AiMTurkSYNTHETIC RESEARCH PRETEST · N=200
Hindsight bias
Ask two matched groups to judge an event, revealing the outcome to only one group.
Design and results
Synthetic responses followed the predicted direction. This is a directional description, not a matched human-participant test.
This study was selected after a twelve-study screening run. The full archive includes studies that did not match predictions.
| Group | n | Result |
|---|---|---|
| Outcome unknown | 100 | 46.3 |
| Outcome known | 100 | 56.75 |
Group B minus A: 10.45.
Bootstrap 95% interval for the difference: 8.4–12.55.
Data and review
Downloads contain synthetic responses. Keep the questionnaire version, codebook and provenance label when analysing them.
Limitations and source
These results can help test questionnaire materials, branching and analysis workflows. They do not estimate human-population effects or replace human validation.
Fischhoff (1975) Original research