Data science: causal inference

I needed a break from a Rails 4 upgrade so I dove into causal inference.

At a high level, Causal Inference is concerned with measuring the non-explicit relationship between “cause” and “effect” as “cause” variables change. Or, more formally,

Wikipedia https://en.wikipedia.org/wiki/Causal_inference Causal inference

Causal inference is the process of drawing a conclusion about a causal
connection based on the conditions of the occurrence of an effect.
The main difference between causal inference and inference of
association is that the former analyzes the response of the effect
variable when the cause is changed. The science of why things occur
is called etiology.

I found this article to be a helpful intro to the topic. And when I wanted to read deeper on the subject, I really enjoyed skimming this open textbook from the Harvard School of Public Health. When trying to find practical application of the topic, I laughed a little too much reading this article analyzing how the Super Bowl, American Idol, and the holidays postively or negatively effected sentiment to thereby influence how loan officers evaluated credit applications (see page 3 to get a gist for the paper). Lastly, Google created an R package, CasualImpact, for analyzing causal inference.

Happy causal inference Friday!