Experiment-dependent priors in psychology and physics

Sometimes conducting an experiment to ascertain the state of a system changes the state of the system being measured. Kahneman & Tversky modelled this effect with support theory. Quantum physics models this effect with probability amplitude mechanics. As this paper shows, probability amplitude mechanics is similar to support theory. Additionally, Viscusi's proposed generalized expected utility model has an analogy in quantum mechanics.     

Some aspects of the reparametrization of statistical models

Definitions are given for orthogonal parameters in the context of Bayesian inference and likelihood inference. The exact orthogonalizing transformations are derived for both cases, and the connection between the two settings is made precise. These parametrizations simplify the interpretation of likelihood functions and posterior distributions. Further, they make numerical maximization and integration procedures easier to apply. Several applications are studied.     

High School Reunions and the Management of Identity

Organized around an intersection of the past with the present, high school reunions confront those who attend with discontinuities in their own lives as well as the lives of others. Based on observations of and interviews with attendees at five reunions, we argue, contrary to many claims about the contemporary segmentation of the self, that reunion goers are able to construct accounts of coherent lives by reference to “true selves” independent of appearances. Although reunion attendees may attempt to manage impressions by controlling information about themselves, these efforts are limited by attendees efforts to sustain convictions of their own integrity. These convictions, however, also depend on accounts, albeit accounts directed inward. Moreover, the maintenance of this conviction depends on the successful “neutralization” of others' judgments.     

The value of a statistical life: A comparison of two approaches

Using an original data set, this article presents the first effort to compare systematically the values of a statistical life obtained from the two main methodologies existing in the literature—the revealed-preference approach (wage-risk analysis) and the contingent-valuation approach—applied to the same sample of individuals. A survey was conducted in the Montreal area with a questionnaire containing two series of questions: one set of contingent-valuation questions and one set of questions pertaining to the information necessary for performing a wage-risk study. The results show that the values of a statistical life obtained from each method are different from each other.This research was supported by the Fonds F.C.A.R. of the Quebec government, the Social Sciences and Humanities Research Council, and by the Fondation Mercure of l'École des Hautes Études Commerciales. The authors want to thank the editor and an anonymous referee for their useful comments.