In a post mortem study, it is demonstrated that linear prediction is as effective as computing a negative hyper-geometric distribution for estimating test norms following matrix sampling from a total test with a highly skewed score distribution, provided the same prediction coefficient is used for all examinee groups. It is also demonstrated…

Selected parameters for a negatively-skewed and a normally distributed normative distribution were estimated in a post-mortem item-examinee sampling investigation. Manipulated systematically were number of subtests, number of items per subtest, and number of examinees responding to each subtest. Each item-examinee sampling procedure was replicated…

A rederivation of Lord's formula for estimating variance in multiple matrix sampling is presented as well as the ways Cronbach's coefficient alpha and the Spearman-Brown prophecy formula are related in this context. (Author/JKS)

Authors conclude that, in the world of real data sample correlation matrices, Guttman's stronger lower bound is not of practical use in determining the effective number of common factors. (Authors/CB)

Author discusses the implications that criterion-referenced tests have for the teacher and for testing programs in the schools. (Author/CB)

Authors present a formula for use when restrictions result from complex or unmeasured variables. (Authors/MB)

Sternberg (1981) presented correlations based on independent samples of about 25 cases each with a statistical comparison in a population parameter of zero and interpreted his results as if with appropriate tests. He needed to use a nonzero population parameter and compare for the difference between the z transformations. (Author/CM)

Sternberg's reply to paper Humphrey criticizes (see TM 507 045) includes: (1) invited position paper and detailed statistical testing would have been inappropriate; (2) Humphrey commits some of "misdeeds" of which he accuses Sternberg; (3) number of cases in the three nonentrenchment experiments are more than adequate for drawing conclusions; and…

A computer program to generate individualized objective test forms for use in a Student Faced Statistics (SPS) course is described. The program features disproportionate sampling from different item domains and enhanced character generation facility for test printing purposes. (Author)

An expanded multiple matrix sampling procedure to collect assessment data for health education courses in upper elementary and junior high schools is outlined. This supplement to the Cooperative Health Education Test provides data at the school building and district levels. (JMF)

This study demonstrates that multiple matrix sampling procedures can be used to collect assessment data efficiently, unabstrusively, and reliably. (MB)

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

A common approach to needs analysis involves the development of a self-report questionnaire administered to members of the intended audience. Such an approach poses a problem however: Concern for data reliability demands polling as many respondents as possible, while concern for cost-effectiveness dictates minimizing the size of the sample. The…

Investigated empirically through post mortem item-examinee sampling were the relative merits of two alternative procedures for allocating items to subtests in multiple matrix sampling and the feasibility of using the jackknife in approximating standard errors of estimate. The results indicate clearly that a partially balanced incomplete block…