Survey experiments have become increasingly popular in political science research.
The Problem: Standard between-subjects designs measure outcomes only once after treatment, requiring large samples for precise estimates.
The Solution: Repeated measures designs collect outcome data multiple times.
But a persistent concern exists: won't gathering more data change results due to consistency pressures?
Our Approach: We tested six different experimental designs directly against each other across multiple studies.
What We Found: Contrary to expectations, repeated measures did not distort findings but significantly boosted precision without sacrificing validity. These methods also provide richer insights into treatment effects and heterogeneity.
The Takeaway: Researchers should consider adopting repeated measures in their survey experiments for more efficient and informative political science research.
