This paper describes structural equation modeling (SEM) in comparison with another overarching analysis within the general linear model (GLM) analytic family: canonical correlation analysis. The uninitiated reader can gain an understanding of SEM's basic tenets and applications. Latent constructs discovered via a measurement model are explored and…

A study investigated the relationship between arts education and academic achievement. Of specific interest was whether teaching the arts for their own sake influenced academic achievement in language arts and mathematics. It was hypothesized that it would influence children's self-efficacy. The sample consisted of 328 children from grades 2-5.…

This study uses an algorithm-based visual display technique (FACES) to provide enhanced detection of multivariate outliers within large-scale data sets. The FACES computer graphing algorithm (H. Chernoff, 1973) constructs a cartoon-like face, using up to 18 variables for each case. A major advantage of FACES is the ability to store and show the…

The work of Joreskog and Sorbom in comparing factor structures of several populations is extended to a more general analysis of covariance structures. Also, more complex constraints on parameters are allowed in this work. (JKS)

A Bayesian framework for making mastery/nonmastery decisions based on multivariate test data is described. Overall, mastery is granted if the posterion expected loss associated with such action is smaller than the one incurred by denying mastery. (Author/JKS)

Virtually all parametric statistical procedures have been shown to be special cases of canonical correlation analysis. This article proposes a logic for conducting stepwise canonical correlation analyses, based upon evaluation of canonical communality coefficients. The procedure is a direct analogue of stepwise multiple regression. (Author/RD)

For multi-subject experiments with all factors repeated (a single independent group of subjects), three multi-subject behavior functions are identified and analyzed. They are based on Shine's actualized and pure single-subject behavior functions extended to multi-subject experiments with some factors nonrepeated (several independent groups of…

Discusses the problem of dichotomous-dependent variables in regression analysis in student attrition studies. Proposes a new method as an alternative to regression analysis. (Author/RC)

Results show that the proposed sampling distribution of the test appears to be appropriate only for sample sizes above 50, and for data where the sample size is 10 times the number of variables. For such cases the power of the test is found to be fairly low. (Author/RL)

Explores relationships between cooperative education work experiences and subsequent benefits. A literature review is followed by a report of data from a survey designed to discover which internship processes are the best predictors of desired outcomes. (CT)

The use of analysis of covariance in conjunction with the multivariate approach to analyzing repeated measures designs is considered for designs involving between- and within-subject factors, one dependent variable, and one observation per subject on the covariate. (Author/RL)

The use of general linear regression methods for the analysis of categorical data is recommended. The general linear model analysis of a 0,1 coded response variable produces estimates of the same response probabilities that might otherwise be estimated from frequencies in a multiway contingency table. (Author/CTM)

Using data from the Current Population Survey and a multivariate approach, compared living arrangements of immigrant children to those of U.S.-born white children with U.S.-born parents. Findings show the "downward assimilation" among some second- and third-generation immigrant children fits a theory of segmented assimilation. Discusses…

Studied the reliability and convergent validity for 27 composite scales and 2 items covering aspects of alcohol use, smoking, marijuana use, and other drug use. Results indicate 23 of the 27 composite scales had at least acceptable reliability and the remaining 4 scales had at least marginally acceptable reliability. At least moderate construct…

Canonical weights and structure correlations are used to construct low dimensional views of the relationships between two sets of variables. These views, in the form of biplots, display familiar statistics: correlations between pairs of variables, and regression coefficients. (SLD)