Group Size and Conformity Rod Bond To cite this version: Rod Bond. Group Size and Conformity. Group Processes and Intergroup Relations, SAGE Publica- tions, 2005, 8 (4), pp.331-354. 10.1177/1368430205056464. hal-00571611 HAL Id: hal-00571611 https://hal.archives-ouvertes.fr/hal-00571611 Submitted on 1 Mar 2011 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. 01 Bond 056464 (bc-s) 30/9/05 1:50 pm Page 331 Group Processes & G Intergroup Relations P 2005 Vol 8(4) 331–354 I R Group Size and Conformity Rod Bond University of Sussex This paper reviews theory and research on the relationship between group size and conformity and presents a meta-analysis of 125 Asch-type conformity studies. It questions the assumption of a single function made in formal models of social influence and proposes instead that the function will vary depending on which social influence process predominates. It is argued that normative influence is likely to be stronger when participants make public responses and are face-to-face with the majority, whereas informational influence is likely to be stronger when participants make private responses and communicate with the majority indirectly. The meta-analysis finds that the relationship differs according to whether public or private responses are obtained and whether an Asch or Crutchfield paradigm is employed. Future research needs to identify how the relationship varies depending on different social influence processes elicited by features of the task and setting. keywords conformity, meta-analysis, social-influence ONE of the oldest questions in social influence rise to conformity and that the function will research concerns the relationship between depend on which process predominates, conformity and the size of the majority. Asch’s depending on particular features of the task (1951, 1955) startling conclusion that majority and setting. size did not have much effect beyond a minimal There are three main parts to this review. number provoked a lively debate, one that con- The first compares theories of group size and tinues to be presented in almost every textbook conformity in terms of the form of the relation- on social psychology. More recent theories, ship proposed and the social influence process such as Latané’s (1981) Social Impact Theory presumed to give rise to it. The second part is a (SIT), mostly disagree with Asch and hold that narrative review of the existing empirical the larger the size of the majority, the larger the evidence for the relationship, including both effect, and the main issue that currently distin- individual studies and previous meta-analytic guishes different models is the precise form reviews, and finds that existing evidence is that the relationship of group size to con- formity is supposed to take. Author’s note In contrast, this paper questions the dominant Address correspondence to Rod Bond, assumption that there will be a single function Department of Psychology, University of that describes the relationship. Instead, it argues Sussex, Brighton BN1 9QH, UK that a number of social influence processes give [email: [email protected]] Copyright © 2005 SAGE Publications (London, Thousand Oaks, CA and New Delhi) 8:4; 331–354; DOI: 10.1177/1368430205056464 01 Bond 056464 (bc-s) 30/9/05 1:50 pm Page 332 Group Processes & Intergroup Relations 8(4) inconclusive. The third part presents a new pressure in which normative and informational meta-analysis that investigates whether the influence are involved to differing degrees. relationship varies depending on two features Nevertheless, his emphasis is on informational of the task and setting: whether the experi- influence in that, for conformity to occur, the mental paradigm entails the participant in face- majority must provide an alternative, and to-face or indirect interaction with the majority apparently incontrovertible, version of reality and whether the participant gives a public or which most participants decide is correct, or private response. Only studies which use Asch’s possibly correct. It is essential, therefore, for (1952a, 1956) line judgment task and where a the majority to be unanimous (Asch, 1955), and unanimous majority confronts a single indi- of sufficient size that their judgments cannot be vidual were selected, thereby exercising greater dismissed as idiosyncratic: control than previous meta-analyses, especially over differences between studies that may be The majority, although limited in size, was repre- confounded with majority size. The aim is to sentative. The minority individual had no reason to suppose that others, not included in the group, demonstrate that the relationship between con- would be more likely to side with him. The given formity and majority size is complex, in that it majority was symbolic of what any portion of cannot be described by a simple function but humanity might perceive. (Asch, 1956, p. 67) instead varies systematically with factors that impact on social influence processes. Increasing the size of the majority will have no additional impact, since the addition of Theories of group size and further individuals who state the same view conformity merely confirms that this is a representative sample. Asch concluded that a majority size of Theories of group size and conformity have three is sufficient for the full impact of the been concerned primarily with describing the group to be felt (Asch, 1955). functional relationship between these variables, More recent theories,1 however, attach greater rather than with how group size relates to the importance to group size and see both norma- psychological processes involved in social influ- tive and informational influence as contribut- ence. Nevertheless, the theories differ in their ing to the effect. In SIT (Latané, 1981; Latané account of the psychological processes that & Wolf, 1981), it is argued that the larger the mediate the relationship, as well as in the form group the greater its impact, not just because of the relationship itself. Most theories draw on the majority provide information about reality dual process theory, according to which con- but also because of the majority’s power to formity is explained by two distinct processes, reward and punish. The relationship between one being normative influence, which reflects the group size and impact is not expected to be group’s power to reward and punish, and the linear, however, and Latané (1981) proposes other being informational influence, which that the additional impact is smaller for each reflects the group’s capacity to provide infor- additional group member, so that the function mation about reality (Deutsch & Gerard, 1955). relating group size to conformity is a negatively Whereas theories of group size and conformity accelerating curve. This is expressed in the have argued that these processes result in a equation single form of relationship, we shall argue in contrast that the form of relationship will I = sN t (1) depend on the relative strength of each of these motives. where I is impact, s is a scaling constant, N is Asch (1951, 1955, 1956) recognizes that the the number of sources of influence, and the processes underlying response to group exponent t is a value less than one. The theory pressure are complex, and his analysis focuses also proposes that impact is a multiplicative on individual differences in response to group function of the strength and immediacy of the 332 01 Bond 056464 (bc-s) 30/9/05 1:50 pm Page 333 Bond group size and conformity influence sources, as well as number, and hence will result in greater conformity. Mullen argues the steepness of the function will reflect that increasing majority size will lead to greater features of the task and setting which affect self-attention, since the minority individual will these parameters, features which may also increasingly stand out as the figure against the reflect whether normative or informational ground provided by the majority, and hence influence predominates. Latané and Wolf increasingly feel the focus of attention. The (1981) demonstrate that SIT provides a good fit Other-Total Ratio (OTR), the number in the to the results from two studies concerned with ‘other subgroup’ (i.e. the majority in conformity conformity and majority size, one by Gerard, studies) divided by the total number in the Wilhelmy, and Conolley (1968) and another by group, quantifies the degree of self-attention. Latané and Davis (cited in Latané & Wolf, Where the minority is a single individual, this 1981). can be expressed by the formula Tanford and Penrod (1984), on the other hand, propose a Social Influence Model (SIM) I = ␣ͩ––––N ͪ (3) entailing an S-shaped function relating group N +1 size and social influence. When the number of targets is a minority of one, the equation is where I = impact, ␣ is an influence constant which might vary from task to task, and N = the I = exp(–4 ϫ exp(–N1.75)) (2) size of the majority. By applying OTR, Mullen proposes that the where I = impact, and N = number of sources of degree of self-attention, and hence conformity, influence.2 They argue that this is superior to will be a negatively accelerating function of SIT in two respects. First, whereas SIT predicts group size. He argues that his model is superior that the first influence source will have the to SIT in that the function is specified a priori, greatest impact, Asch (1951) found that the and a psychological process, self-attention, is second and third sources had greater additional proposed to explain the effect.