take the research idea you have been developing (Will crime rates down if drugs are legalized)
Restate your empirical question.
What type of design do you think is best for your question? Independent Groups? Matched Groups? Nonequivalent Groups? Within Subjects/Repeated Measures?
What is one potential problem with your design (e.g., between groups, within subjects, etc.)? How will you address this? (hint: think about random assignment, matching, and counterbalancing)