Proxy variables and the generalizability of study results
(2023) In American Journal of Epidemiology 192(3). p.448-454- Abstract
- When individuals self-select (or are selected) into a study based on factors that influence the outcome, conclusions may not generalize to the full population. To compensate for this, results may be adjusted, for example, by standardization on the set of common causes of participation and outcome. Although such standardization is useful in some contexts, the common causes of participation and outcome may in practice not be fully observed. Instead, the researcher may have access to one or several variables related to the common causes, that is, to proxies for the common causes. This article defines and examines different types of proxy variables and shows how these can be used to obtain generalizable study results. First of all, the... (More)
- When individuals self-select (or are selected) into a study based on factors that influence the outcome, conclusions may not generalize to the full population. To compensate for this, results may be adjusted, for example, by standardization on the set of common causes of participation and outcome. Although such standardization is useful in some contexts, the common causes of participation and outcome may in practice not be fully observed. Instead, the researcher may have access to one or several variables related to the common causes, that is, to proxies for the common causes. This article defines and examines different types of proxy variables and shows how these can be used to obtain generalizable study results. First of all, the researcher may exploit proxies that influence only participation or outcome but which still allow for perfect generalizability by rendering participation and outcome conditionally independent. Further, generalizability can be achieved by leveraging 2 proxies, one of which is allowed to influence participation and one of which is allowed to influence the outcome, even if participation and outcome do not become independent conditional on these. Finally, approximate generalizability may be obtained by exploiting a single proxy that does not itself influence participation or outcome. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/0d2a5716-cedc-4553-b60b-59c0f5da2330
- author
- Nilsson, Anton LU ; Björk, Jonas LU and Bonander, Carl
- organization
- publishing date
- 2023
- type
- Contribution to journal
- publication status
- published
- subject
- in
- American Journal of Epidemiology
- volume
- 192
- issue
- 3
- pages
- 448 - 454
- publisher
- Oxford University Press
- external identifiers
-
- pmid:36352507
- scopus:85149999695
- ISSN
- 0002-9262
- DOI
- 10.1093/aje/kwac200
- language
- English
- LU publication?
- yes
- id
- 0d2a5716-cedc-4553-b60b-59c0f5da2330
- alternative location
- https://academic.oup.com/aje/article/192/3/448/6811810?login=true
- date added to LUP
- 2023-03-20 12:42:56
- date last changed
- 2024-02-14 15:04:26
@article{0d2a5716-cedc-4553-b60b-59c0f5da2330,
abstract = {{When individuals self-select (or are selected) into a study based on factors that influence the outcome, conclusions may not generalize to the full population. To compensate for this, results may be adjusted, for example, by standardization on the set of common causes of participation and outcome. Although such standardization is useful in some contexts, the common causes of participation and outcome may in practice not be fully observed. Instead, the researcher may have access to one or several variables related to the common causes, that is, to proxies for the common causes. This article defines and examines different types of proxy variables and shows how these can be used to obtain generalizable study results. First of all, the researcher may exploit proxies that influence only participation or outcome but which still allow for perfect generalizability by rendering participation and outcome conditionally independent. Further, generalizability can be achieved by leveraging 2 proxies, one of which is allowed to influence participation and one of which is allowed to influence the outcome, even if participation and outcome do not become independent conditional on these. Finally, approximate generalizability may be obtained by exploiting a single proxy that does not itself influence participation or outcome.}},
author = {{Nilsson, Anton and Björk, Jonas and Bonander, Carl}},
issn = {{0002-9262}},
language = {{eng}},
number = {{3}},
pages = {{448--454}},
publisher = {{Oxford University Press}},
series = {{American Journal of Epidemiology}},
title = {{Proxy variables and the generalizability of study results}},
url = {{http://dx.doi.org/10.1093/aje/kwac200}},
doi = {{10.1093/aje/kwac200}},
volume = {{192}},
year = {{2023}},
}