Carryover detection in switchback experimentation

Switchback experiments assign units to treatment and control over time, yielding more precise causal estimates than fixed designs but risking bias from carryover effects where past treatments influence future outcomes. Existing estimators require specifying an influence period, i.e. an upper bound on carryover duration, often guessed from intuition. We propose a statistical test that detects when this bound is underestimated by comparing estimators with different lag assumptions and rejecting results when significant differences indicate bias. The method has theoretical guarantees under standard assumptions and shows strong performance in simulations with synthetic effects on real-world covariates. This provides practitioners with a practical safeguard against erroneous conclusions in the presence of unaddressed carryover.