18.3. Research Design Concepts#

https://raw.githubusercontent.com/jillxoreilly/StatsCourseBook/main/images/regression6_ExpDes.png
  • What do the following three concepts mean, in the context of conducting experiments? Manipulation, Control, and Random Assignment

  • What is the point of randomization?

  • Why do we need a control group? Why not just compare the outcome before and after manipulation?

  • What is the difference between observational data and an experiment?

https://raw.githubusercontent.com/jillxoreilly/StatsCourseBook/main/images/regression6_payGap.jpeg

Example: A researcher working in an organisation wishes to understand whether a training course on equal opportunities and diversity raises the level of concern about the gender pay gap.

  • Spend a couple of minutes thinking about two possible study designs to address the question: one observational study and one experimental study. How can we infer causality in each case?

  • Do experiments always take place in a lab?

  • In what kind of research scenarios are experiments not possible?

  • What is a natural experiment?