<![CDATA[Although negative marking is intended to discourage random guessing in multiple-choice mathematics assessments, it may systematically influence test-takers’ decisions under uncertainty and thereby affect score interpretation. This study examines how guessing behaviour and risk aversion jointly relate to item-level decision-making and overall performance in negatively marked multiple-choice mathematics tests. Quantitative observational data were obtained from 885 senior secondary school students. Logistic regression models were used to estimate item-level guessing decisions, while linear regression models incorporating interaction terms examined cumulative performance outcomes. Higher expected score gains were significantly associated with an increased probability of guessing, whereas greater risk aversion was associated with reduced guessing and increased omission behaviour. A statistically significant guessing-by-risk-aversion interaction (reported as b = 0.29, p < .001) was interpreted as indicating that the performance benefits associated with guessing diminished as risk aversion increased. These findings demonstrate that scores from negatively marked multiple-choice mathematics assessments may reflect behavioural decision strategies in addition to mathematical competence, thereby introducing construct-irrelevant variance into score interpretation. The study contributes to assessment research by clarifying the joint role of guessing and risk aversion in testing behaviour and underscores the need for assessment designers and score users to evaluate cautiously the validity of negatively marked scores, particularly in high-stakes mathematics testing contexts.]]>
Copyrights © 2026
