The validity of direct versus semi- direct oral tests

The article reports the results of a study which examined the validity of two oral language tests - one direct and one semi-direct. While concurrent validity of the two types of tests (as obtained via correlations in a number of earlier studies) was high, a variety of qualitative analyses indicated that the two tests differed in a number of aspects. Differences were found in the elicitation tasks (in the number and types of functions and topics) and in the language samples obtained (in communicative strategies and in the discourse features). Conclusions are drawn regarding the need to validate tests from multiple perspectives.