General information about supervision

• Aligning expectations: what can you expect from me and what do I expect from you?
• Choosing a topic: what’s right for you?
• Writing a thesis: how to write a thesis in economics.
• Using AI tools: you need to use them, but how?
• Project management: how do you plan and structure the largest project of your life so far?

General tips

• Groups: I recommend working in a group.
• Read other theses: You can find previuos student theses, and it is important to read at least two good ones early on to get a sense of the format, writing style, and structure. Ask the study administraiton for access.
• Literature search: Use google scholar. Find a key paper (possibly with my help), then download and read the abstract of every single paper they cite. Next, read the full introduction of the most relevant of those papers, and fully read the 1-3 most relevant of those.
  • Another approach to literature searching is to search for your key paper and then in google scholar click the “cited by” link. This will list papers that have been written subsequently but which cite your paper. The displayed list is in some sense sorted by “importance” (judged by google).
  • Journals: There are many journals out there, and some are completely irrelevant and full of junk. Of course, as a student you have no way of knowing the good from the bad. And in practice, the relative rankings of journals is evolving. But as a starting point, consult a ranking list for economics journals, e.g. REPEC, and make sure that the journal you are reading a paper from shows up among at least the top 200 journals. For working papers, you need to make sure that the source is reputable: e.g. authors are from reputable great universities, or published in a good working paper series, like IZA, CEPR, NBER. (Ask your supervisor if you are in doubt.)
• Data work: When you are struggling, make sure to simplify your problem to the minimal version that reproduces the issue you are struggling with. Don’t learn to use Pandas’ merge command on two 1GB datasets: use two datasets with 3 rows each, then scale up. This tip may seem simple/stupid, but I promise you that you will ignore it and you will regret ignoring it. Try to minimize the pain you inflict on yourself in this regard.
• Know thy self – getting stuck: Getting stuck is a part of the process. Learn to understand and identify your own signals. Do you procrastinate with instagram/shopping/netflix/cleaning your apartment? Regardless, know that when you smash your brain against a problem that it cannot solve, your brain will try to rebel against you by procrastinating. It will win out eventually, so you must give up before you waste too much time. Instead, find a different task you can switch to. For example, make a rule where you switch tasks if you have been stuck with the same problem for over 30min (e.g. go from coding to writing or reading).

Dos and don’ts