A tester’s tale: The process of constructing reliable and valid EAP reading tests

Abstract

For many years, universities have provided English for Academic Purposes (EAP) to assist international students in their bid to enter degree programmes. While students have the option to sit standardised international tests like IELTS or TOEFL, some students prefer doing a course rather than a one-off test. This in turn has led to EAP programmes needing to write in-house tests.

Recent research (Huang, 2018) has highlighted the problems that EAP practitioners have in terms of their experience, training, and perceived skills in EAP testing and assessment. An added issue is their dual role as teachers and assessors. This is true for those in the secondary school sector as well (Scarino, 2013).

The ‘construct of reading’ and its importance as a skill to successfully complete university courses, has prompted much inquiry into how it should be assessed (Alderson, 2000; Grabe, 2009; Khalifa & Weir 2009). When EAP practitioners have to assess reading, they can do this via proficiency tests, diagnostic/placement tests, classroom-based assessments, and assessment for learning purposes (Grabe, 2009; Black & Wiliam, 2006). In the context of this seminar, the reading tests were written for proficiency purposes.

This seminar will present the process and findings of constructing reading tests in an intensive university EAP pre-degree programme. The time period covered is about 15 years. The presentation will focus on the process of upskilling staff, writing test specifications, test item writing, and finally the changes to the tests in both test construct and administration as informed by research.