Search

Multiple linear regression is discussed as it relates to several goals of research: Predictability, Parsimony, Replication, and Validity Generalization. law. The emphasis i s upon the per cent of variance accounted for in the criterion under investigation, rather than on statistical significance from random events. Additional remarks concerning curvilinear relationships and data snooping are also presented. Dingman's Canons of Repmducibility are discussed within the framework of multiple linear regression and the goals of research. I These goals are presented with the development of a well established physical.

The inclusion in a multiple regression model of a predictor variable which is highly correlated with other prediotor variables is usually not recommended. The argument is that the new predictor variable is accouihing for variance which has already been accounted for in the model. The following discussion is a defense for the procedure of including highly correlated predictor variables under certain circumstances. First, highly correlated variables can be used when there is a requirement that the predictor variables account for a certain number of group membership vectors. A second and more important situation occurs when there is theoretical or empirical justification for the inclusion of such a variable. The following discussion is limited to a very specific kind of highly correlated variable-one that contains the squared elements of one of the original variables.

The Illinois Test of Psycholinguistic Abilities (ITPA) was administered to 80 Ss from four cultural systems of varying degrees of modernization in Sierra Leone. Extensive univariate results were reporbed by Karr (1969). This short note indicates the utility of muliivariate analyses. Multiple dis- criminant analysis was applied to the nine ITPA subtests, resulting in two significant discriminsant functions. The first function separated the groups along the hypothesized modernization continuum. The second function in- dicated one way in which bhe hypothesized levels of modernization did not hold up under empirical investigation.