It is more common for educational and psychological data to be nonnormal than to be approximately normal. This tendency may lead to bias and error in point estimates of the Pearson correlation coefficient. In a series of Monte Carlo simulations, the Pearson correlation was examined under conditions of normal and nonnormal data, and it was compared…

In large-scale educational surveys, a latent regression model is used to compensate for the shortage of cognitive information. Conventionally, the covariates in the latent regression model are principal components extracted from background data. This operational method has several important disadvantages, such as the handling of missing data and…

There is a lack of research on the effects of outliers on the decisions about the number of factors to retain in an exploratory factor analysis, especially for outliers arising from unintended and unknowingly included subpopulations. The purpose of the present research was to investigate how outliers from an unintended and unknowingly included…

In meta-analysis, it is common to have dependent effect sizes, such as several effect sizes from the same sample but measured at different times. Cheung and Chan proposed the adjusted-individual and adjusted-weighted procedures to estimate the degree of dependence and incorporate this estimate in the meta-analysis. The present study extends the…

In exploratory common factor analysis, extension analysis refers to computing the relationship of the common factors to variables that were not included in the factor analysis. A new extension procedure is presented that gives correlations without using estimated factor scores. Advantages of the new method are illustrated. (SLD)

Various indices for measuring agreement among several raters on the presence or absence of a trait can be interpreted as intraclass correlation coefficients. Such a reformulation clarifies the relationships among the measures, simplifies the computations involved, and permits simple significance tests to be carried out. An illustrative example is…

A comparison of measures of association for 2x2 data was carried out by computer analysis. For each of 1,539 tables, 14 measures of association were calculated and evaluated. A measure based on the odds-ratio (Chambers, 1982) was most accurate in capturing the rho underlying a majority of the tables. (Author/BW)

A short description of cross-lagged correlation technique (CLC) in its most common form, a sketch of recent major criticisms of CLC as a test for causality, and a moderate appraisal of CLC as an exploratory technique are offered. (Author/PN)

A review of the development of canonical correlation is presented and insufficiencies in applications from the literature are summarized. A list of analytical procedures for use in canonical correlation is detailed. The relationship between analysis and interpretation is emphasized. An example is presented. (JKS)

META-RAND, a Statistical Analysis System macro that generates random values used for calculating correlations in a meta analysis, is described. Output consists of uncorrected and corrected mean and variances of "r" and several measures of the degree of homogeneity or heterogeneity in the data. (SLD)

Seven approaches to averaging reliability coefficients are presented. Each approach starts with a unique definition of the concept of "average," and no approach is more correct than the others. Six of the approaches are applicable to internal consistency coefficients. The seventh approach is specific to alternate-forms coefficients. Although the…

Sixteen measures of reliability for two-category nominal scales are compared. Upon correcting for chance agreement, there are only five distinct indices: Fleiss's modification of A-sub-1, the phi coefficient, Cohen's kappa, and two intraclass coefficients. Recommendations for choosing an agreement index are made based on definitions, magnitude,…

The equivalence of one-way multivariate analysis of variance and canonical correlation is well known. The usefulness of canonical correlation analysis is performing factorial multivariate analysis of variance is demonstrated. (Author)

While Spearman's rho and Kendall's tau are equally powerful rank-order correlation coefficients under conditions of normality, they have different metrics. Applied to the same data, tau is smaller in absolute value, often no more than two-thirds of the size of rho. This difference in correlational metric appears to need emphasis. (Author/CM)

Reanalysis with updated exploratory and confirmatory factor analytic methods of correlation matrices for 21 structure-of-intellect divergent production tests from the University of Southern California Aptitude Research Project suggests that the results cannot be replicated and support a higher order 2-factor oblique model of covariance among test…