Estimated Type I error rates and power are reported for six tests of variance. Normal and non-normal distributions of a two-group design were investigated. Tests developed by M. B. Brown and A. B. Forsythe and by R. G. O'Brien were considered most useful. Implications are discussed. (SLD)

Considers problems which arise when researchers do not have the optimally large sample sizes desired in structural equation modeling. Discusses the ways in which small sample size affects assessment of model fit. Provides a new estimator that may be beneficial for use in small-sample situations. (Author/RH)

A Monte Carlo study was conducted to compare the efficiency of Listwise deletion, Pairwise deletion, Allvalue, and Samemean methods in estimating the correlation matrix from data that had randomly occurring missing values. The four methods were compared in a 3x3x4 factorial design representing sample size, proportion of incomplete records in the…

A normally distributed data set of 1,000 values--ranging from 50 to 150, with a mean of 50 and a standard deviation of 20--was created in order to evaluate the bootstrap method of repeated random sampling. Nine bootstrap samples of N=10 and nine more bootstrap samples of N=25 were randomly selected. One thousand random samples were selected from…

Tracking and two statistical methods (probability weighting and sample selection modeling) were studied as ways to minimize bias attributable to sample attrition in school-based studies. Data on student smoking from 30 middle schools illustrate that sample weighting yields the best results, with estimates superior to sample selection and much less…

The stability of Rasch test item difficulty parameters was investigated under varying conditions. Data were taken from a mathematics achievement test administered to over 2,000 Australian students. The experiments included: (1) relative stability of the Rasch, traditional, and z-item difficulty parameters using different sample sizes and designs;…