Statistical Workflow

stats
lm
Published

November 10, 2022

A statistical workflow is generally the same for most analyses, and thinking your way through this workflow can help you decide what kind of analysis you’d like to do.

  1. Decide on your question(s)
    • Helps guide exploration and models
  2. Preliminary data exploration
    • Helps determine answers to questions and guide model development
  3. Decide on a model type
    • Simple linear model, generalized linear model, mixed model, etc.
    • What package to use for this model
  4. Run model
  5. Check model diagnostics/assumptions
    • As required, go back to Step 3 (Model type)
  6. Consider transformations if required
    • If so, go back to Step 4 (Run Model)
  7. Interpret model
    • Summary tables
    • ANOVA tables
    • Post-Hoc analyses
  8. Create figures to visually show your model results