Michael Andrades
@michaelandrades.bsky.social
45 followers 56 following 19 posts
Scientist and no-performance cyclist.
Posts Media Videos Starter Packs
Reposted by Michael Andrades
redebrrepro.bsky.social
A Rede Brasileira de Reprodutibilidade está buscando um(a) novo(a) Diretor(a) Executivo(a) com ênfase em pesquisa e educação! A chamada está aberta para pesquisadores de qualquer área, com remuneração baseada na tabela de valores praticados pela FAPESP, com início em novembro/dezembro de 2025!
Chamada Aberta para Diretoria Executiva!
A Rede Brasileira de Reprodutibilidade está buscando um(a) novo(a) Diretor(a) Executivo(a) com ênfase em pesquisa e educação!
www.reprodutibilidade.org
michaelandrades.bsky.social
@filipespeck.bsky.social tesouros de Priamo? Será? No Neues Muesum, em Berlim. Um abraço para vcs do podcast Noites Gregas!
michaelandrades.bsky.social
My humble contribution to #Fesbe2025 at Unicamp @unicampoficial.bsky.social Grateful to be part of it!

#meta-analysis
#transcriptome
michaelandrades.bsky.social
🎯
davidmcraney.bsky.social
Reminder: The smarter a person is and the more education they earn, the better they will become at rationalizing and justifying their incorrect beliefs if so motivated..

Critical thinking is a skill, like playing guitar. It must be learned and practiced (and the person must want to do both).
michaelandrades.bsky.social
You're right. Still, the text says the reviewers can opt to be acknowledged.
michaelandrades.bsky.social
Recognizing reviewers is great (long overdue!), but could it backfire?

1️⃣ Discourage honest critiques due to fear of backlash?

2️⃣ Might it push reviewers to go easy, just to get name recognition, lowering the rigor?

Thoughts?
michaelandrades.bsky.social
Brandon Stell puts it powerfully - and accurately: 'We value publications more than what's inside them.'
It's not something we wanted to hear, but it's something we needed to.
michaelandrades.bsky.social
Great! Will you be broadcasting it online?
Reposted by Michael Andrades
michaelandrades.bsky.social
Belkin et al. demonstrated that male and female hearts respond differently to a preventive diet against dilated cardiomyopathy. Looks like sex matters in #cardiology, huh? See doi.org/10.1016/j.jm...

@reproducibilitea.org @redebrrepro.bsky.social @ukrepro.bsky.social
michaelandrades.bsky.social
Animal research needs a reality check. 🐭🔬
Standardized experiments with genetically identical, same-age, single-sex animals boost internal validity - but at what cost? Lack of heterogenization reduces reproducibility and real-world relevance.
#Reproducibility #AnimalResearch #Science
A researcher measures identical mice emerging from an industrial production line.
michaelandrades.bsky.social
Que análise sobre o voto e países soberanos na democracia moderna. Não tem jeito, o sistema precisa de atualização e o Miguel Lago finaliza com umas poucas ideias. Pena que não investiu mais nessa abordagem. Estou curioso para ler o A Construção de um Estado para o Século XXI. a.co/d/9pyveTN
michaelandrades.bsky.social
What do you mean by 'visual sloppiness'?
michaelandrades.bsky.social
I suspect it's a holdover from the era of physical paper-submitted manuscripts, when high-definition figures were printed separately—almost like an archaeological relic. Sometimes we update the system but not the practices.
michaelandrades.bsky.social
Why do journals insist on requiring figures at the end of the file?
Haven't editors reviewed enough manuscripts to realize it's much easier to check a fig. when it's near the paragraph where it is cited?
Make reviewers' lives easier - you're already not paying them!
#ScientificPublishing
#PeerReview
michaelandrades.bsky.social
Do you think state-funded institutes should be maintained to audit certain state-funded projects, or would changing policies to eliminate harmful incentives (such as overvaluing high-impact journals) be a more cost-effective way to improve science?
michaelandrades.bsky.social
Before the AI era, Clark et al. completed a full systematic review in just 2 wks by dividing the process into parts and automating them.
www.sciencedirect.com/science/arti...

I’m not naive, but this demonstrates that we can improve data synthesis with better training datasets and ethical oversight.
A full systematic review was completed in 2 weeks using automation tools: a case study
Systematic reviews (SRs) are time and resource intensive, requiring approximately 1 year from protocol registration to submission for publication. Our…
www.sciencedirect.com
michaelandrades.bsky.social
AI promises to speed up #meta-analysis, but not without risks. Transparency, reproducibility, and preventing misuse are key concerns. www.nature.com/articles/d41...

Are we ready to trust AI to synthesize decades of science? Or will it just make a mess of the literature?

#AI #Science #Metascience
Can AI review the scientific literature — and figure out what it all means?
Artificial intelligence could help speedily summarize research. But it comes with risks.
www.nature.com