- Step 1: Begin with a collection of subjects (e.g. 60 pupils).
- Step 2: Devise a method to randomise that is purely mechanical (e.g. flip a coin, draw names out of a hat, or use Excel.
- Step 3: Assign pupils from one selection (e.g. "Heads") to be in the control Group. Assign pupils from the other selection (e.g. "Tails") to receive the approach you want to evaluate.