I am looking at the relationship between corruption and environmental degradation and got a negative coefficient in my stata regression. I have used fixed effects model, robust results, and a dataset of 124 countries.
Based on previous literature, I predicted a positive coefficient for corruption (when co2 emissions is the dependent variable)
Any suggestions on how I can justify/give a reason for these unexpected results?
There are two main possibilities here. 1) Check that you did not commit a mistake, if other studies have indeed and consistently found different effects. 2) Try to make sense of the result. Can it be due to a control you include that other researchers have not considered? Can it be due to a sample-selection process? Etecetera. Ultimately, the issue is theoretical.
Regards
So previousely you showed that CO2 is positively correlated with corruption, now you show that environmental degradation is negatively correlated with corruption. What is unexpected? Where is your problem?
What proxies are you using? More fundamentally, what makes you think corrupt governments cannot be environmentally sensitive? Here is the real problem: We live in a scholarly world too keen to use statistics. It is better to develop a frame of the physics of these social phenomena before dashing to statistics. That is, ask, what is actually happening? We call it case study.
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning