🔍 Selection bias: Non-random sampling hurts generalizability → Fix: randomize recruitment. 🏥 Berkson bias: Hospital-only samples ≠ community → Fix: recruit beyond hospital or adjust. ⏳ Attrition bias: Unequal loss to follow-up distorts results → Fix: strong follow-up, intention-to-treat. 🧠 Recall bias: Retrospective studies; cases remember exposure differently → Fix: records/short recall window. 📏 Measurement bias: Poor instruments/protocols → Fix: standardized tools, placebo where relevant.