Abstract:

Placing students into groups according to their corresponding proficiency level in a foreign language allows to create equal learning conditions and, in its turn, reduction of student and faculty frustration and achievement of better learning outcomes. Since the University embarked on online English placement test, we have been working on its effective version. This study is the continuation of a preceding research presented in INTED2018. That study brought test limitations to light and suggested some ideas for improvement. The results were reported to the university administration who took measures on redesigning placement test construct and content. Current research covered a much greater number of students (596) and allowed to compare the performance of two testee groups on the same placement test but with different test settings (venue, instructions, time, etc). In one group in comparison with the previous year the number of students whose level of language proficiency according to Common European Framework of Reference (CEFR) was determined objectively increased by 15%. We assumed this accuracy to be attributed to ideal test setting conditions. The hypothesis was rejected because in the second group we didn’t experiment with test setting, but the accuracy of level determination was even higher. Statistical data and their analysis led to some presuppositions, their rejection/confirmation and conclusions. First, the majority of students accessed the test from their mobile devices which prevented them from doing some listening tasks due to unfriendly graphical user interface. Second, check of test reliability and content validity proved to work well for A0-A2 levels (CEFR). Because the distribution of A0-C1 levels wasn’t even in two groups, the accuracy statistics didn’t look credible. Third, test content rather than test settings is more important for level determination. Findings led to some further modifications and improvements of existing test.

Keywords:

English Placement Test (EPT), online education, CEFR.