Topic development in the English oral proficiency interview for international teaching assistants at the University of Illinois at Urbana-Champaign

The present study examines how a topic is initiated and is developed and expanded for additional questions in order to probe an International Teaching Assistant (ITA) candidate’s oral proficiency in an unscripted interview test. The University of Illinois at Urbana-Champaign (UIUC) had employed SPEAK (Speaking Proficiency English Assessment Kit), a computer-mediated test, to assess ITA candidates’ speaking ability for decades. However, with increasing criticism about the SPEAK in terms of face validity, authenticity and technical problems, a new oral test for ITAs was developed. The new test, the English Proficiency Interview (EPI hereafter), began to be administered in the Summer of 2010, and is conducted via the format of an unscripted interview. The EPI does not provide an interviewer with pre set questions in advance. Topics are initiated spontaneously the on test site and negotiated in the course of discourse. Questions are formulated based on the response that the test taker had provided during the interview. Thus, each interview in the EPI consists of an individualized form of discourse and has its own path in topic development. Findings of discourse analysis are presented using 7 EPIs, which were selected from interviews gathered in the first field trial of the new test in 2009. The interview process is illustrated with excerpts taken from actual interviews. Analysis of the data revealed that the interviewer initiated a topic in the following category: a) initiating a topic from the candidate’s field of study, and b) initiating a topic from the candidate’s interest and life in general. In addition, some features of both conversation and interview are discussed as the EPI is found to have both. Finally, future studies are suggested with reference to topic development across test takers’ language proficiency level. This study is primarily descriptive and suggestive rather than conclusive