We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Methodology

Title: The magnitude and direction of collider bias for binary variables

Abstract: Suppose we are interested in the effect of variable $X$ on variable $Y$. If $X$ and $Y$ both influence, or are associated with variables that influence, a common outcome, called a collider, then conditioning on the collider (or on a variable influenced by the collider -- its "child") induces a spurious association between $X$ and $Y$, which is known as collider bias. Characterizing the magnitude and direction of collider bias is crucial for understanding the implications of selection bias and for adjudicating decisions about whether to control for variables that are known to be associated with both exposure and outcome but could be either confounders or colliders. Considering a class of situations where all variables are binary, and where $X$ and $Y$ either are, or are respectively influenced by, two marginally independent causes of a collider, we derive collider bias that results from (i) conditioning on specific levels of, or (ii) linear regression adjustment for, the collider (or its child). We also derive simple conditions that determine the sign of such bias.
Subjects: Methodology (stat.ME); Statistics Theory (math.ST)
Journal reference: Epidemiologic Methods. 2019. Vol 8, Issue 1
DOI: 10.1515/em-2017-0013
Cite as: arXiv:1609.00606 [stat.ME]
  (or arXiv:1609.00606v3 [stat.ME] for this version)

Submission history

From: Trang Nguyen [view email]
[v1] Fri, 2 Sep 2016 14:09:48 GMT (923kb,D)
[v2] Fri, 28 Apr 2017 23:10:27 GMT (1255kb,D)
[v3] Mon, 14 Jan 2019 20:01:30 GMT (1263kb,D)

Link back to: arXiv, form interface, contact.