Kevin J. Black @kjbinstl@qoto.org

Should statistical thresholds for deciding research "significance" be retired?

The reliance on p-value thresholds has contributed to the replication crisis as scientists are often pressured to have ps < 0.05 to be able to publish, and thereby continue their careers in academia.

"We hope that methods sections and data tabulation will be more detailed and nuanced. Authors will emphasize their estimates and the uncertainty in them"

via https://doi.org/10.1038/d41586-019-00857-9
@academicchatter #Science #SciComm

@NeuroSchnell @schoppik @NicoleCRust @brembs

As a reviewing editor, and as scientist and paper author, I have no idea who the “readership” is. Readers of one paper in a journal are extremely unlikely to want to read most papers from the same journal. Scientists read papers, not journals. A journal’s contents is a collection of misfits. There isn’t and there can’t be a coherent theme, as this would have to be so narrowly defined that it would, in essence, be personalised to each reader.

Scientists generally want their research to be openly accessible to the public.

So what's stopping them?

One study of Berkeley faculty (N = 479) found that 71% supported open access to research, yet only 18% of articles published by the group were open access.

"a journal having no cost to publish in was more important than having no cost to read"

via https://www.arl.org/wp-content/uploads/2023/01/2023.01.18-open-access-publishing-a-study-of-uc-berkeley-faculty-views-and-practices.pdf
#OpenScience #Science #AcademicChatter #SciComm @academicchatter
h/t @oatp

Sending this off to @biorxivpreprint very soon, but you can read it here first.

We've added GPT-3 revision suggestions to #manubot. The point where it suggested changes that fixed an incorrect reference to a mathematical symbol was 🤯.

Check out more at the link! #science #publishing #academia #ai #aiart

https://greenelab.github.io/manubot-gpt-manuscript/

Non-profit journals as a solution to open science?

Publishers charge researchers incredible amounts to remove paywalls from their own papers...and make unbelievable profits from doing so. Several have an >30% profit margin.

Initiatives such as @PeerCommunityIn provide an alternative: open access, non-profit journals that are free to publish in.

https://peercommunityin.org/
https://peercommunityjournal.org/
#OpenScience #Science #AcademicChatter @academicchatter
Graph (2013): @alexh https://alexholcombe.wordpress.com/2013/01/09/scholarly-publishers-and-their-high-profits/

When scholars without expertise in infrastructure reform encounter the goals of the reform movement, such as e.g., replacing academic journals,

https://doi.org/10.5281/zenodo.5526634

they sometimes reflexively defend the status quo that served them well, e.g.:

“You ‘open’ wackaloons are forcing your revolution on us, against our academic freedom!”

I was asked to write a *one-page* cheat sheet with the evidence to counter such a reflex:

https://docs.google.com/document/d/1vjbohSsf-qy8ZkCvQ9FRlF9k1N12oSLu8qwXoKxjD2o/edit?usp=sharing

Can you/we improve on that?
#openscience

Just saved myself a lot of space in this application I'm writing with the DOI shortener tool from the DOI Foundation 👍

Non-professionals can be trained to deliver empirically supported psychosocial interventions that are effective for many people: https://www.statnews.com/2022/12/13/ecuador-lessons-increasing-access-mental-health-care/ 🤝 #mentalhealth #psychology #social #socialwork #psychiatry #health #wellness #research

Starting Jan 17, all veterans will be able to access emergency mental health care free of charge at any Veterans Affairs medical facility or outside clinic. https://www.militarytimes.com/news/pentagon-congress/2023/01/13/va-to-pay-for-all-emergency-mental-health-care-starting-next-week/ 👏🏻 #mentalhealth #military #veterans #emergency #health #healthcare #va #ptsd #suicide #suicideprevention #hope #psychology #psychotherapy #psychiatry #socialwork #goodnews

Just (re)discovered Casey Greene's super "preprint similarity search tool". It did a great job for our recent #preprint. It uses a language learning model to find like #publications to help you pick your target #journals.

https://greenelab.github.io/preprint-similarity-search/