The Linear Model in Attitude Measurement: An Example and Some Comments

Publisher Website
Google Scholar
Abstract
RAMSAY and Case (1970) propose a promising method for at- titude measurement based on the attitude models of Fishbein (1967) and Tucker (1960). They suggest the use of linear multiple regression analysis for predicting evaluative judgments of stimuli which have been scaled on several properties. The regression weights are interpreted as evaluations of the varioua properties of the stimuli, Ramsay and Case outline the advantages of the method and present some data intended to provide an example of its applica- tion. The data show that a linear combination of scaled pwpertiea of nations can successfully predict evaluative judgments of the nations. It is important to note, however, that the data do not show that the regression weights are valid measures of the evalua- tiom of the properties. An attempt by Ramsay and Case to validate the regression weigh& by correlating them with direct evaluations of the properties resulted in correlations of 0.15,0.61,025, and 0.18. Since regression weights are estimated from data, it is necessary to show that their interpretation as evaluations of properties is 1 justified. The present paper presents an example of the use of the linear model for attitude measurement emphasizing the validation

This publication has 5 references indexed in Scilit: