COMPARISON OF FOUR PROCEDURES FOR EQUATING THE TESTS OF GENERAL EDUCATIONAL DEVELOPMENT

The adequacy of equipercentile, linear, one-parameter (Rasch), and three-parameter logistic item-response theory procedures for equating twelve forms of five tests of General Educational Development were compared. All equating was horizontal and examinees were administered two forms to be equated. Test dimensionality was examined and cross-validation analyses were used in the equating methods comparison. The three-parameter and equipercentile methods produced unacceptable equating results, probably because only 200 examinees were administered each of eleven of the forms. The one-parameter (Rasch) and linear methods produced more stable results. In evaluating the results of this study in conjunction with other comparisons, it was concluded that equating method adequacy depends on a variety of factors such as test characteristics, equating design, and sample size. Linear equating was suggested as the method of choice for horizontal equating of achievement tests when a common sample equating design is used with small sample sizes.