Multiple linear regression is used to model the effects of violating statistical assumptions on the likelihood of making a Type I error. This procedure is illustrated for the student's t-test (for independent groups) using data from previous Monte Carlo studies in which the actual alpha levels associated with violations of the normality…

The level of standardized effect sizes obtained by chance and the use of significance tests to guard against spuriously high standardized effect sizes were studied. The concept of the "protected effect size" is also introduced. Monte Carlo methods were used to generate data for the study using random normal deviates as the basis for sample means…

One set of approaches to the problem of clustering with dichotomous data in cluster analysis (CA) was studied. The techniques developed for clustering with binary data involve calculating distances between observations based on the variables and then applying one of the standard CA algorithms to these distances. One of the groups of distances that…

This study explored two methods that are used to assess the dimensionality of item response data. The paper begins with a discussion of the assessment dimensionality and the use of factor-analytic procedures. A number of problems associated with using linear factor analyses to assess dimensionality are also considered. A procedure is presented for…

Describes Monte Carlo simulation studies and their application in vocational education research. Explains study design and analysis as well as use and evaluation of results. (SK)

Describes the development and instructional use of a Microsoft Excel spreadsheet template that facilitates analytical and Monte Carlo risk analysis of investment decisions. Discusses a variety of risk assessment methods followed by applications of the analytical and Monte Carlo methods. Uses a case study to illustrate use of the spreadsheet tool…

Outlines an approach to differential item functioning (DIF) analysis using hierarchical linear regression that makes it possible to combine results of logistic regression analyses across items to identify consistent sources of DIF, to quantify the proportion of explained variation in DIF coefficients, and to compare the predictive accuracy of…

Investigates the commonly suggested rounding rule for addition and subtraction including its derivation from a basic assumption. Uses Monte-Carlo simulations to show that this rule predicts the minimum number of significant digits needed to preserve precision 100% of the time. (Author/KHR)

Two forms, pure-rank and mixed-rank, of a nonparametric, general, linear model-based statistic that can be used to test several hypotheses are presented. A Monte Carlo study was used to investigate the distributional properties of these forms, and their use is discussed. (SLD)

Presents an approach to multidimensional unfolding that reduces the occurrence of degenerate solutions and conducts a Monte Carlo study to demonstrate the superiority of the new method to the ALSCAL and KYST nonmetric procedures for student preference data. (SLD)

Clarifies several issues regarding the effects of autocorrelated errors on Type I error in ordinary least-squares models. Demonstrates through Monte Carlo simulation the conditions under which distortion in Type I error is less than predicted by asymptotic theory. Suggests a recently developed small-sample method for time-series analyses. (SLD)

Used Monte Carlo methods to compare several measures of item-parameter drift, manipulating numbers of examinees and items and numbers of drift items. Overall, Lord's chi square (F. Lord, 1968) measure was the most effective in identifying items that exhibited drift. Discusses the usefulness of other methods. (SLD)

Investigated the power and Type I error rate of the likelihood ratio goodness-of-fit statistic (LR) in detecting differential item functioning (DIF) under F. Samejima's (1969, 1972) graded response model. Monte Carlo study results show conditions under which the LR or Mantel Haenszel procedures have adequate power to detect DIF. (SLD)

Presents a hierarchical Bayes approach to modeling parameter heterogeneity in generalized linear models. The approach combines the flexibility of semiparametric latent class models that assume common parameters for each subpopulation and the parsimony of random effects models that assume normal distributions for the regression parameters.…

Monte Carlo methods were used to evaluate an EM algorithm used for the correction of missing data in latent class analysis. Findings regarding bias in parameter estimates suggest practical limits for the utility of the EM algorithm in terms of sample size and nonresponse rate. (SLD)