Successive unfolding of family preferences

Abstract

A technique to scale preferences in relation to an externally derived stimulus configuration, called Successive Unfolding, is described. Four steps are involved: (1) computing a matrix of inter-stimulus distances; (2) using ALSCAL to obtain a stimulus configuration from the matrix of distances; (3) using Carroll’s regression procedure to solve for subject ideal points; and (4) using this starting configuration to scale preference rank orders in ALSCAL. The technique is used to analyze family preference data. Results suggest that a number preference, a sex preference, and a balance preference are the components contributing to overall family preferences. Race and sex differences are portrayed by locating subject ideal points along these dimensions. Finally, the relationship of Successive Unfolding to previous techniques for measuring family preferences is discussed, and the decision- making process modeled by Successive Unfolding is outlined.