Conjoint Analysis of Residential Preferences

A methodology for studying the effect of different locational factors on individual residential preferences is developed. Observations on the overall preferences of individuals are used to construct profiles of prospective residents. The Conjoint Analysis (CA) method is used for this purpose. Preferences for alternative development proposals are measured with a multidimensional composite index of Residential Location Utility (RLU). Criteria for the selection of the dimensions of the index and the form of the compositional rule are postulated. A vector of the RLU is estimated by the use of ordinal regression in conjunction with linear programming. The methodology was implemented in the study of residential preferences of the population of Assistant Professors of Stanford University. The data collection technique is described and results of the analysis are presented. The members of the population are classified according to their demonstrated behavior in selecting a residential place. An informal validation of the results is made by comparison with direct statements by the respondents on their preferences.