• most highly controlled method of hypothesis testing
  • goal: Isolation of a cause-and-effect relationship
  • independent variables (IVs) are predicted to impact the dependent variables (DVs)
  • manipulation of IVs, then measurement of DVs while controlling all other confounding variables
  • random assignment of participants to different conditions (= different versions of IVs)
    • e.g. treatment vs control group
    • identical groups
      • even tho people cannot be cloned we have statistical unbiased data
      • only applicable to larger sample sizes
        • how many people have long hair? sample size of 2. results: 100%, 50%, 0%
        • with higher sample size the results gets closer to reality
    • all same probably to be selected
    • no prior knowledge about unobservable information necessary

Typical Session

  • motivation, research question, hypothesis, instructions, recruitment, (ethics approval)
  • participant check-in
  • (informed consent)
  • control: side-talk, mobile phones, computer use
  • instructions distributed (read aloud)
    • people know that the instructions are common knowledge
      • important for econ
    • people need to listen to it - cannot miss important details
  • answer questions privately
  • comprehension test, practice rounds maybe
  • SAVE THE DATA
  • post-experimental questionnaire
  • participants paid and dismissed