This research sought to determine whether any measure(s) used in the Carnegie Foundation's classification of Doctoral/Research Universities contribute to a greater degree than other measures to final rank placement. Multilevel Modeling (MLM) was applied to all eight of the Carnegie Foundation's predictor measures using final rank…

A growing controversy surrounds the strict interpretation of statistical significance tests in social research. Statistical significance tests fail in particular to provide estimates for the stability of research results. Methods that do provide such estimates are known as invariance or cross-validation procedures. Invariance analysis is largely…

The purposes of this paper are to provide an explanation of the way in which suppressor variables operate in multiple regression, and to provide illustrations of their effects with actual social science data. Examples are reviewed in which one predictor accounts for more of the variance in the dependent variable when another predictor is included,…

This study of 368 female undergraduates examined self-efficacy and role model influence as predictors of career choice across J. L. Holland's (1997) 6 RIASEC (Realistic, Investigative, Artistic, Social, Enterprising, Conventional) types. Findings showed that levels of self-efficacy and role model influence differed across Holland types. Multiple…

Discusses multiple regression, a very popular statistical technique in the field of education. One of the basic assumptions in regression analysis requires that independent variables in the equation should not be highly correlated. The problem of multicollinearity and some of the solutions to it are discussed. (Author)

Examines the problem definition behavior of 48 White families (two natural parents and adolescent child) discussing simulated problematic situations. Analyzes the effects of family background, interpersonal relations, and discussion process variables. Interprets the importance of interpersonal ties in terms of family interaction issues. Contains…

Often quantitative data in the social sciences have only ordinal justification. Problems of interpretation can arise when least squares multiple regression (LSMR) is used with ordinal data. Two ordinal alternatives are discussed, dominance-based ordinal multiple regression (DOMR) and proportional odds multiple regression. The Q[superscript 2]…

Reviews several potential problems that often invalidate statistical tests of social theories. Discusses how there may be very serious identification problems in the estimation of the relationship between executions and homicides, and that serious questions may be raised regarding the validity of any particular estimation of this relationship.…