Enjoyable for some, stressful for others? Physiological and subjective indicators of achievement emotions during adaptive versus fixed-item testing

Abstract

In light of the increasing use of computerized adaptive testing, we investigated how adaptive testing impacts test-takers’ subjective emotional experiences and their psychophysiological arousal. Applying a within-person design (N = 89), we compared participants’ affective states while working on an adaptive and a fixed-item test of numerical reasoning ability. During both tests, we continuously recorded participants’ skin conductance response. In addition, they filled in a self-report questionnaire after each of the three item blocks per test, assessing discrete achievement emotions (joy, pride, anger, boredom, frustration, and anxiety) and perceived level of task difficulty. As expected, participants showed higher levels of psychophysiological arousal in the adaptive compared to the fixed-item test, indicating that the adaptive test was more stimulating, independent of emotional valence. For subjective achievement emotions, we expected disordinal interaction effects between test type and ability (objective control experience) and between test type and relative perceived difficulty of the two tests (subjective control experience). This was supported for relative perceived difficulty, as participants indeed reported more joy and pride, and less frustration, anxiety, and anger on whichever test they subjectively perceived as easier. Meanwhile, no main effects of test type and no interaction between test type and ability were found. This is in line with the control-value theory and shows that it is not the adaptivity of a test that influences subjective emotional experience, but rather how difficult the adaptive test is perceived by test-takers compared to a fixed-item test. Directions for future research and implications for practice are discussed.