Stanford University
Since the publication of Henri Tajfel's 1982 edited volume Social Identity and Intergroup Behavior, most conformity research has been framed by a new perspective: social identification. Social identity theory posits that conformity stems from psychological processes; that is, being a member of a group is defined as the subjective perception of the self as a member of a specific category (Abrams et al., 1990; Mackie & Cooper, 1984; Tajfel, 1982; Tajfel & Turner, 1986; Turner, 1982, 1985; Turner et al., 1987; Wilder 1990). Indeed, it is the perception of group membership that guides all other conformity processes (Mackie & Cooper, 1984).
The effects of social identification are pervasive and powerful. The findings in classic and recent research indicate that people with group social identity (1) perceive themselves to be more similar to each other (Allen & Wilder, 1975, 1979; Mackie, 1986); (2) are more likely to act cooperatively (Abrams et al., 1990; Back, 1951); (3) feel a stronger need to agree with group opinion (Deutsch & Gerard, 1955; Mackie et al., 1992; Wilder 1990); (4) perceive in-group messages to be of higher quality (Brock, 1965; Mackie et al., 1990); and (5) conform more in both behavior and attitude (French & Raven, 1959; Wilder and Shapiro, 1984).
In addition to being pervasive, social identification is such a powerful force that researchers have been able to manipulate social identity by changing minimal cues. These cues about social identity operate even when the interaction is held constant. For example, Wilder (1990) manipulated in-group and out-group social identity (1) by having subjects wear badges with the group's name and (2) by putting subjects in a room labeled with the group's name. This manipulation induced subjects to be more influenced by messages ostensibly written by in-group members, even when the messages were identical in all conditions. A study by Mackie (1986) uses another cue, group interdependence, to generate vastly different responses to an identical situation. To manipulate intergroup competition, subjects were told they could win twelve dollars by conforming to the procedural rules of the experiment. All subjects listened to an identical taped discussion from ostensible group members, yet subjects in the intergroup competition condition perceived group norms to be more extreme and showed greater attitude polarization.
If social identification is such a powerful cue--as the recent research indicates--then the pervasive effects of social identification might well be generated even when one of the participants in the interaction is not a human--that is, when one of the interactants is a computer. Indeed, using computers to study human social phenomena has shown promising results recently. Nass and colleagues, following the Computers Are Social Actors paradigm (Nass, Steuer & Tauber, 1994), have demonstrated that a number of social rules and psychological processes which guide human-human interaction apply equally to human-computer interaction.
By using a computer as a participant in a group interaction, our study puts the power and pervasiveness of social identity to the test by investigating two key questions: (1) Can researchers manipulate affiliation (create a sense of teamwork or groupfulness) between humans and computers? and (2) Will the effects of human-computer affiliation be similar to the documented effects of human-human affiliation? Given the pervasiveness and power of social identity, and given the implications of the Computers Are Social Actors paradigm, we expect this experiment on human-computer affiliation to produce the same findings as the human-human affiliation studies mentioned above.
Team condition. To manipulate the social identity of "teamness" (presence of affiliation), we simply told "team" subjects that they were part of the "blue team" and that they would interact with a teammate called the "blue computer." In addition, we manipulated interdependence by telling each "team" subject that he or she would be evaluated as a team with the computer.
Individual condition. We explained to each "individual" subject (absence of affiliation) that he or she would be interacting with a computer but would be working as an individual--a "blue individual" working with a "green computer." We also told "individual" subjects that they would be evaluated on the basis of their individual work alone.
In both conditions we explained that the computer did not necessarily have all the requisite information.
Affiliation Assignment. Once subjects completed their initial ranking of items, the experimenter brought them into a room with various computers. The subjects were informed that they would now have a chance to interact with a certain computer about each of the 12 items. At this point, the subjects were told either (1) they were working as a team with the computer or (2) they were working as an individual.
Ranking Exchange. Before the interaction with the computer began, subjects entered their ranking of items onto their own screen and wrote down the rankings from the computer they would be interacting with. Unknown to the subjects, the computer's rankings were systematically related to each subject's ranking. For example, if a subject ranked an item as number 2, the computer would automatically rank that item as number 5 and so on. Because the computer's ranking depended entirely on the subject's ranking, the subject's and the computer's rankings were equally dissimilar in each condition.
Interaction. The experimenter then guided the subjects through a practice interaction with the computer, in which the subjects exchanged information about a practice desert survival item. Subjects typed their ideas into what was designated as their own screen. The computer then presented its information about each item on a different screen. For example, when the flashlight was the survival item under discussion, the text from the computer would read, "The flashlight is the only reliable source of signalling after dark. This is a very important item for survival." During the experiment, subjects exchanged information with the computer on each of the 12 desert survival items. The computer presented identical information in both conditions.
Finally, subjects filled out two questionnaires with 10-point Likert scales. The first questionnaire assessed each subject's response to the interaction with the computer. The second one assessed each subject's response to the computer. We used these measures to determine the subjects' attitudes toward the interaction and the computer itself.
Affiliation was an index of two items: thinking of self as part of a group and thinking of self as a partner with the computer (Cronbach's alpha = 0.69).
Perceived similarity was an index of six items: perceived similarity of approach, perceived similarity of suggestions, perceived similarity of interaction style, perceived similarity of initial rankings, perceived similarity of final ranking to the computer's initial ranking, and perceived similarity of final rankings to the computer's hypothetical final ranking (alpha = 0.68).
Cooperation was an index of three items: cooperation with the computer, desire to reach agreement with the computer, and responsiveness to the computer's suggestions (alpha = 0..86).
Openness to influence was an index of eight items: openness to influence from the computer, receptivity to the computer's suggestions, dependence on the computer's suggestions, acceptance of the computer's advice, agreement with the computer, responsiveness to the computer's suggestions, trust in the computer's information, and desire to reach agreement with the computer (alpha = 0.93).
Perceived information quality was an index of three items: relevance of the computer's information, helpfulness of the computer's information, and insightfulness of the computer's information (alpha = 0.92).
[Insert figure here]
Consistent with the results from human-human studies, team subjects perceived themselves to be more similar to the computer than did the individual subjects, t(26) = 2.26, p < .05. In addition, as predicted by human-human studies, team subjects perceived themselves as more cooperative than did individual subjects, t(26) = 3.78, p < .001. Furthermore, findings from human-human studies were replicated when team subjects perceived themselves to be more open to influence from the computer than were the individual subjects, t(26) = 4.34, p < .001.
The data also show significant results about the subjects' perception of the information from the computer. Consistent with the results from human-human studies, team subjects perceived the computer's information to be of higher quality than did individual subjects, t(26) = 2.96, p < .01. In addition, team subjects perceived the information on the computer to be friendlier (a single item) than did individual subjects, t(26) = 1.93, p < .05.
Finally, the data show significant behavioral conformity. Consistent with the results from human-human studies, team subjects conformed their rankings more to the position advocated by the computer than did individual subjects, M = 16.3 vs. M = 21.9, t(26) = 2.24, p < .02.
This study found that manipulating human affiliation with a computer is surprisingly easy: simply tell subjects (1) that they are part of a team with a computer and (2) that they will be evaluated as a team. By successfully manipulating social identity, this study demonstrates that humans who work in teams with computers display the same sorts of attitudes and behaviors as when working in teams with other humans. In sum, compared to individual subject, team subjects saw themselves as more similar to the computer, saw themselves as more cooperative, were more open to influence, thought the computer's information was of higher quality, and found the computer's information to be friendlier.
In addition to generating significant differences in attitude, this experiment also shows that subjects behave differently because of perceptions of affiliation with a computer--team subjects are more likely to conform to the computer's suggestions.
Although similar conformity attitudes and behaviors have been demonstrated ever since Sherif (1936), such effects have never been observed when the group consisted solely of a person and a computer--a dyad that presents the sparsest form of social interaction possible.
Allen, V. L. & Wilder, D. A. (1979). Group categorization and attribution of belief similarity. Small Group Behavior, 10, 73-80.
Allen, V. L. & Wilder, D. A. (1975). Categorization, belief similarity, and intergroup discrimination. Journal of Personality and Social Psychology, 32, 971-977.
Allen, V. L. (1965). Situational factors in conformity. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 2, pp. 133-176). Orlando, FL: Academic Press.
Back, K. W. (1951). Influence through social communication. Journal of Abnormal and Social Psychology, 46, 9-23.
Berkowitz, L. (1957). Effects of perceived dependency relationships upon conformity to group expectations. Journal of Abnormal and Social Psychology, 55, 350-54
Brock, T.C. (1965). Communicator-recipient similarity and decision change. Journal of Personality and Social Psychology, 1, 650-654.
Cartwright, D. & Zander, A. (1968). Group Dynamics: Research and Theory. New York: Harper & Row.
Deutsch, M. & Gerard, H. B. (1955). A study of normative and information social influences upon individual judgement. Journal of Abnormal and Social Psychology, 51, 629-636.
French, J. R. P., & Raven, B. H. (1959). The bases of social power. In D. Cartwright (Ed.), Studies in social power (pp. 150-167). Ann Arbor: University of Michigan.
Horwitz, M. & Rabbie, J. M. (1982). Individuality and membership in the intergroup system. In H. Tajfel (Ed.), Social identity and intergroup relations (pp. 241-274). Cambridge, England: Cambridge University Press.
Kelman, H. C. (1961). Processes of opinion change. Public Opinion Quarterly, 25, 57-78.
Lafferty, J.C., & Eady, P.M. (1974). The desert survival problem. Plymouth, Michigan: Experimental Learning Methods.
Lewin, K. (1948). Resolving social conflicts. New York: Harper.
Mackie, D. M. (1986). Social identification effects in group polarization. Journal of Personality and Social Psychology, 50, 720-728.
Mackie, D. M., Gastardo-Conaco, M. C., & Skelly, J. J. (1992). Knowledge of the advocated position and the processing of in-group and out-group persuasive messages. Personality and Social Psychology Bulletin, 18, 145-151.
Mackie, D. M., Worth, L. T., & Asuncion, A. G. (1990). Processing of persuasive in-group messages. Journal of Personality and Social Psychology, 58, 812-822.
Mackie, D.M. and Cooper, J. (1984). Group polarization: The effects of group membership. Journal of Personality and Social Psychology, 46, 575-585.
Nass, C., Steuer, J. & Tauber, E. (1994). Computers are social actors. Paper presented to CHI'94 conference of the ACM/SIGCHI Boston, MA.
Shaw, M. E. (1981). Group dynamics: The psychology of small group behavior. New York: McGraw-Hill Book Company.
Sherif, M. (1936) The psychology of social norms. New York: Harper.
Spears, S. C. (1989). Controlling for exposure to intragroup communication: Effects of heterogeneous social perspectives and dependence in intragroup behavior and attitudes. Unpublished doctoral dissertation, Stanford University.
Tajfel, H. & Turner, J. C. (1986). An integrative theory of intergroup conflict. In S. Worchel & W. Austin (Eds.), Psychology of intergroup relations (pp. 2-24). Chicago: Nelson-Hall.
Tajfel, H. (1982). Social identity and intergroup behavior. Cambridge, England: Cambridge University Press.
Tajfel, H., Billig, M., Bundy, R., & Flament, C. (1971). Social categorization and intergroup behavior. European Journal of Social Psychology, 1, 149-178.
Turner, J. C. (1982). Toward a cognitive redefinition of the social group. In H. Tajfel (Ed.), Social identity and intergroup behavior (pp. 15-40). Cambridge, England: Cambridge University Press.
Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive theory of group behavior. In E.J. Lawler (Ed.), Advances in group processes: Theory and research. vol. 2, pp. 77-121. Greenwich, CT: JAI Press.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D. & Wetherell, M. S. (1987). Rediscovering the social group. Oxford, England: Basil Blackwell.
Wilder, D. A. (1990). Some determinants of the persuasive power of in-group and out-groups: Organization of information and attribution of independence. Journal of Personality and Social Psychology, 59, 1202-1213.
Wilder, D.A. & Shapiro, P.N. Roles of out-group cues in determining social identity. Journal of Personality and Social Psychology, 47, No. 2 342-348.