Joreskog's model of cogeneric tests is used to analyze agreement between raters. Raters are treated as measuring instruments. The model of cogeneric tests, of which classical parallelism and tau-equivalence are shown to be special cases, is applied to teachers' ratings of students' responses on open-end questions. (Author/RC)

A discussion of good test criteria reviews the basic concepts of test theory, examines four types of validity, outlines the concept of reliability and its coefficients and limitations, makes suggestions for gauging test quality, and demonstrates use of the standard error of measurement for estimating the likelihood of misgrading. (MSE)

Issues concerning equating test scores are discussed in an introduction, four papers, and two commentaries. Equating methods research, sampling errors, linear equating, population differences, sources of equating errors, and a circular equating paradigm are considered. (SLD)

The Mann-Whitney U Test provides an alternative to the parametric test when applied to the kinds of ordinal data typically generated by semantic differential and Likert-type scales. The application of this test to survey research problems in business education is discussed, and a computer program is given for the use by the business education…

This review of issues about correcting for attenuation concludes that the basic difficulty lies in being able to identify and equate sources of variance in estimates of validity and reliability. Recommendations are proposed for cautious use of correction for attenuation. (Author/CM)

A sample of the literature on test speededness is reviewed; methods of assessing speededness are presented and criticized; the assumptions that underlie these methods are questioned, and alternate, multiple-administration methods are suggested. The importance of the effect of time limits is discussed. (Author/CTM)

A dozen different approaches that yield 13 reliability indices for criterion-referenced tests were identified and grouped into three categories: threshold loss function, squared-error loss function, and domain score estimation. Indices were evaluated within each category. (Author/RL)

Statistical analysis tools in language testing are described, chiefly classical test theory and item response theory. Computer software for statistical analysis is briefly reviewed and divided into three tiers: commonly available; statistical packages; and specialty software. (Author/VWL)

e mean length of utterance in the Sentence Repetition Task grew from approximately two to three words, and the number of omissions of articles, prepositions and modifiers significantly decreased. After 3;0 years old, omissions of free function words practically disappeared. The results of Study 2 showed that mean length of utterance, omission of…

The primary purpose of this research is to determine if there exists covert confounding in student perceptionnaires which significantly affect instructors ratings. Eight calculus classes were randomly selected. Each student completed a Texas A & M University (TAMU) student perceptionnaire the first day of class, envisioning having completed the…

Two independent studies were conducted to investigate possible differences in General Aptitude Test Battery (GATB) aptitude M resulting from the use of different test equipment (wooden vs. plastic apparatus.) As part of a ten-year longitudinal study of Vocational Development being conducted in the Department of Vocational Education at The…

The absence of suitable external criteria is a recurrent problem for test, battery, and inventory developers in selecting items or tests for inclusion in final operational instruments. This report presents a computing algorithm developed for use when no adequate external selection criterion is available. The algorithm uses a multiple linear…

In determining how reliable is reliable enough and how much error can be tolerated in criterion-referenced testing, the following relationships hold: (1) the more specific an objective is, the fewer the items required to reliably measure it; (2) the more specific the objectives are, the more objectives required to cover a given span of the…

An important determinant of equating process efficiency is the correlation between the anchor test and components of each form. Use of some monotonic function of this correlation as a measure of equating efficiency is suggested. A model relating anchor test length and test reliability to this measure of efficiency is presented. (Author/DWH)