Effects of examinees’ perceptions of interlocutor proficiency on paired oral assessment results

Abstract

This study aimed to explore the effects of examinees’ perceptions of interlocutors’ proficiency on their test results in paired oral tests. A group of 22 Korean English as a foreign language (EFL) high school students participated in the study. They were divided into three groups based on high, intermediate, and low English proficiency and were interviewed twice: once before the test and once after. Each examinee was paired with and tested with three interlocutors with high, intermediate, and low English proficiency levels. Raw scores from each rater were analyzed to examine potential interactions among the factors using the bias analysis from the Many-Facet Rasch Model. The results showed that the examinees’ perceptions about the effect of interlocutor proficiency on their test results were not consistent nor reliable enough to influence the test results. This study provides evidence that the paired format can be used as a valid and reliable tool to assess EFL learners’ speaking ability regardless of interlocutor proficiency levels, which can be assumed to influence examinees’ perceptions of the oral test results.