The Use of Presnloothing and Postsnloothing to Increase the Precision of Equipercentile Equating

Abstract

The effectiveness of smoothing in reducing sample- dependent errors in equipercentile equating of short ability or achievement tests is examined. Fourteen smoothers were examined, 7 applied to the distribu tions of scores before equating and 7 applied to the resulting equipercentile points. The data for the study included both results of simulations and results ob tained in the operational administration of a large test ing program. Negative hypergeometric presmoothing was more effective than the other presmoothers. Among the postsmoothers, both orthogonal regression and cubic splines were effective, especially the latter. The use of smoothing methods must be considered in light of their costs (increases in average signed devia tions) and benefits (decreases in root mean square de viations). For many purposes, the benefits of smooth ing with the negative hypergeometric may outweigh its costs.