Culturally comparative studies about test anxiety often use adaptations of the two-dimensional Test Anxiety Inventory (TAI; Spielberger, 1980). To capture multiple facets of test anxiety, however, no multidimensional instrument with transcultural validity exits. In response, the current study reports on the cross-validation of an English version of the four-dimensional TAI-G, which had been introduced as an extended German adaptation of the TAI (Hodapp, 1991, 1996). First year university students of a multi-ethnic South African (N=102) and German sample (N=183) were given parallel versions of the TAI-G in English and German. Multigroup confirmatory factor models were conducted to examine structural and parametric invariance. Indicating high construct validity, a second-order model of test anxiety fits the data well for both samples. Associations with test anxiety correlates suggested high criterion validity. Scale structure and psychometric properties of the English TAI-G replicated findings of German samples suggesting transcultural applicability of the measure.
KEYWORDS: test anxiety, cross-cultural measurement, construct validity, multigroup analysis.