Abstract
Owen's (1969) Bayesian tailored testing method is introduced along with a brief review of its deriva tion. The characteristics of a good item bank are outlined and explored in terms of their influence on the Bayesian tailoring process. The results clearly demonstrate importance of a good item bank; one having a sufficient number of items with high dis crimination, low guessing probability, and a uniform distribution of difficulty.

This publication has 3 references indexed in Scilit: