Drawing on repeat test takers to study test preparation practices and their links to score gains. (October 2020)

• Record Type: Journal Article
• Authors: Knoch, Ute; Huisman, Annemiek; Elder, Cathie; Kong, Xiaoxiao; McKenna, Angela
• Other Names: Green Anthony guest-editor.; Van Moere Alistair guest-editor.
• Is Part Of: Language testing. Volume 37:Number 4(2020)
• Page Start: 550
• Page End: 572
• Publication Date: 2020-10
• Subjects: Repeat test takers -- test impact -- test preparation -- validation -- washback; Language and languages -- Ability testing -- Periodicals; Language and languages -- Examinations -- Periodicals; 407.6
• DOI: 10.1177/0265532220927407
• Languages: English
• ISSNs: 0265-5322

Abstract

A key concern of washback research in language testing is with the value of test preparation for facilitating learning and improving test performance. Although test takers may draw on a wide range of preparation activities, the majority of research studies examining test preparation have taken place in classroom settings, leaving self-access approaches largely unexamined. The aim of the current study was to (a) explore possible links between self-access test preparation activities and improved test performance and (b) examine how repeat test takers adjust their test preparation activities from test sitting to test sitting while preparing for the Pearson Test of English (Academic). The study involved the collection and analysis of interviews from 60 recent repeat test takers. The interview data were coded for themes and sub-themes and analyzed for the kind of test preparation activities in which learners engaged, and how these changed over time. The interviews showed that the test takers were strategic in their preparation, by changing their approaches depending on their previous test results. The largest number of significant improvements was identified for speaking, where test takers engaged in a variety of strategies, some of which were construct-irrelevant. The findings are discussed in relation to test validity and washback.