|
Confirming the disconfirmed and the conjunction fallacies| old_uid | 5994 |
|---|
| title | Confirming the disconfirmed and the conjunction fallacies |
|---|
| start_date | 2009/01/16 |
|---|
| schedule | 14h-16h |
|---|
| online | no |
|---|
| summary | Atkinson et al. (2009) show that it is possible for a conjunction to be confirmed robustly by some evidence to a greater extent than either of its conjuncts individually. They posit that this formal result connects to a distinct, non-traditional class of conjunction fallacy cases. I argue against this claim in two ways. First, formally, I argue that any case meeting their formal requirements is crucially distinguishable from conjunction fallacy cases. Consequently, while it would be a fallacy for experimental participants to rank a conjunction as more probable than one of its conjuncts in such cases, it would be of a different type than the conjunction fallacy. Second, experimentally, I argue that the setup that Atkinson et al. suggest would not test for the presence of a fallacy at all. I introduce a "corrected" version of their experimental setup that would constitute such a test. Then, I present some initial experimental data that shows that humans are impressively sharp and indeed far from committing a fallacy in such cases.
Atkinson, D., Peijnenburg, J., and Kuipers, T. (Forthcoming, 2009). How to Confirm the Disconfirmed: On Conjunction Fallacies and Robust Confirmation. Philosophy of Science |
|---|
| responsibles | Drouet, Martin |
|---|
| |
|