Key points are not available for this paper at this time.
Differences in attitude structure as a function of political involvement are examined using a structural equation procedure with latent variables. In addition, criticisms of an earlier Judd and Milburn (1980) paper are examined. Findings suggest that the less involved are less likely to express political attitudes and that when they do their expressions contain more residual variance or error. At the same time, the attitudes of both the involved and uninvolved show evidence of a single underlying ideology that is remarkably stable in all groups. In arriving at these conclusions, a number of points are made concerning the procedures to be followed in comparing latent variable models between groups.
Judd et al. (Thu,) studied this question.