Research Methods, Evidence-Based Practice, and Statistics
This domain tests study design, measurement, bias, statistical inference, and the integration of evidence into clinical decisions. Expect items on PICO questions, validity, diagnostic accuracy, and interpretation of p values and confidence intervals.
Core rules
- PICO frames a clinical question: population, intervention, comparison, outcome. Systematic reviews and randomized controlled trials generally rank above cohort, case-control, and case reports for intervention questions.
- Internal validity concerns confounding, selection, measurement, and attrition bias. Reliability is consistency; validity is accuracy of measurement for its purpose.
- Diagnostic accuracy uses sensitivity, specificity, predictive values, and likelihood ratios. Predictive values depend on prevalence.
- Inferential tests include t tests, ANOVA, chi-square, correlation, and regression. A p value is the probability of data at least as extreme under the null; it is not the probability that the null is true. Report effect sizes and confidence intervals.
Worked example
Among 80 people with a disorder, 60 test positive; among 20 without it, 10 test positive. Sensitivity = 60/80 = 75%. Specificity = 10/20 = 50%. Positive predictive value = 60/(60+10) = 86% in this sample. Reasoning: sensitivity and specificity are properties of the test across disease status; predictive values change when prevalence changes, so do not transport PPV to another setting without adjustment.
Traps
Do not infer causation from correlation. Do not treat statistical significance as clinical importance. Do not ignore confidence intervals or sample size. Do not confuse sensitivity with PPV.