Question: When might a quasi-experimental design be chosen over an experimental design? What considerations must a researcher make when deciding which approach to take? What are the threats to validity and reliability is this design trying to control for?