Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • adjusting for clustering

    Dear Statalist,

    Is this plan legal in statistics?

    logit ipv i.studyarm, vce(cluster studyarm) or

    I want to calculate the odds ratio for occurrence of "ipv" in the studyarm (control, intervention1, intervention2). but the odds ration needs to be adjusted for clustering since allocation of study arm was assigned to clusters. such that control was at site A, intervention1 was at siteB and intervention2 was at siteC.

    Thanks

  • #2
    If I understood right your query, you don’t have clusters here, but 3 (quite different in terms of intervention as well as site) arms. Besides, and still considering these arms are not clusters, but just supposing they were, three clusters would be too few.
    Best regards,

    Marcos

    Comment


    • #3
      Thanks Marcos

      Comment


      • #4
        The reference for clustered standard errors: http://jhr.uwpress.org/content/50/2/317.refs (version without paywall available on the author's webpage)

        Comment

        Working...
        X