THREAD: Here is what I have learned from being an Assistant Professor at @NYUStern for two years! Will go through the tripod of service, teaching, and research.
First, some meta-advice: all advice is either 1) here’s what I did, it worked for me, you should do it too; or 2) I didn’t do this, I wish I had, learn from my mistakes. This is a mix of the two!
1. SERVICE - Show up to the office, keep your door open, sign up for all of the visiting speakers, and go to as many dinners as you can.
I get a lot of spam; but I respond to every email from people who seem like they have read my CV. If an email takes less than 2 min, I generally do that right away; and try to always stay at zero inbox
Sometimes people are confused how to approach seminar and conferences presentations when they start. But it’s actually easy! If you got a job, you know how to give a job talk. And all seminars are just job talks in disguise! Just try to approach them as seriously.
2. TEACHING - People often minimize the role of teaching, but it’s important to get it right early on.
Teaching electives is underrated. Teach an elective in a growing area, and in some sense your slot is “free” because someone always needs to do it. There are synergies between teaching and research; it’s more fun to teach more substantive content; and it adds to your market value
I was lucky because I inherited an elective course; but spending a month to shape a new course in your area is a valuable use of your time.
I like to spend a few minutes in the beginning of class discussing news stories and how they relate to course material; people seem to like that.
I also like to spend a bunch of time in class asking people to work in class on exercises. I haven’t yet used http://menti.com/  but this looks like a great way to keep track of student answers to these questions.
3. RESEARCH — Get your job market paper out ASAP!
I try to optimize around latencies in the publication process. Sometimes those are early on, ie buying data, but big ones are submission and sitting at a journal. Get those out of the way so they are at least sitting on other people’s inboxes.
Aside from that; I try to spend time on projects in rough proportion to the marginal product. Generally, that means more time on the project the further it is in the process. If R&R, drop everything else and just do this immediately. If close to submission, etc.
That’s the ideal; but more commonly I write down a daily dashboard of tasks and work on whatever most excites me. I’ve found that getting work done steadily is more important than fine tuning the right work with the right day; but I try to look back every week or so and check.
Always be thinking about new projects and ideas; talk to everyone about them and keep track of your thoughts: but in working be ruthless in killing bad ideas or projects and focusing your time on the small number of great ones. NB I am horrible at this.
Here’s what you need for a good paper:
1. Good question. This can be reshaped, especially in the abstract and first couple of pages. Senior colleagues are great in helping to refine the focus. This is essential for getting past an editor screen.
2. Good execution. Really this goes back to the first couple of years of grad school — follow best practices in your field and just execute well.
3. Is the result *novel*. Take three closest papers to yours; would your results be surprising given what you know from those papers? This is different from *does an existing paper do this already exactly.* This is the piece that I’ve found most challenging.
People are *massively* more likely to pick up my phone call or respond to my email. Think like a journalist. Want to figure out how people make decisions in a given area? Try asking them!
I think some people know me as someone with cool data; but I actually think people > data. With the right team, you can figure out the right dataset or empirical approach an dkeep it going.
In the long run really a lot is opening up in terms of data access; and being a professor opens even more doors. Think about what would be the *ideal* dataset or way to approach a problem; and try as hard as you can to get there. If you can’t, abandon the project and move on.
I’ve changed a lot of my views along the way, and am open to thoughts from others on any of this.
We are all very privileged to be given the opportunity to devote our lives to scholarship and teaching. I’ve been very fortunate to have a host of people who have helped me along the way, and it’s an increasingly and enjoyable part of my own job to help along others as well.
You can follow @arpitrage.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: