Tuesday, April 18, 2017

Calling All Econometrics Instructors

Please, please, please read this

It is an essay written by Josh Angrist and Jörn‐Steffen Pischke (the creators of Mostly Harmless Econometrics) calling for a paradigm shift in econometric teaching. I think that if we changed the way we taught econometrics, students at all levels would be better equipped to write good papers and maybe even more importantly, better equipped to judge the quality of different papers long after they have graduated. Maybe most importantly, those of us teaching applied micro courses could spend less time teaching econometrics and more time on content...ok, maybe that's just me being lazy.  ;)  

But seriously, I'd love to hear from folks teaching these courses. Do you believe that because we have changed "how we use econometrics we need to change the way we teach econometrics"? Why should students spend so much time thinking about "functional form, whether error terms are independent and identically distributed, and how to correct for serial correlation and heteroskedasticity" while researchers spend very little time thinking about them? Is econometrics "better taught by example than abstraction"? (Quotes are taken from the article.)
 

Sunday, April 9, 2017

How to Succeed in Academia: Be Curious

Well, here it isLasse Pedersen, from NYU, has produced a definitive guide to pretty much every aspect of my job. 

Favorite pieces of advice: 
1. "It’s about the same amount of time to write an unimportant paper as an important one." (p. 3)  This is very true. I can't tell you how often I have started "a quick little paper" only to spend a significant amount of time working on it. It's important to be choosy about topics. That said, better to work on something rather than nothing. Lots of times what I thought would be quick and easy but unimportant turned out to be complicated but more interesting. 
2. "Become a world expert on a literature." (p. 5)  Because that's just cool. Also, it's good for your letter writers to be able to say that you're the world's expert on X. But mostly, it's just cool. 
3. "The best papers often occur where theory meets empirics (i.e., they relate to both even if they only contribute to one of the dimensions)" (p. 7) 
4. "Talk about your research with lots of people." (p. 10)  I would add that talking about your research with experts will give you a better idea of where your research fits in the literature, talking with policy makers may give you a better sense of what is important and doable, and talking with your grandmother will give you practice explaining your ideas in easy to understand language. All of these are critical. 
5. On the importance of writing: "If a paper with an interesting idea gets rejected, it is often because it is so badly written that the editor/referee fear that its execution will be sub-standard even after several costly rounds of revision." (p. 17). Yes! Yes! Yes! 
6. "During seminars, draw the right questions by making your contribution clear." (p. 24) 
7. Discussing a paper at a conference: "A discussion is a service to the audience, not to the author and not an evaluation."  Yes! Yes! Yes! Of course, give the author all of your comments, but not necessarily in front of everyone. 

And my very favorite piece of advice: "Be Curious!" (p. 35) I think it's really, really excellent to have a job that that is the main overarching guide towards success (in the long run).  Thank you, Lasse, for reminding me of that. 




Saturday, April 1, 2017

Completely New To Stata? Start here.

Do you need to write a paper for a class and don't know how to start the data work? I recommend going through these modules provided by UCLA. If you find better ones, let me know, but remember that Stata is mostly about learning by doing. Start coding!