Chapter 18 Rotter and Mischel: Cognitive Social Learning Theory 551one way of defining personality, but it is a sterile taxonomy that fails to explain
behavior (Mischel, 1990, 1999, 2004; Mischel et al., 2002; Shoda & Mischel, 1998).
For many years, research has failed to support the consistency of personality
traits across situations. Hugh Hartshorne and Mark May, in their classic 1928
study, found that schoolchildren who were honest in one situation were deceitful
in another. For example, some children would cheat on tests but not steal party
favors; others would break rules in an athletic contest but not cheat on a test. Some
psychologists, such as Seymour Epstein (1979, 1980), have argued that studies
such as Hartshorne and May’s used behaviors that are too specific. Epstein con-
tended that, rather than relying on single behaviors, researchers must aggregate
measures of behavior; that is, they must obtain a sum of many behaviors. In other
words, Epstein would say that even though people do not always display a strong
personal trait, for example, conscientiousness, the sum total of their individual
behaviors will reflect a generally conscientious core.
However, Mischel (1965) had earlier found that a three-person assessment
committee, which used aggregated information from a variety of scores, could not
reliably predict performance of Peace Corps teachers. The correlation between the
committee’s judgment and the performance of the teachers was a nonsignificant
0.20. Moreover, Mischel (1968) contended that correlations of about 0.30 between
different measures of the same trait as well as between trait scores and subsequent
behaviors represented the outer limits of trait consistency. Thus, these relatively
low correlations between traits and behavior are not due to the unreliability of the
assessment instrument but to the inconsistencies in behavior. Even with perfectly
reliable measures, Mischel argued, specific behaviors will not accurately predict
personality traits.
Person-Situation Interaction
In time, however, Mischel (1973, 2004) came to see that people are not empty
vessels with no enduring personality traits. He acknowledged that most people have
some consistency in their behavior, but he continued to insist that the situation has
a powerful effect on behavior. Mischel’s objection to the use of traits as predictors
of behaviors rested not with their temporal instability but with their inconsistency
from one situation to another. He saw that many basic dispositions can be stable
over a long period of time. For example, a student may have a history of being
conscientious with regard to academic work but fail to be conscientious in cleaning
his apartment or maintaining his car in working condition. His lack of conscien-
tiousness in cleaning his apartment may be due to disinterest, and his neglect of
his car may be the result of insufficient knowledge. Thus, the specific situation
interacts with the person’s competencies, interests, goals, values, expectancies, and
so forth to predict behavior. To Mischel, these views of traits or personal disposi-
tions, though important in predicting human behavior, overlook the significance of
the specific situation in which people function.
Personal dispositions influence behavior only under certain conditions and in
certain situations. This view suggests that behavior is not caused by global personal
traits but by people’s perceptions of themselves in a particular situation. For exam-
ple, a young man who typically is very shy around young women may behave in