@vk_wilde You can’t introduce collider bias by adjusting for a baseline (pre-randomisation ) variable, so we’re good there. If anyone reading is interested in covariate adjustment: https://t.co/G3owwTcx5g
RT @andersgranholm: 3) If a baseline covariate selected based on clinical knowledge turns out to not be predictive of the outcome, this has…
3) If a baseline covariate selected based on clinical knowledge turns out to not be predictive of the outcome, this has limited negative effects and risk are outweighed by potential gains (see https://t.co/13R6QLyztV by @Brennan_Kahan @tmorris_mrc et al)
RT @jmwooldridge: If the mean function (logistic or exponential) is correctly specified and separate means are estimated for treatment and…
RT @jmwooldridge: If the mean function (logistic or exponential) is correctly specified and separate means are estimated for treatment and…
RT @jmwooldridge: If the mean function (logistic or exponential) is correctly specified and separate means are estimated for treatment and…
RT @jmwooldridge: If the mean function (logistic or exponential) is correctly specified and separate means are estimated for treatment and…
RT @jmwooldridge: If the mean function (logistic or exponential) is correctly specified and separate means are estimated for treatment and…
RT @jmwooldridge: If the mean function (logistic or exponential) is correctly specified and separate means are estimated for treatment and…
RT @jmwooldridge: If the mean function (logistic or exponential) is correctly specified and separate means are estimated for treatment and…
RT @jmwooldridge: If the mean function (logistic or exponential) is correctly specified and separate means are estimated for treatment and…
RT @jmwooldridge: If the mean function (logistic or exponential) is correctly specified and separate means are estimated for treatment and…
If the mean function (logistic or exponential) is correctly specified and separate means are estimated for treatment and control, logit (binary or fractional) and Poisson (weakly) improve over SDM for the ATE. They are always consistent even if the mean is
@SolomonKurz Interested to hear the answer. In the context of logistic regression, covariate adjustment leads to increased power (as in Type I error) can increase but can lead to larger SEs as well. https://t.co/CAsqqDRZLb Could the same happen when using
@f2harrell @TheSGEM @SZehtabchi @drjohnm @VPrasadMDMPH @First10EM @thenntgroup @ProfBEEM @BEEMcme @SAEMEBM Big benefits for power with covariate adjustment in RCTs as well. That is an excellent point Frank, while a lot of unadjusted analyses are being done
@dr_michaelmarks @ADAlthousePhD @DrToddLee @gushamilton For 'but-wait-no-confounding', I think it's useful to see what's on offer in trials, e.g. this paper Tom linked 2/ https://t.co/0RZ9qyGki8
Reduciendo sesgos con analisis ajustados
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @tmorris_mrc: I’ve worked on covariate adjustment for a long time with @Brennan_Kahan. The papers that have received most attention are…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
On point
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @vipuljairath: Thank you for highlighting our paper on this topic @Brennan_Kahan @tmorris_mrc
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
Thank you for highlighting our paper on this topic @Brennan_Kahan @tmorris_mrc
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
Say twice ❤️
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
RT @GuyattGH: It’s standard knowledge that adjusted analysis is crucial in observational studies, but many are unaware that covariate adjus…
@doc_BLocke @arthur_alb1 @BenjaminAbella @kurzmc @sarahmperman Not a statistician, and this is a secondary analysis so perhaps not squarely applicable, but it’s common and generally good practice to adjust for baseline covariates w prognostic implications
RT @tmorris_mrc: I’ve worked on covariate adjustment for a long time with @Brennan_Kahan. The papers that have received most attention are…
RT @tmorris_mrc: I’ve worked on covariate adjustment for a long time with @Brennan_Kahan. The papers that have received most attention are…
RT @tmorris_mrc: I’ve worked on covariate adjustment for a long time with @Brennan_Kahan. The papers that have received most attention are…
I’ve worked on covariate adjustment for a long time with @Brennan_Kahan. The papers that have received most attention are on 1. the power gains available https://t.co/bsrRQno4fp 2. why you sometimes need to do it in trials https://t.co/4Fb4WBrOs7 3/
RT @TayaCollyer: 👉Adjustment for known prognostic covariates led to large increases in power for most outcomes 👉Power decreases on adjust…
RT @kaz_yos: The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies | Trials | Ful…
RT @kaz_yos: The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies | Trials | Ful…
RT @kaz_yos: The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies | Trials | Ful…
RT @kaz_yos: The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies | Trials | Ful…
RT @TayaCollyer: 👉Adjustment for known prognostic covariates led to large increases in power for most outcomes 👉Power decreases on adjust…
RT @kaz_yos: The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies | Trials | Ful…
RT @TayaCollyer: 👉Adjustment for known prognostic covariates led to large increases in power for most outcomes 👉Power decreases on adjust…
👉Adjustment for known prognostic covariates led to large increases in power for most outcomes 👉Power decreases on adjustment for non-prognostic covariates. I look forwarding to deploying this reference in future arguments about this topic 😅 #statstw
RT @kaz_yos: The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies | Trials | Ful…
RT @kaz_yos: The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies | Trials | Ful…
The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies | Trials | Full Text https://t.co/363AP1aezb
RT @felbalazard: @mblum_g @YODAProject It is the same mechanism in clinical trials where covariate adjustment leads to increased statistica…
@mblum_g @YODAProject It is the same mechanism in clinical trials where covariate adjustment leads to increased statistical power as shown by @Brennan_Kahan @tmorris_mrc in this great paper https://t.co/OHZ1v8qywa
RT @jd_wilko: @danlewer A bit too complex to cover in a tweet, but a good paper: https://t.co/zLlWMeFmEr
RT @jd_wilko: @danlewer A bit too complex to cover in a tweet, but a good paper: https://t.co/zLlWMeFmEr
RT @jd_wilko: @danlewer A bit too complex to cover in a tweet, but a good paper: https://t.co/zLlWMeFmEr
RT @jd_wilko: @danlewer A bit too complex to cover in a tweet, but a good paper: https://t.co/zLlWMeFmEr
@danlewer A bit too complex to cover in a tweet, but a good paper: https://t.co/zLlWMeFmEr
@C_Mellenthin @NicoleGbel3 In a trial, you adjust for power and precision - nothing to do with confounding. https://t.co/zLlWMeFmEr
RT @jd_wilko: @I_M_Stratton Lots of confusion on this topic for sure. There's a considerable literature on this. Two examples here. Indeed,…
RT @jd_wilko: @I_M_Stratton Lots of confusion on this topic for sure. There's a considerable literature on this. Two examples here. Indeed,…
@I_M_Stratton Lots of confusion on this topic for sure. There's a considerable literature on this. Two examples here. Indeed, if you stratify randomisation, you *must* adjust to get the right answer: https://t.co/zLlWMeFmEr https://t.co/5H08Hk4miV
RT @stroke_methods: ⭐️Adjustment for baseline covariates in the analysis of randomized controlled trials⭐️ ✅ Leads to increased power whe…
⭐️Adjustment for baseline covariates in the analysis of randomized controlled trials⭐️ ✅ Leads to increased power when covariates are prognostic ✅ Covariates need to be prespecified in the protocol More here⬇️ https://t.co/dvJXqOckEv
RT @tmorris_mrc: @mendel_random @ADAlthousePhD @ProfMattFox @TheLancet There are issues about how best to adjust but not about whether you…
@mendel_random @ADAlthousePhD @ProfMattFox @TheLancet There are issues about how best to adjust but not about whether you should bother. IDK if the trials our simulation study used fit your criteria but here is an example of the the power advantages based
@KertViele @DrSamuelBrown @timothygirard I wonder if there's a sweet spot between adjustment and enrichment. For covadjust I always like https://t.co/OHZ1v88X7A
RT @felbalazard: @FrankSchilbach @paulgp @economeager @jmwooldridge @tmorris_mrc @Brennan_Kahan And actually the topic you should care the…
@FrankSchilbach @paulgp @economeager @jmwooldridge @tmorris_mrc @Brennan_Kahan And actually the topic you should care the most about is adjusting your analysis on baseline covariates predictive of your trial's outcome as it can lead to large gains in power
RT @kaz_yos: The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies https://t.co/3…
RT @CollJamie: A nice reference relevant to @MikeLaValley8's talk this morning at the #ACR20 clinical research conference. There is a nice…
A nice reference relevant to @MikeLaValley8's talk this morning at the #ACR20 clinical research conference. There is a nice post here with some more references: TOPIC: Covariate Adjustment in RCT https://t.co/gqQfbVc4c9
The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies https://t.co/363AP1aezb
RT @MaartenvSmeden: @missD_ta There is a lot of work done on this. The TL;DR: conditioning on baseline covariates that are prognostic for t…
@missD_ta There is a lot of work done on this. The TL;DR: conditioning on baseline covariates that are prognostic for the outcome is generally a good thing, e.g. see: https://t.co/LHmKJsuDN7, and other papers written by @Brennan_Kahan @tmorris_mrc and @ESt
RT @Brennan_Kahan: @chrismainey @statstwitbot We look at this a little in our paper here: https://t.co/G59zMp0Jb0 Broadly, covariate adju…
@chrismainey @statstwitbot We look at this a little in our paper here: https://t.co/G59zMp0Jb0 Broadly, covariate adjustment increases power if the covariates are associated with outcome. If not, there will be a negligible loss in power, provided the sam
RT @jd_wilko: Some nice simulations trying to work out what sort of increases in power we might expect to see in practice (@Brennan_Kahan @…
Some nice simulations trying to work out what sort of increases in power we might expect to see in practice (@Brennan_Kahan @tmorris_mrc ): https://t.co/x3Mn1SSnqw
@tmorris_mrc You just work on the stats papers there pretty boy and let me handle the marketing, hmmmkay? https://t.co/zD78dnRRLe