J. Barnette and J. McLean (1996) proposed a method of controlling Type I error in pairwise multiple comparisons after a significant omnibus F test. This procedure, called Alpha-Max, is based on a sequential cumulative probability accounting procedure in line with Bonferroni inequality. A missing element in the discussion of Alpha-Max was the…

A Monte Carlo study was conducted using the Statistical Analysis System IML computer program to compare the multivariate analysis of variance (MANOVA) simultaneous test procedures of Roy's Greatest Root, the Pillai-Bartlett trace, the Hotelling-Lawley trace, and Wilks' lambda, in terms of power and Type I error under various conditions, including…

Three techniques for estimating Kuder Richardson reliability (KR20) coefficients for incomplete data are contrasted. The methods are: (1) Henderson's Method 1 (analysis of variance, or ANOVA); (2) Henderson's Method 3 (FITCO); and (3) Koch's method of symmetric sums (SYSUM). A Monte Carlo simulation was used to assess the precision of the three…

The analysis of variance, Welch, and Brown and Forsyth tests for mean equality were compared using Monte Carlo methods. The tests' rates of Type I error and power were examined when populations were nonnormal, variances were heterogeneous, and group sizes were unequal. Recommendations for use are presented. (Author/JKS)

Monte Carlo techniques were employed to compare the familiar F-test with Welch's V-test procedure for testing hypotheses concerning a priori contrasts among K treatments. The two procedures were compared under homogeneous and heterogeneous variance conditions. (Author)

Based on a Monte Carlo simulation, this study is designed to investigate the power of the Kruskal-Wallis's H-test compared to the power of the F-test for three equal moderate sample sizes drawn at random from distributions of common or different shapes but for which the population distributions have equal variances. The distributions are the…

Compares three methods of analyzing pairwise treatment differences in a multi-treatment experiment via computer simulation techniques. Under the equal n condition, the robustness of the conventional Tukey Wholly Significant Difference test (WSD) to heterogeneous variances was contrasted with two alternate techniques. Under unequal n conditions,…

This study compared three approaches to the two-factor experiment with repeated measures on one factor: (1) the conventional mixed model analysis of variance, (2) the Greenhouse-Geisser conservative analysis of variance, and (3) multivariate extensions of analysis of variance. Computer simulated data were used in a total of 96 sets of covariance…

The Type I error and power properties of the 2x2x2 analysis of variance (ANOVA) and tests developed by McSweeney (1967), Bradley (1979), Harwell-Serlin (1989; Harwell, 1991), and Blair-Sawilowsky (1990) were compared using Monte Carlo methods. The ANOVA was superior under the Gaussian and uniform distributions. The Blair-Sawilowsky test was…

The Random Normal Scores Test (RNST) has been suggested as a powerful alternative to the use of the parametric analysis of variance (ANOVA) test when the underlying population is non-normally distributed. The major support for this suggestion rests on asymptotic theory. An empirical analysis of the RNST performed under the F and Chi-square…

A Monte Carlo Study was conducted to evaluate six models commonly used to evaluate change. The results revealed specific problems with each. Analysis of covariance and analysis of variance of residualized gain scores appeared to substantially and consistently overestimate the change effects. Multiple factor analysis of variance models utilizing…

Conducted a Monte Carlo computer simulation of the ANOVA design to detect job differences. The design proved reasonably powerful in detecting differences. A second study used Monte Carlo methods to analyze the viability of techniques for determining job differences. Offers guidelines for selecting statistical techniques. (Author/RC)

The purpose of this study is to help define the precise nature and limits of the tolerable range in which a researcher may be relatively confident about the statistical validity of his or her research findings, focusing specifically on the statistical validity of results when violating the assumptions associated with the one-way, fixed-effects…

Believing that examinee-by-item interaction should be conceptualized as true score variability rather than as a result of errors of measurement, Lu proposed a modification of Hoyt's analysis of variance reliability procedure. Via a computer simulation study, it is shown that Lu's approach does not separate interaction from error. (Author/RL)

The ability of 3-way INDSCAL and ALSCAL models to recover true structure in proximity data based on 2-dimensional configurations varying in number of subjects (15 and 20) and stimuli, amount of error, and monotonic transformation is examined. INDSCAL outperformed metric and nonmetric ALSCAL in all conditions. (SLD)