What is it about?

We often choose in ways that do not maximize our long-term gains. For example, we may choose a small immediate reward over a larger reward that is available only after delay, a decision that may not be best for us in the long run. Think of the difficulty of saving for a big expenditure or giving up smoking. This same phenomenon occurs across species and can be investigated in a wide variety of ways. Prior work has shown that signals that predict an upcoming reward or a reduced delay to that reward are, in themselves, rewarding. Our work offers a detailed explanation of how a signal for good news provides a small immediate reward that leads pigeons to make suboptimal choices in a procedure that has puzzled researchers for almost 50 years.

Featured Image

Why is it important?

The choice behavior of pigeons is worth investigating in its own right, but we also point to the generality of the basic principles underlying suboptimal choice. The same dynamics have been shown to influence decisions made by other species (e.g., rats, monkeys, and humans). The advantage of working with pigeons is that the choice situations can be laid out in a consistent and controlled manner. This has allowed us to be very precise in the predictions of the choices pigeons will make. The more we know about the factors that influence choice across all species, the better able we will be to facilitate behavior that is optimal in the long run.


We have worked together for several years to explore and try to understand suboptimal choice behavior. In a workshop held in 2021 we mathematically formalized our hypothesis that signals for good news based on delay reduction reinforce choice. In the following months we tested the fit of our model to a large set of data as documented in the article.

Roger Dunn
San Diego State University

Read the Original

This page is a summary of: Suboptimal choice: A review and quantification of the signal for good news (SiGN) model., Psychological Review, March 2023, American Psychological Association (APA), DOI: 10.1037/rev0000416.
You can read the full text:



The following have contributed to this page