2022
DOI: 10.1214/21-aoas1578
|View full text |Cite
|
Sign up to set email alerts
|

Causal inference for time-varying treatments in latent Markov models: An application to the effects of remittances on poverty dynamics

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1

Citation Types

0
1
0

Year Published

2022
2022
2023
2023

Publication Types

Select...
3
2

Relationship

0
5

Authors

Journals

citations
Cited by 5 publications
(3 citation statements)
references
References 61 publications
0
1
0
Order By: Relevance
“…Following work by Bartolucci et al (2016) and Tullio and Bartolucci (2022), we reformulate the LMM in terms of the potential outcomes framework for defining meaningful estimands in LMMs. Under conditional exchangeability, we define the ATE at 𝑡 as…”
Section: Causal Estimands For Latent Markov Modelsmentioning
confidence: 99%
See 1 more Smart Citation
“…Following work by Bartolucci et al (2016) and Tullio and Bartolucci (2022), we reformulate the LMM in terms of the potential outcomes framework for defining meaningful estimands in LMMs. Under conditional exchangeability, we define the ATE at 𝑡 as…”
Section: Causal Estimands For Latent Markov Modelsmentioning
confidence: 99%
“…Lanza et al (2013) and Clouth et al (2022) and proposed different approaches for using IPW to estimate the causal effect of an exposure on class membership in a cross-sectional LCA. The approach by Lanza and colleagues was extended for longitudinal settings with multiple (time-constant) exposures (Bartolucci et al, 2016) and time-varying exposures (Tullio & Bartolucci, 2022). Recently, Bartolucci et al (2023) proposed a new causal latent transition model inspired by the difference-in-differences approach (Imbens & Wooldridge, 2009).…”
Section: Introductionmentioning
confidence: 99%
“…More specifically, Lanza et al (2013) proposed a one-step method where a latent class model with treatment as the only covariate is estimated on a dataset that has been weighted with the inverse of the propensity score. This approach has been extended for longitudinal settings (Bartolucci et al, 2016;Tullio & Bartolucci, 2019 and for LCA with distal outcomes (Bray et al, 2019;Schuler et al, 2014;Yamaguchi, 2015). Recently, Clouth et al (2022) proposed a three-step approach where the measurement model is estimated on the unweighted data and IPW is introduced in the third step in which the treatment effect is estimated.…”
Section: Introductionmentioning
confidence: 99%