Abstract
An extremely efficient MCMC method for Bayesian adaptive testing with polytomous items is explained both conceptually and in mathematical detail. Results from extensive simulation studies with different item pools, polytomous response models, calibration sample sizes, and test lengths are presented. In addition, the case of adaptive testing from pools with a mixture of dichotomous and polytomous items is addressed.
Original language | English |
---|---|
Pages (from-to) | 427-449 |
Number of pages | 23 |
Journal | Behaviormetrika |
Volume | 47 |
Issue number | 2 |
Early online date | 21 May 2020 |
DOIs | |
Publication status | Published - 1 Jul 2020 |
Keywords
- UT-Hybrid-D
- Bayesian optimality
- Item calibration
- MCMC algorithm
- Polytomous response models
- Adaptive testing