Implementation and Measurement Efficiency of Multidimensional Computerized Adaptive Testing

Abstract

Multidimensional adaptive testing (MAT) procedures are proposed for the measurement of several latent traits by a single examination. Bayesian latent trait estimation and adaptive item selection are derived. Simulations were conducted to compare the measurement efficiency of MAT with those of unidimensional adaptive testing and random administration. The results showed that the higher the correlation between latent traits, the more latent traits there were, and the more scoring levels there were in the items, the more efficient MAT was than the other two procedures. For tests containing multidimensional items, only MAT is applicable, whereas unidimensional adaptive testing is not. Issues in implementing MAT are discussed.

Item Selection in Multidimensional Computerized Adaptive Testing—Gaining Information from Different Angles

“…More research is needed to explore the performance of each item selection rule under various conditions. Factors that will be of interest includes within-item multidimensionality versus between-item multidimensionality (Wang & Chen, 2004); high correlation between dimensions versus low correlation; all dimensions are desirable versus some dimensions are desired and the rest are nuisance (Mulder & van der Linden, 2009;Veldkamp & van der Linden, 2002). In addition, we only considered two-dimensional structures in our simulation studies.…”

Bayesian adaptive estimation: The next dimension

“…In a similar multidimensional extension of adaptive latent trait estimation (Wang & Chen, 2004) similar issues are raised: radically different trials may cause anxiety or fatigue. However, Wang and Chen (2004) note that such a Fig.…”

Multidimensional Computerized Adaptive Testing of the EORTC QLQ-C30: Basic Developments and Evaluations

“…At the same time, it is well known that several domains of HRQOL are highly correlated. With a few exceptions [3][4][5][6][7] the CAT procedures developed and used have been unidimensional i.e., have been developed for a single domain. Using such unidimensional CAT to measure HRQOL necessitates a CAT procedure for each domain.…”