DR-301f · Module 2
Anchoring & Recency Effects
3 min read
Anchoring bias causes the first data point you encounter on a topic to disproportionately influence your final assessment. If the first article you read estimates a market at $5B, every subsequent estimate is evaluated relative to that anchor. An estimate of $8B seems high. An estimate of $3B seems low. But if the first article had estimated $12B, $8B would seem low and $3B would seem absurdly low. The anchor changed, and your entire calibration shifted with it. Recency bias is the mirror: the most recent data point you encountered disproportionately influences your assessment. Both are mitigated by the same technique — collecting all data points before forming an assessment.
Do This
- Collect all relevant data points before forming a position — do not evaluate as you go
- Randomize the order you review sources when possible — sequence should not determine weight
- Look at the full range of estimates, not just the first and last
- When presenting ranges in your output, note the full spread and the methodology behind outliers
Avoid This
- Form a preliminary conclusion after the first source and then "test" it against subsequent sources
- Over-weight the most recently published estimate because it feels most "current"
- Discard outlier estimates without investigating why they differ from the cluster