#bias-assessment

Identify and control systematic bias in research design. Common biases: 1. Selection bias: non-random sample not representative of population. Mitigation: probability sampling, quota sampling, post-stratification weights. 2. Information bias: systematic error in data collection. Mitigation: standardized instruments, blinded assessments, multiple informants. 3. Recall bias: differential accuracy of memories between groups. Mitigation: prospective design, objective records, shorter recall periods. 4. Confirmation bias: seeking information that confirms hypotheses. Mitigation: preregistration, blinded analysis, adversarial collaborations. 5. Publication bias: selective reporting of positive results. Mitigation: study registries, reporting negative results. Assessment tools: Newcastle-Ottawa Scale for observational studies, Cochrane Risk of Bias tool for RCTs. Sensitivity analysis: test robustness of findings to different assumptions about bias.