Responses on the TOEFL© test may reflect both the influence of the examinees' native language and their level of English proficiency. The aim of this study was to appraise the effect of these examinee variables on the structure of the test. The interrelations among TOEFL items, using all of the information provided by the various responses to the items (the four alternatives, omitted, and not reached), were analyzed by three‐way multidimensional scaling for samples of examinees systematically varying in native language and level of English proficiency. Four dimensions were identified: three corresponded to the sections of the test, and the fourth was an end‐of‐test phenomenon. The dimensions were predominantly defined by easy items and were most salient for low‐scoring examinees. The salience of the dimensions did not differ for the various language groups, except for the end‐of‐test dimension. Major conclusions were that the TOEFL's construct validity is supported, the test's interpretation varies with the examinees' English proficiency, easy and difficult items differ in their potential for diagnosis and global screening, and the dimensionality of the TOEFL and of competence in English depend on the examinees' English proficiency.

Multidimensional scaling was used to analyze item response data for the Test of English as a Foreign Language (TOEFL) to uncover the dimensions underlying the test. Four dimensions were identified for samples varying in native language and level of English proficiency: 3 corresponded to the test's sections and 1 was an end-of-test phenomenon. Dimensions were predominantly defined by easy items and were most important for low-scoring examinees. The dimensions' importance did not differ across language groups, except for the end-of-test dimension. Major conclusions were that (a) the TOEFL measures the intended constructs; (b) the test assesses the same constructs in each language group, but the constructs are more differentiated for low-scorers; and (c) easy and difficult items differ in what they measure. Multidimensional scaling appears to be a useful method for item-level analyses of test structure.

No abstract

No abstract