Criterion-related Validation Of CET-6

CET-6(College English Test-Band6), a dominant English test for non-Englishmajors in China, is widely accepted to assess learners’ English ability by learnersthemselves, school teachers, and employers in society. Since CET reform began in2005, a great many efforts have been invested into test validation, to ensure the highquality of the test. But to roughly review the related studies, most of them concentrateon content validity, some referring to construct validity and face validity, withcriterion-related validity not being sufficiently investigated. Therefore, I set thestandpoint on criterion-related validity to validate CET-6, and choose IELTS (theInternational English Language Testing System), a world’s proven English languagetest with high recognition, as the criterion.Different from many domestic comparative studies of two tests, this studyfocuses mainly on test results, i.e. authentic scores of CET-6and the IELTS, andfollowed by a qualitative analysis. With a sample of53sets of scores on the both tests,statistical analysis is conducted to establish the score relationships of the two tests,including total scores and reading section scores. It is found that there is a strongrelationship of total scores on CET-6and the IELTS, but a relatively low relationshipof the reading section scores. Score linking of total scores has been further providedand the factors that lead to the lack of correspondence on reading section have beenexplored in the follow-up comparative analysis.The obvious consistency of total score on the two tests can strengthen theconfidence for this large standardised domestic English test. However, the weakcorrelation of reading scores indicates that validation and examination of readingsection is essential. It is hoped that the results of this study will provide implicationsfor test constructors, test users, and researchers.