ORがcollapsibleなときもあるのか(本筋ではないけど) On the collapsibility of measures of effect in the counterfactual causal framework #iron勉強メモ https://t.co/0ILUL8ZZBs
@psotob91 @arthur_alb1 2) It is possible to compute a weighted average of the conditional RRs to arrive at the marginal RR. But the weights depend on unseen potential risks, so the marginal RR cannot be computed from them in practice. See https://t.co/8wuU
RT @MHitchingsEpi: @EpiEllie oo i actually went and followed the proof in Anders's paper and it comes out pretty quickly that it is collaps…
RT @EpiEllie: And the winner of the day, @MHitchingsEpi, coming through with a confirmed proof that Vaccine Efficacy is collapsible. (Sta…
And the winner of the day, @MHitchingsEpi, coming through with a confirmed proof that Vaccine Efficacy is collapsible. (Stay tuned for a thread on why this is a helpful thing to know & how it can help us make sense of how well the COVID vaccines migh
@EpiEllie oo i actually went and followed the proof in Anders's paper and it comes out pretty quickly that it is collapsible (when VE is estimated by 1-RR). I assume from a case-control study it wouldn't be. ok time for bed https://t.co/yoDx3xO8ED
RT @AndersHuitfeldt: (1/2) Our paper on collapsibility is now available at https://t.co/rTerzN5KD4 . The results in this paper may be quite…
@yudapearl @ESteyerberg @tmorris_mrc @Crowther_MJ @f2harrell You can define it based on purely statistical notions, or based on causal notions (eg do(x)). See for example definition 2 in our paper at https://t.co/szhaaRSDPB . I argue that you get simpler,
@omaclaren @stephensenn @statsepi @ashtroid22 @Epi_D_Nique @healthstatsdude Also: Let me point out that if you construct the group level risk ratio from individual level risk ratios, I believe the weights assigned to anyone with Y(a=0) =0 is 0. See our col
RT @AndersHuitfeldt: (1/2) Our paper on collapsibility is now available at https://t.co/rTerzN5KD4 . The results in this paper may be quite…
On the collapsibility of measures of effect in the counterfactual causal framework https://t.co/fUML9kokbB
On the collapsibility of measures of effect in the counterfactual causal framework https://t.co/oqHTTjAmEC
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @EpiEllie: Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This…
RT @AndersHuitfeldt: (1/2) Our paper on collapsibility is now available at https://t.co/rTerzN5KD4 . The results in this paper may be quite…
RT @AndersHuitfeldt: (1/2) Our paper on collapsibility is now available at https://t.co/rTerzN5KD4 . The results in this paper may be quite…
Have you heard #statstwitter going on about the non-collapsibility of odds ratios and wondered what the heck they mean? This nice paper explains what that means, why it matters, and why the odds ratio isn’t the only measure you should worry about!👇🏼👇🏼
RT @AndersHuitfeldt: (1/2) Our paper on collapsibility is now available at https://t.co/rTerzN5KD4 . The results in this paper may be quite…
RT @AndersHuitfeldt: (1/2) Our paper on collapsibility is now available at https://t.co/rTerzN5KD4 . The results in this paper may be quite…
Retweeting for #AbbeysEpiReadingList
Very interesting!
RT @AndersHuitfeldt: (1/2) Our paper on collapsibility is now available at https://t.co/rTerzN5KD4 . The results in this paper may be quite…
RT @AndersHuitfeldt: (1/2) Our paper on collapsibility is now available at https://t.co/rTerzN5KD4 . The results in this paper may be quite…
RT @AndersHuitfeldt: (1/2) Our paper on collapsibility is now available at https://t.co/rTerzN5KD4 . The results in this paper may be quite…
(1/2) Our paper on collapsibility is now available at https://t.co/rTerzN5KD4 . The results in this paper may be quite obvious to methodologists, but it clarifies the difference between different types of collapsibility which have different properties.