Presents a critical analysis of research explaining dropout in distance education, and suggests the direction of future studies needed to form new theories regarding dropout. Highlights include discussions of mediated communication, audio teleconferencing, and problems with current research methodology and conceptual models. (LRW)

Notes that problems of measurement in educational research are often addressed by using the terms validity and reliability. Discusses the meaning given to these terms and how they relate to the aims of the measurement process in educational research. (RKM)

Describes observation research conducted by the author in a state institution for the mentally retarded. Discusses the moral and ethical dilemmas that occurred when the author witnessed routine abuse to patients by the institution's attendants. (BSR)

Challenges the typical treatment of causal effects in longitudinal data, arguing that models should be conceptualized and tested in ways that directly reflect prior assumptions as to the trait- or state-like nature of the variables. Examples demonstrate that meaningful longitudinal studies of state variables can be conducted without assuming their…

Examined difficulty of obtaining valid informed consent from participants in research with older persons. Of 470 nursing home residents asked to participate in nonrisk study, 406 agreed. High consent rate was attributed to perception of research significance, presentation format, and absence of risk or pain. (Author/NB)

Seven papers on technical and practical issues in equating are presented. Problems related to the use of conventional and item response theory equating methods, using pre- and post-smoothing to increase equipercentile equating's precision, and linear equating models for common-item nonequivalent-population design are discussed. (SLD)

Hypothesizes that, because education is a multivariant process, research attempting to test hypotheses about univariant data will find "nonsignificant" results; several nonsignificant variables may cumulatively or in combination have a significant effect; and many "nonsignificant" findings result from inadequate research design and data analysis.…

Comments on Bell's article "The Principles of Economics from Now until Then," responding to the author's two general premises as well as citing specific arguments. Concludes that Bell's premises are based on a small and unrepresentative number of textbooks. Many textbooks that present material according to the recommendations were…

Examined matching-to-sample procedures that might result in emergence in preschool children of conditional behavior that was never explicitly taught. Children were asked to match a novel and a familiar picture to a spoken word which was English or nonsense. (SKC)

Presents brief descriptions of research methodologies by social studies educators. Offers examples of the potential research questions each method may suitably address. Concludes by listing several ways to improve the quality of research conducted by each method. (Author/GEA)

Explores the differences in estimating the impact of school desegregation on residential integration. Definitions of residential integration and the formulae by which they are calculated will substantially affect conclusions. (LHW)

A review of the conceptual and methodological considerations in assessment and treatment of children's fears and phobias is presented. Areas discussed include the definitional problems in the field, analogue and clinical research, experimental research methods, and criteria for the evaluation of outcome in fear and phobia research. (Author/DWH)

Thissen and Wainer (EJ 284 848) suggested that the introduction of a prior distribution for the lower asymptote may alleviate problems of a large standard error of the location parameter of the three parameter logistic model. The correctness of this suggestion is demonstrated in detail. (Author/BW)

Although the primary responsibility rests with the authors of articles reporting nonsignificant results to demonstrate the worth of the results by discussing the power of the tests, consumers should be prepared to conduct their own power analyses. This article demonstrates the use of power analysis for the interpretation of nonsignificant…

An analysis of all the articles dealing with human or social characteristics or interaction appearing in "Educational Administration Quarterly" during the 1970s determined that inadequate consideration was given to the role of women in both the conduct of the research reported and the form in which the reports were presented. (PGD)