Reliability of classroom-based assessment as perceived by university managers, teachers, and students

Abstract

Classroom-based summative tests are frequently developed and implemented by university teachers who, in the majority, lack the necessary language assessment literacy to cope with this task in a highly professional manner. In universities without test development units, decision makers do not question the test scores obtained on teacher-constructed tests, and invalid tests cannot serve reliable benchmarks of students’ learning outcomes. Nonetheless, urged by accountability, university managers have to disregard the possibility of reporting misinformed decisions. The authors of this chapter presents the research findings about assessment practices in the universities in Ukraine with the focus on summative test reliability issues. The collected data are based on surveying the perceptions of university managers heading foreign languages departments, teachers working for such departments, as well as students learning languages in several universities across the country. As a result of data interpretation, the most common procedural aspects of test construction and administration as perceived from the three perspectives of those engaged in language instruction were established. The implications of the study outline ways to enhance the quality of teacher-constructed tests and ensure reliability of summative assessment.