Cross-Cultural Validation of Item Complexity in a LLTM-Calibrated Spatial Ability Test

Abstract

In the “Linear Logistic Test Model” (LLTM; Fischer, 1983 ), item difficulties of intelligence/ability test items can be modeled in terms of hypothesized facets of complexity. This feature of the LLTM makes it possible to classify cross-cultural bias on a conceptual level that is beyond simple detection of DIF items. In the present study, a spatial ability test, the “Three-Dimensional Cube Comparison Test” (3DC; Gittler, 1990 ), was administered to university students, 384 from the United States and 307 from Austria. Compared to subjects from Austria, US students worked faster but scored considerably lower overall. It is suggested that these differences in test-taking behavior were probably caused by motivational factors. After differences in motivation were accounted for, responses were analyzed using the LLTM and no cross-cultural differences were found in the hypothesized complexity facets of the 3DC items. However, a sufficient number of warm-up items is crucial to ensure the cross-cultural equivalence of the 3DC test.