Test-Takers' Performances on and Perceptions of Two Different Modes of Online Speaking Tests

A direct test of English speaking is important to evaluate what learners can do in real-life situations. However, due to challenges in test administration, especially with a large number of test-takers, a direct speaking test may not be feasible in many contexts and thus indirect tests, such as conversational cloze tests, are mainly used. In response to this problem, this study utilized communication technology to create speaking tests with two different delivery modes: Real-Time Interview with a human interviewer (RTI) and Pre-Recorded Video (PRV). The tests were given to a group of 40 first-year university students to complete, followed by a perception questionnaire and a group interview to collect data about test-takers' perceptions of the tasks. Results showed that the participants performed significantly better on the PRV test tasks and they perceived both tasks positively. The strongest quality of both test tasks, as perceived by the participants, was authenticity. While the RTI tasks were perceived to significantly have more impact and interactiveness than the PRV tasks, the test-takers shared in the interview that they felt more comfortable and less anxious while completing the PRV tasks.