Susie Ellul

11 posts

Susie Ellul banner
Susie Ellul

Susie Ellul

@SusieEllul

Biostatistics @MCRI_for_kids

Katılım Aralık 2019
47 Takip Edilen26 Takipçiler
Susie Ellul retweetledi
Margarita Moreno-Betancur
Margarita Moreno-Betancur@_MargaritaMB·
NEW PREPRINT! With skewed outcomes, the causal difference in medians might be the most meaningful estimand. Ever wonder how to estimate it via IPW or g-computation? Read👉 arxiv.org/abs/2207.05940 Led by @_DaisyShepherd @CEBU_Melbourne @vicbiostat #causaltwitter #epitwitter
Daisy Shepherd@_DaisyShepherd

PREPRINT: Confounding-adjustment methods for the difference in medians. arxiv.org/abs/2207.05940 An updated version is now available with the inclusion of some lesser-known g-computation implementations. @_MargaritaMB @baersighting @CEBU_Melbourne

English
2
12
35
5K
Susie Ellul retweetledi
Clinical Epi & Biostatistics Unit (CEBU)
Clinical Epi & Biostatistics Unit (CEBU)@CEBU_Melbourne·
🚨To stop REGRESSION ABUSE, @CEBU_Melbourne runs a course twice-yearly covering how to distinguish between the 3 types of questions and use regression appropriately for #causalinference. Register here for the virtual offering in Oct 2022: tinyurl.com/y4jn4dyc
Maarten van Smeden@MaartenvSmeden

Keynote by John Carlin at #ISCB43 that really resonates on the abuse of regression models, with great examples from his own work now recognized to be Table 2 fallacy

English
1
7
10
0
Susie Ellul
Susie Ellul@SusieEllul·
Great, very clear presentation on MI in case-cohort studies when we are looking at binary outcomes 👍Important practical implications discussed by @MK_Middleton Look forward to seeing a preprint soon :) #ISCB43
Susie Ellul tweet media
English
1
0
17
0
Jiaxin Zhang
Jiaxin Zhang@JiaxinZhang96·
@SusieEllul Thank you so much Susie! I'm surprised that you were awake at that time 🥰
English
1
0
2
0
Susie Ellul
Susie Ellul@SusieEllul·
Excellent talk by @JiaxinZhang96 on recoverability of casual effects under multivariate missing mechanisms 👏 Missing out in person and resting a very sore fractured ankle but so glad to be able to join in the fun remotely! #ISCB43
Susie Ellul tweet media
English
1
0
10
0
Susie Ellul retweetledi
Ollie Hines
Ollie Hines@hines8·
Nice to see my paper with @SVansteelandt @karlado on mediation analysis published. tldr: Non-parametric inference is nice, but here a semi-parametric model means we can avoid density specification and still be robust rdcu.be/ckYEi
English
0
5
13
0

Keşfet

@SVansteelandt @_MargaritaMB @CEBU_Melbourne @_DaisyShepherd @rush_099 @JiaxinZhang96 @vicbiostat @MK_Middleton