When a decision-making group comes into discussion intending to choose among a set of possible courses of action, any disagreement among the members of the group regarding the best option normally results in some form of social influence. We assume that group members enter discussion having formulated a “pre-discussion preference” for a particular option, that these members leave discussion with a “post-discussion preference” among the options, and that the group as a whole reaches a decision about the best option. Social influence occurs when there is a difference between individual member pre-discussion and post-discussion preferences and/or when the group decision differs from what would be predicted based on the members’ pre-discussion preferences.
Research relevant to social influence within groups has a long history. In fact, perhaps the earliest quantitative research study published in a speech communication journal examined the process of preference convergence due to group discussion (Simpson 1939). Although conducted in relatively unrealistic settings in which aggregates of research participants only voiced judgments in one another’s presence, two early studies of visual perception deserve special note. Sherif (1935) placed participants in an inherently ambiguous circumstance, and observed that voiced comments at the early stages of the task resulted over time in the emergence of a consensual framework that was applied during subsequent judgments.
This study instantiates one of two avenues for social influence in groups, informational influence, in which persuasion occurs as the result of the content of arguments made for and against various options during discussion. Asch (1951), in contrast, used a relatively unambiguous situation and noted that research participants believing themselves to be in disagreement with a majority often were persuaded by that majority, even though the participants’ original judgments were obviously correct. In post-study interviews, many of these participants stated that, no matter what they saw, the very presence of a disagreeing majority was persuasive. These findings reflect the second pathway for social influence in groups, normative influence, in which persuasion is due to minority members coming to believe that the majority must be correct due solely to their majority status. Normative influence should not be confused with compliance, in which minority members voice the opinions of the majority but do not in actuality come to agree with that majority view. By the end of the 1950s, these and other studies had led to three empirical generalizations: that group discussion usually leads to convergence of members’ preferences, that the group decision reflects that convergence, and that group minorities normally are persuaded to conform to the majority point of view.
These three empirical generalizations are evident, for example, in research on jury decision-making (Davis 1973). As data based on real jury deliberations are at best difficult to obtain and in many cases illegal to use for research, workers have turned to mock juries, experimental simulations of jury deliberations usually based on convenience samples but occasionally using people called for jury duty but not assigned to a real case. Studies comparing pre-discussion preferences with group decisions reveal preference convergence and majority rule, with a two-thirds majority for either guilt or innocence usually decisive for a group decision. Mock jury studies of communication content during deliberation have been rare.
An important step forward in our understanding of social influence in groups began in the early 1960s with the employment of a relevant research paradigm, the choice dilemma. In choice dilemmas, group members must choose between two possible courses of actions, one relatively attractive but with some likelihood of failure, the other sure to succeed but with less attractive consequences. Early studies led researchers to infer that groups made riskier decisions than individuals, but this inference was the incorrect consequence of a research artifact. What was actually discovered was that groups make more extreme decisions than individuals.
This group polarization effect can technically be defined as the tendency for group decisions to be in the same direction as but more extreme than the mean of the members’ pre-discussion preferences. In addition, the mean of members’ post-discussion preferences also shift toward the extreme, evidence that actual social influence rather than compliance is occurring. By the late 1970s, two theoretical positions, one relying on normative influence and the other on informational influence, came to dominate attempts to explain group polarization.
The normative influence explanation was an extension of Festinger’s social comparison theory, in which, upon discovering where one another stands on the issue, group members shift toward the majority view. This generally leads to group polarization due to the fact that, at least in this research paradigm, group majorities are usually more extreme than minorities. The informational influence explanation, persuasive arguments theory, hinges on the notion that discussion leads members to hear arguments favoring their pre-discussion preference of which they were previously unaware, in turn leading them to support that preference even more strongly. This account, however, only works if the content of discussion is biased toward the majority preference; if discussion content reflects both sides, then arguments favoring each side would counterbalance one another such that no preference shift would occur. Research evidence does indicate a discussion bias in favor of the majority point of view. Both the social comparison and persuasive arguments explanations boast considerable empirical support, and any comprehensive model of group polarization must include both processes (Lamm & Myers 1978).
Two major research programs exploring group polarization have been instigated by communication scholars. First, Boster and co-workers discovered that the inherent tendency of items to elicit risky versus cautious preferences predicted talk time supportive of each option, but that majority view predicted perceived strength of arguments; as a consequence, Boster and Mayer (1984) proposed a multi-stage model of polarization including both normative and informational influence components. Second, both social comparison and persuasive arguments explanations rely heavily on pre-discussion knowledge and preferences to predict group decisions; in response, structurational theorists (Meyers & Seibold 1990) have attempted to demonstrate the sufficiency of discussion independent of pre-discussion processes. These efforts have been indecisive but resulted in research on the relationship between categories of discussional argument and group decision.
Hidden Profiles And Minority Influence
The next major advance in our knowledge about social influence in groups was again the result of another research paradigm, the hidden profile. Hidden profiles exist when research participants are given information about different options, manipulated so that the information each participant received favors the same option but, taken in total, all of the information provided favors a different option. This can be accomplished by, for one of many possible examples, supplying each member of three-person groups with the same four items of information favoring option A but two unique items of information favoring option B. We would assume that each participant would begin the discussion supporting option A. If all the information came up during discussion, then group members would eventually become aware of all six “unshared” items of information and change their support to the “objectively correct” option B.
However, as with group polarization, research has repeatedly demonstrated that discussion is biased toward information relevant to the originally favored option, so that group members continue to favor and generally choose the “objectively incorrect” option. This occurs if for no other reason than the fact that an item of information is more likely to surface if many people are aware of it. In addition, however, shared items of information are more likely repeated during discussion, an indication that shared items of information are considered more credible (Wittenbaum et al. 2004). As with group polarization phenomena, the manner in which decisions are made in the hidden profile paradigm likely includes both normative and informational influence components. Regarding normative influence, there is strong evidence that the larger the proportion of group members aware of an item of information, the greater the influence of that item on the group’s decision. In the case of informational influence, there is counterbalancing evidence concerning the impact of unshared items of information mentioned during discussion on group decisions. Winquist and Larson (1998) proposed a parallel process model in which informational and normative influence have independent effects on postdiscussion preferences, with the situation determining which of the two has the larger impact.
As noted earlier, majorities are usually successful in persuading minorities to conform to their point of view. However, exceptions have long been noted, and during the 1960s procedures were developed that produce a relatively high proportion of group decisions in which majorities give way to minorities. This discovery inspired the growth of a research literature on minority influence. Although majorities find it easy to discount the arguments of a sole deviant, minorities of two or more people are more difficult to ignore, particularly if the members of that minority remain steadfast in their voiced opinions and provide reasons that are consistent with one another. Research has shown that, along with occasionally convincing majorities to change their mind, minorities can be successful in improving the overall level of critical thinking in a group, such that additional options beyond those originally supported by the majority and minority may be uncovered (Nemeth 1986).
- Asch, S. E. (1951). Effects of group pressure upon the modification and distortion of judgments. In H. Guetzkow (ed.), Groups, leadership, and men. Pittsburgh: Carnegie, pp. 177–190.
- Boster, F. J., & Mayer, M. (1984). Choice shifts: Argument qualities or social comparisons. In R. N. Bostrom (ed.), Communication yearbook 8. Beverly Hills, CA: Sage, pp. 393 – 410.
- Davis, J. H. (1973). Group decision and social interaction: A theory of social decision schemes. Psychological Review, 80, 97–125.
- Lamm, H., & Myers, D. G. (1978). Group-induced polarization of attitudes and behavior. In L. Berkowitz (ed.), Advances in experimental social psychology. New York: Academic Press, vol. 11, pp. 227–299.
- Meyers, R. A., & Seibold, D. R. (1990). Perspectives on group argument: A critical review of persuasive arguments theory and an alternative structurational view. In J. A. Anderson (ed.), Communication yearbook 13. Newbury Park, CA: Sage, pp. 268 –302.
- Nemeth, C. J. (1986). Differential contributions of majority and minority influence. Psychological Review, 93, 23 –32.
- Penrod, S., & Hastie, R. (1979). Models of jury decision making: A critical review. Psychological Bulletin, 86, 462 – 492.
- Sherif, M. (1935). A study of some social factors in perception. Archives of Psychology, 27(135).
- Simpson, R. H. (1939). The effect of discussion on intra-group divergences of judgment. Quarterly Journal of Speech, 25, 548 –552.
- Winquist, J. R., & Larson, J. R., Jr. (1998). Information pooling: When it impacts group decision making. Journal of Personality and Social Psychology, 74, 371–377.
- Wittenbaum, G. M., Hollingshead, A. B., & Botero, I. C. (2004). From cooperative to motivated information sharing in groups: Moving beyond the hidden profile paradigm. Communication Monographs, 71, 286 –310.