College students' behavior on multiple choice self-tailored exams in relation to metacognitive ability, self-efficacy, and test anxiety

The purpose of this study was to observe college students' behavior on five self-tailored, multiple choice exams throughout a semester in relation to: (a) metacognitive ability, (b) self-efficacy expectations, and (c) test anxiety. Additionally, the effect of a self-tailoring procedure on exam scores and content validity of the tests was observed. Self-tailored testing was defined as an option in which students selected up to five questions they wanted to omit from being scored on an exam. Students' metacognitive ability was defined as the percentage of incorrectly answered questions out of the total number omitted.; Ninety-nine college students from two sections of an educational psychology undergraduate course participated in this study. Eighty students completed the study; seventy-one used an option to omit questions on all exams. Before taking exam 1, students answered measures of self-efficacy and test anxiety. After completing each of the five course exams, students marked on the back of their answer sheet up to five questions they wanted to be omitted from scoring. After exam 5, students answered a questionnaire that addressed their perception of the self-tailoring procedure.; MANOVA, repeated measures ANOVA, Pearson correlations, t-test and one-way ANOVA were conducted. Students made a statistically significant increase in their scores on all exams by using the questions omitting procedure. There was a statistically significant linear increase of percentages of incorrectly answered questions out of the total number omitted across five exams. Frequency of items that students omitted from scoring were significantly negatively correlated with item difficulty values. The content validity of the test was affected on two out of five exams based on cognitive level of items and on three out of five exams based on chapter coverage. Students' self-efficacy expectations and test anxiety were not related to the likelihood to apply the self-tailoring procedure or to the degree of success students had in applying the procedure. The study provided a new perspective on self-tailored tests in college classroom with implications for teaching, assessment, and students' metacognitive abilities.