Control Group

• Control Group:

  • By including a group who undergo the same conditions (except...) as the experimental group we can control for numerous possible confounders
  • For within-subjects designs this might be a control condition
• Basic parallel experimental design (pre test-post test)

  • one group is the experimental group that is affected by the independent variable/manipulation
  • one group is the control group that is unaffected by the independent variable/manipulation
  • both groups have the same baseline for the dependent variable - if not, you're not controlling for confounding variables
  • you can solve this by recruiting people going into the experimental or control group from the same sample population, so it is ensured they have the same baseline

  
• Basic parallel experimental design (pre test-post test)

  • you need to ensure the control and experimental group are as equal as possible during the baseline/pre-test
  • this can be done via
  • matching groups
  • use the same group for the different conditions
  • randomisation