Category: Adult Anxiety - Social

PS9- #A20 - Reassurance Seeking and Spoiled Answers on Academic Tests

Saturday, Nov 18
9:45 AM – 10:45 AM
Location: Indigo Ballroom CDGH, Level 2, Indigo Level

Keywords: Social Anxiety

Safety behaviors are involved in the maintenance of many forms of anxiety (see Helbig-Lang & Petermann, 2010). Extending this work to test anxiety, Knoll, Valentiner, & Holzman (in press in Assessment) found that test–related reassurance seeking behaviors were associated with lower ACT scores. However, the reassurance seeking scale appears to include adaptive behaviors (e.g., “I need to review my answers more than once before I turn in an exam, or else my grade will be worse”). Although double-checking test answers is associated with higher test scores on average (Kruger, Wirtz, & Miller, 2005), it appears to lead to lower test scores for about 6% of test takers (Liu, Bridgeman, Gu, Xu, & Kong, 2015). The current studies examine whether reassurance seeking leads to spoiled answers on academic tests.


In Study 1, Introductory Psychology students (N=59; 54% male; 61% White) were recruited to participate in this study. The course was structured around three non-cumulative exams, i.e., after the 5th week and during the 10th and 15th weeks of the semester. After receiving feedback on the 2nd exam and before taking the 3rd exam, students completed the 10-item Test-Related Reassurance Seeking Scale (alpha = .66; Knoll et al., in press) and a 5-item test anxiety scale (alpha = .92; Taylor & Deane, 2002). Two researchers examined the physical Scantrons from the 3rd exam for erasure marks to identify the numbers of spoiled answers, corrected answers, and irrelevant changes. They were able to do so reliably (all absolute IRRs > .84). Disagreements between raters were resolved by consensus.


As hypothesized, reassurance seeking predicted the number of spoiled answers on the 3rd exam (r = .30, p =.016), but not the numbers of corrected answers or irrelevant changes, or overall scores on the 3rd exam. Test anxiety was significantly correlated with reassurance seeking (r = .41, p < .001), but not with any of the outcome variables. These results provide evidence that reassurance seeking is associated with spoiled answers.


Study 2 is in progress. Approximately 600 Introductory Psychology students have completed the reassurance seeking scale after receiving feedback on the 1st exam but before taking the 2nd exam. Scantrons for the second exam and other course data will be obtained to examine the replicability of the findings from Study 1. These results may inform advice about double-checking test answers and interventions to reduce test anxiety and increase test performance.

Anna Snyder

Northern Illinois University

Alexia Kingzette

Northern Illinois University

Aaron Ford

Northern Illinois University

David Valentiner

Northern Illinois University

M. A. Britt

Northern Illinois University