Robust estimation of encouragement design intervention effects transported across sites

Kara Rudolph, Mark J. van der Laan

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

We develop robust targeted maximum likelihood estimators (TMLEs) for transporting intervention effects from one population to another. Specifically, we develop TMLEs for three transported estimands: the intent-to-treat average treatment effect (ATE) and complier ATE, which are relevant for encouragement design interventions and instrumental variable analyses, and the ATE of the exposure on the outcome, which is applicable to any randomized or observational study. We demonstrate finite sample performance of these TMLEs by using simulation, including in the presence of practical violations of the positivity assumption. We then apply these methods to the ‘Moving to opportunity’ trial: a multisite, encouragement design intervention in which families in public housing were randomized to receive housing vouchers and logistical support to move to low poverty neighbourhoods. This application sheds light on whether effect differences across sites can be explained by differences in population composition.

Original languageEnglish (US)
Pages (from-to)1509-1525
Number of pages17
JournalJournal of the Royal Statistical Society. Series B: Statistical Methodology
Volume79
Issue number5
DOIs
StatePublished - Nov 1 2017
Externally publishedYes

Fingerprint

Average Treatment Effect
Robust Estimation
Maximum Likelihood Estimator
Poverty
Observational Study
Instrumental Variables
Robust Estimators
Positivity
Demonstrate
Design
Maximum likelihood estimator
Robust estimation
Average treatment effect
Simulation

Keywords

  • Causal inference
  • External validity
  • Instrumental variables
  • Policy intervention
  • Targeted maximum likelihood estimation
  • Transportability

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Robust estimation of encouragement design intervention effects transported across sites. / Rudolph, Kara; van der Laan, Mark J.

In: Journal of the Royal Statistical Society. Series B: Statistical Methodology, Vol. 79, No. 5, 01.11.2017, p. 1509-1525.

Research output: Contribution to journalArticle

@article{096ffeaf83484377ae9c67f1bb8f92c2,
title = "Robust estimation of encouragement design intervention effects transported across sites",
abstract = "We develop robust targeted maximum likelihood estimators (TMLEs) for transporting intervention effects from one population to another. Specifically, we develop TMLEs for three transported estimands: the intent-to-treat average treatment effect (ATE) and complier ATE, which are relevant for encouragement design interventions and instrumental variable analyses, and the ATE of the exposure on the outcome, which is applicable to any randomized or observational study. We demonstrate finite sample performance of these TMLEs by using simulation, including in the presence of practical violations of the positivity assumption. We then apply these methods to the ‘Moving to opportunity’ trial: a multisite, encouragement design intervention in which families in public housing were randomized to receive housing vouchers and logistical support to move to low poverty neighbourhoods. This application sheds light on whether effect differences across sites can be explained by differences in population composition.",
keywords = "Causal inference, External validity, Instrumental variables, Policy intervention, Targeted maximum likelihood estimation, Transportability",
author = "Kara Rudolph and {van der Laan}, {Mark J.}",
year = "2017",
month = "11",
day = "1",
doi = "10.1111/rssb.12213",
language = "English (US)",
volume = "79",
pages = "1509--1525",
journal = "Journal of the Royal Statistical Society. Series B: Statistical Methodology",
issn = "1369-7412",
publisher = "Wiley-Blackwell",
number = "5",

}

TY - JOUR

T1 - Robust estimation of encouragement design intervention effects transported across sites

AU - Rudolph, Kara

AU - van der Laan, Mark J.

PY - 2017/11/1

Y1 - 2017/11/1

N2 - We develop robust targeted maximum likelihood estimators (TMLEs) for transporting intervention effects from one population to another. Specifically, we develop TMLEs for three transported estimands: the intent-to-treat average treatment effect (ATE) and complier ATE, which are relevant for encouragement design interventions and instrumental variable analyses, and the ATE of the exposure on the outcome, which is applicable to any randomized or observational study. We demonstrate finite sample performance of these TMLEs by using simulation, including in the presence of practical violations of the positivity assumption. We then apply these methods to the ‘Moving to opportunity’ trial: a multisite, encouragement design intervention in which families in public housing were randomized to receive housing vouchers and logistical support to move to low poverty neighbourhoods. This application sheds light on whether effect differences across sites can be explained by differences in population composition.

AB - We develop robust targeted maximum likelihood estimators (TMLEs) for transporting intervention effects from one population to another. Specifically, we develop TMLEs for three transported estimands: the intent-to-treat average treatment effect (ATE) and complier ATE, which are relevant for encouragement design interventions and instrumental variable analyses, and the ATE of the exposure on the outcome, which is applicable to any randomized or observational study. We demonstrate finite sample performance of these TMLEs by using simulation, including in the presence of practical violations of the positivity assumption. We then apply these methods to the ‘Moving to opportunity’ trial: a multisite, encouragement design intervention in which families in public housing were randomized to receive housing vouchers and logistical support to move to low poverty neighbourhoods. This application sheds light on whether effect differences across sites can be explained by differences in population composition.

KW - Causal inference

KW - External validity

KW - Instrumental variables

KW - Policy intervention

KW - Targeted maximum likelihood estimation

KW - Transportability

UR - http://www.scopus.com/inward/record.url?scp=85000501907&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85000501907&partnerID=8YFLogxK

U2 - 10.1111/rssb.12213

DO - 10.1111/rssb.12213

M3 - Article

AN - SCOPUS:85000501907

VL - 79

SP - 1509

EP - 1525

JO - Journal of the Royal Statistical Society. Series B: Statistical Methodology

JF - Journal of the Royal Statistical Society. Series B: Statistical Methodology

SN - 1369-7412

IS - 5

ER -