Discusses some models, including the Petersen Novick Model (TM 502 259) regarding fair selection procedures. (DEP)

When test data from different sources conflict, the truth is often found somewhere in the middle. As this article shows, reports of achievement trends can be heavily influenced by the particular definition of achievement represented and by testing conditions. Recent progress in science and mathematics must be carefully nurtured in the nineties.…

Two problems in the investigation of predictive bias in tests, the effect of unreliability of the predictors, and the effect of excluding a predictor from the regression equation on which there are preexisting group differences, are discussed. (Author)

The validity of interpretations of National Assessment of Educational Progress (NAEP) achievement levels is evaluated by focusing on evidence regarding three types of discrepancies: (1) between standards; (2) among descriptions of achievement levels; and (3) between assessments and content standards. All of these discrepancies raise serious…

It is argued that consequential validity is a concept worth considering. The solution to defining "validity" is not to narrow the concept, but to allow for the differential prediction provided by tests in different circumstances. Consequences of the uses and interpretations of test scores are central to their evaluation. (SLD)

The term "criterion-referenced measurement" has been so frequently used that accumulated surplus meaning has sometimes swamped the fundamental concept. Stripping away extraneous interpretations affirms the contemporary relevance of the concept. Substituting the term "construct-referenced" underscores the need for better…

Generalizability of performance-based assessment scores across raters and tasks is examined, focusing on implications of generalizability analyses for specific uses and interpretations of assessment results. Although it seems probable that assessment conditions, task characteristics, and interactions with instructional experiences affect the…

Political and educational consequences of standard setting for criterion-referenced tests are discussed. Suggestions for setting standards are given. (JKS)

Many of the recent appeals for educational improvement place great reliance on standardized tests as instruments of reform. The appropriateness and interpretation of these tests remain unresolved issues. Concepts of cognitive psychology should be integrated into the development of intelligence and achievement tests. (Author/VM)

During the past 6 years, under a contract from the National Center for Education Statistics, a Technical Review Panel has overseen and conducted a series of research studies addressing a range of validity questions relevant to the various uses and interpretations of the National Assessment of Educational Progress (NAEP). Study topics included: (1)…

Techniques to identify the degree to which a response pattern is unusual and the pattern's relationship to examinee background are presented. Sato's caution index and modified caution index use clusters of items for comparisons of observed performance outcomes and do not require the use of item response theory. Correlation between the indices was…

This study investigated the adequacy of the Rasch model in equating existing standardized tests with groups of examinees not widely separated in ability. With the exception of one test pair and one grade level, the Rasch model using the anchor test procedure provided a reasonably satisfactory means of equating. (Author/CTM)

The most fundamental technical quality issue raised by proposals for a national examination system is that of validity. Types of evidence needed to judge the validity of test use and the interpretation of results are discussed, and examination content, fairness, impact, generalizability, and comparability are explored. (SLD)

Four main approaches to customized testing are described, and their resulting scores' valid uses and interpretations are discussed. Customized testing can yield valid normative and curriculum-specific information, although cautious application is needed to avoid misleading inferences about student achievement. (SLD)