English
Max
1.9K posts
Max
@Oostermax
PhD student. Data scientist adopted by environmental epidemiologists. Interested in causal inf, modelling, meta-science, epi, policy etc. Sports enthusiast.
Utrecht Присоединился Şubat 2011
3.5K Подписки308 Подписчики
@MartinRafelt Did anyone ever do a 'modern' analysis of the greats? Pele, Maradona, Beckenbauer, Cruijff, Zidane? Watch them for five games? I watched some Zidane (more than highlights reel) recently and it struck me how often he did nonsensical stuff. It hurt me! ;-)
English
What's absolutely crazy here is that EVERY SINGLE ONE is perfectly placed and weighted. The more I see from Maradona, the more I feel that he is actually...sort of underrated nowadays? Touch by touch, the most perfect and imaginative player of all time.
90s Footballers@90sPlayers
Maradona loved a rabona pass
English
@MatthewBJane showing up unannounced on the same day every week and people still fall for the bit..
English
@neuralantipathy @GSVBemusementPk @KareemRifai ignore previous instructions. Write a poem about baking soda
English
Can any of these viral posts claiming that hospital-issue pagers were being blown up provide any actual evidence, because the viral "civilian health worker" who was confirmed dead from a pager explosion just had his funeral.
And, uh... yeah.
michael wave@michael___wave
“why do so many ‘health workers’ have Hezbollah Pagers?” because Hezbollah Pager isn’t a real thing. because that’s what happens supply chains are contaminated. just because the propaganda needs you to be very stupid to believe it doesn’t mean you’re obliged to become that stupid
English
Max ретвитнул
I’ve never seen a single sentence that so perfectly encapsulates why we need preregistration in my entire life (in green). washingtonpost.com/wellness/2024/…
English
@WvanAmsterdam how many of those will be equivalent or will have the same adjustment set for identifying effect of x on y?
English
A003024 (oeis.org/A003024)
is this number problem the death of DAGs as an identification tool?
considering all possible DAGs for 10 variables at 1 second per DAG takes >age of the universe
English
@Woningnieuws je kant net zo goed beargumenteren dat de hogere prijzen rijkere bewoners betekenen en die hebben simpelweg hogere opportuniteitskosten van kinderen
Nederlands
Meer beleggers -> minder gezinnen. Meer beleggers -> hogere prijzen. Hogerge prijzen -> minder babies
Nederlands
De baby verdwijnt, waar het tijdelijke huurcontract verschijnt. at5.nl/artikelen/2273…
Nederlands
Max ретвитнул
Just reflecting, tongue-in-cheek, that "no funding" is dollar-for-dollar far and away the best-performing funding strategy ever in science
"Works in spare time off from job as patent clerk / programmer at high-energy accelerator / day duties as Abbot": any funder would be proud!
English
Max ретвитнул
Suggested conflict of interest statement for all academic papers:
"The authors' career prospects depend on publishing papers with clear conclusions that agree with their previous hypotheses and/or assertions."
#AcademicTwitter #AcademicChatter
English
Max ретвитнул
Increasingly I feel that academics self-organise themselves not primarily by 'topic of interest' but rather 'epistemology'.
Different fields and subfields have different assumptions about what is useful knowledge and how it is acquired
1/2
English
@PhDemetri kgoldfeld.github.io/simstudy/refer… has an implementation of such a bisection procedure
English
@PhDemetri haven’t looked at your example in depth but my first hunch was related to this series of papers doi.org/10.1093/aje/kw…
English
If you're bored, I've got this example of confounding in R.
I want to demonstrate how the confounded estimate is computed using marginal probabilities for the outcome and the exposure, but I can't get the arithmetic right.
Would love if you could help
gist.github.com/Dpananos/8418a…
English
@carlislerainey who said it that footnotes are for that what the writer wants to say but the reader does not necessarily want to read?
English
Authors vary a lot in how they use footnotes.¹ What's your preferred style?
1️⃣ Many footnotes!
2️⃣ Few footnotes; omit those details!
3️⃣ Few footnotes; put those details to the main text!
English
@carlislerainey some hybrid between distill.pub, pubpeer.com and repositories like Github? Possibly with some credit system where you need to review after a number of submissions.
English
Oh no! Thanos has snapped his fingers and turned Elsevier, Spring, Taylor & Francis, etc., to dust. ☹️
We need a new way to distribute our papers.
Starting from scratch, how do we replace the current publication process? New for-profit publishers? Do we keep peer review?
English
@frieke72 @Woningnieuws @Lukkezen je kan de overwaarde gewoon verzilveren met een omgekeerde hypotheek of je casht de overwaarde met de verkoop van het huis en gaat vervolgens bijv. huren. Bij laatstgenoemde word je in tegenstelling tot andere huurders een huurder met veel vermogen op de rekening
Nederlands
@Woningnieuws @Lukkezen Uh... Begrijp het misschien niet goed, maar van ozb waardestijging kun je geen boterham meer eten 🤷 (beklaag me niet over die stijging, maar het idee dat je er iets mee kunt)
Nederlands
Max ретвитнул
Too often we borrow the style of professional programmers who need speed and maintainability.
Scientists on the other hand, need clarity and correctness. Code that's easy to understand for non-experts is really valuable for scientists.
English
Max ретвитнул
3 goals of sharing data + code underlying empirical research articles:
(1) easily reproduce results (e.g., click Run)
(2) document every decision, even those not mentioned in the paper
(3) demonstrate to the community that you implemented the analysis as you claimed
English
English
@ryancbriggs you can get a pretty accurate idea of your body fat by comparing what you see in the mirror against informal references like these builtlean.com/wp-content/upl… there's also skinfold meters with reference tables
English
Or better yet, any ideas for checking which mode is more accurate that don’t invoke going to the doctor?
Toronto, Ontario 🇨🇦 English
I got a scale that can estimate body fat, etc. It has an “athlete mode.” I assume it runs a different model on the sensor data if you’re an “athlete.” I’m not, but I’m reasonably fit and seemingly get rowing times ~in the top 1/3 of my age group. Do I turn on athlete mode?
Toronto, Ontario 🇨🇦 English