Validating a CEFR-informed Open Access English Placement Test

点击翻译按钮获取中文摘要

In 2022 a slightly revised version of the course placement test with a 40-minute time limit was trailed. 247 of the Japanese university students who took the test on their smartphones through the Moodle LMS, gave permission for their test results to be used for research purposes. As with the first pilot, strong statistical evidence from a Cronbach's alpha of 0.9 and a Rasch person separation statistic of 2.91 was found to support the test function of dividing students into three ability groups for course placement purposes. However, item analysis revealed some problems with item quality and with the ordering of test items when ranked by CEFR-J wordlist levels. Therefore, further revision and creation of test items is needed before collecting criterion validity data and setting cut scores.