Topic: Crossover Designs

Why Preliminary Carryover Tests Can Mislead

Why Replicate Crossover Designs Are Used

Withdrawals and Incomplete Two-Period Records

Within-Subject Comparisons in Crossover Studies

Worked Four-Period Full-Replicate Layout

Worked Three-Period Partial-Replicate Layout

Worked Two-Period Schedule Audit

Treatment-by-Period Interaction

Treatment-Period Balance in Crossover Schedules

Treatment-Specific Variability Questions and Design Choice