The all-possible subset approach is recommended as an alternative over stepwise methods for selecting the best set of predictor variables for multiple regression. Several criteria are available for selecting the best subset model. These are compared with the principal component regression (PCR) method to investigate their usefulness for subset…

The possibility is considered that research on intelligence is entering or is about to enter a time of crisis. First, it is suggested that the decline of the psychometric paradigm as the primary means for studying intelligence was due in part to the failure of users of the paradigm to meet in a highly successful way four challenges that confronted…

Asserting that the majority of child abuse research has excluded the victim, other than as an incident, this article recommends that the problem of child abuse could be better understood if the variables related to the abused child could be adequately and empirically derived. (JC)

This note proposes an alternative implementation of the regression method which should be slightly faster than the principal components methods for estimating missing data. (RC)

In experimentation where change in X produces a change in Y, the researcher must identify, measure, and control intervening variables by experimental design or statistical procedure. (Author)

The decline of the psychometric paradigm for studying intelligence was due in part to its failure to meet four challenges. On the surface, users of the information-processing paradigms seem successfully to have met these challenges, but at a deeper level, the level of success is not so great. (Author/BW)

Four recent factor analytic studies of the Litwin and Stringer Organizational Climate Questionnaire (LSOCQ) are compared. Although there is somewhat more intra- than inter-organizational replicability of factors, both comparisons raise considerable doubt about the validity of the Litwin and Stringer instrument. (Author)

Examines research currently being conducted by some stakeholders in the name of valid distance learning research. Asserts that such research essentially repeats the mistakes of prior media studies that use the long-discredited media comparison approach (which compares distance courses with traditional campus-based courses). (EV)

The Center for Research on Education Outcomes (CREDO) at Stanford University conducted a large-scale analysis of the impact of charter schools on student performance. The center's data covered 65-70% of the nation's charter schools. Although results varied by state, 17% of the charter school students have significantly higher math results than …

This article is a reflection on an area of particular interest in the current research environment, but which has not yet been explored satisfactorily in the education literature: the evaluation of educational research. The particular focus is on the UK context, but the article is informed by comparative evidence from six countries (gathered…

This repeated measures factorial design study compared the results of two procedures for identifying multivariate outliers under varying conditions, the Mahalanobis distance and the Andrews-Pregibon statistic. Results were analyzed for the total number of outliers identified and number of false outliers identified. Simulated data were limited to…

Although it is widely known that special assumptions are needed for univariate analysis of repeated measures data, researchers seldom examine their data for violation of these assumptions. This paper reviews ways in which repeated measures analyses are usually handled and describes limitations of these methods. A design with two within subject…

Although the mail survey has been widely criticized as a viable research methodology, it remains the most feasible approach for retrieving data from large, widely-dispersed populations. This paper suggests that there exists a readily accessible body of pragmatic recommendations which, if adhered to throughout the instrument design and data…

Presents a new approach for estimating missing observations together with the results of a Monte Carlo study of the relative strengths and weaknesses of this technique and three other available methods. These techniques are then examined with respect to their ability to use incomplete data to estimate the correlation matrix obtained using a full…

Four potential biases incurred by using a sample of wives of male respondents as a sample of women for sex comparisons of the occupational attainment process were examined and empirically assessed. Using data from the Wisconsin longitudinal study of high school seniors, the extent and nature of two potential biases were considered: (1) a sampling…