Construct validation of Bachman and Palmer's (1996) strategic competence model over time in EFL reading tests

Abstract

This article reports on a large-scale study that aims to validate the theory of strategic competence proposed by Bachman and Palmer (1996) through the use of structural equation modeling (SEM). The present study examines the relationship of test-takers' long-term strategic knowledge (i.e., trait strategies) and actual strategy use (i.e., state strategies) to second language (L2) reading test performance over time. The data were gathered on two occasions (during the mid-term and final examination periods). Five hundred and sixty-one Thai university test-takers answered a trait strategy use questionnaire prior to the mid-term and final reading achievement tests and, immediately after completing each test, they answered a state strategy use questionnaire. It was found that (1) trait metacognitive strategy use (MSU) directly and strongly affects trait cognitive strategy use (CSU); (2) trait CSU does not greatly affect state CSU; (3) trait MSU directly affects state MSU in a specific context, which in turn directly affects state CSU; and (4) state CSU directly affects a specific language test performance to a varying degree.