Optimal Sequential Designs for On-Line Item Estimation
Abstract
Replenishing item pools for one-line ability testing requires innovative and efficient data collection designs. Based on a theoretical framework for generating exact D-optimal-designs for selecting individual examinees, and for consistently estimating item parameters, this article presents a sequential procedure for on line item calibration. These procedures were derived for general, dichotomous item response models, using Welch (1982) for exact n-point D-optimal designs and Stefanski and Carroll (1985) for consistent estimators. In simulations, these designs appear to be considerably more efficient than random seeding of items.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1991
- Accession Number
- ADA242933
Entities
People
- Douglas H. Jones