The researchers could have used a matched pairs design in order to create 'experimental twins'. The researchers would first have to identify a variable which may affect the dependent variable - the one being measured - for example, age or gender. Then the participants would have to be assessed according to this variable and therefore creating pairs of individuals. The researchers could then place one of these 'twins' in one condition and the other in the second condition in order to overcome confounding participants variables in an independent groups design.