Randomization inference on cell effects under absorbing treatment onset

Abstract

In multiple-baseline experiments, staggered adoption designs, and stepped-wedge trials, every unit switches once from baseline to treatment at a randomized onset time and remains treated afterward. We label these designs as absorbing onset. Because such experiments often involve only a few heterogeneous units, randomization tests of the sharp null of no effect on any unit at any period are common. Rejecting this null, however, implies that some effect exists somewhere, not that it is large, positive for most unit-periods, or persistent. We study what randomization tests can establish for bounded nulls and quantile nulls on unit-period cell effects in absorbing onset designs. Under the assumption that a cell's potential outcome depends on its current treatment status and not on when treatment began, we establish finite-sample validity for both tests and develop the power theory for the bounded-null tests with a fixed number of units $N$ and a growing series length $T$ that admits $K_T$ onsets. The $p$-value against a fixed alternative decays as $K_T^{-N}$, not in $T$, with matching lower bounds, so the onset window, not the series length, determines the power. We also treat cases in which this assumption fails partially or completely, which motivates a sensitivity analysis.