Success-Failure Attributions in Competitive Groups: Αn Exception to Egocentrism

Two laboratory experiments were conducted to investigate success/failure attributions within competing groups. In both studies, attributions to the own team or to opponents were egocentric in that members of winning teams assigned responsibility for success primarily to their own team whereas members of losing teams assigned responsibility for the loss primarily to the opponents. Within-team attributions, however, revealed a reverseegocentric pattern. Members of winning teams assigned primary responsibility to their teammates, and losing team members accepted primary responsibility for the loss themselves. Attribution patterns, which were consistent across both studies and for both males and females, were interpreted as reflecting a team-enhancing strategy or norm. Article: Investigations of cognitive phenomena, especially causal attributions, currently dominate social psychology research. The switch from more behavioristic studies of situational factors and their effects to an emphasis on the individual's thoughts and interpretations has its counterpart in sport psychology, but investigations of attributions in sport settings are quite limited in number and in scope. Much of the attribution research, including research on success/failure attributions in competition, emanates from the model of Weiner, Frieze, Kukla, Reed, Rest, and Rosenbaum (1971). Weiner et al. identified the four standard causal attributions of ability, effort, luck, and task difficulty and proposed a two-dimensional classification system with attributions classified as internal (ability and effort) or external (luck and task difficulty) and stable (ability and task difficulty) or unstable (effort and luck). Following that basic paradigm, researchers have consistently observed that success is attributed internally more than failure, which tends to be attributed to external factors (e.g., Fitch, 1970; Frieze & Weiner, 197I; Wolosin, Sherman, & Till, 1973; Wortman, Costanzo, & Witt, 1973). This tendency to take personal credit for success and shift blame for failure to external factors is generally interpreted as an egocentric or selfenhancing bias. More recently, the same tendency toward self-enhancement in success; failure attributions has been observed with sport teams. Iso-Ahola (1975, 1977) and Roberts (1975, 1978) investigated win; loss attributions with Little League baseball; Forsyth and Schlenker (1977b) examined causal attributions following tennis matches; and Bird and Brame (1978) studied female intercollegiate basketball players. In all cases the results concurred with the nonsport literature in that winners or successful competitors gave more internal attributions than losing or unsuccessful competitors. In the studies noted thus far, egocentrism has been determined from the relative use of internal (ability and effort) and external (luck and task difficulty) attributions following success and failure. Schlenker and his colleagues have, however, recently investigated egocentrism in terms of attributions of responsibility within the group. Typically, Schlenker asks members of problem-solving groups to rate their personal responsibility for success or failure outcomes. Results consistently indicate that group members accept more personal responsibility for group success than for group failure (Forsyth & Schlenker, 1977a; Schlenker, 1975; Schlenker & Miller, 1977a, 1977b; Schlenker, Soraci, & McCarthy, 1976). In addition to Schlenker's work, Mynatt and Sherman (1975) observed that members of failing groups took less responsibility for the outcome than members of successful groups or individuals who either succeeded or failed. Also, Wolosin, Sherman, and Till (1973) reported that individuals assigned more responsibility to themselves than to their partners when they succeeded on a cooperative group task, but assigned more responsibility to the partner when they failed. Assignment of responsibility within groups has not been examined directly with sport teams, but the studies of Iso-Ahola (1977), Roberts (1978), and Bird and Brame (1978) hint that similar egocentric biases occur. Roberts (1978) observed that self-serving strategies were used with self-attributions, but that information-processing was used with team attributions. In Iso-Ahola's study, team failure did not decrease subjects' evaluations of individual ability and effort, but did decrease those evaluations of the team. Bird and Brame (1978) similarly noted that members of losing teams rated individual effort higher than team effort. Although these findings suggest team members take more personal credit for success than failure, such an egocentric bias was not directly examined. Instead, the studies used separate ratings of the four attributions for self and team. The current study departs from the standard causal attribution measures and adopts Schlenker's approach to examine attributions within the group by asking individuals to assign primary responsibility for success and failure to themselves or to their teammates. The overall investigation included two separate laboratory experiments. Rather than the four standard attributions of ability, effort, luck, and task difficulty, two teamrelated forced-choice attribution items were used in both studies. First, individuals were asked to assign primary responsibility for success or failure to their own team or to the opponents. Previous research consistently indicates that success is attributed more internally than failure. Thus, it was hypothesized that group members attribute success primarily to their own team and failure primarily to the opponents. The second attribution measure requested individuals to assign primary responsibility for success or failure to themselves or to their teammates. Egocentric attribution patterns similar ιo those found by Schlenker were hypothesized, with individuals takingsimilar ιo those found by Schlenker were hypothesized, with individuals taking credit for success but assigning responsibility for failure more to teammates. Experiment 1 Method Subjects and design. All 64 college females who participated in Experiment were volunteers from an introductory kinesiology class. All subjects participated in an initial session consisting of 20 individual trials on the maze task prior to the actual experiment. On the basis of the individual performance times obtained in the initial session, 32 two-person groups were formed and then assigned to the winning or losing condition. Tusk. The motor maze, which consists of an aluminum maze mounted on a two-dimensional teeter-board, was the group task used for Experiment 1. The object of the task is to move a steel ball around the maze from start to finish as quickly as possible while avoiding numerous cul-de-sacs. Two handles located 011 adjacent sides of the maze tilt the maze hoard forward and backward and side-to-side, respectively, and allow the ball to move through the maze. Performance on the maze task is operationally defined as the time taken to complete the maze to the nearest .1 second. Although the maze can he operated by one individual, the maze was used as a group task in the current study with each of the two group members using one handle of a single maze task. The group task required the joint efforts of both group members and considerable cooperation because the two handles must be operated together to successfully negotiate the maze. Ρrocedures. All subjects performed 20 individual trials on the maze task in an initial session approximately 1 week prior to the actual experiment. On the basis of these individual performance times, subjects were matched with partners so that the groups represented a range of average ability levels. The initial session and matching procedures were used to examine the relationship between individual ability and group performance and those results are discussed elsewhere (Gill, 1979). Two groups of the same average ability level were scheduled at the same time for Experiment I. One group was randomly assigned to the winning condition and the other to the losing condition. The experimenter explained the operation of the group maze task and the competition set-up to all four subjects together, and each group was allowed one practice trial. The two groups were then assigned to separate experimental rooms where they could not see or hear each other. The experimenter was located in a third room connected to each maze room with a one-way mirror. The experimenter had a master console connected to each of the mazes that monitored progress of the ball through each maze with photoelectric cells and recorded the time taken to complete each maze to the nearest .1 second. All four subjects wore earphones during the competition, and after each trial the experimenter announced the two team times and the winner for that trial. In reality, the team times were contrived so that the group assigned to the winning condition won 15. 16, or 17 of the 20 trials. After all 20 trials were completed, final win/loss totals were announced, and the subjects remained in the experimental rooms to complete the postcompetition questionnaire. Questionnaire. An initial item on the questionnaire asked subjects to rate their team's overall performance as successful or unsuccessful. Two forcedchoice items were then used to assess subjects' perceptions of the causes for that performance outcome. First, respondents were asked to indicate whether the success or lack of success was due more to their team or their opponents, and a second item asked whether the success or lack of success was due more to themselves or their partner. Names were not requested and subjects were assured their responses would not be seen by anyone except the experimenter. Results The initial item of the questionnaire was essentially a check on the win/loss manipulation. Generally, subjects rated their team's performance as successful or unsuccessful in line with the win; loss manipulation, χ2(1) = 43.31,unsuccessful in line with the win; loss manipulation, χ2(1) = 43.31,ρ < .001 (all chi-squares reported in this study are corrected for continuity). All 32 members of winning teams rated their team's performance as successful but so did five members of losing teams. The informal observation of the experimenter and the responses of these five individuals to an open-ended attribution item indicated that they did not think they had won the group competition but rather that they had tried hard, performed well, and considered themselves successful despite the loss. Such a response is understandable and the discrepancy between win/ loss outcomes and perceptions of success/failure is worthy of more extensive investigation. In this study, however, responses to the attribution choices depended on the successful/ unsuccessful rating, and these five subjects were eliminated from the remaining analyses. Responses to the own team vs. opposing team choice were significantly affected by the win; loss condition, χ2(1) = 7. 1 2. p < .01. Of the 32 members of winning teams, 26 (81%) indicated their own team was more responsible for the success, whereas 15 (56%) of the 27 losing team members indicated that the opposing team was more responsible for the lack of success. The responses follow an egocentric pattern with more responsibility assigned to the own team for a win and to the opposing team for a loss. Responses to the self versus partner choice were, however, quite different. The self versus partner choice was significantly affected by the win/loss condition, χ condition, χ 2 (1) = 3.88,p < .05, but the responses did not reflect an egocentric pattern. Instead, 20 (63%) of the 32 members of winning teams gave more credit to the partner while l8 (67%) of the losing team members placed more blame on the self. The results are, then, directly opposed to the predicted egocentric responses. Experiment 2 The rather serendipidous finding of Experiment 1, namely that self versus partner attributions not only failed to exhibit an egocentric pattern but actually exhibited a reverse-egocentric pattern, prompted Experiment 2. The primary purpose of Experiment 2 was to replicate the group competition situation of Experiment 1 to determine if the observed attribution choices were reliable findings. Also. Experiment 2 included both male and female subjects. Although the literature on sex differences in attributions is not conclusive, some evidence suggests that females are less egocentric than males (Deaux, 1976; Deaux & Farris, 1977; Forsyth & Schlenker, I977ó; McHugh, Duquin, & Frieze, 1978; Nicholls, 1975; Wiegers & Frieze, 1977). Thus, including only females in Experiment I may have produced the nonegocentric tendency to accept personal responsibility for a loss and assign credit to the partner for success. Method Much of the method for Experiment 2 was identical to Experiment I, and only the differences are discussed in detail. All subjects (32 males and 32 females), who were volunteers from an introductory kinesiology class as in Experiment I, were randomly matched with a partner of the same sex forming 32 two-person groups. Equal numbers of male and female groups were then randomly assigned to the winning (win 16 of 20 trials) and losing (win 4 of 20 trials) conditions and scheduled for a group competition session with an opposing group of the same sex. The group maze task and the experimental set-up for group competition were identical to Experiment I. Experiment 2 did not include an initial practice session, and therefore, the group competition was the subjects' first exposure to the maze task. All other experimental procedures were identical to those of Experiment 1. Results To avoid the problems of Experiment I with the successful/ unsuccessful choice, respondents were simply asked to indicate whether their team had won or lost. All 64 subjects correctly indicated that their team had won or lost in line with the manipulation, χ 2 (1) = 60.06, p< .001. As in Experiment 1, win/loss significantly affected the own team versus opponents choice, χopponents choice, χ 2 (1) = 12.26, ρ < .001. Again, an egocentric pattern emerged with 24 (75%) of the members of winning teams indicating their own team was primarily responsible, and 23 (72%) of the losing team members assigning primary responsibility to the opponents. (See Table 1 for the results of both experiments.) Male and female members of winning teams exhibited exactly the same pattern, but males on losing teams were slightly more external than females. Only two (13%) males indicated their own team was primarily responsible for the loss, whereas 7 (43%) of the females did so. Nevertheless, neither sex nor the sex by win/ loss interaction significantly influenced the own versus opposing team choice when the total chi-square was partitioned as described by Winer (1971). Responses to the self versus partner choice also confirmed the findings of Experiment 1 with win; loss significantly influencing the choice, x2(1) = 6.31, p .05. The pattern was reverse-egocentric with 20 (63%) members of winning teams assigning primary responsibility to the partner, and 23 (72%) members of losing teams accepting primary responsibility themselves (see Table 2). Males and females exhibited nearly identical attribution patterns and neither sex nor the sex by win/loss interaction influenced the self versus partner choices. In sum, results of Experiment 2 confirmed those of Experiment 1 suggesting that the reverse-egocentric self versus partner attributions were reliable findings. Furthermore, Experiment 2 indicated that males and females alike were egocentric in terms of the own team versus opponents' attributions, but reverseegocentric in terms of attributions within the team. Discussion The findings of the current investigation were consistent across two laboratory experiments and held for both males and females. Team members exhibited the hypothesized egocentric pattern by assigning responsibility for success primarily to their own team and by assigning responsibility for failure to the opposing team. Contrary to previous findings and predictions, however, attributions to the self or to teammates not only were not egocentric, but, in fact, demonstrated a reverse-egocentric pattern. Team members consistently gave credit to their partners for success, but assigned responsibility for failure to themselves. The results are especially surprising in light of the fact that the egocentric pattern of success /failure attributions and the assignment of responsibility within groups is one of the most consistently supported findings in current social psychology literature. Although previous research suggests egocentric attributions occur within groups, that literature also provides some clues as to why such patterns might not be observed. Some of Schlenker's work, which provides the strongest support for egocentric attributions within groups, also suggests factors that may reduce these egocentric tendencies. Schlenker, Soraci, and McCarthy (1976) observed egocentric attributions, but less perception of being better than others than previous work had found. Schlenker et al. speculated that the face-toface contact and communication in their groups might have minimized the egocentric differences in accepting credit and blame. Schlenker and Miller (19776) subsequently examined cohesiveness as an influence on egocentric attributions. High-cohesive groups in the Schlenker and Miller study did not show the same egocentric attribution patterns shown by the low-cohesive groups. In both laboratory experiments, group members were involved in highly cooperative tasks that required considerable interaction, and this group interaction and cohesiveness may well have reduced egocentric tendencies. The current results, however, indicate that egocentrism was not only reduced but reversed. The tendency of individuals to credit their teammates for success and accept blame for failure themselves sets the current study apart from previous work on group attributions. This study is not the first one to report nonegocentric attributions for success/failure in competition. Scanlan (1977) observed that individual competitors on the maze task attributed failure to the self more than success. Scanlan (Scanlan, 1977; Scanlan & Passer, 1978) interprets such attributions as examples of "good winner" and "good loser" norms. Good winners do not downgrade their opponents and good losers accept the loss without blaming others. Similar norms apply to teams in competition. Even individual superstars do not take full credit for success. Doubtlessly, everyone has heard a football running back credit his teammates on the line for a successful game. Furthermore, a good team player does not blame teammates for a loss. Such norms are not exactly the same as the good winner and good loser norms noted by Scanlan. Instead, the norms that are proposed to operate in the current study are team norms. Egocentrism, as used in previous studies, and the contrary good winner and good loser norms discussed by Scanlan focus on the individual. Egocentrism reflects a self-serving or self-enhancing bias, but the current results suggest a "team-serving" or "team-enhancing" bias. That is, the team is the focal point of the individual's attributions. Crediting teammates for success and not blaming them for failure maybe team-enhancing in that positive interpersonal relationships and cohesiveness are maintained. Discriminant analysis results from the Bird and Brame (1978) investigation provide indirect support for the notion that the team is the focal point for attributions in basketball teams. Of the four attributions that discriminated between winning and losing teams, three were team attributions, and the most powerful discriminator was team ability. Of course, the speculation that the reverse-egocentrism in the current study results from a team-enhancing biasis just that--speculation. The study was not designed to determine why such effects occurred. In fact, thereverse-egocentric attributions were not anticipated. Nevertheless, the fact that these unique attributionalpatterns were observed, and were observed for both males and females in two separate studies, is notable. Thefindings underscore the importance of considering alternative approaches and varied measures of attributionsand the assignment of responsibility within groups or teams to fully understand the cognitive processesoperating in team competition. ReferencesBird, Α.Μ., & Brame, J.M. Self versus team attributions: A test of the "l'm OK, but theBird, Α.Μ., & Brame,J.M. Self versus team attributions: A test of the "l'm OK, but the team's so-so" phenomenon. ResearchQuarterly, 1978, 49, 260-268.Deaux, K. The behavior of women and men. Monterey, Ca.: Brooks: Cole, 1976. Deaux, K., & Farris, Ε.Attributing causes for one's own performances: The effects of sex,Deaux, K., & Farris, Ε. Attributing causes forone's own performances: The effects of sex, norms, and outcome. Journal of Research in Personality, 1977, 11,59-72.Fitch, G. Effects of self-esteem, perceived performance, and choice on causal attributions. Journal ofPersonality and Social Psychology, 1970, 16, 311-315.Forsyth, D. R., & Schlenker, B. R. Attributing the causes of group performance: Effects of performance quality, task importance, and future testing. Journal of Personality, 1977, 45, 220-236. (a)Forsyth, D. R., & Schlenker, B.R. Attributional egocentrism following performance of a competitive task.Journal of Social Psychology, 1977, 102, 215-222. (b)Frieze, I., & Weiner, B. Cue utilization and attributional judgments of success and failure. Journal ofPersonality, 1971, 39, 591-605.Gill, D.L. The prediction of group motor performance from individual member abilities. Journal of MotorBehavior, 1979, 11, 113-122.Iso-Ahola, S. A test of the attribution theory of success and failure with Little League baseball players. In C.Bard, M. Fleury, &J.; Salmela (Eds.), Mouvement: Actes du 7esymposium en apprentissage psycho-moteur etpsychologie du sport. Québec: Association des professionnels de l'activité physique du Québec, 1975.Iso-Ahola, S. Effects of team outcome on children's self-perception: Little League baseball. Scandinavian Journal of Psychology, 1977, 18, 38-42.McHugh, M.C., Duquin, Μ.Ε., & Frieze, McHugh, M.C., Duquin, Μ.Ε., & Frieze, 1.1-I. Beliefs about successand failure: Attribution and the female athlete. In C.A. Oglesby (Ed.), Women and sport: From myth to reality.Philadelphia: Lea & Febiger, 1978.Mynatt, C., & Sherman, S.J. Responsibility attribution in groups and individuals: A direct test of the diffusionof responsibility hypothesis. Journal of Personality and attainment Social Psychology, 1975, 32, 111 1-1 118.Nicholls, J. Causal attributions and other achievement related cognitions: Effects of task, outcome, attainmentvalue and sex. Journal of Personality and Social Psychology, 1975, 31, 379-389.Roberts, G.C. Win-loss causal attributions of Little League players. ln C. Bard, M. Fleury, & J. Salmela (Eds.),Mouvement: Actes du 7esymposium en apprentissage psycho-moteur et psycholgie du sport. Québec:Association des professionnels de ]'activité physique du Québec, 1975. Roberts, G.C. Children's assignment of responsibility for winning and losing. In F.L. Smoll & R.E. Smith(Eds.), Psychological perspectives in youth sports. Washington, D.C.: Hemisphere, 1978.Scanlan, Τ.Κ. The effect of success-failure on the perception of threat in a competitiveScanlan, Τ.Κ. The effectof success-failure on the perception of threat in a competitive situation. Research Quarterly, 1977, 48, 144-153. Scanlan, Τ.Κ., Τ.Κ., & Passer, M. W. Anxiety-inducing factors in competitive youth sports. In F.L. Smoll &R.E. Smith (Eds.), Psychological perspectives in youth sports.Washington, D.C.: Hemisphere, 1978. Schlenker, B.R. Group members' attributions of responsibility for prior group performance. RepresentativeResearch inn Social Psychology. 1975, 6, 96108.Schlenker, BR., & Miller, R.S. Egocentrism in groups: Self-serving bias or logical information processing.Journal of Personality and Social Psychology, 1977, 35, 755-764. (a)Schlenker, B.R., & Miller, R.S. Group cohesiveness as a determinant of egocentricperceptions in cooperative groups. Human Relations, 1977, 30, 1039-1055. (b) Schlenker, BR., Soraci, S., Jr., &McCarthy, B. Self-esteem and group performance as determinants of egocentric perceptions in cooperativegroups. Human Relations,1976, 29, 1163-1176.Weiner, B., Frieze, 1.H., Kukla, D., Reed, L., Rest, S., & Rosenbaum, R.M. Weiner, B., Frieze, 1.H., Kukla, Δ.,Reed, L., Rest, S., & Rosenbaum, R.M. Perceiving the causes of success and failure. Morristown, N.J.: GeneralLearning Press, 1971. Wiegers, R.M., & Frieze, 1.H. Gender, female traditionality, achievement level, andcognitions of success and failure. Psychology of Women Quarterly, l977, 2, 125-137. Winer, B.J. Statisticalprinciples in experimental design (2nd ed.). New York: McGraw-Hill, 1971.Wolosin, R.J., Sherman, S.J., & Till, Α. Effects of cooperation and competition onWolosin, R.J., Sherman, S.J.,& Till, Α. Effects of cooperation and competition on responsibility attribution after success and failure. Journalof Experimenial Social Psychology, 1973, 9, 220-235.Wortman, C.B., Costanzo, P.R., & Witt, T.R. Effects of anticipated performance on the attributions of causalityto self and others. Journal of Personality and Social Psychology, 1973, 27, 372-38l.