Revisiting the Middleton Alienation Scale: In Search of a Cross-Culturally Valid Instrument
The article contributes to the issue how to deal with measurement non-variance. I address a well-established scale by Middleton (1963) which was created to measure alienation. However, unlike commonly treated in literature, there is evidence that the scale is two-dimensional, and consists of the measures of anomie and alienation. I use the data from two datasets, where the scale was most recently applied, World Values Survey (2011), and Euromodule (1999-2002), for a set of diverse countries representing Western, post-communist, and Eastern societies. The datasets are analyzed separately. Results of confirmatory factor analysis followed by multigroup comparisons give evidence that for a set of countries the two-dimensional scale is applicable or preferable. Full measurement invariance is reached for Russia and Kazakhstan in the World Values Survey, and Slovenia, Switzerland, Germany, Austria, Turkey, and South Korea in the Euromodule. Approximate measurement invariance using Bayesian statistics enabled to reach approximate scalar invariance in Russia and Kazakhstan in the World Values Survey and in Slovenia, and Switzerland in the Euromodule dataset. Additionally, to be sure that the two dimensions are indeed distinct, I used a set of indicators to predict each of the factors.