Critiqued and reanalyzed a recent study of the validity of five Minnesota Multiphasic Personality Inventory (MMPI) alcoholism scales. The original study concluded that these scales were invalid in terms of classifying patients in regard to alcoholic status. Reanalysis provides evidence for the validity of four of the five scales. (Author/JAC)

Focuses on Binomial Effect Size Display (a concept helpful in interpreting size of an experimental effect). Suggests that both statistical significance and explained variance be reported in experimental research and that values for the latter concepts can be viewed with a more accurate perspective by using BESD. (Author/JN)

The assumption of equal treatment implementation is questioned. Through the reanalysis of data from a nutrition supplementation program evaluation, the power of the analysis of treatment effects is shown to increase when data on the level of treatment implementation is included. (Author/CM)

A reliability study is illustrated in which subjects are judged on a dichotomous trait by different sets of judges, possibly unequal in number. A kappa-like measure of reliability is proposed, its correspondence to an intraclass correlation coefficient is pointed out, and a test for its statistical significance is presented. (Author/CTM)

A process is described for multiple comparisons when covariates are involved in the analysis. The method can be accomplished with considerable ease whenever pairwise comparisons are involved. More complex contrasts require the use of full and restricted models of variance. (CTM)

Raw scores for the clinical scales of MMPI can be changed to T scores by linear transformation. If the underlying distributions of raw scores are not distributed normally, one cannot assume the traditional point of significance. Suggests that clinicians consider more conservative interpretation for some scales. (Author)

When reporting results, researchers must not change predetermined significance levels. Such attempts to make results more significant are statistically inaccurate, illogical, and unethical. American Psychological Association standards for reporting significance should be more explicit. (CP)

A procedure for approximating attained significance levels of exact conditional tests is proposed. The procedure utilizes a sampling from the null distribution of tables having the same marginal frequencies as the observed tables. (Author/JKS)

The negative bias of biserial correlations in the presence of guessing is demonstrated by Monte Carlo studies, and another estimator is described that is free of such bias. The usual biserial-correlation assumptions and assumptions about guessing are discussed. Brogden's coefficient of selective efficiency and the triserial correlation are…

A Monte Carlo technique was used to investigate the effect of the differences in covariate means among treatment groups on the significance level and the power of the F-test of the analysis of covariance. (Author/GDC)

The importance of measuring the size of an effect for fixed effects factorial analysis of variance designs is emphasized. Technical issues in such measurement are considered and examples are provided. (Author/JKS)

This article describes the organization and results of the 1990 New Zealand Children's Census. This census, which preceded a National Census, provided a powerful tool for raising statistical awareness nationwide. (Author)

An Edgeworth approximation for accurate significance probabilities for the Wilcoxon two-sample test is substantially simplified. A method is developed that allows quick calculations of very accurate probabilities. Exact formulas are given for most of the remaining cases, and tables are presented comparing the new simplification to likely…

One source of bias, regression artifacts, in program evaluations that use regression discontinuity design is identified, and its importance is discussed. A method is suggested for its correction. Although use of this method does not solve the problem of selection bias, it does improve the reliability of nonexperimental methods. (SLD)

The performance was studied of five small-sample statistics--by F. M. Lord, W. Kristof, Q. McNemar, R. A. Forsyth and L. S. Feldt, and J. P. Braden--that test whether two variables measure the same trait except for measurement error. Effects of non-normality were investigated. The McNemar statistic was most powerful. (TJH)