Spontaneous Trait Generation: A New Method for Identifying Self-Schemas


  • Study 1 was conducted by Rebecca L Goss under the direction of the first two authors partial fulfillment of the requirements for the Master of Arts degree in psychology at Mount Holyoke College. The research described in Studies 2 and 3 was conducted under the direction of the first two authors by Adrienne Weible and Lynn Letoumeau, respectively, as part of their senior honors theses at Mount Holyoke College

  • Portions of this manuscript were presented at the 92nd annual meetmg of the American Psychological Association, August 1984, in Toronto, Canada, and at the meetmg of the New England Social Psychology Association, April 1986, University Massachusetts, Amherst, MA. The authors wish to thank the students who helped with various aspects of the research, including recruiting and running SubjectS and assisting with data analysts and clerical tasks. We are grateful to Paul Barrows, Robin Bliss, Lauren Brier, Pat Ferrie, Nina Gladir, Veronica Harris, Betsy Harrison, Elizabeth Hilty, Ann Liebergesell, Jennifer McGonigal, Nancy Smith, Jeanne Soderberg, Nina Szap, and Anne Valk for their help with these tasks. We would also like to thank Gail Hornstein and Mary Potter for their comments on an earlier version this manuscript.

concerning this article should be addressed to Francine M Deutsch, Department of Psychology and Education, Mount Holyoke College, South Hadley, MA 01075


ABSTRACT In three studies subjects spontaneously generated lists of self-descriptive traits using a new measure of self-schemas the trait generation task In Study 1 subjects had higher ratings, faster reaction times, and better recall to self-generated traits than to traits generated by another subject, demonstrating that the trait generation task does identify self-schemas In Study 2 subjects generated a list of traits characteristic of their self-ideal They had faster reaction times to self-ideal traits than to traits generated as part of another subject's self-ideal, suggesting that traits in the ideal self are also represented as self-schemas In Study 3 freshman subjects generated self-descriptive traits early in the academic year and again several months later Comparisons of their reaction times and ratings at the two points in time suggest that during a time of transition self-knowledge develops from an undifferentiated to a more differentiated form Because the identification of self-schemas using the trait generation task, unlike standard measures, is not confounded with extremity of self-rating, it was possible to show that the effects obtained for self-schemas could not be attributed to self-perceived extremity in a particular domain