Demonstrates that some commonly used indices can be misleading in their quantification of reliability. The effects are most pronounced on gain or difference scores. Proposals are made to avoid sources of invalidity by using a procedure to assess reliability in terms of upper and lower limits for the true scores of each examinee. (Author/JDH)

The problem of predicting universe scores for samples of examinees based on their responses to samples of items is treated. The measurement model categorizes items according to the cells of a table of test specifications, and the linear function derived for minimizing error variance in prediction uses responses to these categories. (Author/JKS)

McCauley and Colberg described a theory of transportability and provided data to demonstrate the feasibility of their approach. It is argued that the transportability notion resembles earlier cross cultural work and does not add new insights into cross cultural comparison. Their statistical checks do not preclude the possibility of bias.…

Latent trait and latent class analyses of Likert-type data are compared. Key similarities and differences between these methods are described and illustrated by applying a latent trait model and a latent class model to the analysis of a set of "life satisfaction" data. (Author/NSF)

Considers the relationship between coordinate estimates in components analysis and multidimensional scaling. Reports three small Monte Carlo studies comparing nonmetric scaling solutions to components analysis. Results are related to other methodological issues surrounding research on the general ability factor, response tendencies in…

This study extended prior Law School Admission Council (LSAC) research related to the item response theory (IRT) local item independence assumption into the realm of classical test theory. Initially, results from the Law School Admission Test (LSAT) and two other tests were investigated to determine the approximate state of local item independence…

Two linearly constrained models based on the Rasch model are discussed. Necessary and sufficient conditions for the existence of unique conditional maximum likelihood estimators are derived. Methods for hypothesis testing within this framework are proposed. (Author/JKS)

The need to use images and imagery in language in schools, and to recognize and measure such aptitudes are discussed. Tests of imagination are presented, and the San Francisco Art Institute Study of predictors of ability in art illustrates methods and implications for measurement of creativity in educational programs. (CM)

Reliability coefficients of linear combinations of observed scores have anomalous properties which have led to difficulties in the investigation of difference scores and gain scores in test theory. Discrepancies between classical results and correct results obtained from more general formulas, which allow for correlated errors, are examined…

Examined the validity of data generated by the School Readiness for Reforms: Leader Questionnaire using an iterative procedure that combined classical and Rasch rating scale analysis for the analysis of responses from 167 leaders. The combined approach yielded comprehensive diagnostic information on the quality of the instrument's five subscales.…

Describes the development of a test designed to allow meaningful and widespread computerized testing of various spatial factors. Examines the differences between traditional paper and pencil and computerized versions of the same test. Compares an interactive test designed to measure a working memory factor to the computerized version of…

Compares results of Successive Perception Test I (SPT) for the study population of freshman engineering students to their results on the group-administered Purdue Spatial Visualization Test: Visualization of Rotations (PSVT) and the individually administered Haptic Visual Discrimination Test (HVDT). Concludes that either visual and haptic…

Examined the influence of deviant scale items on item parameter estimates of focal scale items and person parameter estimates through a comparison of item response theory (IRT) and classical test theory (CTT) models. Used Monte Carlo methods to explore results from a pilot investigation of job attitude data. Discusses implications for researchers…

Multilinear formula scoring (MFS) is reviewed, with emphasis on estimating option characteristic curves (OCSs). MFS was used to estimate OCSs for the arithmetic reasoning subtest of the Armed Services Vocational Aptitude Battery for 2,978 examinees. A second analysis obtained OCSs for simulated data. The use of MFS is discussed. (SLD)

The book reviewed is a compendium of current thinking about measurement theory and test use. It includes content by 26 authors at 3 levels: (1) accessible to educators, policy makers, and graduate students; (2) suited for technical students; and (3) written for qualified measurement specialists. Strengths and weaknesses are noted. (SLD)