The present study addressed the effects of cognitive styles on answer-switching practices across online and traditional tests. After completing Ehrman and Leaver’s (2003) cognitive style questionnaire, a sample of upper intermediate students took pen-paper and online versions (59 test-takers each) of an already validated teacher-developed test of English. The data from think-aloud and erasure analyses revealed significantly more frequent total and right-to-wrong changes in the traditional and online tests, respectively. Multiple regression values explained more than 50% of right/wrong-to-wrong, wrong-to-right, and overall answer-changing variance based on cognitive styles in the pen-paper exam. However, the regression results from online performance analyses could not prove the power of thinking styles in predicting answer-changing strategies. Fisher’s exact tests showed significantly different answer-changing strategies adopted by field-dependent, leveler, analog, concrete, and impulsive individuals in the traditional test but no significant differences between the behaviors of individuals with different cognitive styles in the online exam. Based on the present findings, online and pen-paper platforms may require different test-taking strategies. Language instructors and test developers can use these findings to align their instructional and assessment practices with various cognitive styles and testing environments.
