Professor of Sociology
University of Wisconsin - Madison
Logan has completed development and testing of a new, two-sided probit method for estimating preferences of men and women for characteristics of marital partners when data consist only of the characteristics of married partners and single men and women. He published results with collaborators Hoff (Statistics, UW Madison) and Newton (University of Washington) in Journal of American Statistical Association (J.A. Logan, P.D. Hoff and M.A. Newton, “Two-Sided Estimation of Mate Preferences for Similarities in Age, Education, and Religion,” June, 2008. Vol. 103, no. 482: pp. 559-569). This new method allows for posterior predictive simulations of effects of changes in the characteristics of potential mates in the marriage market. It is expected to have application in employment and other two-sided matching problems in addition to marriage. Though the method requires fewer assumptions than competing approaches, it is computationally demanding and is not expected to be widely used in its current form