In adolescence research, the treatment of measurement reliability is often fragmented, and it is not always clear how different reliability coefficients are related. We show that generalizability theory (G-theory) is a comprehensive framework of measurement reliability, encompassing all other reliability methods (e.g., Pearson "r,"…

Because test scores are ordinal not cordinal attributes, the average test score often is a misleading way to summarize the scores of a group of individuals. Similarly, correlation coefficients may be misleading summary measures of association between test scores. Proper, readily interpretable, summary statistics are developed from a theory of…

Educational test theory consists of statistical and methodological tools to support inferences about examinees' knowledge, skills, and accomplishments. The evolution of test theory has been shaped by the nature of users' inferences which, until recently, have been framed almost exclusively in terms of trait and behavioral psychology. Progress in…

In almost all applications of measurement there is some sort of response by a human subject. Almost always, the response scale is ordinal, but almost always it is treated as if it were an interval measure. Methods for treating data ordinally are currently being developed in three areas: ordinal analysis for questionnaire responses, ordinal…

This paper traces the course of the consequences of viewing test responses as simply providing dichotomous data concerning ordinal relations. It begins by proposing that the score matrix is best considered to be items-plus-persons by items-plus-persons, and recording the wrongs as well as the rights. (Author/CTM)

Investigates the general factor and uniqueness characteristics of the individual tests of the Woodcock-Johnson Test of Cognitive Ability-Revised (WJTCA-R). Only 2 of the 19 WJTCA-R tests examined had low general factor loadings, while 2 had low uniqueness. All other tests had medium or high uniqueness. Discusses implications for clinical…

The literature on criterion referenced testing is full of discussions concerning whether classical measurement techniques are appropriate, whether variance is necessary, whether new indices of reliability are needed, and the like. What appears to be lacking, however, is a clear and simple discussion of why the problems occur. This paper suggests…

The thesis of this paper is that the decision to use one of three approaches to unit-of-analysis in educational research should be based on substantive considerations, not statistical factors. In addition to the commonly used "total analysis" (regression analysis across individuals), the within and between analyses are inherent in the…

This text was written to help the reader acquire a base of knowledge about classical psychometrics and to integrate new ideas into that framework of knowledge. The material is organized into five units: (1) introduction to measurement theory; (2) reliability; (3) validity; (4) item analysis in test development; and (5) test scoring and…

The Kuder Richardson-20 Formula is shown to be a special case, where each examinee is given sufficient time to answer each item, of a more general formula where each examinee may not be allowed the necessary time. The formula is extended to allow two scores, knowledge and speed, to be extracted from each examinees test score. Using a sample of 82…

The family of Weibull distributions was investigated as a model for the distributions of response times for items in computer-based criterion-referenced tests. The fit of these distributions were, with a few exceptions, good to excellent according to the Kolmogorov-Smirnov test. For a few relatively simple items, the two-parameter gamma…

A procedure is demonstrated for constructing tables showing, for each score on a commercial reading achievement test, the percentage of real-world materials that the testee is likely to comprehend with at least a criterion level of proficiency, the percentages of students in a local or national sample who can competently comprehend a given…

Three approaches to the construction of achievement tests are compared: construct, operational, and empirical. The construct approach is based upon classical test theory and measures an abstract representation of the instructional objectives. The operational approach specifies instructional intent through instructional objectives, facet design,…