PreprintarXiv licencearXiv
A LATE-Targeting Experimental Design for Multi-Arm RCTs
Abstract
We study the design of multi-arm experiments that identify local average treatment effects under imperfect compliance. With three or more arms, conventional randomization fails to identify margin-specific effects even under strong extensions of monotone compliance. A choice-theoretic analysis unifies these extensions and motivates a two-stage design: subjects report a preferred arm and are randomized conditional on it. If treatment is confined to the assigned or reported arm, one-sided noncompliance is restored conditional on the report, identifying the effect between the two for their compliers. We discuss the design's applicability and feasibility in relation to the choice model.
Related
- Experimental investigation and multi-response design of FFF-printed PLA honeycomb sandwich panels for flexural performance and mass efficiency
- BE-BOIN: A Dose Optimization Design Accommodating Backfill and Late-Onset Toxicity
- Synthetic Controls for Experimental Design
- FairBED: Incorporating Fairness into Bayesian Experimental Design
- Computationally Efficient Experimental Design with Generalized Posteriors
BibTeX
@misc{kaji2026latetargeting,
title = {{A LATE-Targeting Experimental Design for Multi-Arm RCTs}},
author = {Tetsuya Kaji},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.25667},
eprint = {2610.25667},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.25667},
}
RIS
TY - UNPB
TI - A LATE-Targeting Experimental Design for Multi-Arm RCTs
AU - Tetsuya Kaji
PY - 2026
DA - 2026/10/20
PB - arXiv
DO - 10.48550/arxiv.2610.25667
UR - https://arxiv.org/abs/2610.25667
AB - We study the design of multi-arm experiments that identify local average treatment effects under imperfect compliance. With three or more arms, conventional randomization fails to identify margin-specific effects even under strong extensions of monotone compliance. A choice-theoretic analysis unifies these extensions and motivates a two-stage design: subjects report a preferred arm and are randomized conditional on it. If treatment is confined to the assigned or reported arm, one-sided noncompliance is restored conditional on the report, identifying the effect between the two for their compliers. We discuss the design's applicability and feasibility in relation to the choice model.
ER -