The Structure and Measurement of Social Identity

Abstract
Drawing on the work of the Survey Research Center of the U of Michigan on party identification (A. Campbell, G. Gurin, & W. Mitter, THE VOTER DECIDES, Evanston, Ill: Row, Peterson, 1954, & A. Campbell, P. Converse, W. Miller, & D. Stokes, THE AMERICAN VOTER, New York, NY: Wiley, 1964) & the Likert approach to attitude measurement, a procedure is developed for measuring the concept of social identity among a number of individuals with a single instrument. Results of the Twenty Statements Test research were used to compile a wide-ranging sample of 40 stimulus groups (H. Mulford & W. Salisbury, II, “Self-conceptions in a General Population,” in SYMBOLIC INTERACTIONS; A READER IN SOCIAL PSYCHOLOGY, J. G. Manis & B. N. Meltzer, Eds, Boston, Mass: Allyn & Bacon, 1967). A sample group of 146 Uc & graduate students at Temple U, Philadelphia were given the questionnaire during a class period. The sexual, racial, & religious make-up of the sample was calculated. Pearson product-moment correlations were measured from the raw data matrix. To extimate the N of meaningful factors present, the Cattell eigen value technique was used. The r matrix was factored & refactored with an iteractive principal axis routine to determine the 7 factors: ‘Jewish’ identity, ‘black M’ identity, ‘sex’ identity, ‘respectable member of the community’ identity, ‘upper status’ identity, ‘blue-collar Roman Catholic’ identity, & ‘white race’ identity. All proved reliable under Cronbach’s alpha test for measurement equivalence, except the ‘blue-collar Roman Catholic’ identity. Applications of the instrument to the study of policy compliance & group support are suggested. 3 Tables. Modified HA.