Concurrent and Predictive Validity of Computer-adaptive Freshman English Test for College Freshman English in Taiwan

Abstract

Undoubtedly, it is always a continuing effort for colleges and universities to seek a reliable assessment tool for placing students into appropriate college Freshman English levels in order to maximize its instructional effectiveness. The assessment tool, however, is required of fulfilling cardinal principles of test validity, reliability, and practicality. This study was designed to determine the level of criterion-related validity in terms of concurrent and predictive validity of a computer-adaptive Freshman English Test (CAFET) employed as a Freshman English placement test at a university in central Taiwan. Two other English proficiency tests, the National Joint College Entrance Exam on English Proficiency (NJCEEE) and Taiwan’s General English Proficiency Test Intermediate Level (GEPT-I1) were used as criterion variables for measuring the CAFET’s concurrent validity. In addition to evaluating the CAFET’s internal quality, individual and/or combined predictive power of the CAFET, NJCEEE, and GEPT-I1 for Freshman English course achievement of participants of School Year 2010-2011 was also investigated. Statistical procedures of Pearson’s Product Moment correlation coefficients, one-way ANOVA, and simple and multiple regressions were performed. The results supported usability of the CAFET as an appropriate tool to place students into proficiency levels of the Freshman English program, especially when practicality is a prime consideration. Implications and suggestions for future research are presented.