The background essay introducing this special issue on leadership describes potential initiatives aimed at educational leadership research and development. The research supporting the invited journal articles was conducted at two national education leadership research centers. (MLF)

A method is proposed for calculating the sample size needed to achieve acceptable statistical power with a given test. The minimally important difference significant (MIDS) criterion for sample size is explained and supported with recommendations for determining sample size. The MIDS criterion is computationally simple and easy to explain. (SLD)

Describes characteristics of second-generation research in early intervention, that is, greater specificity, the use of more rigorous scientific standards, and an increased understanding of biosocial relationships than in initial investigations. Examines studies conducted by Utah State University's Early Intervention Research Institute in the…

Examines Richard Clark's analysis of the role of media (IR 529 139) from the perspective of an instructional designer. Clark's statements about interchangeability of media attributes, research evidence, and the influence of learner preconceptions are challenged. Suggestions for future research are provided. (Contains six references.) (KRN)

Offers a rebuttal to Giyoo Hatano's commentary on a study on sociomoral development in the kindergarten classroom. It is suggested that the danger in restricting evaluation to stated goals is that unintended negative effects will not be recognized. (LB)

A review of 12 studies comparing computer-assisted (CAI) and traditional instruction in adult basic and secondary education (ABSE) found that, in 10 of 12, CAI results were at least as good as traditional and CAI also aided retention, self-confidence, privacy, feedback, and faster learning. CAI's place in ABSE is confirmed, but it is not a miracle…

Argues educational researchers must recognize the need for a dialectic between theory and method. Proposes a quasi-historical methodology based on a theory of social action. Suggests a means for understanding the meanings of actions. Outlines a schema for the interpretation of text and offers researchers a form of disciplined inquiry. (Author/NL)

Provides a brief history of research in word processing and writing, describes the use of word processing as a writing aid, and discusses the problems and issues in research design in the area of word processing instruction. (MG)

Four approaches to estimating interrater reliability in early childhood special education research are illustrated and compared: correlation, comparison of means, percentage of agreement, and generalizability theory techniques. Generalizability theory techniques are proposed as a method for estimating the amount of variance attributable to…

Argues that writing instructors have a significant role to play in producing important computer-writing research. Asserts that the insights these instructors gain from teaching inform their research and guide their research practices by allowing them to distinguish subtle patterns and questions worthy of investigation. (PRA)

Tests hypothesis that people recall their expectations in ways that are consistent with outcomes. Finds that autobiographical recall may be generally reconstructive and potentially biased. Canadian students reported their preparations and expectations for an undergraduate examination. Observes that more than 90 percent of the students did not…

Design, analysis, and use of an interview task to identify preservice teachers' conceptions of teaching science are discussed. A copy of the task ("Interview-about-instances of science teaching") is included, and selected interview responses are presented and discussed. Results are discussed with regard to implications for research and…

Research over the past 15 years concerning parametric and nonparametric statistics and 10 recently developed nonparametric techniques for testing interactions in experimental design are reviewed. The new nonparametric techniques are robust, powerful, versatile, and easy to compute. A fabricated dataset is used to illustrate the technique. (TJH)

Comparative research on teachers' perspectives and classroom behavior must focus both on external systematic and institutional influences and on individual differences in teacher personality and background. One such systematic study of French and English elementary school teachers identified "meta-national" similarities and international, local,…

A review of 208 studies of children with learning disabilities (LD) published in 10 major journals between 1988 and 1990 examined the extent to which recommendations concerning research design and reporting made in 1980 and 1984 have been implemented. The review found continuing problems, including imprecise subject descriptions, heterogeneous…