Consequences of Test Anxiety on Adaptive Versus Fixed Item Testing

Abstract

We investigated the effects of test anxiety on test performance using computerized adaptive testing (CAT) versus conventional fixed item testing (FIT). We hypothesized that tests containing mainly items with medium probabilities of being solved would have negative effects on test performance for testtakers high in test anxiety. A total of 110 students (aged 16 to 20) from a German secondary modern school filled out a short form of the Test Anxiety Inventory (TAI-G; Wacker, Jaunzeme, & Jaksztat, 2008 ) and then were presented with items from the Adaptive Matrices Test (AMT; Hornke, Etzel, & Rettig, 1999 ) on the computer, either in CAT form or in a fixed item test form with a selection of items arranged in order of increasing item difficulty. Additionally, half of the students were given a short summary of information about the mode of item selection in adaptive testing before working on the CAT. In a moderated regression approach, a significant interaction of test anxiety and test mode was revealed. The effect of test mode on the AMT score was stronger for students with higher scores on test anxiety than for students with lower test anxiety. Furthermore, getting information about CAT led to significantly better results than receiving standard test instructions. Results are discussed with reference to test fairness.