Working paper on stereotype vulnerability and attribution theory


Legal education generally calls upon and values a narrower range of intellectual capacities and problem-solving styles than are needed to do the work of lawyers. Similarly, legal education calls upon and values a narrower range of intellectual capacities and problem-solving styles than law students generally possess.

We posit that some law students, as a matter of self-concept, identify readily with ways of working that are prized and explicitly developed in law schools. By the same token, other law students disidentify with ways of working that are dominant in legal education. We believe that this disidentification is due, at least in part, to “stereotype vulnerability” —a disabling anxiety triggered by being called upon to perform tasks one is expected, as a result of group membership, to perform poorly. This second group of students is likely, we believe, to identify with ways of working that are unexplored in legal education but central to legal practice. In an educational context in which students are usually assigned tasks to which they are stereotype vulnerable, these students will have few opportunities to perform well. In an educational context in which stereotype vulnerable students are asked to perform tasks with which they identify comfortably as well as tasks as to which they are stereotype vulnerable, they should be more confident and better able to take on anxiety-provoking challenges. At the same time, they should be able to take leadership roles in the domains in which they work most comfortably. Our project is principally designed to make all students more versatile by developing the full range of skills necessary to effective and responsible practice. For the second group, described above, the project may have the additional benefit of alleviating stereotype vulnerability.


Attribution theory

Attribution theory refers to a varied body of scholarship exploring the causes that individuals attribute to various events. Extensions of the basic theory suggest that when subjects perform challenging tasks in a domain in which they expect to perform poorly, they attribute their difficulties to inability, develop a sense of hopelessness with regard to their ability to perform in the domain in question, and perform poorly. Early work on attribution in achievement contexts focused on locus of control: causes were said to be either internal, as with ability or effort, or external, as with luck or difficulty of task, and individuals were thought to make internal or external attributions with consistency (Rotter, 1966). Subsequent research suggested that causes are more usefully described along multiple dimensions and that they are likely to be context dependent, rather than characteristic of individuals’ personalities. Weiner’s comprehensive model of attribution (1971) suggests that the locus, stability, and controllability of attributions affect associated emotions as well as subsequent expectations and performance. For example, when poor performance on a test is attributed to an internal, stable, and controllable cause such as lack of preparation, the poor performance may induce shame and study. However, when poor performance is attributed to an internal, stable, and uncontrollable cause such as lack of ability, the poor performance may induce discouragement and lack of effort.

Subsequent research focused on the relationship between expectations and attributions. The general model proposes that when individuals’ expectations are met, they attribute their performance to stable and internal causes, such as ability, but when individuals’ expectations aren’t met, they attribute their performance to unstable causes such as luck or lack of effort (Feather & Simon, 1971, 1973; Gilmor & Minton, 1974). In a study demonstrating the general model, Feather (1969) gave 167 college students an anagram test. Before the test, students rated their confidence from low to high on a 5-inch scale; afterwards they made attributions about their performance on a 5-inch scale with luck at one end and ability at the other. Feather found that students who were confident and passed and students who were unconfident and failed were likely to attribute their performance to ability, while students who were confident and failed and students who were unconfident and succeeded were likely to attribute their performance to luck. Feather also found that the female students were more likely to have lower expectations than the male students and that the female students who succeeded were more likely to attribute their success to luck than were the male students who succeeded.

In a further exploration of the effects of gender on expectations and attributions, Deaux and Farris (1977) gave an anagram test to groups of males and females under two conditions. In the first condition, the test was described as one which men typically perform more successfully. In this condition the male participants had higher expectations of success, evaluated their performances more positively, and attributed their performances to ability more often than did the female participants. However, when the test was described as one which women typically perform more successfully, male and female participants did not differ with regard to expectations or evaluations of performance, and female participants had only a slightly greater tendency to attribute their performance to luck.

Based on the findings of this study and others (Feather & Simon, 1975; Frieze, Fisher, Hanusa, McHugh & Valle, 1978; Rosenfeld & Stephan, 1978), Deaux (1976) concludes that gender differences in expectations and attributions arise primarily from stereotypes about tasks. Given that women generally have low expectations for success when completing “male” tasks, they tend to attribute their success to luck and to attribute their failure to lack of ability. However, men, who generally expect success, often attribute their failure to luck and their success to ability. Additional research suggests that the same biases about tasks and expectations hold for observers (Deaux & Emswiller, 1974; Hanson & O’Leary, 1983). For example, observers attribute expected success to ability and unexpected success to luck. Thus, when women succeed at tasks they are stereotypically expected to perform poorly, observers attribute their success to good luck.

Research at the University of Pennsylvania Law School (Guinier, Fine, & Balin, 1994) resonates with research linking expectations, attributions, and gender. Interviews with female students at the school suggest that unlike male students, “many women...internalize a relatively weak performance on a single exam as evidence of personal failure” (87). In one woman’s words: “Guys think law school is hard, and we just think we’re stupid” (44). Such differences in attribution may reflect differences in expectations due to stereotypes about the intellectual capacities involved in legal education.

Over time, the tendency to expect failure and to attribute it to lack of ability can produce a condition known as learned helplessness. In one study demonstrating a link between attributions and helplessness, Dweck and Reppucci (1973) asked 40 fifth grade boys and girls to arrange colored blocks to match pictures presented them by two experimenters. The success experimenter gave them pictures they could reproduce with the blocks; the failure experimenter gave them pictures they could not reproduce. After a number of trials, the failure experimenter began to present pictures which the children could reproduce. After experiencing failure, some children’s performances worsened; they began to act helplessly, as if it were not possible for them to succeed. However, another group of children persisted in their efforts and were mastery-oriented. Dweck and Reppucci also found that the helpless and mastery-oriented children differed significantly on their responses to the Intellectual Achievement Responsibility scale (Crandall, Katkovsky, & Crandall, 1965). The helpless children had a greater tendency to attribute their performance to ability, while the mastery-oriented children had a greater tendency to attribute their performance to effort.

To better understand the effects of helplessness on performance, Diener and Dweck (1978) monitored the problem-solving strategies and verbalizations of an additional sample of fifth-graders. Before attempting the experimental task, the participants responded to the IAR scale (Crandall, Katkovsky, & Crandall, 1965) and were identified as helpless or mastery-oriented given the extent to which they attributed their performance to ability or effort. The participants were then given a three-dimension, two-choice discrimination task. Before any experiences of failure, the problem-solving strategies and verbalizations of the helpless and mastery-oriented children were similar. However, following experience of failure, the helpless children began to use increasingly less effective problem-solving strategies and to verbalize solution-irrelevant statements, including attributions of their performance to lack of ability. By contrast, the mastery-oriented children verbalized self-instructing and self-monitoring statements and used more effective problem-solving strategies. When pressed by the experimenter to make attributions, the mastery-oriented children attributed their difficulties to effort, luck, the experimenter not being fair, or the difficulty of the task.

Subsequent scholarship has linked learned helplessness to individuals’ theories of intelligence. In their review of the literature, Dweck and Bempechat (1983) explain that children with entity theories believe that intelligence is a fixed commodity, while children with instrumental-incremental theories believe that intelligence can be increased through learning. Research has shown that children’s theories of intelligence have significant implications for their goals and performance (Bandura & Dweck, 1981; Elliott & Dweck, 1981). Children who believe that intelligence is an expandable resource tend to enter achievement contexts with learning goals and to be mastery-oriented whether they have high or low expectations of success. Children who think of intelligence as a fixed entity enter achievement contexts with performance goals; they want to show or avoid showing their intelligence. In situations in which they have high expectations for success and feel they can obtain a favorable competence judgment, they demonstrate mastery-orientations. However, in situations in which they have low expectations for success and want to avoid a negative judgment, they become helpless.

Notably, theories of intelligence are context specific and can be influenced by characteristics of the situation, including the theory of the teacher (Dweck, Tenney, & Dinces, 1982). Aspects of legal education, including conventional classroom discourse and the absence of periodic assessment, appear to support fixed entity theories of intelligence and performance goals. When professors ask questions and students answer, classroom discourse imitates written exams on which students are expected to perform. Graded and irregular participation and the expectation that students know the right answer the first time further encourage students to perform, rather than test the limits of their understanding. In addition, testing students once at the end of year eliminates possible connections between assessment and learning. In the absence of opportunities to work at what they don’t know, students may implicitly assume that legal thinking isn’t learned: one either “gets it” or not.

For students with low expectations for success or learning goals, such contexts may be particularly anxiety-provoking and/or painful. Consistent with this hypothesis, women at the University of Pennsylvania Law School reported “that when speaking feels like a ‘performance,’ they respond with silence rather than participation, especially when the Socratic method is employed to intimidate or to establish a hierarchy within large classes” (Guinier, Fine, & Balin, 1994, 46).


Stereotype vulnerability

Claude Steele’s work on stereotype vulnerability also demonstrates that expectations may have drastic effects on performance. In a series of creative experimental studies, Steele and his colleagues have shown that when subjects performing challenging tasks are aware that their ability is being gauged in a domain in which members of the subjects’ group are generally thought to perform poorly, they feel anxious about confirming or being judged by the stereotype, and their anxiety interferes with their performance. Steele calls this condition of anxiety “stereotype vulnerability.” Steele believes that students who are stereotype vulnerable tend to disengage from work in the anxiety-provoking domain to avoid ego assaults that may cause them to feel helplessly incapable.

In his initial research on stereotype vulnerability, Steele found that when equally able male and female students took a difficult English test, they performed equally, but when equally able male and female students took a difficult math test, the female students performed significantly less well than the male students. Steele hypothesizes that because the test was difficult, the female students experienced some frustration, lending support to stereotypes about their ability and arousing anxiety about confirming such stereotypes. In a subsequent study, when equally able male and female students took a difficult math test after having been told that it is a test on which men and women perform equally, the male and female students performed equally. In this case, Steele hypothesizes that the female participants did not invoke the stereotype to explain their frustration and were not anxious about confirming it. Similarly, black students performed less well than equally able white students on a challenging standardized test when told it was a measure of ability, but performed equally well when told performance was unrelated to ability. Steele and his colleagues also found that participants expecting to take the test which measured ability exhibited greater stereotype activation, greater concern about their performance, and a greater reluctance to have their racial identity associated with their scores. Furthermore, Steele found that stereotype vulnerability could be induced by asking students to identify their races in a pre-test questionnaire, even when they had been told that the test was not a measure of ability.

Finally, Steele argues that stereotype vulnerability is a universal phenomenon and is not associated with any particular group. In one study, white males who had been told that Asians outperformed all other ethnic groups on a challenging math test performed less well than white males who had not been told that Asians performed better on the test.

Steele’s follow-up work at the University of Michigan suggests that many of the detrimental effects of stereotype vulnerability can by offset by “wise schooling,” a term Steele uses to refer to practices which help students who are stereotype vulnerable feel less vulnerable. The 21st Century Program, which provides housing, challenging workshops, and group activities for approximately 250 first-year students, exemplifies many characteristics of wise schooling: honorific recruitment emphasizes students’ strengths; workshops advertised as challenging help students attribute failure to the difficulty of the task and attribute success to their abilities; defining ability as an expandable resource motivates students to continue working despite initial setbacks; and, group work, particularly in groups which are integrated with regard to race and gender, helps participants see that everyone has strengths and weaknesses and that problems are shared. In its first few years, participants in the program have outperformed non-participants by as much as a grade, and black participants have performed as well as white participants. Notably, preliminary data suggest that participants still feel stereotype vulnerable; Steele hypothesizes that the program works by keeping students from disengaging, despite their feelings of vulnerability.



Given the literature on stereotype vulnerability and attribution theory, we hypothesize that some students’ performance at law school is negatively affected by their own and others’ preconceptions. Our admissions policies assure that all of our students are capable of functioning at extremely high levels across the range of lawyering skills. Yet, female students and students of color may have low expectations of success and may be stereotype-vulnerable in some or all of their classes. On the other hand, women and people of color may be strong in cognitive domains that legal education tends most to neglect: interpersonal, intrapersonal, and narrative intelligence and interactive reasoning. Male and majority group students may be stereotype vulnerable with respect to these neglected domains. Indeed, we suspect that this stereotype vulnerability may account for the neglect of interpersonal, intrapersonal, narrative, and interactive cognitive processes in law school education. We hypothesize that stereotype vulnerability will be reduced for all students by work that draws upon the full range of capacities relevant to good lawyering in a context which promotes shared work and a conception of ability as an expandable resource. Few if any law students will be stereotype vulnerable with respect to all of the intellectual domains relevant to good lawyering. Successful experiences with the mixed use of relevant intellectual capacities to solve legal problems in contexts which support learning goals should serve to reduce anxiety. The confidence that comes from believing that one is, and is perceived as being, competent in certain relevant domains should facilitate engagement in the anxiety-inducing domains, allowing development of skills from which a student has tended to disengage.




Bandura, M. and Dweck, C.S. (1981). Children's theories of intelligence as predictors of achievement goals. Unpublished manuscript, Harvard University.

Crandall, V.C., Katkovsky, W., and Crandall, V.J. (1965). Children's belief in their own control of reinforcements in intellectual-academic achievement situations. Child Development, 36, 91-109.

Deaux, K. (1976). Sex: A perspective on the attribution process. In J.H. Harvey, W.J. Ickes, and R.F. Kidd (Eds.), New directions in attribution research, Vol. 1 (pp. 335-352). Hillsdale, NJ: Erlbaum.

Deaux, K. (1984). From individual differences to social categories: Analysis of a decade's research on gender. American Psychologist, 39(2), 105-116.

Deaux, K. and Emswiller, T. (1974). Explanations for successful performance on sex-linked tasks: What is skill for the male is luck for the female. Journal of Personality and Social Psychology, 29, 80-85.

Deaux, K. and Farris, E. (1977). Attributing causes for one's own performance: The effects of sex, norms, and outcome. Journal of Research in Personality, 11, 59-72.

Diener, C. and Dweck, C. (1978). An analysis of learned helplessness: Continuous changes in performance, strategy, and achievement cognitions following failure. Journal of Personality and Social Psychology, 36(5), 451-462.

Dweck, C. and Bempechat, J. (1983). Children's theories of intelligence: Their consequences for learning. In S.G. Paris, G.M. Olson, and H.W. Stevenson (Eds.), Learning and motivation in the classroom. Hillsdale, NJ: Lawrence Erlbaum Associates.

Dweck, C. and Reppucci, N.D. (1973). Learned helplessness and reinforcement responsibility in children. Journal of Personality and Social Psychology, 25(1), 109-116.

Dweck, C., Tenney, Y., and Dinces, N. (1982). Unpublished data, Harvard University.

Elliott, E.S., and Dweck, C. (1981). Children's achievement goals as determinants of learned helpless and mastery-oriented achievement patterns: An experimental analysis. Unpublished manuscript, Harvard University.

Feather, N.T. (1969). Attribution of responsibility and valence of success and failure in relation to initial confidence and task performance. Journal of Personality and Social Psychology, 13, 129-142.

Feather, N.T. and Simon, J.G. (1971). Causal attribution for success and failure in relation to expectations of success based upon selective or manipulative control. Journal of Personality, 39, 527-541.

Feather, N.T. and Simon, J.G. (1973). Causal attributions for success and failure at university examinations. Journal of Educational Psychology, 64, 46-56.

Feather, N.T. and Simon, J.G. (1975). Reactions to male and female success and failure in sex-linked occupations: Impressions of personality, causal attributions, and perceived likelihood of different consequences. Journal of Personality and Social Psychology, 31, 20-31.

Fiske, S. and Taylor, S. (1984). Social Cognition. Reading, MA: Addison-Wesley.

Frieze, I.H., Fisher, J.R., Hanusa, B.H., McHugh, M.C., and Valle, V.A. (1978). Attributions of causes of success and failure as internal and external barriers to achievement. In J.L. Sherman and F.L. Denmark (Eds.), The psychology of women: Future directions in research (pp. 519-552). New York: Psychological Dimensions.

Gilmor, T.M. and Minton, H.L. (1974). Internal and external attribution of task performance as a function of locus of control, initial confidence, and success-failure outcome. Journal of Personality, 42, 159-174.

Guinier, L., Fine, M., and Balin, J. (1994). Becoming gentlemen: Women's experiences at one Ivy League law school. University of Pennsylvania Law Review, 143(1), 1-110.

Hanson, R.D. & O'Leary, V.E. (1983). Actresses and actors: The effects of sex on causal attributions. Basic and Applied Psychology.