In Savalei's (2011) simulation that evaluated the performance of polychoric correlation estimates in small samples, two methods for treating zero-frequency cells, adding 0.5 (ADD) and doing nothing (NONE), were compared. Savalei tentatively suggested using ADD for binary data and NONE for data with three or more categories. Yet, Savalei's…

Parallel analysis (PA) assesses the number of factors in exploratory factor analysis. Traditionally PA compares the eigenvalues for a sample correlation matrix with the eigenvalues for correlation matrices for 100 comparison datasets generated such that the variables are independent, but this approach uses the wrong reference distribution. The…

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…

Grading on the curve is a common practice in higher education. While there are many critics of the practice it still finds wide spread acceptance particularly in science classes. Advocates believe that in large classes student ability is likely to be normally distributed. If test scores are also normally distributed instructors and students tend…

A general procedure is provided for comparing correlation coefficients between optimal linear composites. It allows computationally efficient significance tests on independent or dependent multiple correlations, partial correlations, and canonical correlations, with or without the assumption of multivariate normality. Evidence from Monte Carlo…

The use of U-statistics based on rank correlation coefficients in estimating the strength of concordance among a group of rankers is examined for cases where the null hypothesis of random rankings is not tenable. (Author/BW)

A Monte Carlo evaluation of four test statistics for comparing dependent zero-order correlations was conducted with four sample sizes and three population distributions. Results indicate that choice of optimal test statistic depends on sample size and distribution, and predictor intercorrelation and effect size or magnitude of the…

Proposes a procedure for generating multivariate nonnormal distributions. The procedure, an extension of the Fleishman power method (A. Fleishman, 1978), generates the average value of intercorrelations much closer to population parameters than competing procedures for skewed and heavy tailed distributions and small sample sizes. Reports Monte…

A modification of Horn's parallel analysis is introduced that is based on the Monte Carlo simulation of the null distributions of the eigenvalues generated from a population correlation identity matrix. This modification reduces the tendency of the parallel analysis procedure to overextract or to extract poorly defined factors. (SLD)

Provides an overview of standard and modified normal theory and asymptotically distribution-free covariance and correlation structure analysis techniques and details Monte Carlo simulation results on Type I and Type II error control. Demonstrates through the simulation that robustness and nonrobustness of structure analysis techniques vary as a…

Few methods have been tried and used to graphically represent more than two variables. This poster session showed a new method for representing three continuous variables on a single scatterplot using the THREEDE computer program. Two variables are represented as a normal bivariate distribution. The third variable is represented by a symbol, e.g.…

A comparison is made of the power and actual alpha levels of three tests of homogeneity for independent product-moment correlation coefficients using Monte Carlo methods while selectively studying sample size and varying the number of correlation reliabilities. How robust these are in applied work is discussed. (SLD)

Intraclass correlation and Cronbach's alpha are widely used to describe reliability of tests and measurements. Even with Gaussian data, exact distributions are known only for compound symmetric covariance (equal variances and equal correlations). Recently, large sample Gaussian approximations were derived for the distribution functions. New exact…

Validity generalization meta-analysis (VG) examines the extent to which the validity of an instrument can be transported across settings. VG offers correction and summarization procedures designed in part to remove the effects of statistical artifacts on estimates of association between criterion and predictor. By employing a random effects model,…

By employing a concomitant variable, researchers can reduce the error, increase the precision, and maximize the power of an experimental design. Blocking and analysis of covariance (ANCOVA) are most often used to harness the power of a concomitant variable. Whether to block or covary and how many blocks to be used if a block design is chosen…