--- name: psychology-evidence-appraisal description: Use when a learner asks whether a psychological study, experiment, therapy claim, theory, or popular finding is supported by empirical evidence. --- # Appraise empirical psychology Evaluate what the study design and wider evidence can support. Distinguish a historical theory, a single result, a replicated finding, and an accepted clinical recommendation. ## Workflow 1. Translate the claim into a testable question: construct, participants, comparison, outcome, and setting. Check that the measured outcome matches the claim made about it. 2. Inspect the original study or the best available report. Use `get_pubmed_abstract` or `get_open_access_sections` when available, and label the exact access level. Record design, sample and recruitment, measures, controls, analysis, and reported results only where the inspected text actually supplies them. Note whether the report is a preprint, peer-reviewed article, review, or historical account. 3. Assess threats to inference: selection bias, confounding, small or unrepresentative samples, unreliable measures, multiple testing, missing data, and selective reporting where relevant. Distinguish correlation from causal evidence. 4. Look for independent replication, systematic reviews, or credible later criticism. Check whether the finding generalizes to the people or setting the user asks about. Treat a famous experiment's cultural influence separately from its evidentiary strength and ethics. 5. State the evidence level in ordinary language and name the uncertainty. An abstract may support a narrow report of its stated findings, but may omit crucial methods or limitations. Use effect sizes, confidence intervals, and study counts only when verified from inspected sources; do not invent numerical precision. ## Deliverable Lead with a direct answer: what the evidence supports, what it does not establish, and why. Cite inspected sources near the claims. For advanced learners, include methodological details and competing explanations; for beginners, explain the same limits with a concrete example. Do not diagnose an individual or turn an educational answer into personal treatment advice.