Concludes that the rhetorical specification approach to developing essay examination topics (giving writers information about the purpose of a composition, its audience, speaker, and subject) may not be the best way to elicit good writing. (JL)

Asserts that some student products, performances, and skills are not suited to evaluation using paper-and-pencil tests. Describes three alternative methods for evaluating student achievement using rating scales, checklists, and anecdotal records. (PGD)

For tests used to make pass/fail decisions, the relevant standard error of measurement (SEM) is the SEM at the passing score. If the test is highly stratified, this SEM should be estimated by a split-halves approach. A formula and its derivation are provided. (Author)

The purpose of this study was to determine the extent to which Crockett's Role Category Questionnaire (RCQ) is confounded by verbal ability. Participants included 285 undergraduates from a large southern university. The results show the RCQ to be independent of the verbal abilities measures. (Author/BW)

The assumption of equal treatment implementation is questioned. Through the reanalysis of data from a nutrition supplementation program evaluation, the power of the analysis of treatment effects is shown to increase when data on the level of treatment implementation is included. (Author/CM)

The emphasis placed on test scores in college admissions is discussed. The need for colleges to periodically analyze and define their admissions policy is recommended. (Author/AL)

The sensitivity and simplicity of Wald's sequential analysis test in monitoring a preventive health care program are discussed. Data exemplifying the usefulness and expedience of employing sequential methods are presented. (Author/GK)

The use of Bayesian and empirical Bayesian techniques to summarize results from parallel randomized experiments is illustrated using the results of eight such experiments from an SAT coaching study. Graphical techniques, simulation techniques, and methods for monitoring the adequacy of model specification are highlighted. (Author/JKS)

It is argued here that the choice of the appropriate method for calculating least squares analysis of variance with unequal sample sizes depends upon the question the experimenter wants to answer about the data. The different questions reflect different null hypotheses. An example is presented using two alternative methods. (Author/BW)

Believes extra attention must be focused on language and cultural differences in preparing Individualized Educational Programs. Reports observations from working with Spanish speaking exceptional children in Costa Rica. Describes pilot study conducted to attempt to discover problems associated with administration of Wechsler Intelligence Scale for…

Subjected the standardization data for the Wechsler Adult Intelligence Scale-Revised (WAIS-R) and the original Wechsler Adult Intelligence Scale (WAIS) to principal-factor analysis. A two-factor solution was adopted for each scale. The stability of the two factors, Verbal Comprehension and Perceptual Organization, was high both within and between…

Describes two broad-band factor scales of marital distress constructed to supplement existing profile scales of the Marital Satisfaction Inventory. The two new scales, labeled Disaffection and Disharmony, both discriminated between normative and clinical samples. Distinct distributions support the concept of two separate, interactive components of…

Although Model A, the only norm-referenced evaluation procedure in the Title I Evaluation and Reporting System, requires no data other than the test scores themselves, it introduces two sources of bias and involved three test administrations. Roberts' two-test procedure offers the advantages of less bias and less testing. (RL)

The purpose of this article is to review the essential features of the Perkins-Binet Tests of Intelligence for the Blind, and to discuss problems and limitations in the test kit and manual. Recommendations are also made for smoother administration of both forms of the test. (Author/BW)

Issues related to the decision of the number of factors to retain in factor analyses are identified. Three widely used decision rules--the Kaiser-Guttman (eigenvalue greater than one), scree, and likelihood ratio tests--are investigated using simulated data. Recommendations for use are made. (Author/JKS)