ERIC Number: ED127329
Record Type: Non-Journal
Publication Date: 1976-Mar
Reference Count: N/A
Equating Reading Tests With the Rasch Model. Management Report.
Rentz, R. Robert; Bashaw, W. L.
To determine if Rasch Model procedures have any utility for equating pre-existing tests, data from the equating phase of the Anchor Test Study (ATS) were reanalyzed. This management report summarizes the work completed in the project, describes the differences in this project and that of the ATS, and presents recommendations and conclusions to the U.S. Office of Education. The tests involved included seven reading test batteries, each having one to three levels and two forms, and each having a vocabulary and comprehension subtest. There were 28 form-level combinations possible. Therefore, the concern was the simultaneous equating of 28 tests for each of vocabulary, comprehension, and total scores. The obvious difference between this project and the ATS is that this project used Rasch Model test calibrations and equating methods while the ATS used a variety of equipercentile and linear model methods. Other important differences are outlined. Finally, some conclusions and recommendations relative to any future equating effort that might be undertaken are made. Topics discussed include technical issues as well as design and cost considerations. (RC)
Publication Type: Reports - Research
Education Level: N/A
Sponsor: National Center for Education Statistics (DHEW), Washington, DC.
Authoring Institution: Georgia Univ., Athens. Educational Research Lab.