Impact of Inclusion or Exclusion of Repeaters on Test Equating
Author:
- 29 July 2011
- journal article
- research article
- Published by Taylor & Francis in International Journal of Testing
- Vol. 11 (3) , 215-230
Abstract
This study examined the effect of including or excluding repeaters on the equating process and results. New forms of two tests were equated to their respective old forms using either all examinees or only the first timer examinees in the new form sample. Results showed that for both tests used in this study, including or excluding repeaters in the new form sample had very little impact on the final equating results. These findings do not provide enough evidence to support routine exclusion of repeaters from the equating sample before conducting the equating. Results of some additional analyses showing the impact on random equating error due to the reduction in sample size caused by excluding repeaters are presented. Finally, some examples of simple analyses to assist the testing practitioner when making a judgment about whether including or excluding repeaters in the equating sample is reasonable are provided.Keywords
This publication has 5 references indexed in Scilit:
- Test Equating, Scaling, and Linking2004
- Small‐Sample Equating With Log‐Linear SmoothingJournal of Educational Measurement, 1993
- A Comparison of Bivariate Smoothing Methods in Common-Item Equipercentile EquatingApplied Psychological Measurement, 1991
- NOTES ON THE USE OF LOG‐LINEAR MODELS FOR FITTING DISCRETE PROBABILITY DISTRIBUTIONSETS Research Report Series, 1987
- The Effects of Repeaters on Test EquatingEducational and Psychological Measurement, 1978