A general framework for deliberate randomization

Abstract

This study presents the General Deliberate Randomization (GDR) model as a tool to analyze risk, time and social preferences in risky environments. Traditional models typically use either an ex-post approach, assessing deterministic profiles and combining profiles over risk, or an ex-ante approach, assessing risk for individual elements and integrating them into deterministic preference. The GDR model combines these approaches using a linear weight parameter. It simplifies to deterministic preference in risk-free scenarios and to risk preference when only one element is present. The GDR model provides a unified approach to account for both deliberate randomization and correlational aversion/seeking across risk, time and social settings.

Date of Award	2024
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology
Supervisor	Songfa ZHONG (Supervisor)