Determining how learners use MOOCs effectively is critical to providing feedback to instructors, schools, and policy-makers on this highly scalable technology. However, drawing inferences about student learning outcomes in MOOCs has proven to be quite difficult due to large amounts of missing data (of various kinds) and to the diverse population…

A common reason for not participating in intimate partner violence (IPV) research is thought to be fear for one's safety. However, little is known about those who do not participate due to safety fears. To better characterize this population, we investigated correlates of being "not safe" to answer the optional IPV module in the 2006…

Velicer (1978) defines suppression situations in terms of the relations between part correlation and zero order validity. Conger (1974) defines suppressor in terms of regression weights. The present work clarifies the difference between the two approaches: Velicer's suppressor circumscribes only a subset of the suppression situations as defined by…

The perfunctory use of factor scores in conjunction with regression analysis is inappropriate for many purposes. It is suggested that factoring methods are most suitable for independent variable sets when some consideration has been given to the nature of the domain, which is implied by the predictors. (Author/BW)

Compares approaches to modeling ordinal outcome variables, including assumptions, interpretations, and limitations. Explores how the multiple regression approach with ordinal level data can compromise the understanding of the effects of the independent variables and of the ordinal level response. Provides applications with data from a multisite…

A method is proposed for overcoming the problem of multicollinearity in multiple regression equations where multiplicative independent terms are entered. The method is not a ridge regression solution. (JKS)

In almost all applications of measurement there is some sort of response by a human subject. Almost always, the response scale is ordinal, but almost always it is treated as if it were an interval measure. Methods for treating data ordinally are currently being developed in three areas: ordinal analysis for questionnaire responses, ordinal…

Factor scale scores are sometimes used as weights to create composite variables representing the variables included in a factor analysis. If these composite variables are then used to predict some dependent variable, serious theoretical and methodological problems arise. This paper explores these problems and suggests strategies for circumventing…

This interim report is concerned with the analysis of educational experiments and quasiexperiments where alternative teaching methods are applied to intact classes or where alternative programs are set up in samples of schools or communities. Generally, such studies have used the class, the school, or the community as the unit of sampling with the…

Evaluation Research Society members (n=160) assessed the attributes of research that produce credibility. The study focused on the acceptability of no-difference findings. Randomization, statistical power, number of measures reported, and equivalent groups were tested to determine their influence on acceptability of both no-difference and…

The author examined the prevalence of multiple operational definitions of outcome constructs and an estimate of the incidence of Type I error rates when univariate procedures were applied to multiple variables in educational psychology. Multiple operational definitions of constructs were advocated and wider use of multivariate analysis was…

The effects of using different data analysis methods on estimates of treatment effects of educational programs were investigated. Various regression models, such as those recommended for Title I program evaluations, were studied. The first effect studied was the amount of bias that might be expected to occur in the various settings. Results…

Discusses the concept of temperament interactions in the context of statistical interaction. Categorizes temperament interactions that involve temperament as an independent variable, as a dependent variable, or as both. Describes use of hierarchical multiple regression for the analysis of temperament interactions. (Author/CI)

The present paper contributes to research methodology as practiced in the field by providing practitioners with a concise statement of the problems in analyzing unbalanced designs, and by clarifying the conditions under which the analysis alternatives are most appropriate. The purposes of the present paper are: (1) to examine the analysis…

With even the simplest bivariate regression, least-squares solutions are inappropriate unless one assumes a priori that reciprocal effects are absent, or at least implausible. While this discussion is limited to bivariate regression, the issues apply equally to multivariate regression, including stepwise regression. (Author/CTM)