Hypothesis Testing Design Chain

validated scientific chain v1.0.0 cc-by-sa

Block ID: f4b02706-482b-4174-a239-2fe4a3de0eb1

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

Template

Design a statistical test for the research question in {research_question}, committing to each decision before the next. (1) Translate the question into a precise null and alternative hypothesis about a defined population parameter — vague hypotheses guarantee uninterpretable results. (2) Choose the primary outcome measure and justify it over alternatives; declare it primary now, because outcomes chosen after seeing data are exploratory, not confirmatory. (3) Select the test appropriate to the data type and design (paired versus independent, parametric versus rank-based), stating the assumptions the test requires and how each will be checked. (4) Set alpha and justify one-sided versus two-sided — one-sided requires a pre-committed directional rationale. (5) Perform a power analysis: state the minimum effect size worth detecting, the assumed variance and its source, and compute required sample size at the target power. (6) Pre-specify the handling of outliers, missing data, and multiple comparisons. Document all six decisions before data collection begins.

Design a statistical test for the research question in {research_question}, committing to each decision before the next. (1) Translate the question into a precise null and alternative hypothesis about a defined population parameter — vague hypotheses guarantee uninterpretable results. (2) Choose the primary outcome measure and justify it over alternatives; declare it primary now, because outcomes chosen after seeing data are exploratory, not confirmatory. (3) Select the test appropriate to the data type and design (paired versus independent, parametric versus rank-based), stating the assumptions the test requires and how each will be checked. (4) Set alpha and justify one-sided versus two-sided — one-sided requires a pre-committed directional rationale. (5) Perform a power analysis: state the minimum effect size worth detecting, the assumed variance and its source, and compute required sample size at the target power. (6) Pre-specify the handling of outliers, missing data, and multiple comparisons. Document all six decisions before data collection begins.

Variables

Name	Type	Required	Trust level
research_question		yes

hypothesis-testingexperimental-designstatisticschain-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