When foreign language programs have had problems administering a face-to-face oral interview due to time constraints, logistics, and training for interviewers and raters, one solution was to utilize technology. The Oral Simulated Interview System (OASIS) for Japanese was designed for eliciting speech samples in order to assess speaking ability that meets the specific syllabus objectives including the master of linguistic features and content. This study examined its methodological comparability with the face-to-face oral interview in a classroom testing situation.
OASIS and the face-to-face oral interview were administrated to thirty stratified subjects from 125 university students enrolled in the second semester of a first-year Japanese course. Thirty subjects were required to take both tests in sequence. Then, sixty speech samples were scored by two raters. In addition, questionnaires were distributed to examinees and raters concerning their attitude toward both tests.
The study found that interrater reliability was high, while correlation between the two tests was moderate. While both examinees and raters noted the lack of interaction on OASIS, examinees felt less nervous and more in control of time on OASIS. Also, raters were positively impressed by the efficient rating procedures on OASIS.
The results claimed OASIS did not involve authentic human interaction, as the face-to-face oral interview did; however, it could be a practical alternative when the face-to-face oral interview was not feasible. Based on these findings, the relevance and implications for teaching Japanese in a tertiary setting will be addressed.