Heuristics

1. Simple Heuristics That Make Us Smart Gerd Gigerenzer Max Planck Institute for Human Development Berlin

3. If one of two objects is recognized and the other is not, then infer that the recognized object has the higher value. The heuristic is successful when ignorance is systematic rather than random, that is, when lack of recognition correlates with the criterion. Ecological Rationality Recognition Heuristic

5. Wimbledon 2003 Frings & Serwe (2004) ATP Entry Ranking 50% 60% 70% 66% 68% 69% ATP Champions Race Seedings Recognition Laypeople Recognition Amateurs Correct Predictions

6. Wimbledon 2003 Frings & Serwe (2004) ATP Entry Ranking 50% 60% 70% 66% 68% 69% 66% 72% ATP Champions Race Seedings Recognition Laypeople Recognition Amateurs Correct Predictions

7. The Less-is-More Effect The expected proportion of correct inferences c is is the number of recognized objects is the total number of objects is the recognition validity, and is the knowledge validity A less-is-more effect occurs when  >  where n N  

8. Number of Objects Recognized (n) 0 50 100 50 80 75 70 65 60 55 Percentage of Correct Inferences (%)  =. 5  = .6  = .7  = .8

10. Gaze heuristic

11. Gaze heuristic

12. Gaze heuristic

13. Gaze heuristic

29. Ecological Rationality

31. Ecological Rationality Take The Best Tallying Weight Cue Cue 1 2 3 4 5 1 2 3 4 5 Martignon & Hoffrage (1999), In: Gigerenzer et al., Simple heuristics that make us smart. Oxford University Press compensatory non-compensatory

32. 0 10 20 30 40 50 60 70 80 90 100 0-24 25-48 49-72 73-96 97-120 121-144 145-168 Choices predicted by Take The Best (%) Non-compensatory Feedback Compensatory Feedback Reinforcement learning: Which heuristics to use? Feedback Trials Rieskamp & Otto (2004)

34. Robustness

36. 55 60 65 70 75 Take The Best Tallying Multiple Regression Minimalist Fitting Prediction Robustness Accuracy (% correct) Czerlinski, Gigerenzer, & Goldstein (1999)

37. Czerlinski et al. (1999): Multiple regression (MR) improves performance (76%) but so does Take The Best (76%) Hogarth & Karelaia (2004): Take The Best is more accurate than MR if: - variability in cue validities is high - average intercorrelation between cues ≥ .5 - ratio of cues to observations is high Akaike’s Theorem: Assume two models belong to a nested family where one has fewer adjustable parameters than the other. If both have, on average, the same number of correct inferences on the training set, then the simpler model (i.e., the one with fewer adjustable parameters) will have greater (or at least the same) predictive accuracy on the test set. Robustness: What if predictors are quantitative?

38. Design

39. Chest Pain = Chief Complaint EKG (ST, T wave ∆'s) History ST&T Ø ST  T  ST  ST  &T  ST  &T  No MI& No NTG 19% 35% 42% 54% 62% 78% MI or NTG 27% 46% 53% 64% 73% 85% MI and NTG 37% 58% 65% 75% 80% 90% Chest Pain, NOT Chief Complaint EKG (ST, T wave ∆'s) History ST&T Ø ST  T  ST  ST  &T  ST  &T  No MI& No NTG 10% 21% 26% 36% 45% 64% MI or NTG 16% 29% 36% 48% 56% 74% MI and NTG 22% 40% 47% 59% 67% 82% No Chest Pain EKG (ST, T wave ∆'s) History ST&T Ø ST  T  ST  ST  &T  ST  &T  No MI& No NTG 4% 9% 12% 17% 23% 39% MI or NTG 6% 14% 17% 25% 32% 51% MI and NTG 10% 20% 25% 35% 43% 62% The heart disease predictive instrument (HDPI) See reverse for definitions and instructions

40. Coronary Care Unit regular nursing bed chief complaint of chest pain? Coronary Care Unit regular nursing bed yes ST segment changes? any one other factor? (NTG, MI,ST  ,ST  ,T) yes yes no no no Fast and frugal classification: Heart disease Green & Mehr (1997)

41. .0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 .0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 Sensitivity Proportion correctly assigned False positive rate Proportion of patients incorrectly assigned Physicians Heart Disease Predictive Instrument Fast and Frugal Tree Emergency Room Decisions: Admit to the Coronary Care Unit?