A cross-cultural comparison of item analysis data on the revised ITPA

Abstract

Item difficulty levels were determined for seven of the subtests of the revised ITPA using a preliminary normative sample of Australian children. It was clear from the results that uneven grading of difficulty levels led to problems of premature application of ceilings on some subtests. Similar analyses then were carried out on the original North American standardisation data, thus allowing direct comparisons of facility indices. The results highlighted a number of cross-cultural differences in item difficulty levels, thus emphasizing the need for considerable caution when using the revised ITPA in Australia.