If you're interested in the USF undergraduate economics program, the MS in Applied Economics (MSAE), or the MS in International and Development Economics (MSIDEC), please see our department's Linktree, here.
Otherwise, I'm going to be using this space to provide resources, especially what I regard as best practices, for current students.
So you wanna learn econometrics
In general all statistics is the same. There are some data. There is an estimator. You look at the estimator and you think. If you're lucky you get less wrong or something igidk. In any event, if you would like to get better at it, I recommend working through the following.
Introductory materials, basic philosophy
- USF's very own Mike Jonas has some very friendly YouTube videos covering introductory undergraduate econometrics students
- My slides on economic research designs and the "identification revolution" in economics are here. Also, of course, & so forth.
- Shalizi's Data over Space and Time is an excellent intro course, as is his Undergraduate Advanced Data Analysis / ADAEPOV , depending on your idea of "intro". His blog is an inspiration to esoteric nerds everywhere, notably recently in the domain of shoggothim.
- ISLR is excellent, and widely loved for a reason, plus it has a ton of supporting material.
- McElreath's Statistical Rethinking is more or less guaranteed to make you a better statistician.
- Heiss's classes are excellent and markdown-rich, see e.g. on impact evaluation and measurement
- ... you should probably sign up for free weekly emails from the National Bureau of Economic Research.
- ... you could also just read through everything Gelman has ever blogged.
Keeping up with the epistemological Joneses
- Goldsmith-Pinkham is providing a constant stream of data public goods, notably in his Applied Empirical Methods course.
- Facure's Causal Inference for the Brave and True is excellent, and perhaps the best current treatment of causal machine learning methods .
- Not to be that guy, but these days every time I come across a new estimator, perverse sample construction issue, or horrifying development in the difference-in-differences literature, I like to run it by Claude, and then if need be chat up a little demo in R or Python using Claude Code to figure out how it works. More on that to come very soon.
- On that note: how does one best use Claude Code? This is, of course, evolving, but Sant'Anna has some great setup notes here, Antonio Mele has some nice tools here, and Goldsmith-Pinkham has some general advice here.