Development and validation of a scale to measure test-wiseness in EFL/ESL reading test takers

Researchers investigating how EFL/ESL students take reading tests typically note that testees use a combination of prior knowledge, analysis of the text and accompanying questions, and test-taking skills. Test-wiseness (TW) is an important source of test content/construct invalidity since it examines students' ability to answer correctly by exploiting weaknesses in test design. The purpose of the study reported here was to design a valid and reliable test of TW in EFL/ESL students. A 33-item multiple-choice instrument with four subscales was constructed and trialled on several groups of ESL students. The test was validated by analysis both of item form and content and of testees' explanations of their answers. Statistical item analyses suggest that the test has reasonable internal consistency and that two subscales may have psychological reality. The findings indicate that students are differentially skilled in test taking and that the scores of some learners may be influenced by skills which are not the focus of the test, thus invalidating their results. The test of test-wiseness in ESL students (TOTWESL) described here can be used to diagnose inexperienced or underachieving language test takers, and (by exposure and sensitization) to confront their weaknesses.