Article ID: | iaor20071553 |
Country: | Netherlands |
Volume: | 61 |
Issue: | 3 |
Start Page Number: | 205 |
End Page Number: | 249 |
Publication Date: | Nov 2006 |
Journal: | Theory and Decision |
Authors: | Hogarth Robin M., Karelaia Natalia |
The effectiveness of decision rules depends on characteristics of both rules and environments. A theoretical analysis of environments specifies the relative predictive accuracies of the ‘take-the-best’ heuristic (TTB) and other simple strategies for choices between two outcomes based on binary cues. We identify three factors: how cues are weighted; characteristics of choice sets; and error. In the absence of error and for cases involving from three to five binary cues, TTB is effective across many environments. However, hybrids of equal weights and TTB models are more effective as environments become more compensatory. As error in the environment increases, the predictive ability of all models is systematically degraded. Indeed, using the datasets of Gigerenzer