The Boundaries of Instance Based Learning Theory for Explaining Decisions from Experience
Abstract
Most demonstrations of how people make decisions in risky situations rely on decisions from description, where outcomes and their probabilities are explicitly stated. But recently, more attention has been given to decisions from experience where people discover these outcomes and probabilities through exploration of the problems. More importantly, risky behavior depends on how decisions are made (from description or experience), and although Prospect Theory explains decisions from description, a comprehensive model of decisions from experience is yet to be found. Instance-Based Learning Theory (IBLT) explains how decisions are made from experience through interactions with dynamic environments (Gonzalez, Lerch, and Lebiere, 2003). The theory has shown robust explanations of behavior across multiple tasks and contexts, but it is becoming unclear what the theory is able to explain and what it does not. The goal of this chapter is to start addressing this problem. I will introduce IBLT and a recent cognitive model based on this theory: the IBL model of repeated binary choice; then I will discuss the phenomena that the IBL model explains and those that the model does not. The argument is for the theorys robustness but also for clarity in terms of concrete effects that the theory can or cannot account for.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2013
- Accession Number
- AD1018094
Entities
People
- Cleotilde Gonzalez
Organizations
- Carnegie Mellon University