Part 1
Supposing you want to see how student perception about civic engagement change in college. You collect data on first year students in 2010, and then collect data from the same students in 2014, their senior year. Suppose you have a perception score (on a scale of 1 through 5, 5 being the best) for 2010 and 2014 of each student, and other student characteristics.
A) What kind of data structure are you dealing with?
B) Does your data have limited dependent variable? If yes, what should you do?
C) How would you estimate the model?
Part 2
What if you just collect data on seniors in 2014 from all 4-year colleges of North Carolina about their perception on civic engagement?
D) What kind of data structure are you dealing with?
E) What should you do to estimate the model? What estimation issues are you going to face?