Contrary to Anderson (EJ 187 936), his rule for equation identification is a necessary but not sufficient condition; furthermore, the choice of two-stage or ordinary least squares depends on results and not on methodological properties of estimators. Modification of Anderson's rule and a means for choosing between estimates is offered. (Author/CP)

Technical assistance given to facilitate implementation of continuous progress programs in a Georgia elementary school system is discussed in terms of long-range goals for staff development and general objectives for a continuous progress model and its mainstreaming program. (DS)

Practical methods for implementing true experimental designs in evaluation settings in which such designs are rarely used are presented. Particular attention is paid to educational settings. (Author/JKS)

Differences in approaches to research into student learning are analyzed in terms of differences in the conception of six aspects of the research process. It is argued that underlying various strategies there is a variation in perspective, description, conceptualization, relations of categories, comprehension, and application of findings.…

True experimental designs are thought to provide internally valid results. In this investigation of the evaluations of five interventions, a source of internal invalidity is identified when self-report measures are used. An alternative approach is presented and implications of the findings for evaluation research are discussed. (JKS)

The developmental, defensive, and role-playing theories of parental identification were tested on high school seniors. Results supported the developmental (warmth) hypothesis but were inconsistent with the defensive and role-playing hypotheses. Previous studies on identification are cited. (Author/SJL)

Discusses the purposes, trends, problems, and implications of the media attitude research published in AV Communication Review during its 25 years. Abstracts are included for 68 such studies and references are cited directing the reader to studies originally published elsewhere but abstracted in AVCR.

Advantages and disadvantages of nonexperimental paradigms in mathematics education are discussed. Suggestions are made for action toward establishing such designs as quality research. (MP)

Asserts that the emphasis on explaining individual differences in political orientations among preadults accounts for "the disappointing contribution of political socialization research to political understanding." Argues that "possessing an explicit model of democratic citizenship would provide greater political meaning to empirical data."…

An experiment is reported on the effects of a moral education program based on cognitive-developmental theory, at the high school level. Students were pre-tested and post-tested on Kohlberg's Moral Judgment Scale. (Editor)

Advocates that the Current Population Survey (CPS) be redesigned to permit valid longitudinal studies, that the CPS include questions to evaluate poverty programs, that a common sampling framework be designed for the CPS and operating agencies serving the poor to assure that links between CPS and other data can be established, and that samples of…

Intends to identify some of the problems associated with evaluation of a province-wide program, to describe how these problems were dealt with in the particular case under review, and to suggest how evaluators and their clients might approach problems of a similar nature. (Author/RK)

Focuses on selecting learning experiencing which allow students direct involvement in the discussion of issues most important to them. Reports research findings from a 1971 study of marital role expectations of high school students in a large Alberta city and outlines teaching strategies which utilize this data. (Author/RK)

Compared 26 large-group research grants (PO1) and 66 individual research grants on mental retardation, focusing on success rates, productivity, costs, impact, publications, and outcome. Results found no overwhelming evidence that one funding mechanism is more effective than the other; however, PO1s had a higher success rate. (Author/CR)

Monte Carlo methods were used to estimate the power of randomization tests used with single-case designs involving random assignment of treatments to phases. Simulations of two treatments and six phases showed an adequate level of power when effect sizes were large, phase lengths exceeded five, and autocorrelation was not negative. (SLD)