Considers both validity and reliability simultaneously in selecting the best short forms (SFs) of the Wechsler Adult Intelligence Scale-Revised (WAIS-R). Results indicate that incorporating reliability as a criterion has a dramatic impact on the obtained best SFs. (LLL)

Reports the validities and reliabilities of two short forms of the Wechsler Adult Intelligence Scale (Revised) (Vocabulary and Block Design, and Arithmetic and Picture Arrangement) for each of nine age groups, together with standard errors of estimate and measurement. Results support the use of these forms for their intended purpose. (BH)

An adjustment to the Rand statistic, which will allow comparison across different levels for number of clusters found within a classification, is proposed. This adjustment permits the user to compare several different classifications with respect to classification agreement, while correcting for the contribution of chance to any observed…

A rationale for, and a relatively simple method of estimating the reliability of an observational method are described. A separate "reliability study," while not required, may be employed when examining data. Empirical findings show that high frequencies of occurrence are not necessary prerequisites for the reliable measurement of behavior.…

Maximally reliable composites found in canonical reliability when expressed in the form of a canonical factor analysis solution are shown to have highly desirable data reduction properties. Theoretical relationships among canonical factor analysis, principal components analysis and canonical reliability analysis are emphasized. (Author/JKS)

Too many researchers speak of "the reliability of the test," thus indicating their basic misunderstanding of reliability. This paper explains classical reliability and the score features that influence coefficient alpha. It explains when coefficient alpha can be negative, even though it is conceptually a variance-accounted-for statistic.…

Restriction of range occurs whenever design or circumstances abbreviate the values of one or both variables being correlated. Range restriction has an impact on score validity, score reliability, and statistical power. As a result, researchers need a more sophisticated concept of the relationship between variance and correlation than the simple…

Because reliability is a function of scores, and not tests per se, it is inaccurate to hold that a given test will yield scores with the same reliability across samples. Therefore, score reliability should always be reported and interpreted in both measurement and substantive studies. In an effort to facilitate this outcome, this paper is intended…

The Delphi method is a means of structuring group communication process so that a group of experts can gather information or forecast future problems effectively. A primary objective of a Delphi study is to obtain consensual and consistent opinions from a group of experts in two or more successive rounds on a given research subject. Consensus and…

Many researchers fail to understand that reliability is a function of scores, not tests. This paper provides an explanation of the distinction as well as a description of the reliability generalization meta-analysis technique. Reliability generalization meta-analysis can provide a way to aggregate test score reliability coefficients from prior…

The purpose of this paper is to illustrate how displaying disattenuated correlation coefficients along with their unadjusted counterparts will allow the reader to assess the impact of unreliability on each bivariate relationship. The paper also demonstrates how a proposed new "what if reliability" analysis can complement the conventional null…

Discusses weaknesses in Weiss' arguments and techniques for the nonlinear predictor of academic achievement. (MB)

Results of this study indicate that the correlation between half-test scores over repeated splits, over persons, and over repeated testings resulting in different sets of observed scores, is given by Kuder-Richardson Formula 21. (RF)