Summary: A new study shows that mathematical models can shape human decisions more effectively than traditional psychological strategies.
Researchers introduced “choice engineering,” a data-driven alternative to “choice architecture,” which relies on heuristics and intuition. In a global academic challenge, models like CATIE—designed to reflect real behavioral trends—outperformed both intuitive strategies and standard machine learning tools.
The success of these computational models suggests that behavior can be precisely influenced using systematic, optimized approaches.
Key Facts:
- Choice Engineering Defined: Uses computational models to influence behavior with precision, unlike intuition-based nudging.
- CATIE Model Success: CATIE outperformed both Q-learning models and psychological strategies in influencing choices.
- Ethical Caution: While effective, the technique demands robust ethical frameworks for real-world use.
Source: Hebrew University of Jerusalem
A new study published in Nature Communications demonstrates that mathematical models can be more effective than psychological intuition when it comes to influencing human decisions.
Led by Prof. Yonatan Loewenstein from Safra Center for Brain Sciences (ELSC) at Hebrew University, in collaboration with Dr. Ohad Dan from Yale University and Dr. Ori Plonsky from the Technion, the research introduces a novel concept: choice engineering.
The study draws a distinction between two approaches to influencing behavior.
The first, known as choice architecture, has gained widespread popularity since one of its pioneers, Richard Thaler, was awarded the Nobel Prize in Economics in 2017—with behavioral insights (“nudge”) teams emerging in governments around the world. Choice architecture relies on psychological principles—such as primacy, anchoring, or intuitive heuristics—to subtly steer decisions.
The second approach, proposed by the researchers, is choice engineering: a method that uses computational models and optimization techniques to systematically shape behavior with precision.
To put these approaches to the test, the team launched an academic competition where international academic teams were tasked with designing an incentivization mechanism (“reward schedule”) that would get people to choose one of two objectively equal-value options.
More than 3,000 participants took part in the experiment, each exposed to one of several reward strategies. Some were built on intuition and psychological insights, while others were crafted using computational models.
The most effective schedule was based on a computational model called CATIE (Contingent Average, Trend, Inertia, and Exploration), designed by Dr. Ori Plonsky together with Prof. Ido Erev from the Technion.
The model integrates multiple behavioral tendencies into a unified predictive framework. This CATIE-based strategy significantly outperformed those based on the widely used machine-learning model Q-learning, and those informed by qualitative intuition alone.
“Our study shows that just as engineers use mathematical models to build bridges or design aircraft, we can use models of learning and decision-making to influence behavior—reliably and efficiently,” said Prof. Loewenstein.
The findings demonstrate that behavior can be engineered with surprising accuracy when guided by well-calibrated models.
Moreover, the study offers a new method for evaluating cognitive models—not only by their explanatory power, but also by their effectiveness in shaping real-world decisions.
The implications are far-reaching. In fields ranging from education and public health to digital design and policy-making, choice engineering could enable the development of empirically optimized, scalable interventions.
At the same time, the researchers note that ethical frameworks will be essential to guide the responsible application of these tools.
As a proof of concept, this study underscores the emerging potential of mathematical modeling in the cognitive sciences—not just for understanding behavior, but for actively guiding it.
About this math and human behavior research news
Author: Danae Marx
Source: Hebrew University of Jerusalem
Contact: Danae Marx – Hebrew University of Jerusalem
Image: The image is credited to Neuroscience News
Original Research: Open access.
“Behavior engineering using quantitative reinforcement learning models” by Yonatan Loewenstein et al. Nature Communications
Abstract
Behavior engineering using quantitative reinforcement learning models
Effectively shaping human and animal behavior is of great practical and theoretical importance. Here we ask whether quantitative models of choice can be used to achieve this goal more effectively than qualitative psychological principles.
We term this approach, which is motivated by the effectiveness of engineering in the natural sciences, ‘choice engineering’.
To address this question, we launched an academic competition, in which teams of academic competitors used either quantitative models or qualitative principles to design reward schedules that would maximally bias the choices of experimental participants in a repeated, two-alternative task.
We found that a choice engineering approach is the most successful method for shaping behavior in our task. This is a proof of concept that quantitative models are ripe to be used in order to engineer behavior.
Finally, we show that choice engineering can be effectively used to compare models in the cognitive sciences, thus providing an alternative to the standard statistical methods of model comparison that are based on likelihood or explained variance.
