The Difference between Attitudes and Nonattitudes in the Mass Public: Just Measurements?

In: American journal of political science: AJPS, Band 29, Heft 2, S. 291-307

Abstract

Two explanations have been proposed in the literature for the instability & low coherence of responses to policy issues in mass surveys: the fallibility of measurement instruments & the prevalence of nonattitudes. These explanations are tested with data from a survey of 184 adult residents of Long Island, NY, conducted Apr, 1980. Multiple magnitude scaling measures of attitudes are employed & a measurement model with latent variables is proposed. Response reliabilities are estimated jointly for 4 policy issues & for party identification & liberal/conservative self-location, using LISREL. Both the response reliabilities for the items & the correlations among the latent variables are higher in the sophisticated half of the sample than in the unsophisticated half. These findings support the nonattitudes hypothesis, though not ruling out the existence of defects in the measurement instruments. 3 Tables, 2 Figures, 36 References. Modified HA