Currently I am working on my undergraduate thesis. I am studying the effect of salary differentials between the EC staff and between EU member states on the representation of nationality in the EC bureaucracy. So:
Independent variable: salary differential (the difference between the net disposable income of an average EC employee and the net disposable income on country-level of an EU member state, given in Purchasing Power Standard)
Dependent variable: representation ratio of nationality (the percentage of EC staff of a certain nationality divided by the percentage of that nationality in the total EU population, where a ratio of 1.0 means equal representation, below 1.0 underrepresentation and above 1.0 overrepresentation).
Since there are 28 EU member states, I'll have 28 observations. To increase the number of cases, I have the data for a period of 5 years (2013-2017), giving 28 x 5 = 140 cases. My teacher told me that these observations of course are not independent. My question now is: what statistical model do I use to analyse this data?
I am only familiar with OLS regression analysis. So I really some help in deciding what model to use.
I hope you guys can help me out!