Applications of Item Response Theory to Analysis of Attitude Scale Translations

摘要

Methods of detecting item bias developed from a logistic item response theory (IRT) are generalized to analyze the fidelity of foreign language translations of psychological scales. These IRT methods are considered as alternatives to traditional sample dependent methods. Transformed item charcteristic curves generated in the original and target languages, rather than item parameters from two languages, are examined for significance of differences. Data from a Spanish translation of the Job Descriptive Index are used to illustrate the method. It is argued that equivalent item characteristic curves across the original and translated items of a scale produce equivalent instruments in both languages, and nonequivalent item characteristic curves pinpoint differences between the two versions of the scale. (Author)