Notes that because estimates of effect sizes are often either misreported or not reported at all, meta-analysts must use conversion formulas that allow estimates of effect sizes from information available. Focuses on formulas that convert "F" in ANOVA, a statistical test, to eta-squared, "d," or the correlation equivalent. Demonstrates that the…

Current practices of reporting a single mean intercoder agreement in meta-analysis leads to systematic bias and overestimates reliability. An alternative is recommended in which average intercoder agreement statistics are calculated within clusters of coded variables. Two studies of intercoder agreement illustrate the model. (SLD)

A distribution theory is derived for a G. V. Glass-type (1976) estimator of effect size from studies involving paired comparisons. The possibility of combining effect sizes from studies involving a mixture of related and unrelated samples is also explored. Resulting estimates are illustrated using data from previous psychiatric research. (SLD)

In contrast to the average effect sizes of the approach to metanalysis that can be thought of as literature synthesis, an effect-size surface is proposed as a function of scientifically relevant factors, estimated by extrapolating a response surface of observed effect sizes to a region of ideal studies. (SLD)

Kendall's Tau is often considered equivalent to Spearman's Rho as an ordinal measure of correlation in spite of its different metric. Formulas for converting Tau to Rho are reviewed; and a table of corresponding values is presented for Tau, Rho, and several related indices. (SLD)

Parametric measures to estimate J. Cohen's effect size (1966) from a single experiment or for a single study in meta-analysis are investigated. The main objective was to examine the principal statistical properties of this effect size--delta--under variance homogeneity, variance heterogeneity with known variance ratios, and for the Behrens-Fisher…

Combining information to estimate standardized partial regression coefficients in a linear model is discussed. A combined estimate obtained from the pooled correlation matrix is proposed, and its large sample distribution is obtained. The method is generalized to handle a random effects model in which correlation parameters vary across studies.…

A meta-analysis of studies conducted between 1974 and mid-1987 on sex differences in mathematical tasks is presented. Methods used include estimations of: (1) parameters for a random effects model, and (2) coefficients for a linear-regression equation, all based on effect sizes calculated from each study. Differences have been decreasing. (TJH)

The observed probability "p" is the social scientist's primary tool for evaluating the outcome of statistical hypothesis tests. The small-sample accuracy of nonnull asymptotic distributions of several functions of "p" was studied. Implications for use of the approximations are discussed. (SLD)

Results obtained in a metanalysis using indirect alternative methods of computing standardized mean difference effect sizes (SMDESs) were compared with results obtained from direct calculations of SMDESs. In most cases, alternative methods yield very comparable results and are beneficial to use, especially when sample sizes are large. (SLD)

The use of statistical methods to combine the results of independent empirical research studies (metanalysis) has a long history, with work mainly divided into tests of the statistical significance of combined results and methods for combining estimates across studies. Methods of metanalysis and their applications are reviewed. (SLD)

A meta-analysis of research on the relationship between inspection time (IT) and IQ was performed. The analysis was designed to determine whether a non-zero relationship between IT and IQ exists, estimate the size of this relationship, and test whether IT is ontogenetically related to "g" (intelligence). (TJH)

Statistical methodologists have sometimes criticized the use of conventional statistics in meta-analysis, and in recent years a number of them have advocated the use of a special new statistical methodology for research synthesis. An examination of recent books describing this methodology shows that it is seriously limited in its applicability to…

Analyses for results of a series of studies examining intercorrelations among a set of as many as p+1 variables are presented. Several estimators of a pooled or average correlation vector and its variances are derived for cases in which some studies do not report complete correlation matrices. A test of the homogeneity (consistency) of the…