• type of randomised controlled/clinical trial
• problem with multiple treatment conditions when using repeated measures sequence of interventions may bias results
• crossover design used subjects as own controls
• all subjects receive the different interventions, but order is randomised and a ‘washout’ period may be used between interventions to prevent the carryover of effects.
• two methods
1) Randomise order of intervention
2) Crossover – half subjects receive one first and the other second