Week 12: Bayesian Decision Theory

DSAN 5650: Causal Inference for Computational Social Science
Summer 2026, Georgetown University

Class Sessions
Author
Affiliation

Jeff Jacobs

Published

Wednesday, August 5, 2026

Open slides in new window →

Schedule

Today’s Planned Schedule:

Start End Topic
Lecture 6:30pm 7:00pm Why Should We Use Bayes’ Rule? →
6:45pm 7:10pm Why Should We Think Causally About Decisions? →
7:10pm 8:00pm Heterogeneous Treatment Effects →
Break! 8:00pm 8:10pm
8:10pm 9:00pm Causal Forests →

Why Should We Use Bayes’ Rule At All?

Assume Big Dave has the following beliefs about the likelihood of events \(P\) and \(Q\):

\(p\) \(q\) \(\Pr(P = p, Q = q)\)
0 0 1/4
0 1 1/4
1 0 1/4
1 1 1/4

But decides not to use Bayes’ Rule to update after learning \(Q = 1\): he just decides… he likes 0.6, so \(\Pr(P = p \mid Q = 1) = 0.6\)

The Dutch Ticket

  • If Q is true, this ticket entitles the bearer to $1 if P is true and nothing otherwise.
  • If Q is false, this ticket may be returned to the seller for a full refund of its purchase price.

Why Should We Focus On Causality When Making Our Decisions?

  • Savage’s Rule
  • \(\leadsto\) Jeffrey’s Rule (no relation)
  • \(\leadsto\) Causal Decision Theory!

References