Posts

DAGs and Control Variables
When analyzing causal relationships, it is very hard to understand which variables to condition the analysis on, i.e. how to “split” the data so that we are comparing apples to apples.
Jul 10, 2023 10 min read
DAGs and Control Variables
Double Debiased Machine Learning (part 1)
Causal inference, machine learning and regularization bias In causal inference, we often estimate causal effects by conditioning the analysis on other variables. We usually refer to these variables as control variables or confounders.
Jul 10, 2023 11 min read
Double Debiased Machine Learning (part 1)
Double Debiased Machine Learning (part 2)
In the previous part of this blog post, we have seen how pre-testing can distort inference, i.e., how selecting control variables depending on their statistical significance results in wrong confidence intervals for the variable of interest.
Jul 10, 2023 14 min read
Double Debiased Machine Learning (part 2)
Experiments on Returns on Investment
An introduction to the delta method for inference on ratio metrics. When we run an experiment, we are often not only interested in the effect of a treatment (new product, new feature, new interface, …) on revenue, but in it’s cost-effectiveness.
Jul 10, 2023 15 min read
Experiments on Returns on Investment
Experiments, Peeking, and Optimal Stopping
In the decade preceding the Second World War, there was a massive increase in industrial production of war materials, so there was a need to ensure that products, especially munitions, were reliable.
Jul 10, 2023 17 min read
Experiments, Peeking, and Optimal Stopping
From Causal Trees to Forests
How to use regression trees to do policy targeting. In my previous blog post, we have seen how to use causal trees to estimate heterogeneous treatment effects of a policy. If you haven’t read it, I recommend starting there first, since we are going to take the content of that article for granted and start from there.
Jul 10, 2023 16 min read
From Causal Trees to Forests
Goodbye Scatterplot, Welcome Binned Scatterplot
When we want to visualize the relationship between two continuous variables, the go-to plot is the scatterplot. It’s a very intuitive visualization tool that allows us to directly look at the data.
Jul 10, 2023 10 min read
Goodbye Scatterplot, Welcome Binned Scatterplot
How to Compare Two or More Distributions
A complete guide to comparing distributions, from visualization to statistical tests Comparing the empirical distribution of a variable across different groups is a common problem in data science. In particular, in causal inference the problem often arises when we have to assess the quality of randomization.
Jul 10, 2023 19 min read
How to Compare Two or More Distributions
Mean vs Median Causal Effect
An introduction to quantile regression. In A/B tests, a.k.a. randomized controlled trials, we usually estimate the average treatment effect (ATE): effect of a treatment (a drug, ad, product, …) on an outcome of interest (a disease, firm revenue, customer satisfaction, …), where the “average” is taken over the test subjects (patients, users, customers, …).
Jul 10, 2023 13 min read
Mean vs Median Causal Effect
Omitted Variable Bias And How To Deal With It
In causal inference, bias is extremely problematic because it makes inference not valid. Bias generally means that an estimator will not deliver the estimate of the true effect, on average.
Jul 10, 2023 11 min read
Omitted Variable Bias And How To Deal With It