{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreid7nwvhj6u73phc4oymuoo5maor733eaufkqgvj6qd2qfozk5l3qq",
    "uri": "at://did:plc:wwyqal4cnqhuwyacdj7rqq3n/app.bsky.feed.post/3mmaicocry232"
  },
  "path": "/t/collider-in-rct-subgroup-analysis/28689?page=2#post_23",
  "publishedAt": "2026-05-19T21:12:08.000Z",
  "site": "https://discourse.datamethods.org",
  "textContent": "Well, I roll the rationale into the conclusion as follows:\n\na) if one decides the third variable is prognostic, then it cannot be either post-randomization nor be gating a post-randomization variable\n\nb) a mediator is clearly a post-randomization variable - induced by the treatment (I call this an induced mediator)\n\nc) an effect modifier is NOT a post-randomization variable but clearly gates such a variable (I call this a conditional mediator aka proxy for an induced mediator)\n\nNow all three are prognostic for the outcome so they only differ by one of the above. My point is that each needs a specific analysis and that if a variable either gates a post-randomization variable or is a post-randomization variable then in both cases it is a mediator in some sense. So the first conclusion is that a variable can only be an effect modifier _if it influences at least one mediator in a causally sufficient representation of all pathways from treatment to outcome_. The second conclusion is that the mediator collides on sample selection with the treatment so mediators should not enter the outcome model to avoid bias.",
  "title": "Collider in RCT Subgroup Analysis"
}