Small Sample Estimation in Dichotomous Item Response Models: Effect of Priors Based on Judgmental Information on the Accuracy of Item Parameter Estimates

By H. Swaminathan, R. K. Hambleton, S. Sireci, Dehui Xing, Saba M. Rizavi. 2003

52 citations recorded by the source corpus.

Large item banks with properly calibrated test items are essential for ensuring the validity of computer-based tests. At the same time, item calibrations with small samples are desirable to minimize the amount of pretesting and limit item exposure. Bayesian estimation procedures show considerable promise with small examinee samples. The purposes of the study were (a) to examine how prior information for Bayesian item parameter estimation can be specified and (b) to investigate the relationship between sample size and the specification of prior information on the accuracy of item parameter estimates. The results of the simulation study were clear: Estimation of IRT model item parameters can be improved considerably. Improvements in the one-parameter model were modest; considerable improvements with the two- and three-parameter models were observed. Both the study of different forms of priors and ways to improve the judgmental data used in forming the priors appear to be promising directions for future research.

Subfields: Computer Networks and Communications, Management Science and Operations Research, Social Psychology

View Paper: https://www.semanticscholar.org/paper/e01898936a65963da3214b10fc8b6e270c698c98

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

  1. Máaarcia D. Branco
  2. Heleno Bolfarine
  3. Jorge L. Bazán

In Management Science and Operations Research

  1. Máaarcia D. Branco
  2. Heleno Bolfarine
  3. Jorge L. Bazán

Citations in the corpus