SOCIAL REFERENTS AND SELF EVALUATION IN EXAMINATIONS.
Technical rept. no. 7 on Dynamics of Social Influence,
CALIFORNIA UNIV LOS ANGELES
Pagination or Media Count:
Students were asked after each of three examinations to estimate 1 what score they would obtain on that examination, 2 their minimal goal the lowest score with which they would be satisfied, 3 the average score obtained by students in that class. Data from 562 respondents are analyzed to determine the bases for self-evaluation. Before making their estimates on each of the exams, 272 students in a Referent condition were given the average score of male pre-medical students. The remaining 290 respondents, in the No Referent condition, were not given any such information. With no social referent, and no experience on the first examination, correlation between predicted and obtained scores is very low r .21. Accuracy of prediction increases on successive examinations, reaching a correlation of .42 on the third examination. Where there is no social referent and no experience, there is autistic distortion--a tendency to overestimate ones performance. Such distortion is particularly evident among students who score poorly. Autism is reduced on successive examinations, and particularly when a social referent is provided. The social referent results in much greater accuracy in the first examination, compensating for lack of experience. Other factors in self evaluation and evaluation of others are also examined. Author