#statstwitter — Public Fediverse posts

2017 #StatsTwitter is back baby! 😜

RE: https://bsky.app/profile/did:plc:ynhvprwsjim3ttkefxsttf7y/post/3mirhugt2zs2s

Missing at Random (MAR) means that the probability of data being missing is purely random after conditioning on observed data (e.g., in multiple imputation). That is to say, it depends only on the observed data, not the missing values themselves.

#phdchat #statstwitter

dropping in briefly to say that the blue sky vibes are really taking off rn and feel like a twitter replacement in a way that mastodon does not. hope to see you there! #rstats #statistics #academictwitter #statstwitter #pydata

RT @J_my_sci: What statistical model would you use to detect changes in count (poisson) data over time? Assume each timepoint contains counts of the same thing from different individuals. Also want to detect if there is an interaction between time and genotype group #statstwitter #statistics

Yes, this pic is worth million! Please add to it life itself! #AcademicChatter #AcademicTwitter #academia #EpiVerse #statstwitter #phdlife #PhD #SocialEpidemiology https://x.com/dr_asingh/status/1690013031680339970?s=46&t=b_yQj0hburjs9vkASx7q4A

This sorts of call for the use of mixed methods 🤣🤣 #epitwitter #epiverse #BlackEpiMatters #statstwitter
---
RT @xkcd
K-Means Clustering http://xkcd.com/2731
https://twitter.com/xkcd/status/1620059894966788098

Excellent thread on RCT, logistic regression and transportability. #epitwitter #epiverse #statstwitter #AcademicChatter
---
RT @ildiazm
In this post and multiple others, @f2harrell says that a generally correct way to analyze RCTs with binary outcomes is to fit a logistic regression. One of the main claims is that this leads to "one-number" effect measures that are "highly transportable".
https://twitter.com/f2harrell/status/1625569700326…
https://twitter.com/ildiazm/status/1626233895221092352

This cannot be said enough 🎯🎯 #epitwitter #epiverse #BlackEpiMatters #statstwitter
---
RT @MaartenvSmeden
Just because you put a bunch of covariates in a regression model, doesn't mean all of the covariate-outcome associations are somehow magically freed from confounding
https://twitter.com/MaartenvSmeden/status/1627552214892916736

Are one-sided [vs sharp] tests well-defined for permutation/conformal inference type procedures [e.g. this one for synth https://arxiv.org/abs/1712.09089 by @wuthrich_k and co]? I'm having a hard time thinking of a test statistic.
#statstwitter #econtwitter #xp

do the nuisance functionals in the EIF "completely" characterize an estimand from an identification pov?

i.e. to characterize the ATE i only need to think about the propensity score and the conditional expectation of the outcome. is this general / is there some nice way to express it?

#semiparametrics #eif #causalinference #statistics #statstwitter

Odds ratio vs risk ratio: “Estimating the odds ratio is simple but interpreting it is hard. But estimating the risk ratio is challenging but its interpretation is straightforward.” #epitwitter #statstwitter #AcademicTwitter #phdchat https://academic.oup.com/ije/advance-article/doi/10.1093/ije/dyac220/6843281

Data and code sharing: "Compliance with mandatory data and code sharing policies was observed in 48% and 0% of articles, respectively. but, 88% of articles included data availability statements when required." #epitwitter #AcademicTwitter #statstwitter https://bmcmedicine.biomedcentral.com/articles/10.1186/s12916-022-02644-2

🔥Three recent surveys on #causal #machinelearning for treatment or #PolicyEvaluation:
@[email protected] (2021): https://academic.oup.com/ectj/article-abstract/25/3/602/6596870;
Lieli et al (2022): https://link.springer.com/chapter/10.1007/978-3-031-15149-1_3;
Shah et al (2021): https://www.elgaronline.com/view/edcoll/9781788976473/9781788976473.00025.xml
#EconTwitter #CausalTwitter #EpiTwitter #statstwitter

@jon_y_huang @willball12 @mkellyirving why would we call it #EpiTwitter or #StatsTwitter, why not #Epidemiology, #Epiverse, #Epidon, #EpiToot etc?

Hi everyone. I'm Stephen a South African statistician working on HIV/TB and non-communicable diseases in KwaZulu-Natal. I've also been known to do standup comedy when I'm bored and want to spread that boredom

#Introduction #stats #statstwitter #statsmastodon #epitwitter

Like @paulgp, I am contemplating starting an instance for #statstwitter #rstats and #datascience folks

Goals:

(1) local timeline focused on stats and data sci content to help with discovery, and

(2) hands-on, hopefully community-determined moderation to keep a constructive, troll-free environment

Thoughts? #twittermigration #tidytuesday #mastodonmeta

I've seen a lot of people flaunt about how they don't use p-values and they avoid dichotomous thinking, but then their papers discuss how the 95% CI doesn't overlap with 0. #statstwitter