Statistical Experiments
Experimental design connects the scientific question to treatment assignment, independent replication, and the analysis of observed responses. The design determines which comparisons the data can support.
Begin with Design & Mean Comparisons. Use blocking to account for nuisance variation, factorial designs to study interacting factors, response surfaces to explore optimization, and random-effects models for grouped or nested observations.
Design & Mean Comparisons
Experimental units, treatment comparisons, and ANOVA.
Blocking & Nuisance Variation
Blocking and control of nuisance variation.
Factorials & Screening
Factorial designs, interactions, and screening.
Response Surfaces & Optimization
Curvature, response surfaces, and optimization.
Random Effects & Hierarchies
Random effects, mixed models, and nested designs.