Announcement

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

  • How to test The Parallel Trends Assumption when there is a staggered intervention

    Hi. I have 48 weeks of week-level data in an unbalanced panel dataset which tracks the effect of a telemedicine intervention on Covid health outcomes across different states. The telemedicine has a staggered/phased rollout across states starting from week 4 of the data. Not all states were treated, and hence, I have a treatment group and a control group. My goal is to look at the effect of the intervention on the health outcomes through an event-study design, and I need to verify the parallel trends assumption.

    Unfortunately, I am struggling to find the right code to run to test the parallel trends assumption with a staggered intervention rollout. Could any of you please give me a hand? I have a treatment variable which turns on (==1) in different weeks across the treated states based on the week in which the treatment started.

    Any help would be GREATLY appreciated.

  • #2
    Scott:
    see (with related references)
    Staggered Difference-in-Difference Estimation - Tilburg Science Hub
    Verifying Parallel Trends in DiD Analysis - Statalist
    PowerPoint Presentation
    Kind regards,
    Carlo
    (StataNow 18.5)

    Comment


    • #3
      a lot of research relies on the pseudo-treatment

      Comment

      Working...
      X