Abstract

The functioning of a vocabulary testing instrument rests in part on the test-taking actions made possible for examinees by item format, an aspect of test development that warrants consideration in second-language vocabulary research. For example, although iterations of the written receptive vocabulary levels test (VLT) have integrated improvements in lexis sampling and distractor-item creation (i.e., Beglar & Hunt, 1999, Nation, 1983, 1990, Schmitt et al., 2001, Webb et al., 2017), its clustered form-meaning matching format has remained fundamentally unchanged. This study qualitatively explores the influence of this test item format on test-taking actions observed when taking the updated vocabulary levels test (UVLT, Webb et al, 2017). Data from a think-aloud protocol and retrospective interviewing indicated the predominant use of test-taking strategies for answering test items on the UVLT, such as bidirectional matching and elimination of cluster options, and that these actions enabled correct responses for clusters of target vocabulary about which the test taker demonstrated partial or even no knowledge. This evidence at the interface of test taker and test draws attention to the interconnection of estimating learners’ vocabulary knowledge and the action possibilities provided by item format on vocabulary tests. Such affordances are hierarchically structured in a proposed Taxonomy of Test-taking Actions Afforded by Receptive Vocabulary Test Format as a heuristic to evaluate the influences of test format on written receptive vocabulary assessment.

Suggested citation