Stephen G. Harkins

Identifiability as a deterrant to social loafing: Two cheering experiments.

Two experiments tested the extent to which the identifiability of ones individual output moderates social loafing—the reduction of individual efforts due to the social presence of others. In the first stage of Experiment 1, participants were asked to produce noise either alone, in groups of two and six, or in pseudogroups where the individuals actually shouted alone but believed that one or five other people were shouting with them. As in previous research, people exerted less effort when they thought that they were shouting in groups than when they shouted alone. In the second stage, the same people were led to believe that their outputs would be identifiable even when they cheered in groups. This manipulation eliminated social loafing. Experiment 2 demonstrated that when individual outputs are always identifiable (even in groups), people consistently exert high levels of effort, and if their outputs are never identifiable (even when alone), they consistently exert low levels of effort across all group sizes. In concert, these studies suggest that identifiability is an important mediator of social loafing.

Social loafing and social facilitation

Abstract Social facilitation and social loafing have been treated as separate lines of research in the social psychological literature. However, it is argued in the present paper that these two paradigms are closely related; in fact, they are complementary. Viewed from this perspective, the experimental conditions that have been included in loafing and facilitation research fall into three cells of a 2 (Alone vs. Coaction) × 2 (Evaluation vs. No Evaluation) factorial design. In the current research, the complete 2 × 2 design was run in two experiments. In both experiments, consistent with the findings of previous loafing research, with number held constant, participants whose outputs could be evaluated outperformed participants whose outputs could not be, but, inconsistent with descriptions of the loafing effect (e.g., B. Latane, K. Williams, & S. Harkins, 1979 , Journal of Personality and Social Psychology , 37 , 823–832), with evaluation potential held constant, pairs outperformed singles. These data suggest that both social facilitation and social loafing can be accomodated in the same design. It is argued that combining the paradigms in this way refines our understanding of both phenomena.

Social loafing: Allocating effort or taking it easy?

When asked to work both alone and in groups, people exert less effort in groups, a phenomenon we call “social loafing.” Either of two possible strategies could explain this outcome: an allocational strategy where people work as hard as they can overall but conserve their strength for individual trials where work is personally beneficial and a minimizing strategy where the primary motive is to “get by” with the least effort possible. However, an allocational strategy would lead participants who always work in groups to put out as much effort as participants who always work alone, since there is no need to husband strength. Two studies using a sound production task found social loafing even under these conditions, suggesting that allocational strategies are not prevalent. Social loafing seems to occur when people perform together in groups, regardless of whether or not they must also perform alone.

Mere effort and stereotype threat performance effects.

Although the fact that stereotype threat impacts performance is well established, the underlying process(es) is(are) not clear. Recently, T. Schmader and M. Johns (2003) argued for a working memory interference account, which proposes that performance suffers because cognitive resources are expended on processing information associated with negative stereotypes. The antisaccade task provides a vehicle to test this account because optimal performance requires working memory resources to inhibit the tendency to look at an irrelevant, peripheral cue (the prepotent response) and to generate volitional saccades to the target. If stereotype threat occupies working memory resources, then the ability to inhibit the prepotent response and to launch volitional saccades will be impaired, and performance will suffer. In contrast, S. Harkinss (2006) mere effort account argues that stereotype threat participants are motivated to perform well, which potentiates the prepotent response, but also leads to efforts to counter this tendency if participants recognize that the response is incorrect, know the correct response, and have the opportunity to make it. Results from 4 experiments support the mere effort but not the working memory interference account.

Mere effort as the mediator of the evaluation-performance relationship.

The research traditions that have examined the evaluation-performance relationship do not agree on the mediating process(es), nor is there any compelling evidence that favors one account over the others. In the current research, a molecular analysis of performance on the Remote Associates Test was undertaken in an effort to identify the mediating process(es). This analysis suggests that the potential for evaluation leads participants to put greater effort into the prepotent response and that this mere effort alone can account for the typical finding that evaluation improves performance on simple items and debilitates performance on complex ones. Subsequent research will be aimed at testing the generalizability of this account.

Race of source effects in the elaboration likelihood model.

In a series of experiments, we investigated the effect of race of source on persuasive communications in the Elaboration Likelihood Model (R.E. Petty & J.T. Cacioppo, 1981, 1986). In Experiment 1, we found no evidence that White participants responded to a Black source as a simple negative cue. Experiment 2 suggested the possibility that exposure to a Black source led to low-involvement message processing. In Experiments 3 and 4, a distraction paradigm was used to test this possibility, and it was found that participants under low involvement were highly motivated to process a message presented by a Black source. In Experiment 5, we found that attitudes toward the sources ethnic group, rather than violations of expectancies, accounted for this processing effect. Taken together, the results of these experiments are consistent with S.L. Gaertner and J.F. Dovidios (1986) theory of aversive racism, which suggests that Whites, because of a combination of egalitarian values and underlying negative racial attitudes, are very concerned about not appearing unfavorable toward Blacks, leading them to be highly motivated to process messages presented by a source from this group.

The Multiple Source Effect in Persuasion The Effects of Distraction

Holding constant information about the number of sources and number of arguments to which they might be exposed, subjects viewed one source presenting three different arguments, three sources presenting different versions of a single argument, or three sources presenting three different arguments (one each) in favor of a counterattitudinal position. For hayf of the subjects, this message was accompanied by a distraction task. In the single task conditions, replicating Harkins and Petty (1981), three-source/three-arguments subjects were more persuaded than subjects in the other two conditions, but when the message was accompanied by a distractor, this persuasive advantage disappeared. Since distraction also led to disruption of favorable thought production, but left recall unaffected, these data are consistent with the view that the enhanced persuasion found in the multiple-source/multiple-argument condition is the result of additional message elaboration elicited by the combination of different sources and different arguments.

Social loafing and self-evaluation with an objective standard

Abstract It has been found that social loafing, the finding that participants working together put out less effort than participants working alone, can be eliminated by the prospect of evaluation of individual outputs. In this work, the role of the experimenter as evaluator has been emphasized. However, when the experimenter cannot evaluate individual outputs, the participants are also unable to do so. In the present experiment, it was demonstrated that the potential for self-evaluation was sufficient to eliminate the loafing effect. In this study, the participants could evaluate themselves by comparing their performances to an objective standard, the number of signals presented in a vigilance task. Previous research has demonstrated similar effects using a social standard. Taken together, these experiments suggest that the possibility of evaluation by an external source is not required to eliminate the loafing effect; the potential for self-evaluation alone is sufficient to do so. These findings are discussed in relation to current theories of self-evaluation.

Need Threat Can Motivate Performance After Ostracism

Ostracism threatens fundamental needs of belonging, self-esteem, control, and meaningful existence, which should motivate participants to respond to this threat. However, research has yet to examine the role of need threat in producing motivation after ostracism. In the current work, participants completed a “cognitive ability” (antisaccade) task following Cyberball-induced ostracism or inclusion. In two experiments, it was found that when ostracized, participants do not see antisaccade performance as a means of responding to the concerns produced by need threat; they respond only to the social threat, leading to worse performance than included participants (Experiments 1 and 2). However, when participants see an avenue of response (the Cyberball players can compare antisaccade performances), ostracized participants outperform included participants (Experiment 2). Moreover, this effect was mediated by the need for belonging, suggesting that ostracized participants were motivated to elevate their inclusionary status by demonstrating their worth on the cognitive ability task.

Self-Evaluation and Creativity

Lack of evaluation potential leads to the social loafing effect, the finding that participants working together expend less effort than participants working alone. In 1988 Bartis, Szymanski and Harkins replicated this effect, showing that the lack of potential for evaluation by the experimenter led participants to generate fewer uses for a common object when they were asked to generate as many uses as possible. However, their research also demonstrated that this lack of evaluation potential facilitated creativity. Szymanski and Harkins have previously shown that the potential for self-evaluation has the same effect on loafing as experimenter evaluation. When participants can self-evaluate, the loafing effect is eliminated. In the present study, the effect of the potential for self-evaluation on creativity was tested. In a replication of the Bartis et al. study, subjects were asked either to generate as many uses as possible for a common object or to generate uses that were as creative as possible. The results demonstrated that the potential for self-evaluation undermined creativity and facilitated productivity of performance.