Scientist, Physician, Epi/Biostat, Causal Inference, Epistemology. #epitwitter @tfeend.bsky.social

Research Triangle, NC
Joined April 2009
✨NEW ✨ BMJ Research Methods & Reporting article on reading and conducting instrumental variable studies. Thanks to Eleanor Sanderson, @MGLevin, @damrauer, @Tfeend, and @nm_davies. doi.org/10.1136/bmj-2023-078…
1
6
16
840
Tim Feeney retweeted
I’ve had this for ten years. Nothing changes.
508
25,181
331
238,144
5,476,690
Trying to interpret a dataset with no documentation.
1
28
1
445
18,686
Fun fact: Instead of saying "factorial" you can scream the answer and it means the same thing. Example: 26! = "26 factorial" = "TWENTY SIX"
13
47
6
457
18,737
Tim Feeney retweeted
Progress on the Chatbot Assessment Reporting Tool - #CHART! Panel Consensus 2/3 co-chaired with @GordonGuyatt Pictured: our global, multi-d expert panel - apologies to those missed #MedTwitter #AI @juliomayol @Tfeend @eloder @AFlanagin @riazagha @sa_antoniou @GSCollins @nliulab
1
14
1
30
6,249
“This assumption is probably not necessary.”
1
7
121
12,150
Great to hear about causal inference methods to estimate per-protocol effects in randomised trials from @Tfeend Traditional methods used in trials cause bias but methods of observational research can permit causal interpretations #SER2024 @societyforepi
4
10
846
Here's a thought. Instead of cutting libraries... Let's invest in them. #NoCutsToLibraries Sign your name today: on.nypl.org/4dqV8xZ
2
58
2
159
9,271
Instead of just sharing equations, I truly believe that code sharing enhances transparency and reproducibility in our field. 🔄 #OpenScience Equations are compact, but we can do more. and if the code works, sometimes written equations don't. 2/3
2
2
3
1,236
The US just keeps getting worse every single day, huh? There’s just no bottom to this thing. They’ll turn ultra-violent, ultra-fascist militarised police and sniper rifles loose on their own students for daring to protest against Obviously Bad Things, and the media will find some way to not even mention the Obviously Bad Things *AT ALL* and instead centre those who egged this on and who support the Obviously Bad Things as being some oppressed little group. The collective delusion is phenomenal to observe. There are so many people involved in upholding this giant lie — that it is the protestors that are the real evil here, and not Israel’s illegal, unjust and unprecedented campaign of violence that they’re protesting — that this is a level of gaslighting I don’t think I’ve ever seen before. I always said Huxley had the future nailed more than Orwell, but we’re into levels of doublespeak that would make Big Brother recoil. The American Project has failed, and it’s failed so badly it’s destroying half the world with it.
244
2,186
78
7,395
210,201
A paid statistical reviewer job is available at BMJ Open statsjobs.com/job/freelance-… I think all journals should do this. Having confidence that the stats are correct in a paper then allows for proper discussion of the merits of whatever findings it is reporting.
7
8
1,030
Stand against the right-wing assault on diversity! The New York Times reveals an effort to dismantle inclusivity at UNC. Join our coalition to defend free speech and protect the values that make Carolina great. Together, we can make a difference. #FreeUNC shorturl.at/qrUV4
1
4
564
what starts here changes the world
323
10,048
995
38,658
6,394,782
Tim Feeney retweeted
Ars Technica
@arstechnica19 Apr 2024
DARPA’s AI test pilot successfully flew a dogfight against a human trib.al/l4UyzHx
5
26
5
49
15,059
NEW PAPER out today in @BMJ_latest TRIPOD+AI: reporting recommendations for studies developing or validating prediction models for use in healthcare that use #machinelearning methods bmj.com/content/385/bmj-2023… #ArtificialIntelligence #AIstandards #OpenAccess Please share 🙏
10
221
38
439
148,725
I've been making a Shiny app to show how Hamiltonian Monte Carlo works. It starts to look nice and I'm quite happy about it! #shiny #bayes #hmc #mcmc
11
21
1
148
10,951
#OTD 1994 the first of BMJ’s ‘Statistics Notes’ series (‘Correlation, regression and repeated data’) was published. In 1993 BMJ asked Doug Altman to write a few 600-word “filler” pieces on statistics education. 1/2
2
15
1
47
5,734
They were just trying to feed people. Both the IOF and Israeli civilians are blocking aid from entering Gaza and these workers were slaughtered for trying to feed starving families.
🚨Zomi & Chef Oli's last video shows them cooking meals for Gaza's hungry in @WCKitchen. They were killed today by Israeli warplanes in #Gaza , in disregard of international laws, echoing the daily killings of Palestinian civilians in Gaza over the past 177 days. #GazaGenocid
535
5
1,211
61,137
Tim Feeney retweeted
Would Boeing murder a whistleblower? Before this month I'd have said it depends. I didn't think Boeing would kill a middle manager who never worked on the 737 Max & left in 2017. What surprised me was how many more informed observers felt Boeing...might prospect.org/infrastructure/…
89
2,520
274
15,285
2,795,146
128
2,653
253
6,293
709,505