Too smart to take a risk? exploring nonlinearities between cognitive ability and risk preferences
Date
2014-12
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Does being too smart make one risk averse? In this paper, we explore the potential curvilinear relationship between cognitive ability and risk preferences using data collected from survey
experiments held among 400 UP Diliman students. For cognitive ability, we administered Frederick's (2005) Cognitive Reflection Test (CRT), and obtained the Grade Weighted Averages and UP Predicted Grades of the students. For risk preference, we ran the Gneezy and Potters' (1997) investment game, asked hypothetical choice gambles and a self-reported measure of risk preference. We gathered the biological, socio-economic and psychological characteristics of the students as well. Using OLS regressions, we find evidence that nonlinearities do exist between cognitive ability and risk preferences. One's self-rated willingness to take risks initially increases with one's score on the CRT, peaks, and then declines. On the other hand, we find risk preference patterns across cognitive ability to differ between students studying math-oriented courses and non-math-oriented courses using data from the Gneezy and Potters investment game. An inverted U relationship emerges for the former and a U pattern for the latter. We offer as a tentative explanation the different approaches these students take when making decisions in the face of uncertainty.
Description
Keywords
Cognitive Reflection Test, Cognitive Ability, Risk Preference, Nonlinear Relationship, Quadratic, Curvilinear, Choice Gambles, 2D:4D digit ratio