Study Quality Appraisal Chain

validated scientific chain v1.0.0 cc-by-sa

Block ID: d5536fae-c2c8-4cba-a77b-24d9d5d394b2

Community-contributed block. PromptDNA makes no guarantee of output quality or fitness for purpose. User assumes all responsibility for use.

Template

Appraise the study described in {study_description} systematically before accepting its conclusion. (1) Identify the design — randomized trial, cohort, case-control, cross-sectional, case series — and state the causal strength ceiling that design permits regardless of execution quality; a cross-sectional study cannot establish temporality no matter how large. (2) Examine selection: how were subjects recruited, what population do they represent, and who was excluded — then state who the results can generalize to. (3) Assess measurement: were exposure and outcome measured with validated instruments, by assessors blinded where feasible, at appropriate timepoints? (4) Hunt confounding: list plausible third variables and check whether the design or analysis (randomization, matching, adjustment) addressed each. (5) Evaluate the statistics: sample size relative to effect size, appropriateness of tests, whether reported outcomes match what was pre-registered, and multiple-comparison handling. (6) Weigh attrition and missing data. Conclude with a verdict: what the study actually demonstrates, stated more conservatively than its abstract likely does, and the single flaw that most limits it.

Appraise the study described in {study_description} systematically before accepting its conclusion. (1) Identify the design — randomized trial, cohort, case-control, cross-sectional, case series — and state the causal strength ceiling that design permits regardless of execution quality; a cross-sectional study cannot establish temporality no matter how large. (2) Examine selection: how were subjects recruited, what population do they represent, and who was excluded — then state who the results can generalize to. (3) Assess measurement: were exposure and outcome measured with validated instruments, by assessors blinded where feasible, at appropriate timepoints? (4) Hunt confounding: list plausible third variables and check whether the design or analysis (randomization, matching, adjustment) addressed each. (5) Evaluate the statistics: sample size relative to effect size, appropriateness of tests, whether reported outcomes match what was pre-registered, and multiple-comparison handling. (6) Weigh attrition and missing data. Conclude with a verdict: what the study actually demonstrates, stated more conservatively than its abstract likely does, and the single flaw that most limits it.

Variables

Name	Type	Required	Trust level
study_description		yes

critical-appraisalstudy-designevidence-evaluationchain-of-thought

Ratings

0.0

Overall (0)

0.0

Accuracy

0.0

Consistency

0.0

Clarity

0.0

Efficiency

Benchmarks

Not yet self-validated against any benchmark. Automated, evaluative only - not a factor in whether this block was published.

Submitted by James P FounderMod via mcp · 2026-07-18