Data Analysis and Causal Inference · Strata Research & Advisory

Overview

Not every question is answered by knowing what happened, sometimes the real question is what caused it. We help partners get more value from data they already have and, where the question calls for it, apply experimental and quasi-experimental methods to separate real impact from noise.

What we offer

Where this practice area applies.

Randomized controlled trials (RCTs) — randomly assigning who receives a program to identify its causal effect
Difference-in-differences and matching methods — comparing similar groups over time to isolate a program's real effect
Regression discontinuity design — comparing people just above and below a cutoff, like an eligibility threshold
Statistical modeling and regression analysis — finding and quantifying the relationships hiding in your data
Data cleaning, merging, and management
Index construction and scale development — combining multiple measures into a single, reliable score
Data visualization and dashboard design
Secondary and administrative data analysis — getting more value out of data you already have, like government or program records