You need to academically justify every single aspect in your dissertations. From sample selection to methods!
2. You need to identify one main paper that resembles what you want to do and follow it!
3. The minimum you need to have is at least one independent variable which will be part of your main hypothesis and add to your contribution. Success depends on the contribution too.
4. You need to have a model and define dependent and ALL independent variables using references (i.e., citing papers). You need to say where you get them from too!
5. Stick with a simple analysis. For example, just short-run returns should be enough. Or just long-run returns. You don’t need to do both due to time limitations.
6. AVOID field research. There is not time for that! If you decide to do so then it is your own responsibility but I am not recommending it. I would suggest you do only secondary data which will benefit you also in your future jobs. It demonstrates analytical thinking and also that you are able to deal with big datasets, spreadsheets etc. Keep it simple!
7. Ask WRDS if the data you are looking for are there. We are subscribed to a huge lists of datasets and most likely what you are looking for is there!
8. Make sure you use up-to-date data when possible. For example, your sample should run until 2014 or even 2015 (from whenever the data is available).