Describes a program system which was developed for the purpose of making estimates about the nature of the distribution of test scores which will result from an administration of a test composed of items of which estimates of item-difficulty and discrimination are available. (Author/RC)

Responses of male prisoners to the Personal Orientation Inventory were clustered, using hierarchical linkage analysis. Six second-order clusters accounted for all the items. Reliabilities of these clusters were comparable to those of the first-order clusters. Relative validity of cluster scores and scale scores remains to be determined. (RC)

Examines the question of how large a sample must be in order to produce empirical Bayes estimates which are preferable to other commonly used estimates, such as proportion correct observed score. (Author/RC)

Research performed pursuant to a grant from the Charles F. Kettering Foundation.

Using an empirical investigation of alternate item nonresponse adjustment procedures in a National Longitudinal Study (NLS) of missing and faulty data, it is indicated that in some cases imputation can reduce the accuracy of survey estimates. A National Sample of the high school class of 1972 is designed to provide statistics on students moving…

The purpose of this study was to present an empirical correction of the KR21 (Kuder Richardson test reliability) formula that not only yields a closer approximation to the numerical value of the KR20 without overestimation, but also simplifies computation. This correction was accomplished by introducing several correction factors to the numerator…

A project was carried out to determine the degree of content homogeneity that a test item pool must have in order to accomplish successful Rasch calibration. Mathematics item pools were administered to upper elementary children. The items were analyzed under two conditions, with items organized into separate subtests and as a global mathematics…

A basic description of the Rasch model and a brief review of the work previously done on the model is related. Simulated data was used to test goodness of fit to the Rasch model. It was found that data with near perfect fit to the Guttman model provides perfect fit to the Rasch model, though a perfect Guttman scale collapses. Random data also…

A test bias definition, applicable at the item-level of a test is presented. The definition conceptually equates test bias with measuring different things in different groups, and operationally equates test bias with a difference in item fit to the Rasch Model, greater than one, between groups. It is suggested that the proposed definition avoids…

An algorithm for item analysis in which item discrimination indices have been defined for the distractors as well as the correct answer is presented. Also, the concept of convergent and discriminant validity is applied to items instead of tests, and is discussed as an aid to item analysis. (Author/JKS)