The Influence of Dimensionality on Cat Ability Estimation
Ralph J. De Ayala · Educational and Psychological Measurement · 1992
This study examined the effect of dimensionality on the ability estimation of an adaptive test. Two-dimensional data sets were generated which differed from one another in the interdimensional ability association, the correlation among the difficulty parameters, and the presence or absence of confounding of item discriminations with item difficulty. The generated data were used for Bayesian computerized adaptive testing (CAT) simulations (three parameter logistic model), and the CAT ability estimates were compared with the simulees known abilities (θ T s). Results show that the dimensionality of the response data shifts the focus for the minimization of measurement errors from OT (with unidimensional data) to the average of the latent abilities (with bidimensional data).