ERIC Number: ED122603
Record Type: Non-Journal
Publication Date: 1975-Jun-27
Reference Count: N/A
Validity in Proficiency Testing. ELT Documents (75/3).
Some validity criteria were applied to the language proficiency measures used by the British Council, particularly the "Davies Test." Five types of test validity are identified: construct validity; content validity; concurrent validity compared with another test; prediction validity; and face, or apparent, validity. Foreign students studying in Britain were tested for proficiency in English for academic and personal uses. In addition, students' tutors were asked for assessments, and the tester evaluated students' oral and written communicative skills and comprehension. Correlations were made between students' performance assessment prior to departure for Britain and tutors' ratings after arrival, and the first assessment was found about 85 percent accurate. It was concluded that the Davies Test, despite weaknesses of construct, is essentially a valid instrument. Subjective assessments, especially when rigorously administered, were generally valid. Students all improved when in Britain and most passed their threshold of minimum necessary proficiency. It is proposed that a carefully constructed and tested proficiency measure is likely to possess overall validity even if some parts are deficient. (CHK)
Descriptors: Adult Students, English (Second Language), Foreign Students, Language Proficiency, Language Skills, Language Tests, Measurement Instruments, Performance Tests, Test Construction, Test Validity
English Teaching Information Centre, The British Council, 10 Spring Gardens, London SW1A 2BN, England
Publication Type: Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: British Council, London (England). English-Teaching Information Centre.