Comparability of conventional and computerized tests of reading in a second language

Computerization of L2 reading tests has been of interest among language assessment researchers for the past 15 years, but few empirical studies have evaluated the equivalence of the construct being measured in computerized and conventional L2 reading tests and the generalizability of computerized reading test results to other reading conditions. In order to address various issues surrounding the effect of mode of presentation on L2 reading test performance, the present study reviews the literature in cognitive ability testing in educational and psychological measurement and the non-assessment literature in ergonomics, education, psychology, and L1 reading research. Generalization of the findings to computerized L2 assessment was found to be difficult: The nature of the abilities measured in the assessment literature does not necessarily involve language data; mode of presentation studies in the non-assessment literature involving L2 readers are scarce; and there are limitations in the research methodologies used. However, the literature raises important issues to be considered in future studies of mode of presentation in language assessment.