Alternative Approaches to Updating Item Parameter Estimates in Tests With Item Cloning (CT-03-01)
by Cees A. W. Glas, University of Twente, Enschede, The Netherlands
All item-cloning techniques are based on a formal description of a set of parent items and an algorithm to derive a larger set of operational items from them. By using such techniques, item pools for computerized adaptive testing can be created more cost effectively and item presentation can be enhanced by using more uniform item formats.
The item response theory (IRT) mathematical model is usually applied in computerized adaptive testing. In IRT, several statistics, called item parameters, are used to describe individual test items. One of the issues to be addressed in the application of item cloning techniques is the extent to which the item parameters will vary between parent and cloned items. Previous researchers have proposed a mathematical model, referred to in this paper as the item cloning model (ICM), to deal with possible variability of the item parameters introduced by item cloning.
The current study compares two procedures for updating the parameter estimates for an item bank consisting of cloned items calibrated under the ICM. Results from the simulation studies indicated that the first procedure tended to have a minor gain in precision in the parameter estimates. The second method was 8% faster, not a large enough gain to offset loss of precision.