Statistics, experimentation, causal reasoning · insight-1019

Describe a regression discontinuity effect as local to the cutoff and do not extrapolate it to the whole base

Situation
The estimate says a status upgrade raises rebooking by 4 points for members near the threshold, and the sponsor wants to upgrade all 200,000 members.
Action
State the population as the window around the cutoff and the estimand as the effect there. Look for other cutoffs, past rule changes or a small pilot that test elsewhere; treat any extrapolation as an extra assumption backed by its own evidence.
Reason
The design compares units on both sides of one threshold, so it speaks only to the units whose score sits close to it.
Boundary
Threshold-based policies and eligibility rules.

Symptom or risk

Members near the threshold are not typical, and the effect for far-below members could be smaller, larger or reversed.

Settings where effects are known to be homogeneous across the score, which needs outside evidence.

A narrow claim is less exciting but defensible.

Quoting the local estimate in a business case sized on the full base.