performers present at session Sex of subjects" Men Women Both Culture of subjects Eastern Western 14 144 8 8 64 94 3 3 28 12 126 17 148 147 11 8 variables are Note. (d) used college-age samples. as a supplementary procedure. and 114 for both. QWi.

Taken together. and (j) generally examined subjects from Western cultures. (c)studied low-valence groups of strangers.001.44 0.48 964.23 and 0. Inspection of the out- Table 3 Summary of Effect Sizes (ds) Criterion ds with outliers Sample size (n) Mean weighted d (d+t 95% CI for Homogeneity (Q) of ds making up d:b Mean unweighted d 95% CI for mean unweighted d Median d Differences indicating social loafing Value a.50 be labeled as moderate. the available research seems to suggest that there is at least a moderate tendency for individuals to engage in social loafing and to reduce their effort when working on collective tasks.79) ds excluding outliers Sample size (n) n removed outliers" Mean weighted d (d+) 95% CI for d.e.79 Note. (b) did not manipulate task valence.001) and when researchers did not specify a prediction (di+ = 0. producing lower effort levels in the collective condition than in the coactive condition.44 may not necessarily provide the best measure of the overall magnitude of social loafing.20.001. 99 64 (. Bond and Titus's (1983) review of the social facilitation literature reported that effect sizes ranged between -0. 163 0.19/0.00). On the basis of these standards. a Effect sizes were weighted by the reciprocal of the variance.e. many of the coactive-collective comparisons included in our meta-analysis were drawn from situations in which social loafing was not predicted. The several measures of central tendency reported in Table 3 are relevant to this issue.D1.61 0. Thus. by a confidence interval that does not include 0.06.19. Overall Magnitude of Social Loafing The summary provided in Table 3 allows one to determine whether. n = 15. who suggested that ds of 0.76.e.39) 0.39/0. 128/163 (. and ds of 0. Homogeneity (Q) of ds making up d.86. individuals tended to engage in social loafing and exert less effort collectively than coactively. within the middle range of effects in the domain of social behavior. The weighted mean was very similar to the unweighted mean. b Significance indicates rejection of the hypothesis of homogeneity. C liers revealed that among the 64 effect sizes removed. depending on the complexity of the task and the type of performance measure. Effect sizes were stibst~nti~ larger (di+ = 0. (f) used tasks in which individuals' inputs to the collective product were either potentially redundant or completely redundant with the inputs of other group members. p < .SOCIAL LOAFING 695 tion potential across the coactive and collective conditions. 95% CI = -0. In a particularly relevant comparison. (h) tended to have all task performers physically present at the experimental session. engaged in social loafing). However. ds of 0. a categorical model computed for whether social loafing was predicted by the original researcher(s) was highly significant. Most studies were designed to test the potential of one or more factors for eliminating social loafing.00 to 1. it should be recognized that the tendency of researchers to focus on isolating conditions under which social loafing can be reduced or eliminated is likely to lead to an underestimation of the magnitude of the effect.. individuals tended to engage in social loafing. The proportion appears in parentheses. the mean effect size of d = 0.20 be labeled as small. an 5 Contrasts were a priori. One means for interpreting the magnitude of the mean social loafing effect size is to compare it with effect sizes produced by other quantitative reviews in similar domains. As shown by the homogeneity statistic given in Table 3. and the removal of outliers substantially decreased the value of the weighted mean. n = 89) when social loafing was predicted than when social loafing was not predicted (di+ = -0.21. the effect sizes were not homogeneous (i. Positive effect sizes indicate that subjects worked harder coactively than collectively (i. unless otherwise indicated. these figures indicate approximately a four-tenths standard deviation in the direction of reduced collective effort. The post-hoc contrast procedure is modeled after Scheffe's method for multiple comparisons in an analysis of variance (see Hedges & Olkin.005). The implication of the finding that people tend to engage in social loafing when working collectively depends in part on its magnitude. •• p < . Consistent with the CEM and with prior reviewers' conclusions.. d.24 0. .5 The "predicted" and "not predicted" values may represent estimates of upper and lower bounds for the magnitude of social loafing.. 49 were above and only 15 were below the median d computed for the original sample.36/0. The confidence interval for the weighted mean after outlier removal showed that the overall tendency toward decreased collective effort was still significant after outliers were removed.29 125. Finally.80 be labeled as large. social loafing might be regarded as small to moderate in magnitude. In view of these comparisons. 1985). on the whole. QB(2) = 310.70·· 0. CI = confidence interval. n = 59.48 0. Another standard is provided by Cohen (1977).81.29. Q = homogeneity of ds. consistent) across the studies. Therefore. (e) did not manipulate expectations of co-worker performance. (g) had a relatively small median group size. whereas the median was somewhat lower. = mean weighted effect size. contrast p < . Eagly's (1987) overview of a variety of mean effect sizes found in meta-analyses of social psychological research topics suggested that they ranged from about 0. 95% CI = 0. with regard to the robustness of social loafing.09 to 0. (i) typically included both male and female subjects. post-hoc contrast p < .00 value that indicates exactly no difference (i. An overall difference in effort across coactive and collective conditions is shown by a mean effect size that differed significantly from the 0. (d)did not provide an opportunity for group evaluation.41 .70 to 0. Indeed. A very large proportion (39%)of the effect sizes had to be removed to attain homogeneity.95% CI = 0.08 to 0. and larger than social facilitation. d = effect size.

they must actually work harder than they would ordinarily work alone. 1985) and social compensation (Williams & Karau. Contrasts showed that the tendency to loaf was greater when task valence was low than when task valence was either high (p < .696 STEVEN 1. 1991) views. KARAU AND KIPLING D. individuals loafed when they expected their co-workers to perform well or when no expectations were provided. or low (p < . social loafing was eliminated when evaluation potential was not varied across coactive and collective conditions. Individuals did not loaf when group valence was high. contributed to this relative inconsistency.27). Thus. given that the unspecified tasks were likely relatively moderate in valence. Task valence. when the interaction between expectations of coworker performance and task meaningfulness was considered.005. but loafed in all other conditions. individuals loafed less when group valence was high than when it was low (p < . Evaluation potential Consistent with the CEM and with evaluation accounts. In support of this notion. Kerr & Bruun. post-hoc contrast). as has been found in the prior social compensation work. Consistent with the predictions of the CEM. 1983) and Harkins and Petty (1982).00 I). This suggests that high levels of task meaningfulness or personal involvement might eliminate loafing. It is possible that the focus of social loafing research on limiting conditions. WILLIAMS examination of the consistency of findings across studies is worthwhile. do not completely resolve the controversy between the matching of effort (Jackson & Harkins. These results highlight the potential of the CEM for predicting situations in which motivation gains might emerge on collective tasks.00 I. individuals were more likely to loaf when their outputs could only be evaluated in the coactive condition than when their outputs could not be evaluated at all (p < . group members' . post-hoc contrast) than when it was low.11 to -0. as evidenced by the confidence intervals presented in Table 4. Also. Categorical models that yielded significant between-classes effects are presented in Table 4. As hypothesized. This figure is substantially larger than that reported in other metaanalyses of social behavior. they are conceptually distinct. Individuals also loafed less when group valence was either moderate (p < . low vs. However. yet essential to group performance. 95% CI = -1. high contrast. = -0. these results suggest a direct relationship between task valence and individual effort on collective tasks. post-hoc contrast). An examination of confidence intervals shows that individuals loafed when task valence was low or unspecified. these findings suggest that enhancing group cohesiveness or group identity might reduce or eliminate social loafing and that group-level outcomes are indeed valued by individuals in certain situations. a significant motivation gain was found only for the four comparisons in which individuals worked on a meaningful task with co-workers who were expected to perform poorly (d. p < . Uniqueness a/individual inputs. individuals must not only increase their usual collective effort. the tendency to loaf was greater when task valence was unspecified than when task valence was high (p < . Individuals worked just as hard collectively as coactively when their individual inputs to the collective product were unique. as predicted by the CEM and prior theories of self-validation processes in groups. unknown or unclear (p < . the tendency to engage in social loafing decreased as task valence increased. which both suggest that individuals' outcomes are more contingent on their efforts when working with others who are expected to perform poorly. The logic of the CEM suggests that individuals would be unlikely to exert such extraordinary effort unless they viewed the task as meaningful and therefore highly valued a favorable outcome.05. p < .001) or when their outputs could be evaluated in both the coactive and collective conditions (p < . post hoc). Thus.001) or unknown (p < . Finally.005.06. For example.00 I)." 6 Although uniqueness and dispensability are often closely related. but loafed when their inputs were either potentially redundant or completely redundant (ps < . individuals may perceive that their efforts are less instrumental in obtaining valued outcomes when their inputs to the collective product are either dispensable or largely overlapping with those of others. Similarly. outlier analyses showed that the effect sizes were highly inconsistent in magnitude: 39% of these effect sizes had to be removed to attain homogeneity.001) or unspecified (p < .00 I. These findings are consistent with the claim made by Harkins (1987) and others that evaluation potential is a mediator of social loafing. Consistent with the CEM and with the results of Kerr (I983. combined with the wide range of tasks and subject populations used. but not when it was high. the tendency for individuals to engage in social loafing decreased when evaluation potential was held constant across the coactive and collective conditions. Expectations of co-worker performance. Taken together. on an additive rope-pulling task.69. a resolution is possible when the role of task meaningfulness is considered. however. Theory-Relevant Predictors of Social Loafing The study attributes were examined as predictors of differences in individuals' coactive versus collective effort. These findings. In fact. the tendency to engage in social loafing was reduced when participants were provided with a group-level comparison standard. It is also possible that this inconsistency reflects the complex nature of the phenomenon and suggests that social loafing may represent a class of effects with psychologically distinct causal mechanisms. effort levels were not significantly higher collectively than coactively when participants expected their co-workers to perform poorly. but did not loaf when they expected their co-workers to perform poorly (low vs. and significant continuous models are presented in Table 5. Group valence and group-level comparison standards.00 I). evidenced by 79% of the comparisons being in the direction of reduced collective effort. A group member's inputs can be indistinguishable from the inputs of others. Also. unspecified contrast. Nevertheless. The basic pattern of findings is consistent with both the CEM and the social compensation research. The direction of the effect was very consistent. However.0 I).. post-hoc contrast). the tendency to loaf was lower when group valence was high than when it was moderate (p < . Because social compensation represents a motivation gain.00 I).

13.33). This pattern is consistent with the reasoning presented by Jackson and Williams (1985) and provides initial. Kawana. & Moreland. This finding suggests that men may be particularly attentive to the reduced efficacy of their inputs when working in larger groups. 1980.05. post-hoc contrasts). tentative support for the arousal-reduction viewpoint. or college students (p < . suggesting that telling subjects the purpose of the experiment may inhibit their usual tendency to engage in social loafing. Experiment 4). Jackson & Williams. Szymanski. 1985). whereas women may tend to exert more consistent levels of collective effort across various group sizes. However. Williams. because the six comparisons based on adult nonstudents involved diverse sarnples. possibly resulting in smaller error terms and larger effect sizes and (b) tended to focus on documenting the effect rather than isolating moderating variables and were thus more likely to find larger effect sizes across all comparisons. the mean weighted effect size was-O. If this difference were indeed attributable to a greater tendency by men to focus on individualistic outcomes. the magnitude of social loafing was not dependent on whether (a) maximizing or optimizing tasks were used. Simple linear regressions showed that both group size and the number of task performers present were positively related to social loafing. subjects were more likely to loaf when they were not informed that the purpose of the study was to examine effort or performance. four of the studies coded as using a complex task used manipulations of high task difficulty (Bartis. Kerr & Bruun. post-hoc contrast). 1989. individuals did not perform significantly better collectively than coactively when working on complex tasks. b* = . Harkins & Petty. (b) tasks were additive or compensatory. Consistent with the CEM. the magnitude of the effect did not vary depending on whether (a) the coactive-collective comparison was made within-subjects or between-subjects. Thus.14 to ." magnitude of social loafing was also larger in laboratoryexperiments than in field studies.f The responses are completely redundant in form. Moreover. S One comparison involved adult residents who worked on a laboratory brainstorming task (Cooley.001). It is possible that the prior finding may be indicative of a developmental trend whereby children do not become attentive to strategic concerns in task performance until they reach a certain age (see Williams & Williams. Social impact theory and self-attention theory devote particular attention to group size effects relative to other aspects of social loafing. & Latane. the mean weighted effect size was -0. 1989). they may not have been difficult enough to create an inverse relationship between effort and performance. we estimated several multiple regression models to examine the Models Relevant to Evaluating the Generality of Social Loafing The status of subjects produced a significant between-classes effect. In addition. only seven comparisons involving complex tasks were available. as arousal reduction would predict. presumably because they have more group-oriented priorities than do men and individuals in Western cultures. adult nonstudents tended to loaf less than fourth through sixth graders (p < . creating low power to detect a difference. post-hoc contrast). Finally. suggesting that social loafing is robust across tasks. Task complexity Individuals performed better coactively than collectively when working on simple tasks. 1982). interpretation of this time trend is clouded when it is recognized that the earlier studies both (a)tended to use fewer experimental paradigms that were more easily controlled. as shown in Table 5.18.005. 1991). This finding suggests that social loafing does not occur simply because subjects allocate greater effort to coactive trials than to collective trials (see Harkins. post-hoc contrast).01. Latane. . The latter trend is relatively uninterpretable. cognitive. the magnitude of social loafing was larger in studies that used samples of only male subjects than for studies that used either mixed samples (p < . Multiple Regression Models Although most of the theory-relevant predictor variables were relatively independent (rs ranging from -.09. post-hoc contrast) or college students (p < .00 I). albeit for different reasons. 1981).0 I. and one comparison involved Japanese business managers working on a noise-production task (Williams. Gender and culture. Although these tasks were not simple. post-hoc contrast) or samples of only female subjects (p < . the group size effect reported earlier was only reliable for comparisons involving groups of men (b = . for a discussion). Also as predicted. 7 However. and (c) physical. 1984. & Harkins. Fichman. 1988. the magnitude of social loafing was larger for subjects from Western cultures than for subjects from Eastern cultures. but performed at least as well collectively as coactively when working on complex tasks (p < . For these studies. Finally. For the remaining three comparisons that directly manipulated complexity (Griffith.36. 1981. by definition. & Williams. but the effort of each member still contributes to the total group performance. p < .SOCIAL LOAFING 697 Group size. a simple linear regression showed that date of publication was negatively related to the magnitude of social loafing such that effect sizes were larger for earlier studies. Contrasts showed that third graders and younger students tended to loaf less than either fourth through sixth graders (p < . These provocative findings suggest that both women and individuals in Eastern cultures are less likely to engage in social loafing. Interestingly.05. Most of the theories discussed in the introduction would predict group size effects. or perceptual effort was measured. Effect sizes were larger for studies that combined more individual inputs into the collective product and for studies that had more individuals perform the task at each session. four comparisons involved either American or Chinese adult managerial trainees working on an in-basket exercise (Earley. the CEM would predict that men would be more attentive to strategic concerns than women when working on collective tasks. Yet.05. evaluative. Social loafing was also greater for studies reported in journal articles than for dissertations and master's theses or for unpublished documents (ps < . Impact of Other variables on Social Loafing The trend toward decreased collective effort was enhanced when a deceptive cover story was used.00 I). Consistent with this reasoning.

22 0.18 0.41 0.42 0.57 0.28*** 72.08 0.13 809.44 -0.47 771.51 0.49 3.08*** 3.17 616.05 355.31 *** 0.19 0.25 0.11 *** 118 23 14 8 0.39*** 140 23 10.36 0.79*** 18.06 0. Mean weighted effect size (di+) Homogeneity within each class (Qwi)" Variable Evaluation None potential" and class n Lower Upper Coactive condition only Coactive and collective conditions Unclear Task valence High Moderate or no basis for determining Low Group valence High Moderate Low Unknown or unclear Opportunity for group evaluation Present Absent Expectations of co-worker performance High Moderate Low Uniqueness of individual contributions Unique Potentially redundant Completely redundant Sex of subjects" Men Women Both Culture of subjects Eastern Western Complexity of task Simple Complex Unclear Status of subjects Third grade students or younger Fourth through sixth grade students Junior high or high school students College students Organizational employees Setting of study Laboratory Field Type of cover story used Effort or performance related Other Publication form Journal article Meeting paper Dissertation or master's Unpublished document thesis Positive effect sizes indicate that subjects exerted more effort coactively than collectively (i. C Calculated on effect sizes that were..90 -0.41 -0.49 0.64 0.32*** 2.30 0.55 -0.54*** 14 141 8 16.19 0.16 0.48 -0.45 0. KARAU AND KIPLING D.56 0.39 0.52*** 30 128 5 45.37 0.42 -0.22 0.20 -0.35 0.46 0.58 0.62*** 52.44 -0.39 0.46 0.08 0.01.43 0.50 0.32* 0.52 8.16 0.05 0.25 0.32*** 12.66 655.42 0.15 17.32 0.35 794.10 0.33 0.44*** 15 148 21.68 0.45 0.37*** 133 7 23 29.19*** 9 11 127 16 42.47 0.47 -0. engaged in social loafing).17 0.44 0.10 65.61*** 39 23 113 15.18 -0. *** p < .49 168. b Evaluation potential refers to the conditions under which individual outputs can be evaluated.28 0. partitioned by sex of subject.41 0.D3 0.53 0.48 0.14 0.38 0.25 0.48*** 25.22 0.03 0. a Significance indicates rejection of the hypothesis ofhomogeneity.49 0.38 0.96*** 845.46 0.43 0.76*** 0.50 0.59 0.75*** 66.40*** 0. ** p < .54 0.44 0.12 0.08 0.02 0.14 0.59 614.86*** 70.45 0.17 0.30*** 107.22 0.75*** 261.42 -0.34 0.51 0.001.57 0.32 0.14*** 0.25*** 0.14 0.13 0.47 -0.14*** 589.65 0.55 0. * p < .33 0.49 0.58** 24.50*** 855.698 Table 4 STEVEN 1.34*** 96. WILLIAMS Categorical Models for Objective Effort Effect Sizes Between-classes effect 136.32 801.99 5.24 0.85*** 0.19 -0.46*** 64.50 103.05.28 -0.02 0.08 0.e.51 0.48 129.20 -0.50 0.54*** 853. .45 0.45 0.45 -017 0.96*** 95% CI for d.24*** 12 151 20.22*** 5 115 27 16 91.77*** 8 64 91 29.01 0.68*** 12.25*** 5 5 6 141 6 15. whenever possible.53 0. CI = confidence interval.46*** 154.54 1.52 0.36 0.10 0.32 -0.83*** 112 51 21.22*** 34.51 0.14** 0..31 0.86*** 743. Note.11 0.62*** 339.39 0.91*** 719.

00. expectations of co-worker performance. This finding suggests that participants are either unaware of the fact that they engage in social loafing or are unwilling to report that they engage in social loafing. the effect became even smaller and was not significantly different from 0. Positive effect sizes indicate higher effort coactively than collectively.19 . In the multiple regression model.21** -0. this model was moderately successful in accounting for variability in the objective effort effect sizes. and evaluation potential producing especially large effects.02*· 0. ported effort were also included so that supplementary analyses could be conducted. expectations of co-worker performance. subjects' self-reports were in the direction of social loafing. Although none of these models was correctly specified.28 . although the magnitude of this effect was very small and was only significantly different from 0.67** -.12 -. The relatively large number of categorical and continuous variables that produced significant one-way models (and the presence of three or more classes for several of the categorical variables) restricted the viability of the multiple regression models because the number of potential predictors was large in relation to the number of effect sizes." Self-Reported Effort Although our meta-analysis was primarily concerned with objective effort measures.07 Note. or coactive and collective conditions (evaluation potential constant across conditions).59. 1 = field..26 . effect sizes based on subjects' self-re- 9 A number of factors had a significant impact on the magnitude of the self-reported effort effect sizes. publication form. (c) maximizing rather than additive tasks were used. individuals were less likely to report that they had engaged in social loafing when (a)the cover story informed subjects that the study was interested in effort or performance.51** 0. setting of study. simultaneous impact of several of the continuous and categorical variables.46 22. a Number of individual inputs combined into collective outcome. (b) a within-subjects rather than a betweensubjects design was used.SOCIAL LOAFING 699 Table 5 Continuous Models for Objective Effort Effect Sizes Simple linear regressions Predictor or outcome Continuous variables Date of publication Group size" Number of task performers present" Categorical variables Publication form" Culture of subjects" Setting of study" Task valence" Expectations of co-worker performance" Evaluation potential" Additive constant MultipleR SE of estimate QEi b -0. n = 156.54** 0. This model entered a number of the most central predictors-namely. n = 163. **p<. As reflected in the R of. Moreover.12 -0. 1 = Western. f 0 = low. categorical variables were dummy coded. *p<. • 0 = low. date of publication.15 -. b* = standardized regression coefficient. task valence. 1 = unpublished document. -o = Eastern. moderate..07 .59 2. one particularly informative model is presented in Table 5. b = unstandardized regression coefficient. 1 = moderate. 1 = coactive only (evaluation potential confounded across conditions) or unclear. Models are weighted least squares simple linear and multiple regressions calculated with weights equal to the reciprocal of the variance for each effect size.15 . . however. the predictors were entered simultaneously. i Significance indicates model not correctly specified..09 607. no basis for determining.01* b* -. For these analyses. 1 = high.33** -0.07 . Only slightly more than half of the reported differences were in the direction of reduced collective effort. All of the predictors in this model were significant. when nonsignificant reports were considered. 0 = none. b n = 156.OOI. b* -. and evaluation potential. or high. "O = journal article. In general. As shown by the measures of central tendency in Table 6. h Refers to the conditions under which individual inputs can be evaluated. culture of subjects.00 when effect sizes were weighted by the reciprocal of the variance and when outliers were excluded.37** 0.01* 0.01* 0. with task valence. eo = laboratory. and (e)studies were conducted more recently.41 ** Multiple regression b -0. group size. or no basis for determining. Further information on these analyses can be obtained from Steven 1.02 . Karau.OI.12 . (d) group size was smaller.

. the effect is not restricted to laboratory studies of American men. social loafing is not restricted to additive. making tasks unique such that individuals feel more responsibility for their work. and making individuals feel that their contributions to the task are necessary and not irrelevant might all serve. KARAU AND KIPLING D. Homogeneity (Q) of ds making 25 3 0. some of which are directly relevant to areas of controversy among prior researchers. mere acquaintance with one's co-workers was not sufficient to eliminate the effect. although the magnitude of the effect was reduced for field studies.g. the specification of these moderating variables has theoretical as well as practical importance. Our meta-analysis shows that individuals tend to engage in social loafing when working in groups. when they expect their co-workers to perform well. when working on tasks that are perceived as low in meaningfulness or personal involvement. maximizing tasks. Third. social loafing is robust across gender. In addition. a Effect sizes were weighted by the reciprocal of the variance. The findings of this review have a number of implications for future research and practice. However. when working with strangers.. For example.54) reduced collective effort Note. under some conditions. individual inputs cannot be reliably identified or evaluated and people are frequently asked to work on mundane or uninspiring tasks. In particular. In particular. and tasks. * p < . the review highlights a number of conditions under which social loafing should be more or less likely to occur.00-0. at least some important aspects of the work are likely to be repetitive or dull (e.04-0.27 33. Social loafing appears to be moderate in magnitude and generalizable across tasks and subject populations.e.29 0. Even within occupations that appear to have a great deal of intrinsic value. our meta-analysis also uncovered a number of novel findings. may have led inevitably to the erroneous view that social loafing is inconsequential.11 0. when a group-level comparison standard is not available. although the effect is smaller for women and for subjects from Eastern cultures.09-0.07 13/24 (. The tendency of social loafing research (but not theory) to focus on limiting conditions for the effect. Nevertheless. Q = homogeneity of ds. enhancing the cohesiveness of work groups. The CEM suggests that social loafing occurs because individuals expect their effort to be less likely to lead to valued outcomes when working collectively than when working coactively.01. Summary of novel findings. some of our findings confirm long-standing predictions. Finally.00 d unweighted d ds excluding outliers Sample size (n) n removed outliers" Mean weighted d (d+) 95% CI for d. d. . monitoring individual performance or making such performance identifiable. b Significance indicates rejection of the hypothesis of homogeneity. Conclusions Overview of findings and implications. and for subjects in Eastern cultures. Because this meta-analysis integrates a rich research literature in which several hypotheses are well-established. Thus. for women. engaged in social loafing). this tendency to focus on moderating variables does have the positive consequence of suggesting a number of ways in which social loafing might be reduced or overcome in natural settings: Providing individuals with feedback about their own performance or the performance of their work group. culture. but instead is restricted to artificially constrained laboratory studies that use trivial tasks. social loafing appears to be a problem that is likely to manifest itself in a variety of settings. it is important to recognize that. CI = confidence interval. b 28 0. or perceptual).01* 0. assigning meaningful tasks.The tendency to engage in social loafing was moderated by a large number of variables. All reports Sample size (n) Mean unweighted d Differences indicating 42 0. However. in many real-world contexts. the CEM suggests that all of these variables have one thing in common.g. Moreover.writing memos or performing accounting and other maintenance activities). First.22 55. and the CEM was quite successful in accounting for the empirical status of these moderators. that is.16 0. rather than on underlying process.700 STEVEN 1.cognitive. Thus. social loafing was still significant under all of these conditions. individuals are more likely to engage in social loafing when their individual outputs cannot be evaluated collectively. although social loafing was eliminated when participants worked with close friends or teammates. C The proportion appears in parentheses. A superficial evaluation of these findings might imply that social loafing is not a very serious problem in the real world. they influence individuals' perceptions of either the instrumentality of their efforts for obtaining valued outcomes or the degree to which outcomes in a particular setting are likely to be valued. Although each moderating variable may be associated with a distinct set of psychological processes. the effect was robust across both maximizing and optimizing tasks and across tasks demanding different types of effort (e. d = effect size. although social loafing was eliminated when participants' scores could be evaluated collectively and when highly meaningful tasks were used. and when their inputs to the collective outcome are redundant with those of other group members. Thus. Positive effect sizes indicate that subjects worked harder coactively than collectively (i. physical. a closer examination reveals that social loafing could indeed have serious consequences for a variety of everyday collective settings. to reduce or eliminate social loafing. Homogeneity (Q) Mean unweighted 95% CI for mean Median d (d+l' of ds making up d. WILLIAMS Table 6 Summary of Self-Reported Effort Effect Sizes Criterion Value ds with outliers Sample size (n) Mean weighted d 95% CI for d. individuals do not necessarily know their co-workers or do not interact with them frequently enough to develop high levels of cohesiveness. = mean weighted effect size.10 up d.10 -0. Second. In many (but not all) work contexts..

compensatory. and task uniqueness have only been directly examined in a small handful of studies. (b) individuals were more likely to loaf when they expected their co-workers to perform well. including the type of cover story used and the setting of the study (laboratory versus field). the model suggests that the effects of group size on collective effort may depend on factors such as relative attention to strategic versus cooperative outcomes. making previous conclusions about these factors tentative. we found that individuals were generally either unwilling or unable to acknowledge that they loafed on collective tasks. committees. and group valence. when inputs are identifiable. Similarly. despite mixed results from a few prior studies. 2. An especially pressing need is to further substantiate social loafing outside of the laboratory in realworld settings such as industry. individuals in Eastern cultures. and others' evaluations of one's individual performance. Identifiability should enhance effort on superficially aversive tasks that are nevertheless important. Future research could profitably proceed at a number of levels. partnerships. or competition). 1991). and discretionary tasks. these predictions could not be examined in the current data set because of an insufficient number of comparisons (or none at all) in crucial comparison cells. satisfaction of other group members. we found strong. group size might actually decrease social loafing when individuals attach greater value to collective outcomes (as might occur when working in highly cohesive groups. Once the effect has been documented in such settings. and task valence. both old and new. although group-level comparison standards. or particular subclasses) prefer tasks on which they can loaf? Is social loafing adaptive in some cases? Does loafing allow individuals to devote their energies more fully to aspects of tasks that are within their particular domains of mastery and expertise? How might social loafing manifest itself within interacting groups such as decision-making teams or juries.SOCIAL LOAFING 701 only a handful of prior studies have directly examined gender. whereas increased group size might increase the magnitude of social loafing when individuals place greater value on strategic considerations (due perhaps to individual differences. and amount of identification with the group. and individuals high in collectivism and need for belonging should be more likely to attach at least moderate importance to outcomes such as group harmony. task meaningfulness. In addition. men. whereas group size did not influence the magnitude of social loafing for women. our meta-analysis confirms the conclusion of many previous researchers that individuals tend to engage in social loafing when working collectively and reveals that the effect is consistently obtained and moderate in magnitude. the effects of identifiability of inputs. group size interacted with gender such that the magnitude of social loafing increased in larger groups for men. Future directions. both in terms of dispositional and situational factors. and other group members' evaluations of one's individual contributions to the group. Kerr & Bruun. However. on intrinsically meaningful tasks. For example. task meaningfulness. consistent effects for these variables across studies. culture. However. However. they may actually increase such perceptions on disjunctive (or conjunctive) tasks for high ability (or low ability) group members (cf. and how might individual efforts be measured in these contexts? Is it necessarily the case that the quality of the work suffers when people loaf? How aware are people of their social loafing in different settings? Does knowledge that one loafs in collectives affect the likelihood of loafing? The CEM generates a number of additional predictions about how various factors might interact to determine effort. The CEM generates similar predictions regarding possible interactions among additional factors such as expectations of co-worker performance (see Williams & Karau. and presence of feedback may interact with factors such as personal relevance of the task. and (c) social loafing was reduced when individuals worked with acquaintances and was eliminated when individuals worked in highly valued groups. importance of the group to the individual. Our findings also support prior conclusions that individuals are more likely to loaf when the potential for evaluation is absent. Thus. the CEM suggests that making individual inputs identifiable or providing a comparison standard or performance feedback should have a stronger impact when this information has direct implications for the individual's self-evaluation. . For example. group success. In addition. Our findings also highlighted the potential of the CEM for generating predictions about interactions among moderating variables. In particular. and may even undermine effort. group valence. and individual differences in collectivism should influence the degree to which individual outcomes are valued relative to group outcomes. task type. In summary. In addition. With regard to prior hypotheses. availability of comparison standards. Among the questions that might be more directly addressed are the following: What are the long-term effects of working collectively? Do people (in general. Increases in group size are likely to lead to decreased perceptions that one's inputs are related to one's outcomes on most additive. others' evaluations of the group's performance. Unfortunately. increases in group size might enhance rather than inhibit performance under many conditions. our meta-analysis revealed that (a)the relative magnitude of social loafing was reduced for women and for subjects from Eastern cultures. the CEM successfully accounted for the empirical status of the vast majority of moderators of social loafing. Additional research could further address the generalizability of social loafing. juries. our meta-analysis contributed a number of new findings and also showed that several previously established findings are reliable. perceived importance of the task to significant others or to reference groups. Among some of the more intriguing implications of the CEM are the following: 1. competitive intragroup recognition. and interpersonal relationships. culture. attempts could be made to determine whether the moderating variables isolated in laboratory studies apply to other contexts as well. or when a strong social desirability or cooperation norm is salient). but should be less likely to enhance effort. Similarly. and individuals low in collectivism and need for belonging should attach greater importance to outcomes such as pay. individuals in Western cultures. Gender. women. expectation of co-worker performance. Thus. cultural or situational norms. and those studies have produced contradictory results. and identifiability. The meta-analysis also identified some key methodological factors influencing the magnitude of social loafing. 1983).

Adult children of alcoholics: Interpersonal trust and social compensation. (1984).. HiIlsdale.. Social loafing and social facilitation: An empirical test of the cognitive-motivational model of performance.. L. Freeman. Davis. L. 763-769. R. B.. Hillsdale. Bond. CA: Sage. Hastorf & A. culture. (I988). Akron. Newbury Park. McGaw. 47. (1982). 445-453. Goethals (Eds). Glass. Deci. K. 109-139. L. 629637. 4. A. Latane. Eagly. 12. (197 I). San Diego. Cates. Griffith.. S. 34. Latane.. Harkins. 1.. In R. For example. Social loafing as a response to an appraisal of appropriate effort. D. Journal of Personality and Social Psychology.702 STEVEN 1.59-78. Broverrnan. Thorne.. 1. November). New York: Springer-Verlag. G. E. University of Toledo. A. Gender stereotypes stem from the distribution of women and men into social roles. The duality of human existence: An essay on psychology and religion. R. WILLIAMS 3. M.. Broverman. Social comparison theory: Self-evaluation and group life. Cialdini. R. Chicago: Rand McNally. D. & Wang. 94. & Luhtanen.. 109-147). Cooley. Evaluation and performance: A two-edged knife. Effects of personal involvement: Thought-provoking implications for social loafing. Dissertation Abstracts International. M. identifiability. Social loafing and personality: The effects of individual differences on collective performance (Doctoral dissertation. W (1988).. Carli. & Titus. and cooperation in the prisoners dilemma game. Sex-role stereotypes: A current appraisal. References Abrams. I.. Intrinsic motivation. R. Group participation in the organization: Social loafing as a limitation of group effectiveness. Crocker.. Hackman. Behavioral and Brain Sciences. T. C. NJ: Erlbaum. Journal of Personality and Social Psychology. R. reward allocation. & Moreland. when individual efforts are dispensable. (I976). (1969). G. Greenwald. & Ostrom. but should hinder collective effort when it relays negative task attitudes or contributes to feelings of dispensability. A. S. Gabrenya. Group performance.. Higgins (Eds). S. (1986). Bartis. Taiwan. WK. Y. Walker. K.. Communication among group members should enhance collective effort when it enhances perceptions of task importance or social responsibility. Wyer & T. The social self: On being the same and different at the same time. Y. Cohen.. For example. Eagly. pp. . Integrating research: A guide for literature reviews (2nd ed. L. 51. D. G.). Personality and Social Psychology Bulletin. Beverly Hills. E E.. WK. Cohen. (1972). C. Paper presented at the annual meeting of the Speech Communication Association. students. Employee reactions to job characteristics. M. Group structural factors and member roles may profoundly influence collective effort in ongoing groups by affecting perceptions of the instrumentality of one's inputs and the value of various outcomes. I 326B.. (1989). K.253-271. 1. (1991). Cooper. Bakan. Journal of Cross-Cultural Psychology. E. 265-292. T. & van de Kragt. & Greenwald. Collins. Basking in reflected glory: Three (football) field studies..ed). L. Unpublished manuscript. Journal of Social Issues. (1987). August). & Hogg. H. (1988). Clarkson. III. OH: University of Akron.. Caporael. Social loafing in cross-cultural perspective: Chinese in Taiwan. 124. Isen (Eds. K. Cognitive social psychology (pp. 368-384. V. L. 1. (1977). R. E. Fordham University). C. Journal of Personality and Social Psychology. (I989). The research reviewed in this article supplies a consistent and robust foundation from which to continue this investigation. H. In R. 565-581. S. & Badura. Journal of Applied Psychology. 1. 55. E. (University Microfilms No. MA: Addison-Wesley. (1986). K. Gabrenya. P. (199 I).. R. P. (1981. M. (1987. L. 60-67. Anderson. 14. 5. Reading. CoIlective self-esteem and ingroup bias. New York: Plenum Press. R. M. B. (1989). Are women more social and men more task oriented? A meta-analytic review of sex differences in group interaction. Geen. New York: Springer-Verlag. Brickner.. Handbook of motivation and cognition (Vol. M. & Blanchard.. G. 1. & Wang. S. Basic and Applied Social Psychology. B.. and individual differences in collectivism and need for belonging should have less impact. A. Srull (Eds). & Steffen. A. Meta-analysis in social research. 366-375.. M. & Lawler. A. M. H. R. In B. New York: Guilford Press. R. Social motivation. Theories of group behavior (pp.). Psychological Bulletin. (I991). The self. Selfishness examined: Cooperation in the absence of egoistic incentives. Statistical power analysis for the behavioral sciences (rev. R. Social loafing among Chinese overseas and Us. & Darley.. S. Brewer.. Breckler. L. Vogel. Motivational facets of the self. 10. B. R. Taipei. Personality and Social Psychology Bulletin. 1. & Schmidt. H. 129-178). (1990). Social facilitation: A meta-analysis of 241 studies. 34. 86-11258). Early. M. 0.. P. Greenwald. R. 0. D. M. especially in cohesive or highly valued groups. S. NJ: Erlbaum. Sorrentino & E. N. (1983). 1. Hallmark. Szymanski. A. (1986). Paper presented at the Asian Conference of the International Association for CrossCultural Psychology. L. & Downs. (1984). Unpublished master's thesis.. Jr. Sex differences in social behavior: A social-role interpretation. Jr. (1982). 3. A. Dissertation Abstracts International. E. Boston. 17. (1975). & Pratkanis. B. Fichrnan. 475-482. L. & Rosenkrantz. L. A. 1. L. Mullen & 0. Orbell. 58. M. 242-251. Dawes. 683-699. Journal of Personality and Social Psychology.... (1990).. 1. Administrative Science Quarterly. CA: Sage.377-399. New York: Elsevier North-Holland. 145-164). & Harkins. Annual Review of Psychology: 42. E. 55. 0. R. Harcum. Self-esteem and direct versus indirect formsofself-enhancement.. R. G. E. Borden. Joumal ofPersonality and Social Psychology. & Wingard. leaders and high-status group members may view their inputs as more instrumental to group outcomes. M. 46. Social identity and self-motivation: Improvements in productivity Unpublished manuscript. S. These are just a few of the questions that future research can attempt to answer. 1. Ego task analysis: An integration of research on ego-involvement and self-awareness. CA: Academic Press. and task uniqueness on task performance. (1966). coalition formation. 1. KARAU AND KIPLING D. Brickner. Goethals. (198I). 28(2). Handbook of social cognition (pp. 259-285. v. T. Sex differences in task and social-emotional behavior. 735-754. (1989). & Smith. & Sloan. Social identity theory: Constructive and critical advances. high levels of effort are unlikely when individual behaviors are not instrumental in obtaining those outcomes.. 21-47).. 49. 1. Brown. 0. In A. 14. H. M. (1983).. (1988). 1.. and norms encouraging social responsibility and hard work within groups should have a positive effect on collective effort. (1987). Basic and Applied Social Psychology. Social loafing and coIlectivism: A comparison of the United States and the People's Republic of China. H. 1. gender. A.. R. Even if outcomes are highly valued. T. (1982). I430B. (1990). L. Journal of Psychology. The effects of evaluation.

& Moreland. 23.. (1985). R.. R. & Onglatco. 49. (I 981). 45. Mullen. & Bruun. (1974). 224-231. Paulus (Ed. social justice. M. & Olkin. & Peckham. 0. & Latane. Latane. N. Jackson. Williams. R. Williams. G.. Paulus. Hillsdale. (1982a. S. (1986). American Psychologist. L. J. Naylor. 1. Journal oj Personality and Social Psychology. L. 819-828. 24. 56. J. 0.. Journal oj Personality and Social Psychology Bulletin. & Latane. & Harkins. Journal of Experimental Social Psychology. empirical. & Latane. (1987). D. The role oj insight in the reduction oj social loafing. S. 1214-1229. 48. K. 10.). B. G. (1987). (1983). American Psychologist. Equity in effort: An explanation of the social loafing effect. Hillsdale. 1199-1206.. May). G. T. Harkins. L. S. The effects of . K. Latane. Journal oj Personality and Social Psychology. Review of personality and social psychology: Group processes (Vol. Seta. 50. 7. L. K. DSTAT Software for the meta-analytic review oj research literatures. B. Zaccaro. & Williams. D. Lepper. Hedges. J. Hsu. Journal of Experimental Social Psychology.. (1980). G. Latane. B. Pritchard. 354-365. K. The phenomenology of being in a group: Integrationsof social cognition and group processes. B. Martin.. Journal of Sport and Exercise Psychology.). 0. Many hands make light the work: The causes and consequences of social loafing. Newbury Park. B. C. 1. Minneapolis. New York: Springer-Verlag. 50. Assimilation and contrast as a function of people's willingness and ability to expend effort in forming an impression. (1987). V. 1. May). & Bruun. individual differences in social loafing on cognitive tasks.. A. Levinger. & Daly. Graves. (1980). B. Journal of Personality and Social Psychology. pp. Cognitive Therapy and Research. (1993). B. S. U. (1985). Social loafing and self-evaluation with an objective standard. T. S. How hard is hard science. M. May). & Williams. (1980). Journal oj Experimental Social Psychology. i6. 371-384. S. 165-172. Social Science. Kravitz. & Greene. McGovern.). 0. R. S. Cambridge. Jackson. A theory cfbehavior in organizations. 297-323. B. Harkins. Joumal of Personality and Social Psychology. Ringelmann revisited: Alternative explanations for the social loafing effect. M. 36. M. & Ruder. Social loafing and social facilitation: New wine in old bottles. & Szymanski. Basic group processes (pp. E. Strength and immediacy of sources: A meta-analytic evaluation of the forgotten elements of social impact theory. N. 443-455. (1978. 1. Petty. S.. (1981). M. (1987). B. Paper presented at the annual meeting of the Midwestern Psychological Association. Doody. G. S. Jackson. K. 934-941. Kerr. (1977). L. MA: Harvard University Press. In P. C J. 27. D. L. Minneapolis. 1458-1466. M. CA: Academic Press. (1991). I I. CA: Academic Press. how soft is soft science? The empirical cumulativeness of research.. K. E. K. Jackson. Learned helplessness and the generality of social loafing. 936-941. G. international Journal of Sport Psychology. 511-513. Garden City. 0. Harkins. Journal of Experimental Social Psychology. 0. 1-18. Journal of Personality and Social Psychology. & Latane. S. M. The dispensability of member effort and group motivation losses: Free rider effects. C. J.SOCIAL LOAFING Hardy.. Attributions of responsibility for collective endeavors. & Forsyth.. S. & Kasmer. L. Paper presented at the annual meeting of the Midwestern Psychological Association. Petty. J. & Latane. M. Chicago. 10. 78-94. 1. v.407-415. (1989). Leary. Journal oj Personality and Social Psychology. Motivation and Emotion. Huddleston. (1983). CA: Sage. San Diego. G. 10. & Szymanski. 42.. Heckhausen. NJ: Erlbaum. Journal of Personality and Social Psychology. Motivation losses in small groups: A social dilemma analysis. Cacioppo. May). Jackson. (1985). The Ringelmann effect: Studies of group size and group performance. Karau. (1985). (1978). Personality and Social Psychology. & Williams. Harkins. & Latane. 343-356. R. (1990). Jackson. D. Effects of task difficulty and task uniqueness on social loafing. & lIgen. M. 37. (1970).. (1985. (1982b. Social loafing on a cheering task. Operationalizing the effect of the group on the individual: A self-attention perspective. Kakuyama. L. 1. & Williams. R. M.. H. G. Ingham.. Harkins. (1982). NJ: Erlbaum. M. Statistical methods for meta-analysis. & Martin. M. 43. 9. The psychology of social impact. V. 81-92. Florida Institute of Technology. S. K. Achievement motivation and its constructs: A cognitive model.. J.. C Masters & W P. Petty. The role of evaluation in eliminating social loafing. (1977). R. Harkins. S. (1965). 38. Obedience: Strength and immediacy oj authority and social loafing Paper presented at the annual meeting of the Midwestern Psychological Association. Kerr. Social loafing and group evaluation. M. & Crelia. Journal of Personality and Social Psychology.. K. The logic oj collective action: Public goods and the theory of groups. Harkins. CA: Sage. (1987). G. G. (1989). S. 167-188). (1987). 27-37. NJ: Erlbaum. S. (1986).. B. Peterson. Review of personality and social psychology (Vol. Smith (Eds. Journal oj Personality and Social Psychology. & Szymanski. 8.). 72. J. (Eds). R. The hidden costs oj reward. (1988). Johnson. E L. Chicago.. Social loafing on difficult tasks: Working collectively can improve performance. Harkins. & Harkins. C. Hedges. A.. Hillsdale. A. Hardy. 1. Williams. (1986). Journal oj Personality and Social Psychology. The effects of diffusion of cognitive effort on attitudes: An information processing view. Newbury Park. N.. E.. 49. and relative deprivation: Theoretical. & Jackson.. Paper presented at the annual meeting of the Midwestern Psychological Association. t».. (1983). 71. G. J. 575584. Levine. Manuscript submitted for publication. Mullen.. K. 44. Social loafing and social facilitation. Group influences on individual task performance. B. The effect of prior knowledge of the social loafing phenomenon on performance in a group. 167-188). 97-120).. P. Social loafing in cheerleaders: Effects of team membership and competition. Americans and Chinese: Purposes and fulfillment in great civilization (2nd ed. Olson. S. K. B. S. In C Hendrick (Ed).. (1983).. R. In C Hendrick (Ed. 822-832. T. 295-322. Journal of Personality and Social Psychology. R. B. 59. Mullen. Social comparison and outcome evaluation in group contexts. 563-569. Ringelmann rediscovered: The original article. (1986). 937-942. Journal oj Personality and Social Psychology.. (I 985). K. and policy perspectives. Matsui. A. Experimenter immediacy and social loafing. In J. & Petty. In defense of social impact theory: Comment on Mullen. J. (1988). NY: Natural History Press. 703 Kerr. M. The effects oj group cohesiveness on social loafing and social compensation. 1:. i6. 176-182. Social comparison. B. J. pp. I. G. D. 109-114. San Diego. B. Unpublished master's thesis.. Effects of goals and feedback on performance in groups. Harkins. B. 457-465. C (1986). Journal of Applied Psychology. (1985). Journal oj Personality and Social Psychology. (1979). Social loafing: Allocating effort or taking it easy? Journal oj Experimental Social Psychology.. The effects oj sexual composition and group size on individual effort... 1. 283-329.

& Latane. Individual contributions to a collective effort: An incentive analysis. E. Experimental studies of the sources of gregariousness. & Williams. 348-361. Weldon. H. (1964). Identifiability as a deterrent to social loafing: Two cheering experiments. 67-81. S. & Miller. 1. & Latane. Quintanar. (1991). Effects of coactor's presence: Social loafing and social facilitation. Berkowitz (Ed. 303-311. San Diego. Waller. J. Paper presented at the annual meeting of the Midwestern Psychological Association. E. V. Jackson. D. K. Basic and Applied Social Psychology.' I. Washington.. & Mustari. Meta-analytic procedures for social research (2nd ed.)... Wrightsman. B. J. and social loafing. B.. Columbus. DC: Hemisphere. 79-86.. E. & Burmont. B. (I988). A. R. Reis.. 53. (1986). 10. NC. 33-48). & Brown. K. H.. (1982). The effects of coactor ability on social loafing Paper presented at the annual Graduate Student Conference in Personality and Social Psychology. 61... K. S.570-581. K. Williams. Zaccaro. J. K. & Pryor. Steiner. Effects of affiliation-related motives on swimmers in individual versus group competition: A field experiment. A look at social loafing in a cognitive task: Individual effort in advertising evaluations. 49. Triandis. 506-520. (1984. Williams. 10. Organizational Behavior and Human Decision Processes. (I984). Psychological Review.. L. Organizational Behavior and Human Decision Processes. R. Taipei. British Journal of Social Psychology. behaviors. 21. A. J. (1960). & Harkins. Stevenson. L. Personality and Social Psychology Bulletin. Paper presented at the First Annual Nags Head Conference. T. & Rubin. G. Cognitive loafing: The effects of accountability and shared responsibility on cognitive effort. Paper presented at the annual meeting of the Midwestern Psychological Association. Rosenthal. K. E. Organizational Behavior and Human Decision Processes. (1989). task responsibility. 36.. Kill Devil Hills. KARAU AND KIPLING D. Journal of Personality and Social Psychology. Williams. Porter. Illusion and well-being: Some social psychological contributions to a theory of mental health. 4821B. Social loafing of Japanese students. August). (1982. Psychology of intergroup relations (pp. CA: Academic Press. (1989). The development of social loafing in America. Journal of Personality and Social Psychology.. A. May). D. 50. M. In L.. H. 1. W. K. Toward a self-evaluation maintenance model of social behavior. New York: Wiley. S. Social loafing: The role of task attractiveness. Rosenthal. L. E. (1988). Stanford. Managerial attitudes and performance. L. C. (1993). Organizational Behavior and Human Decision Processes. Y. (1989). Self-interest and collective action: The economics and psychology ofpublic goods. 0. Tesser. Baca. D. J. Journal of Personality and Social Psychology. Dissertation Abstracts International. & Lawler.. & Wright. Psychological Bulletin. Advances in experimental social psychology (Vol.181-227). 14. Psychological Bulletin. Nida. and perceptions. S. 21.. (1988). (1990).. 330-351. (1959). 96. H. Chicago. H. J.). Shirakashi. Extending the social information processing perspective: New links to attitudes. 121-137. (1988). 36. H.. CA: Stanford University Press. G. S. 891-897. & Gargano. Wheeler. (1987). Williams. 27. (1978). 216-222.. R. & Turner. 35-40. & Frey. Szymanski. San Diego. Kawana. Hiroshima Forum for Psychology. Homewood. K. 15. 90-06796) Stroebe. May). Harkins. Yamaguchi. E. Journal of Abnormal and Social Psychology. R. S. B. Social loafing and self-evaluation with a social standard. CA: Sage.OH. K. (1972). Detroit. & Latane. Sorrentino. M. Decision responsibility. Felt dispensability in groups of coactors: The effects of shared responsibility and explicit anonymity on cognitive effort. Motivating effort under social loafing conditions: The effects of private self-awareness on task effort.. Personality and Social Psychology Bulletin. 3. 57. & Sheppard. D. A. (1987). The self and social behavior in differing cultural contexts. The effects of group cohesiveness on social loafing.. 159-171. 205-246. K. 99. Paper presented at the annual meeting of the American Psychological Association. 61. & Williams. . M. (1989). D. Okamoto. (1985).. The psychology of affiliation. K. Tajfel. R. 0. 193-210. S. Meta-analytic procedures for combining studies with multiple effect sizes. B. In S. & Latane. 704-714. K. 41. B. Washington. 215-222. 5574B. Williams. (1985). Personality and Social Psychology Bulletin. 99-106. 103. Social loafing and swimming: Effects of identifiability on individual and relay performance of intercollegiate swimmers. K. J. E. Personality and Social Psychology Bulletin. Weldon. Taiwan. E. L.. B. W. (198I. Collectivism-individualism in everyday social life: The middle kingdom and the melting pot. S. 0. The social identity theory of intergroup behavior. & Oka. 40. Newbury Park. Social comparison processes: Theoretical and empirical perspectives. 10. 113. 0. 47. Group process and productivity. CA: Academic Press. Schachter. (1968). pp. G. & Gargano.704 STEVEN 1. S. S. Effects of waiting with others on changes in level of felt anxiety. (1981). Shepperd. 579-582.. 0.). IL: Dorsey Press. M. Worchel & W Austin (Eds). Williams. (1991). B.. L. (1985). June). 73-81. C. M. Vroom. R. (I981. Work and motivation. (1980). Cognitive effort in additive task groups: The effects of shared responsibility on the quality of multiattribute judgments. Journal of Personality and Social Psychology. J. & Ford. (Eds.. (I989). (1986). L. Weldon. 40. H. Williams. Chicago: Nelson-Hall. (1981. Japanese Psychological Research. A... identifiability. S. C. Collective effort in small groups in the United States and Japan. Shepperdv J. August). Group diffusion of cognitive effort as a determinant of attribution. DC. Social loafing and social compensation: The effects of expectations of co-worker performance.. M. B. Price. K. J. S. (1990). (1977). Williams. Zalesny. Suls. 330-345. 141-149. Paper presented at the Asian Conference of the International Association for Cross-Cultural Psychology. 400-406. & Karau. Thompson. Productivity loss in performance groups: A motivation analysis. Taylor.. University of Kentucky). Dissertation Abstracts International. WILLIAMS group size on cognitive effort and evaluation. (University Microfilms No. Psychological Bulletin. & Bond. T.. B. Journal of Personality and Social Psychology. Developmental aspects of self-evaluation and social loafing in children (Doctoral dissertation.

& Latane. Group participation in the organization: Social loafing as a limitation of group effectiveness. S. Journal of Personality and Social Psychology. Petty. T. S. & Harkins. (1986). (1991). (1985.. B. T. Effects of personal involvement: Thought-provoking implications for social loafing. (1986). Journal of Personality and Social Psychology. Paper presented at the annual meeting of the Speech Communication Association. S. K. E. G. & Petty. Cooley. P. (1981. University of Toledo. A. R. & Szymanski. (1980). 1. 763-769. Harkins. 5]. S. N. (1978. D. K. The Ringelmann effect: Studies of group size and group performance.. 1.. B. D. T. Journal of Sport and Exercise Psychology. (1985).. & Latane. A.. Journal of Personality and Social Psychology. Social loafing in cross-cultural perspective: Chinese in Taiwan. L. Harkins. Earley. (1989). The role of insight in the reduction of social loafing. 1-18.. The effects of diffusion of cognitive effort on attitudes: An information processing view. & Harkins. Jr.. 38. Obedience: Strength and immediacy of authority and social loafing. S. (1980). M. & Badura. Social loafing among Chinese overseas and Us. T.. JournalofCross-Cultural Psychology. (1982). Kerr. & Ostrom. (1974). Minneapolis. M. Jackson. Williams. B. 1. Y E. (1985).. & Latane. (1987).. L... S.. S. Cognitive Therapy and Research. G. Latane. L. & Latane.. Harkins. Seta. Gabrenya. L. C (1986). 1214-1229. E.. Motivation losses in small groups: A social dilemma analysis. Y E. S. (1985). 563-569. 23. Jr.. K. G. (1988). Jackson. Social loafing on a cheering task. Ingham. Administrative Science Quarterly. 1. The effects of sexual composition and group size on individual effort. A. Social Science. Kakuyama. L.. 1. identifiability. 407-415. G. Chicago. Unpublished manuscript. Unpublished master's thesis. Unpublished master's thesis. L. Paper presented at the annual meeting of the Midwestern Psychological Association. 43. 44. (1990).. R. U. May). K. (1993).SOCIAL LOAFING 705 Appendix Studies Used in the Meta-Analysis Bartis. D.. Cates. Harkins. Social loafing as a response to an appraisal of appropriate effort. 86-11258) Cohen. H. Szymanski. (1983). May). & Wang. (l982b. Kerr. 629637. Harkins. 934-941. Social loafing and collectivism: A comparison of the United States and the People's Republic of China. Fordham University).. C 1.. Many hands make light the work: The causes and consequences of social loafing. 1. Fichman. Social loafing and self-evalua- . (J 988). Ringelmann revisited: Alternative explanations for the social loafing effect. The role of evaluation in eliminating social loafing. Latane. 72. Zaccaro. Harkins.. & Williams. Latane. Latane. ]4. Kerr. Assimilation and contrast as a function of people's willingness and ability to expend effort in forming an impression.. 16.. T. & Onglatco. Taipei. & Peckham. Social loafing on difficult tasks: Working collectively can improve performance. Journal of Experimental Social Psychology. & Wingard. M. Williams.. Social loafing in cheerleaders: Effects of team membership and competition. Florida Institute of Technology. Petty. G. G. 1. Journal of Personality and Social Psychology. 24. Taiwan. S. Jackson. (1983). G. The effects of group cohesiveness on social loafing and social compensation. 109-114. L. (University Microfilms No. Experimenter immediacy and social loafing. 1. 81-92. Akron. Paper presented at the annual meeting of the Midwestern Psychological Association. 49. & Harkins. Hardy. E.. M.. 575-584. Brickner. students. Paper presented at the annual meeting of the Midwestern Psychological Association. N. Harkins. B. Jackson. D. M. Journal of Applied Psychology. Social loafing and group evaluation. B. Social loafing and social facilitation: An empirical test of the cognitive-motivational model of performance. M. The dispensability of member effort and group motivation losses: Free rider effects.. Hallmark. Petty. B. 59. May). 242-251. & Jackson. & Bruun. (1986).. 1. 49. & Latane. 253-271. 45.. L.. S. Individual differences in social loafing on cognitive tasks. Social loafing on an optimizing task: Cross-cultural differences among Chinese and Americans. 1326B. (1982a. Social loafing and social facilitation. Effects of goals and feedback on performance in groups. (1988). & Williams. (1989). C (1989). Personality and Social Psychology Bulletin. 1199-1206. 10. (1987). Personality and Social Psychology Bulletin. K... B. M. G. 10. (1990). K. 1. I 430B. 368-384. E. & Downs. Effects of task difficulty and task uniqueness on social loafing. K. K.. D. Journal of Personality and Social Psychology. K. K. (1979). Journal of Cross-cultural Psychology. Karau. V.. 565-581. OH: University of Akron. B. Harkins. 819-828. M. S. 224-231. 822-832... R. R. Chicago.. R. S. 1. & Wang. tion with an objective standard. B. (1983). W. B. Equity in effort: An explanation of the social loafing effect. Social identity and self-motivation: Improvements in productivity. G. K. Dissertation Abstracts International.457-465. 1. and task uniqueness on task performance. M. Journal of Personality and Social Psychology. S. C. Wang. Evaluation and performance: A two-edged knife. Adult children of alcoholics: Interpersonal trust and social compensation. (1986). 14. Manuscript submitted for publication. Journal of Personality and Social Psychology. (1977). 1. 7. K. 165-172. 49. K.. & Williams. & Szymanski. W. Matsui. November). ]6. Harcum. Journal of Personality and Social Psychology. Social loafing and personality: The effects of individual differences on collective performance (Doctoral dissertation. /0. (1988). & Daly. 1. C 1. Cacioppo. Graves. R. 937-942. 56. 37. K.. & Latane. R. Personality and Social Psychology Bulletin. W. Paper presented at the Asian Conference of the International Association for CrossCultural Psychology. The effects of evaluation. A. A. Jackson. Levinger... 1. Griffith. G (1988). Journal of Personality and Social Psychology. 47. G (1985). Gabrenya. 223-242. K. Harkins. D. N. (1987. 34.. 1]. 78-94. August). B. L. Minneapolis. R. Hardy. 71. Social loafing: Allocating effort or taking it easy? Journal of Experimental Social Psychology. & Latane. Williams.. Journal of Personality and Social Psychology. & Moreland. 1: M. & Bruun. Martin. Paper presented at the annual meeting of the Midwestern Psychological Association. R. 27-37. M. Peterson. Brickner. M. Basic and Applied Social Psychology. Williams. 124. M.. S. S.. S. Learned helplessness and the generality of social loafing. C. T E. Journal of Experimental Social Psychology. Dissertation Abstracts International. 1. (1981). Boston. & Kasmer. Gabrenya. & Crelia. McGovern. The effects of ]0. S. Journal of Personality and Social Psychology. 354-365. G. May). 371-384. Journal of Psychology.

Chicago.. (University Microfilms No. Harkins. M. K. H. Taiwan. Jackson... K. 90-06796) Szymanski. 14. Collective effort in small groups in the United States and Japan. G. S. & Latane. Paper presented at the Asian Conference of the International Association for Cross-Cultural Psychology. The development of social loafing in America. Motivating effort under social loafing conditions: The effects of private self-awareness on task effort. Journal of Personality and Social Psychology. May). S. & Sheppard. Organizational Behavior and Human Decision Processes. Quintanar. S. 1. M. K. 49. 215-222. H. K. Dissertation Abstracts International. 141-149. Williams. R.. (1989). Waller. L. KARAU AND KIPLING 0. D. Hiroshima Forum for Psychology. M. & Mustari. K. 36. K. University of Kentucky). Y. S. D. and social loafing. (1985). S. Social loafing and social compensation: The effects of expectations of co-worker performance. Williams. 3. B. R. The effects of coactor ability on social loafing. Felt dispensability in groups of coactors: The effects of shared responsibility and explicit anonymity on cognitive effort. B.. The effects of group cohesiveness on social loafing. & Gargano. Weldon. Weldon. (1981. 1.. August). Received February 18. Personality and Social Psychology Bulletin. Decision responsibility.706 STEVEN 1. & Latane. C. (1980).. (1982.. A. (I 984. (1981. Developmental aspects of self-evaluation and social loafing in children (Doctoral dissertation. 303-31 I. A.. Shirakashi. K. Group diffusion of cognitive effort as a determinant of attribution. 330-351. E. Williams. K. (1988). Effects of coactor's presence: Social loafing and social facilitation. August). B. (1990).. R. K. Basic and Applied Social Psychology. OH. 41. Kawana. 570-581. & Williams. Williams. NC. G. 5574B. Journal of Personality and Social Psychology. Williams. Kill Devil Hills. Thompson. Organizational Behavior and Human Decision Processes. (1981). & Karau.73-81. WILLIAMS group size on cognitive effort and evaluation. 891-897. Social loafing and self-evaluation with a social standard. June).. E. 159. 27.. Sorrentino.. Paper presented at the annual meeting of the American Psychological Association. Effects of affiliation-related motives on swimmers in individual versus group competition: A field experiment. 36. Nida. Stevenson. (1985). 50. It). (1991). Weldon. 1. DC. Williams. Price. Shepperd. & Latane. & Latane.. 579-582. B. 1. L. B. E. S. Journal of Personality and Social Psychology.. Paper presented at the annual meeting of the Midwestern Psychological Association. Individual contributions to a collective effort: An incentive analysis. Personality and Social Psychology Bulletin. K.. Paper presented at the annual meeting of the Midwestern Psychological Association. K. (1981. Personality and Social Psychology Bulletin. Detroit. 4821B. (1987). 15. D. Williams.. & Harkins. (1978). Yamaguchi. Identifiability as a deterrent to social loafing: Two cheering experiments. task responsibility. Japanese Psychological Research. E. A. & Oka. D. identifiability. D. Baca. 1993 Accepted May II. 1. (1987). Social loafing and swimming: Effects of identifiability on individual and relay performance of intercollegiate swimmers. Williams. Paper presented at the annual Graduate Student Conference in Personality and Social Psychology. 330-345. K. Cognitive loafing: The effects of accountability and shared responsibility on cognitive effort. 40. (1988). 0. 1992 Revision received May 3. L. & Pryor. T.. 704-714. May). B. /0. 35-40. & Gargano. Columbus. Dissertation Abstracts International. Washington. & Williams. Journal of Personality and Social Psychology. C. (1985). (1989). E... 348-361. Okamoto. 1993 • . Organizational Behavior and Human Decision Processes.1 71. (1989). D.. Paper presented at the 1stAnnual Nags Head Conference. K. S.. & Wright. A look at social loafing in a cognitive task: Individual effort in advertising evaluations. & Burmont. 40. G. Taipei. B. Cognitive effort in additive task groups: The effects of shared responsibility on the quality of multiattribute judgments. 1. Social loafing of Japanese students. B. 61. 53.