Criticizes Iris Rotberg's December 1990 "Kappan" article for two points involving age group comparisons and alleged sampling inconsistencies. Some indicators Rotberg proposes are as flawed as those she would replace. Comparative educational research has a vital, continuing role in furthering the United States' efforts to improve its…

Responds to article by Horan, Hanish, Keen, Saberi, and Hird (1993) criticizing Howard and Curtin's (1993) study of productivity and impact in counseling psychology and offering different version of productivity study. Contends that Horan et al.'s questioning of Howard and Curtin's methodological decisions is unfounded and at variance with facts…

Clarifies several issues regarding the effects of autocorrelated errors on Type I error in ordinary least-squares models. Demonstrates through Monte Carlo simulation the conditions under which distortion in Type I error is less than predicted by asymptotic theory. Suggests a recently developed small-sample method for time-series analyses. (SLD)

Describes a procedure for automated-test-forms assembly based on Classical Test Theory (CTT). The procedure uses stratified random-content sampling and test-form preequating to ensure both content and psychometric equivalence in generating virtually unlimited parallel forms. Extends the usefulness of CTT in automated test construction. (Author/SLD)

This study developed a testing approach to increase sample size to approximately 450 scorable items, previously determined as necessary to optimize the reliability of a speech recognition test. The approach involves an interactive computer program which presents words in 50 three-word groups. The approach was evaluated with 100 persons with normal…

The Gibbs sampling algorithms presented by M. H. Seltzer (1993) are fully generalized to a broad range of settings in which vectors of random regression parameters in the hierarchical model are assumed multivariate normally or multivariate "t" distributed across groups. The use of a fully Bayesian approach is discussed. (SLD)

The three-parameter logistic (3-PL) model is described and a derivation of the 3-PL observed information function is presented for a single binary response from one examinee with known item parameters. Formulas are presented for the probability of negative information and for the expected information (always nonnegative). (SLD)

Offers an instructional approach and an exercise that can serve as a useful introduction to stratified random sampling. (ASK)

Explains how schools and students are selected for the National Assessment of Educational Progress (NAEP) and why high participation rates are important. Describes the multistage sampling method, which allows the National Center for Education Statistics to give an accurate picture of student performance while keeping the burden on schools and…

Unemployed managers (n=95) completed the Vocational Interest Survey for Australia. Compared to a sample of secondary students, they had higher interest in vocational activities; they were able to discriminate more dimensions of vocational interests than a nonmanagerial sample did. Multisample norming was recommended. (SK)

Discussion of shelf reading as a method to maintain library books in their proper location focuses on a statistical approach to determine when shelf reading is necessary. Uses sampling to obtain data on misshelved books over time and develop a control chart to assess when shelf reading is necessary. (Author/LRW)

Discusses the need for tools that allow effective analysis of search engine queries to provide a greater understanding of Web users' information seeking behavior and describes a study that developed an effective strategy for selecting samples from large-scale data sets. Reports on Poisson sampling with data logs from the Excite search engine.…

Describes the multistage sampling system used in the 1996 national roadside survey of driver blood alcohol and compares it to the sampling methods used in prior surveys in 1973 and 1986. Although many more drivers were sampled in the most recent survey, the procedures are similar enough to allow comparison of results. (SLD)

Examined the effect of complex sampling of items on the measurement of differential item functioning (DIF) using the Mantel-Haenszel procedure through a Monte Carlo study. Suggests the superiority of the pooled booklet method when items are selected for examinees according to a balanced incomplete block design. Discusses implications for other DIF…

A split-sample replication criterion originally proposed by J. E. Overall and K. N. Magee (1992) as a stopping rule for hierarchical cluster analysis is applied to multiple data sets generated by sampling with replacement from an original simulated primary data set. An investigation of the validity of this bootstrap procedure was undertaken using…