Causality/2
-
16. Regression Discontinuity DesignCausality/2 2025. 3. 30. 15:22
https://matheusfacure.github.io/python-causality-handbook/16-Regression-Discontinuity-Design.htmlWe don’t stop to think about it much, but it is impressive how smooth nature is. You can’t grow a tree without first getting a bud, you can’t teleport from one place to another, a wound takes its time to heal. Even in the social realm, smoothness seems to be the norm. You can’t grow a business in one..
-
15. Synthetic ControlCausality/2 2025. 3. 30. 07:45
https://matheusfacure.github.io/python-causality-handbook/15-Synthetic-Control.htmlOne Amazing Math Trick to Learn What can't be KnownWhen we looked at difference-in-difference, we had data on multiple customers from 2 different cities: Porto Alegre and Florianopolis. The data span 2 different time periods: before and after a marketing intervention was done in Porto Alegre to boost customer depo..
-
14. Panel Data and Fixed EffectsCausality/2 2025. 3. 27. 12:13
https://matheusfacure.github.io/python-causality-handbook/14-Panel-Data-and-Fixed-Effects.htmlIn the previous chapter, we explored a very simple Diff-in-Diff setup, where we had a treated and a control group (the city of POA and FLN, respectively) and only two periods, a pre-intervention and a post-intervention period. But what would happen if we had more periods? Or more groups? Turns out this ..
-
13. Difference-in-DifferencesCausality/2 2025. 3. 27. 08:58
https://matheusfacure.github.io/python-causality-handbook/13-Difference-in-Differences.htmlThree Billboards in the South of BrazilThe problem with billboard and TV ads is that it is hard to know how effective they are. Sure, you could measure the purchase volume, or whatever you want to drive, before and after placing a billboard somewhere. If there is an increase, there is some evidence that th..
-
12. Doubly Robust EstimationCausality/2 2025. 3. 27. 07:25
https://matheusfacure.github.io/python-causality-handbook/12-Doubly-Robust-Estimation.htmlDon't Put All your Eggs in One BasketWe’ve learned how to use linear regression and propensity score weighting to estimate E[Y|T=1]−E[Y|T=0]|X. But which one should we use and when? When in doubt, just use both! Doubly Robust Estimation is a way of combining propensity score and linear regression in a way y..
-
11. Propensity ScoreCausality/2 2025. 3. 26. 07:31
https://matheusfacure.github.io/python-causality-handbook/11-Propensity-Score.htmlThe Psychology of GrowthThe field of positive psychology studies what human behaviours lead to a great life. You can think of it as the intersection between self help books with the academic rigor of statistics. One of the famous findings of positive psychology is the Growth Mindset. The idea is that people can hav..
-
10. MatchingCausality/2 2025. 3. 25. 16:14
https://matheusfacure.github.io/python-causality-handbook/10-Matching.htmlWhat is Regression Doing After All?As we’ve seen so far, regression does an amazing job at controlling for additional variables when we do a test vs control comparison. If we have independence, (Y0,Y1)⊥T|X, then regression can identify the ATE by controlling for X. The way regression does this is kind of magical. To get so..
-
09. Non Compliance and LATECausality/2 2025. 3. 25. 13:32
https://matheusfacure.github.io/python-causality-handbook/09-Non-Compliance-and-LATE.htmlDipping our Toes into a Heterogeneous WorldPreviously, we’ve seen Instrumental Variables through a more traditional lens. IV was seen as some sort of natural experiment we can leverage. In contrast, modern IV practice draws a lot of insight from medical sciences. It partitions the world into 4 kinds of subje..