Run tasks

Statistics

State the hypothesis and run /stats — the skill identifies the design, checks assumptions, chooses and justifies the test, and reports effect sizes with confidence intervals and achieved power.

Start

text
/stats Using pss_trial.xlsx, test whether the biofeedback group reduced PSS-10 more than control. Identify the design, check assumptions, choose and justify the test, and report the effect size with a 95% CI and achieved power.

If the data was already analyzed in the session, you can just say "test the group difference in change scores".

What happens

  • Design recognition — between/within, number of groups, repeated measures, covariates.
  • Assumption checks — normality, variance homogeneity, outliers; the choice between parametric and non-parametric tests is made from the checks and stated.
  • Test — run with the verified tools; multiple-comparison correction when relevant.
  • Effect sizes — Cohen's d, η², odds ratios and the like, with 95% CIs, plus achieved power for the observed effect.
The statistical report: question, hypotheses, assumption checks, test and effect size.
The statistical report: question, hypotheses, assumption checks, test and effect size.

What you get

  • <analysis>_statistical_report.md — question and hypotheses, design, assumption checks, test selection and justification, results, effect sizes, power, interpretation.
  • <analysis>_descriptives.csv, a group plot (boxplot with means and CIs), and statistics_structured.json for downstream use.
Effect sizes summarized in the answer.
Effect sizes summarized in the answer.

Psychometrics

Reliability (Cronbach's α, McDonald's ω), EFA, CFA, SEM, IRT, mixed models, Bayesian estimation and power analysis are available as method skills — ask for them by name. See Neutropic for psychometrics.

A CFA path diagram.
A CFA path diagram.

Walkthrough: Running the statistics.

Data analysis PreviousFigures & illustrationsNext