August 2025 Quantitative Empirical Methods Exam Answers

2026-05-04

Here are my answers to the August 2025 quantitative empirical methods qualifying exam. I hope they are useful in studying for the exam, but be aware that the format is changing from an at-home to an in-person exam, so I would not over-index on past years' exams.

People get psyched out about this exam. Frankly, you need to remember that while the exam is difficult, no single question is too hard for you to answer, especially after a summer of studying. A great way to not learn any of the material and fail the exam is to tell yourself that the exam is too hard and the curriculum is unreasonable.

You can see this by looking at a historical exam. Ask yourself if there are any questions on there that you would seriously not be able to answer in about an afternoon with free access to the internet. You could probably do well on the exam even if there are questions you would not be able to answer, but if you have taken a causal inference course you will likely find all of the problems tractable.

Here are some miscellaneous tips for studying:

• The amount of time it will take you to get ready for the exam will depend on the number of quant classes you have taken before. In my mind, there is no reason to not take the exam first year as long as you know you can dedicate the time to studying for it.
• The best way to be ready for this exam is to stay ready. If you can find a way to work on practice problems / puzzles related to the exam material then you can prepare yourself for the exam with little formal studying. This is especially true for probability questions, but these seem to have become less common in the exams over time.
• I think that study groups are an important commitment device that will help you prepare for the exam. I did four or five zoom calls with other students and I would always study much more in preparation for these calls so I could participate and so that I wouldn't look unprepared.
• Know what a p-value is. This sounds stupid but most people (even applied quantitative researchers) have misconceptions about what a p-value can and cannot tell you. Take the six question quiz in this article and commit the answers to memory. Most tests contain a question that tries to get you to misstate what a p-value can tell you.
• About half of the computer-aided sections ask you to do a simulation study, and the other half ask you to analyze a dataset. You probably analyze a dataset in R every day (or at least every week) so you probably don't need to study for this. Unless you do methods, you probably have much less experience writing simulation studies so this is the part of the exam you should practice. If you are fast, you can do this part of the test in under an hour.