statistical-power
Calculates sample sizes and statistical power for study planning
by k-dense-aiFree
Description
statistical-power is an agent skill for planning studies by calculating sample sizes and statistical power. It answers how many independent experimental units are needed to detect an effect of scientific importance, or what effects a feasible sample could detect. The skill says to choose the inferential goal first, because precision, equivalence, non-inferiority and sequential monitoring each need their own calculations that the bundled superiority-test helpers do not cover. For any given test, four quantities are linked: sample size, effect size, significance level and power. Fixing three of them lets you solve for the fourth, when a solution exists, for a fixed design, analysis and nuisance parameters. The skill covers analytical calculations, which are exact under standard assumptions and approximate otherwise, and simulation or Monte Carlo methods, which need a credible data-generating process and a calibrated planned analysis. It also helps you choose and convert effect sizes, which is often the hardest part. Use it when someone asks how many subjects, samples or replicates are needed, or wants an a priori power analysis.…
Highlights
- Sample size and power calculations
- Analytical and simulation-based methods
- Effect size selection and conversion
- Matches the inferential goal first
- A priori study planning
Ratings & Reviews
No ratings yet
Be the first to rate statistical-power.
People also visit
statistical-power FAQ
What question does it answer?
How many independent units are needed to detect a scientifically important effect, or what effects a feasible sample could detect.
How are the four quantities related?
Sample size, effect size, significance level and power are linked. Fixing three lets you solve for the fourth, when a solution exists.
Does it cover precision or non-inferiority?
Not with the bundled superiority helpers. The skill says those goals need their own calculations.
When should simulation be used?
When there is a credible data-generating process and a calibrated planned analysis, the skill says.
How do I install it?
Run npx skills add https://github.com/k-dense-ai/scientific-agent-skills --skill statistical-power.
Is it checked for security?
Its security audits show passes from Gen Agent Trust Hub, Socket and Snyk.
What are the best statistical-power alternatives?
Popular alternatives to statistical-power include consciousness-council, experimental-design and pydeseq2. See all statistical-power alternatives