ERIC Number: ED235198
Record Type: RIE
Publication Date: 1983-Apr
Reference Count: 0
An Investigation of the Feasibility of Applying Item Response Theory to Equate Achievement Tests.
Cook, Linda L.; Eignor, Daniel R.
The purpose of this study was to examine the feasibility of using item response theory (IRT) methods to equate different forms of three College Board Achievement Tests (Biology, American History and Social Studies, and Mathematics Level II) and one Graduate Record Examinations Achievement Test (Advanced Biology), rather than conventional or equipercentile methods. The criterion for evaluation of the results was scale drift, which is said to have occurred if the results of equating test form A directly to test form D is not the same as that obtained by equating test form A to test form D through intervening forms B and C. The results of three conventional linear equating methods, conventional equipercentile equating with an anchor test, and two IRT equating methods were compared. No linear equating method produced scaled scores that could be considered seriously discrepant from the criterion scores, indicating that they perform quite adequately. The equipercentile method produced the largest total error. The IRT concurrent and characteristic curve transformation methods gave very similar results, and results indicate that it is feasible to use IRT to equate the tests in this study. (BW)
Publication Type: Speeches/Meeting Papers; Reports - Research
Education Level: N/A
Sponsor: Educational Testing Service, Princeton, NJ.
Authoring Institution: N/A
Identifiers: College Board Achievement Tests; Equipercentile Equating; Graduate Record Examinations; Linear Equating Method
Note: Paper presented at the Annual Meeting of the American Educational Research Association (67th, Montreal, Quebec, April 11-15, 1983). This study was supported through Program Research Planning Council funding.