experimental-design

SUMMARY

What it does

This skill provides procedural guidance for designing experiments and studies before data collection. It covers randomization, blocking, stratification, controls, factorial and fractional-factorial designs, design of experiments (DOE), screening designs, response-surface optimization, crossover, repeated-measures, split-plot, cluster-randomized, Latin square, and sequential/adaptive designs. It includes decision trees, reference documents, and Python scripts for generating randomization schedules and DOE matrices. The skill emphasizes avoiding structural mistakes such as pseudoreplication and confounding. It is designed for research agents and requires Python >=3.10 with numpy, pandas, and pyDOE3. The skill is part of the Scientific Agent Skills library by K-Dense Inc. and is licensed under MIT.