Chance-Constrained Automated Test Assembly

Classical automated test assembly (ATA) methods assume fixed and known coefficients for the constraints objective function. This hypothesis is not true estimates of item response theory parameters, which are crucial elements in classical models. To account uncertainty ATA, we propose a chance-constrained version maximin ATA model, allows maximizing α-quantile sampling distribution information function obtained by applying bootstrap on parameter estimation. A heuristic inspired simulated annealing optimization technique implemented to solve model. The validity proposed approach empirically demonstrated simulation study. applicability proven using real responses Trends International Mathematics Science Study (TIMSS) 2015 science test.