The comparison of test equating with covariates using Bayesian nonparametric method

This research is based on the use of covariates to obtain equated scores with Bayesian nonparametric model. As covariates in the study; gender, mathematics self-efficacy scores and common item scores were used. The distances of equated scores obtained by using common items and covariates are calculated from distances to target test. These distances were compared with distances of equated scores obtained from methods based on Item Response Theory to target test. The study was conducted on Canadian and Italian samples of PISA 2012 application. The scope of work, PARSCALE was used for estimation of material parameters, IRTEQ was used for scale transformation, and R software was used for Bayesian nonparametric model. Distributions of equated scores obtained in the Bayesian nonparametric approach changed according to the covariates. When gender, mathematics self-efficacy scores, and common item scores were used as covariates in the model, distance values of obtained equated scores to target test are close to each other but their distributions are different. The closest distribution to target test was achieved when gender and mathematics self-efficacy scores were used together as covariates in the model and the farthest distributions was obtained from item response theory methods. As a result of research, it was determined that model is more informative than the classical methods. Covariates can be used instead of common items and even better in some cases, and that equated scores obtained with model can give closer results to target test.