A well-designed experiment, study, or trial is vital to draw (causal) conclusions from data that are valid and efficient. This workshop is intended to provide a guided exploration of different design choices, corresponding appropriate statistical models, and their consequences and trade-offs to help you plan a stronger study. Topics covered will include:
- Validity and types of studies (observational, quasi-experimental, randomized)
- Cross-sectional vs longitudinal designs
- Clustered or split-plot (hierarchical) designs
- Crossover studies
- Best practices
- Randomization
- Blocking or stratification
- Replication
- Choosing a statistical model to fit the design
Although not required, familiarity with ANOVA and mixed-effects models covered in our workshop Linear Mixed Effects Models is suggested. See also Interpreting Linear Models: Regression and ANOVA.