|
Homo Heuristicus: Decision Making under Uncertainty| title | Homo Heuristicus: Decision Making under Uncertainty |
|---|
| start_date | 2025/01/21 |
|---|
| schedule | 16h30-17h30 |
|---|
| online | no |
|---|
| location_info | Room B10 & on zoom |
|---|
| summary | In well-defined situations with known risks, the axioms of classical decision theory can provide a framework for optimal decision-making. However, when Savage introduced his axioms, he clarified that they apply to situations involving risk but not to those characterized by uncertainty or intractability. Uncertainty refers to ill-defined scenarios where the exhaustive and mutually exclusive set of future states of the world and their outcomes are either unknown or unknowable. Intractability, on the other hand, pertains to well-defined but overly complex situations, such as chess, where finding optimal solutions is impractical. While thinkers such as Knight, Keynes, and Simon had drawn similar distinctions, most theories have traditionally reduced uncertainty and intractability to risk, such as by using second-order probabilities, equal priors, or Bayesian subjective probabilities.
In contrast, I advocate for a genuine theory of decision-making under uncertainty, rooted in the empirical study of decision-making under uncertainty. This approach focusses on three core research areas. The first is descriptive: What heuristics are available in the adaptive toolbox of individuals and organizations, and how do they select between them? The second is prescriptive: Under what conditions do heuristics outperform more complex strategies, and when do the latter succeed? This line of inquiry, known as the study of ecological rationality, examines the alignment between decision-making strategies (heuristic or complex) and the structure of environments. The third area is engineering and intuitive design: How can we create heuristic systems that empower both experts and non-experts to make better decisions?
To advance this agenda, three methodological tools are essential: formal models of heuristics (move beyond vague terms like "System 1 thinking"), competitive testing of heuristics against complex strategies (instead of relying on null hypothesis testing), and assessing the predictive power of heuristics (rather than just fitting them to existing data). Through examples from finance, management, and sports, I demonstrate conditions under which heuristics can predict as accurately but more efficiently than complex strategies. Less can be more. |
|---|
| responsibles | Esposito |
|---|
Workflow history| from state (1) | to state | comment | date |
|---|
| submitted | published | | 2025/01/20 10:32 UTC |
| |
|