The emphasis on the mechanics of research techniques has often overshadowed other aspects of evaluation research that are determiners of research quality. Many problems encountered by program evaluators are not solvable by the information provided in research texts. Some of the practical problems encountered and dealt with during an experimental…

One of a series of sixteen knowledge transformation papers, this paper examines the advantages and disadvantages of longitudinal studies as a method for evaluating vocational programs. First, longitudinal methods are defined for the differences between them and cross-sectional studies are established. The results of a literature search conducted…

The degree to which the chosen units of analysis are likely to produce spurious findings in staged combinations of multiple linear regression procedures are examined. The effects of grouping variables (e.g., classroom, school, and school district) on Procedures such as Coleman's semipartial regression and Mayeske's commonalities, in light of…

Problems encountered in large-scale educational evaluations are discussed using the evaluation of the Emergency School Aid Act as an example. This extensive article focuses primarily on methodological issues. (JKS)

A critical review of the research methodology used in the Wilms study on the effectiveness of public and proprietary vocational training. Focus is on the overall design, data collection, data analysis, and derived conclusions and implications, with the conclusion that the study has some significant design and methodological limitations. (Editor/JT)

The definition of progress as applied to writing research and practice is examined, and three different approaches to understanding the writing of exceptional children are considered: a literal approach, an interpretive approach, and an evaluative approach. The scientific and social implications of each of the approaches are considered.…

Considers possible contributions of developmental psychology to analysis and solution of practical problems by proposing a practical action model comprising the following: nature of the phenomenon or problem; existence of the phenomenon or problem; outcome in case of no intervention; goals for intervention; achievement of goals; and assessing…

Within-class regression is a method, developed in this paper, of comparing a large number of nonequivalent groups. This study indicated that within-class regression was a less biased method of data analysis and will yield more accurate estimates of treatment effects than analysis of covariance. (PN)

Critiqued is the article, "Teacher as Shaman: An Educational Criticism," (Studies in Art Education, v25 n1), containing a case study of an art teacher who employs performance-based ritual to stimulate thinking about art's meaning and history. The problem that ethnographic researchers have conveying credibility when summarizing…

The main purpose of this article is to generate a functional model of evaluation that EMSs can be able to empower online communications characterized by imperative decision making task. The evaluation process of EMSs must merge the multicultural strategies of the theory of "Media Richness," and the ethical concerns of the critical approach. "Media…

Problems with experimental research on mental retardation include: (1) difficulties with defining mental retardation; (2) neglect in demonstrating causal relationships; and (3) reliance on methodology not necessarily appropriate for defining the criteria of retardation. The author advocates abandoning definitions based on intelligence tests and…

Presents both logical and empirical evidence to illustrate that the conventional scoring algorithm for ranking tasks significantly underestimates the initial level of group ability and that Slevin's alternative scoring algorithm significantly overestimates the initial level of ability. Presents a modification of Slevin's algorithm which authors…

The biasing effects of a pretest on subsequent post-test results were investigated in two experimental studies. In general, the results argue for using designs without pretests. (Author/JKS)

Ten articles on the use of multilevel data in educational research are presented. Topics include model specification and building, analytical procedures, aggregation bias, least squares statistics, hierarchical linear models, significance testing, and testing multilevel product-process networks. (TJH)