This article presents and discusses a number of empirical findings concerning the validity of multiple-choice cloze tests as tests of general language proficiency. Foreign students of Dutch at Delft University of Technology were given both a series of separate proficiency tests in listening, reading, speaking, writing and comprehension, and a series of multiple-choice cloze tests. Scores on the multiple-choice cloze tests were found to correlate significantly with each of the proficiency tests. In addition, scores on the multiple-choice cloze test appeared to form a solid basis for prediction of the total scores for Iistening, speaking, reading and writing taken together. Finally, a close, structural similarity was found to exist between curves of average scores on successive multiple-choice cloze tests and the presupposed growth of vocabulary during the language learning process. Together, these findings tend to show that multiple-choice cloze tests constitute a valid instrument for measuring general language proficiency.
|Journal||Tijdschrift voor Onderwijsresearch|
|Publication status||Published - 1987|