Columbia University

Technology Ventures

Efficient and accurate method for evaluating risk-tolerance, time preferences and related behaviors

Technology #m11-041

Questions about this technology? Ask a Technology Manager

Download Printable PDF

License this Technology
Non-Commercial Software License
Image Gallery
Olivier Toubia
Eric Johnson
Managed By
Richard Nguyen

The ability to predict the decisions that an individual makes based on the risk inherent to those decisions is a valuable asset for a wide range of different fields, from consumer finance and banking to medicine. Many different behavioral models have been developed to assess risk-tolerance, but in order to generate data with meaningful predictive power, there is a need for more accurate and efficient methods for applying these models to research subjects, clients, patients, etc. This technology is a method for utilizing behavioral models to evaluate a group of individuals' risk-tolerance. It employs an adaptive questionnaire that automatically selects later questions based on an individual’s previous answers. This method has more predictive power than and is as efficient as its non-adaptive counterparts. It is ideal for use on large populations, such as those taking online surveys.

Dynamic choice design optimizes information provided by each selection

By basing later questionnaire options on an individual’s earlier answers, the method encompassed by this technology optimizes the information provided by each selection an individual makes. When applied to a large number of individuals, the method generates a heterogeneous data set that can capture response error and reveal statistically significant relationships between an individual’s choices and behavior. The method has been used specifically to measure the parameters of two behavioral models: Cumulate Prospect Theory, which describes an individual’s tendency to frame their decisions based on the current status rather than potential outcome, and the Quasi-Hyperbolic Time Discounting model, which describes an individual’s tendency to prioritize immediate risk and reward over future risk and reward. It can in principle be adapted to other behavioral models, as well.

This method has been used to generate statistically significant results for three diverse case studies that evaluated risk-tolerance in homeowners, retirement age individuals, and based on current and prenatal hormone levels. Software implementing the methodology has also been developed.

Lead Inventors:

Olivier Toubia, Ph.D.

Eric J. Johnson, Ph.D.


  • Tool to evaluate a group of individuals' risk-tolerance and related behaviors for research, medicine, finance, and government
  • Evaluation and prediction of a client’s lending potential by financial institutions
  • Evaluation and prediction of a patient’s health behaviors/outcomes including for long-term medical treatments, diet and exercise planning, substance abuse treatment
  • Evaluation and prediction of individual’s work and retirement behavior
  • Method for determining risk tolerance in large populations via online surveys


  • Predicts risk-tolerance more accurately than similar methods
  • Questionnaire can be completed in three minutes on average
  • Takes the same time to complete as less accurate methods
  • Reduces redundant questions

Patent Information:

Patent Pending

Tech Ventures Reference: IR M11-041

Related Publications: