Indeed, #methodsmatter nature.com/articles/s41592-0…

📊 Smarter analysis in #axSpA imaging SPAC cohort (2-year data): ➡️ Integrated models include more patients ( 12–16%) ➡️ Reduce selection bias without losing precision ➡️ ⚠️ Aggregated scores can mislead (↓ spine change, ↑ SIJ change) 📌 Multilevel models > simple approaches 🔗 rmdopen.bmj.com/content/12/1… #Rheumatology #Imaging #MethodsMatter #Research

ICYMI: "Clinical prediction models using machine learning in oncology: challenges and recommendations" --> bmjoncology.bmj.com/content/… #machinelearning #digitalhealth #predictionmodels #methodsmatter

This #paper offers an incredible #methods section detailing how to rear ticks in the #lab and how to dissect them for anatomical study. #tick #science #methodsmatter

Agreed. When I first started working in genomics, I printed and read protocols for all common single-cell genomics assays— I learned a lot. #methodsmatter

Just saw another one TODAY that @MikieSherrill was 5% ahead... #MethodsMatter

NEW paper "Clinical prediction models using machine learning in oncology: challenges and recommendations" --> tinyurl.com/2h2mcy5h #machinelearning #methodsmatter #digitalhealth

🚨Numbers lie if you don’t read their rules. tfr=1.9 is a signal; reality is a system. #UPSC #GS1 #GS2 #GS3 #GS4 #Essay #Population #Demography #HealthPolicy 📉👶🏽📊 what the headline says vs what it means #TFR101 🧮 TFR (1.9) = period synthetic estimate of average births if today’s age-specific fertility rates (ASFRs) persist through a woman’s life. It is not the real completed family size of living cohorts. Replacement ≈ 2.1 (lower if female mortality ↓ and sex ratio at birth normalises). India’s TFR < 2.1 ⇒ sub-replacement nationally; still >2 in a few high-fertility states (e.g., Bihar, UP), and ≤1.6–1.8 in many southern/western/urban regions. #ConceptClarity #PrelimsReady the measurement traps (why “1.9” can mislead) #MethodsMatter 🧪 Synthetic-cohort assumption bias: Today’s 20–24 behaviour ≠ behaviour of the same women at 30–34 tomorrow; tempo (timing) shifts distort TFR. Exclusion rule: TFR ignores births to <15 and ≥50; in low-income contexts with early marriage, under-count risk ↑. Under-recording: concealed teen marriages/births; abortion/miscarriage data gaps; survey discomfort (noted globally in DHS/NFHS fieldwork). Tempo effect: postponement (urban & increasingly rural) from 20–24 → 25–29 / 30–34 lowers period TFR even if lifetime births unchanged. ASFR shape shift (NFHS-3→5): sharp fall in 15–19, 20–24; rise in 25–29 (urban) and 24–30 & 30–34 (rural) ⇒ postponement signature. #Demography #TempoVsQuantum

Descriptive piece, not a rigorous study; 8-mo counts≠rates/CIs; “highly educated” undefined (read: BA); distinguish W-2 vs. 1099 vs. founders. In a depressed jobs market, data matters. #WomensWorkforce #JobsEconomy #MethodsMatter

Women’s exits are real. The method isn’t: 8-mo headcount ≠ rate; “highly educated” undefined; mothers ≠ all women; childcare ≠ proven cause; govt cuts/RTO ignored; 25–54 is dated (misses 45–64); no denominators/CI. Show LFPR/EPOP CIs and clear cohorts. #MethodsMatter

B risk you want to see in world. #DataSecurity #AI $AI is not magic. #MethodsMatter Functional literacy speaks for itself. Builders Build. Local, rooted, personal leverage in cloud relationships win every day over the data tracking/looting, advertising surveillance for masses.

As @emukamel and Yu point out, #methodsmatter. They are spot on here: nature.com/articles/s41586-0…

Artificial vs Actual Intelligence assessment underway... #AI #MethodsMatter

Our paper on combining high-throughput enhancer assays in neurons and phenotype-rich assays in vivo for testing common psychiatric variants finally out! See linked 🧵for overview and this one for updates. #MPRA #methodsmatter @NatureComms nature.com/articles/s41467-0…

#ConsumerAI fraud has been BIG help in identifying security risks & improving our tools for applied literacy. #EDU As we further secure our systems & processes for evaluating human input to creative outcomes, our priority is support for *people* in civil society. #MethodsMatter

The @NIH website, and associated sites, are officially offline. Crucial bioinformatics tools like BLAST, ffq, gget depend on these services. This is completely unacceptable. #methodsmatter

🎓💼 Unsolicited Advice 🎓💼 As the end-of-year rush kicks in 🏃‍♂️, I know early-career researchers often feel the pressure to publish quickly to hit career milestones. Reflecting on my own experience, publishing often feels like the bias-variance tradeoff: 👉 Rush too much, and you risk bias—overlooking critical details. 👉 Over-polish, and you risk variance—spending endless time tweaking without finishing. For reference, my most impactful first-author paper took 5 years to publish. Many of my collaborative projects reflect many years of teamwork and dedication. One thing I’ve learned: publishing with purpose leads to the greatest impact. Taking the time to get it right pays off in the long run. Strive for balanced, meaningful work that leaves a lasting impression. 🚀 #PublishingWithPurpose #MethodsMatter #QualityOverQuantity #HappyHolidays2024 P.S. Your reputation is one of your greatest assets. Protect it with work you’re proud of. Cheers to finishing strong!💪

Sometimes you just have to persevere through negative reviews. Reviewers, please be kind :) #methodsmatter bit.ly/gs2-scholar

the silence of @lpachter and the rest of his lab (phd students included) on the genocide in gaza is concerning given how outspoken they are on comparatively meaningless issues like umaps #moralitymatters #methodsmatter