Overview
This guide addresses common biases affecting product research and experiments within GLIDR (opens in new tab). There are many biases you should watch out for so you don't skew your results.
Cognitive Biases
The article identifies 15 cognitive biases, including:
- Anchoring Effect (opens in new tab): Using initial information to shape later judgments
- Availability Bias (opens in new tab): Assessing likelihood based on how readily examples come to mind
- Confirmation Bias (opens in new tab): Favoring information supporting existing beliefs
- Curse of Knowledge: Difficulty empathizing due to superior subject knowledge
- Halo Effect (opens in new tab): Letting positive impressions obscure negative aspects
- Hindsight Bias (opens in new tab): Falsely believing past outcomes were predictable
- Observer Bias (opens in new tab): Influencing research through knowledge of study objectives
- Overconfidence (opens in new tab): Overestimating personal abilities and underestimating personal risk
- Primacy (opens in new tab)/Recency (opens in new tab) Effects: Overweighting initial or final information
- Self-Fulfilling Prophecy (opens in new tab): Expectations shaping observed behavior
Research Biases
Seven additional research-specific biases are covered:
- Selection Bias (opens in new tab)
- Measurement Bias (opens in new tab)
- Framing Effect (opens in new tab)
- False Positives/Negatives (opens in new tab)
- Omitted-Variable Bias (opens in new tab)
- Planning Effect (opens in new tab)
The article emphasizes consulting individual method sections for bias details and provides extensive external references for deeper learning.