Elea McDonnell Feit (@eleafeit) header
Elea McDonnell Feit (@eleafeit) profile picture

Associate Professor of Marketing @DrexelUniv. Bayesian. Philadelphian. Borough Council for KSQ. Opinions mine.

Philadelphia
Joined January 2010
Photo by Elea McDonnell Feit (@eleafeit)Photo by Elea McDonnell Feit (@eleafeit)Photo by Elea McDonnell Feit (@eleafeit)Photo by Elea McDonnell Feit (@eleafeit)Photo by Elea McDonnell Feit (@eleafeit)Photo by Elea McDonnell Feit (@eleafeit)Photo by Elea McDonnell Feit (@eleafeit)Photo by Elea McDonnell Feit (@eleafeit)Photo by Elea McDonnell Feit (@eleafeit)
Any PhD student in Marketing can tell you that endogeneity and causal inference are important. But do you know just how many in marketing papers use causal inference methods? @DPapies, Peter Ebbes and I took a closer look.
5
14
6
89
19,164
Check out some of the cool events on global trade and sanctions that my colleagues at the new @DrexelCGPA are planning!
Many thanks to our great partners @WIFOat and @WU_Econ for hosting the Second Annual Workshop on sanctions in Vienna on May 27, 2025! 🙏 Please consider submitting your paper on sanctions before the deadline on February 28, 2025. We hope to see you there.
2
258
#Eagles and #Dragons fly together! 💚🦅🤍 💙🐉💛 Drexel will close on Friday to join the rest of #Philly in celebrating the Champions!
236
Goodbye X. You can find me over at BS. Still @eleafeit over there.
5
809
Elea McDonnell Feit retweeted
🎇 PhD student internships in Computational Social Science at Microsoft Research NYC 🎇 These are 12 week, in person internships with some combo of Dan Goldstein, Jake Hofman, & David Rothschild Early (January - April) start dates somewhat preferred jobs.careers.microsoft.com/g…
3
71
6
276
55,609
You want to know the effect of some treatment on an outcome and have a time series of outcomes for each subject pre- and post-treatment. Access to treatment is regional and is arguably quasi-random. Compliance is definitely not random. What analysis would you do?
2
4
649
Elea McDonnell Feit retweeted
On September 13, 1826, Anthony J. Drexel was born in Philadelphia. Happy birthday to our university's founder!
3
34
1,127
Treatment/control incrementally experiments measure the total incrementally of a campaign. That's great if you are trying to decide whether or not to use a particular channel, but...
How many consumers should a campaign reach or what budget should be set? @CaioWaisman and I present a method for these *intensive margin* ad decisions. 📜"Multi-cell experiments for marginal treatment effect estimation of digital ads", soon at MngSci. arxiv.org/abs/2302.13857
2
2
5
1,988
To set advertising budgets, you really need to know the _marginal_ incremetality. This paper shows how to design an experiment to measure marginal incrementality. Super important!
4
282
Elea McDonnell Feit retweeted
How many consumers should a campaign reach or what budget should be set? @CaioWaisman and I present a method for these *intensive margin* ad decisions. 📜"Multi-cell experiments for marginal treatment effect estimation of digital ads", soon at MngSci. arxiv.org/abs/2302.13857
2
9
1
28
4,402
I'm very sad to share the news that my dear colleague @DanielKorschun passed away last week. He brought an endless supply of espresso and optimism to our department that I will miss. You can read more at mailchi.mp/lebow/remembering…
11
461
How is it that I can stream TikTok's all day long, but getting a weather radar animation takes for-ev-er?
3
386
Is diff in diff just a very specific and highly-manual form of matching?
1
2
381
Today’s perspective
12
456
Well, I guess I can cross "learn Mathematica" off my to-do list. ChatGPT 4o can take partial derivatives of complex functions.
4
15
1,361
And... it just wrote a unit test for me.
1
165
If you put in the wrong likelihood to start, it doesn't find the mistake. So, I guess I'm still useful for something...
1
2
215
Except ther I’m the one that put in the wrong likelihood, so maybe not 🤔
1
172
Is this old news for everyone?
1
268
Nothing like the feeling of picking up a project that you set aside months ago and solving the problem you were hung up on.
1
6
521