Empirical Bayes methods are shown to provide a practical alternative to standard least squares methods in fitting high dimensional models to sparse data. An example concerning prediction bias in educational testing is presented as an illustration. (Author)

This report introduces the General Education Development Testing Services (GEDTS) Administrator screening model for adults with learning disabilities. The screening model is designed to identify adults whose educational achievement is significantly below overall potential due to an informational processing deficit and who are thus eligible for…

This report examines the potential of adaptive testing, two?-stage testing in particular, for improving the data quality of the National Assessment of Educational Progress (NAEP). Following a discussion of the rationale for adaptive testing in assessment and a review of previous studies of two-?stage testing, this report describes a 1993 Ohio…

This paper reports changes in the professional knowledge, general knowledge, and communication skills components of the National Teacher Examinations and explains validation procedures. Research efforts to: (1) establish the tests' credibility and usefulness; (2) identify factors the tests measure; and (3) study their relationship to effective…

A study compared college students' performance on complex multiple-choice tests with scores on multiple true-false clusters. Researchers concluded that the multiple-choice tests did not accurately measure students' knowledge and that cueing and guessing led to grade inflation. (PP)

Technical problems in the assessment of learning disabilities (LD) are only a small part of the problems leading to the misidentification of children in this category of handicap. Given all the other motives and purposes served by labeling a child LD in school, apparently sound measurement advice may actually do more harm than good. (LC)

This study attempted to determine whether certain demographic factors, Intelligence Quotient tests, or achievement tests best predict achievement of second through fourth grade students. Findings suggest that the group mental ability test did not add to the knowledge gained from a comprehensive achievement testing program. (Authors/PP)

A model is presented using examinee judgements in detecting ambiguous/misinterpreted items on teacher-made criterion-referenced tests. A computational example and guidelines for constructing domain categories and interpreting the indices are presented. (Author/PN)

The effectiveness of several methods used to identify intellectually superior black children was evaluated. Findings suggest that less commonly used identification methods, such as parents' opinion and the Leiter International Performance Scale, may point out these children more accurately than do traditional measures. (Author/PP)

Recent empirical evidence for the influence of examiner's race on examinee's performance on intelligence tests is reviewed. The current literature, 1966 through 1980, offers little support for the hypothesis that examiner's race has a systematic effect on examinee's performance on intelligence tests. Conceptual and methodological issues are…

Provides rationale for computer-assisted testing and assessment, describes systems currently in use, and examines potential benefits, potential problems, and future implications. Suggests that computer-assisted testing and assessment has potential to either enhance services and resources or to generate problems and confusion for clients and…

Subjecting the comprehension passages of the Gates-McGinitie Reading Test to readability analysis using a multiformula computer program revealed that the instrument can best be used for making comparisons between groups using the same test forms and levels in the same sequence. (FL)

Obtained normative data from 305 children on a Coding Recall Measure adapted from the WISC-R. Although coding recall has face validity as a visual memory measure, there are some contradictions to such clinical use in view of possible lack of representativeness of the sample and limited reliability and validity. (Author/JAC)

Determined the degree to which classification of students (N=512) as formal or concrete based on three Inhelder tasks administered by clinical interview method was mirrored by Longeot and Karplus-Lawson-Renner (KLR) classifications. Results indicate that Longeot and KLR differentiate between concrete/formal operations and are discussed from…

Discusses the meaning of intelligence, the social functions that tests serve, the appropriate use of personality tests, controversies regarding IQ measurement, minimum competency testing, test disclosure, test bias, and "truth in testing." Stresses that testing is as much a social and political issue as it is an issue of scientific measurement.…