Current facets (Pre-Master)

PhD defence of Vitalie Spinu on Thursday 18 April 2013

Promotie Vitalie Spinu op donderdag 18 april 2013

On Thursday 18 April 2013 Vitalie Spinu will defend his PhD thesis entitled 'Nonadditive Beliefs From Measurement to Extensions'. Supervisors are Professors Peter Wakker and Han Bleichrodt. Other members of the Doctoral Committee are Professor Horst Zank (The University of Manchester), Dr. Kirsten Rohde and Dr. Vladimir Karamychev.

Time and location

The PhD defence will take place in the Senate Hall of Erasmus University Rotterdam and will start at 15.30 hrs.

About the dissertation

Spinu’s dissertation deals with subjective decision theories. It looks specifically at the empirical measurement and the mathematical foundation of these theories.  

The research is divided into two parts, one empirical and one theoretical. The empirical part of the dissertation proposes a novel method for elicitation of beliefs and forecasts from experts.
The theoretical part introduces a general method for extending decision models from simple choice alternatives to larger domains of interest. The proposed extension procedure can serve as a general add-on for virtually all presently existing behavioural representations.

Spinu observes that many papers provide their own version of extensions in the context of the particular models they propose. He aims to comprise all these extensions into a general and unifying framework that can be applied to every behavioural decision model.

About Vitalie Spinu

Vitalie Spinu obtained both his bachelor’s and master’s degrees in Actuarial Science from the Academy of Economic Studies Bucharest. In 2008 he started as a doctoral student at the Tinbergen Institute with behavioural decision theory and experimental economics as his primary fields of interest.

His research interests include computational statistics and data analysis, with a primary focus on Bayesian hierarchical modelling. His research has been published in various journals such as the Journal of Mathematical Economics and the Journal of Risk and Uncertainty. He is also a core-developer of Emacs Speaks Statistics – a rich Integrated Development Environment (IDE) for statistical computing languages.

Abstract of 'Nonadditive Beliefs: From Measurement to Extensions'

This thesis deals with empirical measurement and mathematical foundation of subjective decision theories.

The empirical part of the thesis investigates the predictive power of decisions under ambiguity (unknown probabilities) of several well-known models. It provides evidence that source dependent Prospect Theory outperforms its’ most popular alternatives, including subjective expected utility, Choquet expected utility, and three multiple priors theories: maxmin expected utility, maxmax expected utility, and alpha-maxmin expected utility for predicting individual choices made under natural ambiguity.

The theoretical part introduces a general method for extending decision models from simple choice alternatives (prospects) to larger domains of interest. Simple objects are usually functions taking finitely many values, such as simple acts in decision under uncertainty, or simple lotteries in decision under risk. The whole set of objects of interest often includes continuous and sometimes unbounded prospects. The proposed extension procedure is agnostic to the type of objects or the form of the representation functional, and can serve as a general add-on for virtually presently existing behavioural representations.

More information

Spinu conducted his doctoral research in Economics at the Tinbergen Institute. The Tinbergen Institute is a joint research institute of the Erasmus School of Economics, Erasmus University Rotterdam, Faculty of Economics and Business, University of Amsterdam and Faculty of Economics and Business Administration of VU University Amsterdam.

For more information about this ceremony, please contact Ronald de Groot, Communication Officer of the Erasmus School of Economics phone +31 10 408 1762 or by e-mail: rdegroot@ese.eur.nl.