@tomstafford@mastodon.online
@tomstafford@mastodon.online avatar

tomstafford

@tomstafford@mastodon.online

"A vast confusion of vows, wishes, actions, edicts, petitions, lawsuits, pleas, laws, proclamations, complaints, grievances, are daily brought to our ears. "

This profile is from a federated server and may be incomplete. Browse more on the original instance.

tomstafford, to random
@tomstafford@mastodon.online avatar

Uncertainty in peer review. https://doi.org/10.31219/osf.io/fm6st

  • measured within reviewer uncertainty (~ 1/ confidence) in journal paper reviews

  • predictions of greater reviewer uncertainty: gender, if the paper was a protocol

  • NOT predictive: reviewer experience, time taken on review, reviewer nationality, paper version (first submission vs revision), paper length, readability

#metaresearch #metascience #PeerReview

tomstafford, to random
@tomstafford@mastodon.online avatar

TODAY: Launch event for our programme

"an Accelerator For Innovation & Research Funding Experimentation"

A-F-I-R-E !

https://researchonresearch.org/project/a-f-i-r-e/

PaulGrahamRaven, to philosophy
@PaulGrahamRaven@assemblag.es avatar
tomstafford,
@tomstafford@mastodon.online avatar

@PaulGrahamRaven good Monday thoughts! Thanks for writing

tomstafford, to random
@tomstafford@mastodon.online avatar

I just been told by a hotel that they do not offer an inclusive breakfast, but a "delicious hospitality concept" is available.

I'll report back, but I'm wondering if a concept is really going to get me through the day better than some toast and a cup of coffee

tomstafford, to random
@tomstafford@mastodon.online avatar

UK HE shrinking

"This is a live page of all the redundancies and closures taking place across the sector at the moment. "

https://qmucu.org/qmul-transformation/uk-he-shrinking/

Page from QM , listing 50 institutions which are planning for redundancies

tomstafford,
@tomstafford@mastodon.online avatar
tomstafford, to random
@tomstafford@mastodon.online avatar

Researchers want a ‘nutrition label’ for academic-paper facts

"An ‘at a glance’ approach to publication details, such as journal acceptance rates and the number of peer reviewers, would promote transparency, scientists say."

https://www.nature.com/articles/d41586-024-01135-z

tomstafford,
@tomstafford@mastodon.online avatar

@PaulGrahamRaven agree that there's no substitute, but i'm hopeful more transparency in journal publishing is a good thing. Currently it is a massive scandal with public money being ripped off by predatory publishers at one end, and prestige publishers at the other, and the exact value delivered by both unclear

tomstafford, to random
@tomstafford@mastodon.online avatar

"this meme demonstrated the importance of sample size better and more efficiently than any science class I've ever taken"

tomstafford, to random
@tomstafford@mastodon.online avatar

"Rule 1 of cybersecurity is to keep your attack surface small; in LLMs the attack surface appears to be infinite."

Gary Marcus nails it https://garymarcus.substack.com/p/an-unending-array-of-jailbreaking

via Data Machina https://datamachina.substack.com/p/data-machina-248

tomstafford, to random
@tomstafford@mastodon.online avatar

Duolingo call for research proposals

https://drive.google.com/file/d/11Z1kr-LSR2n9Si_MQseUMxIHAagM45qH/view

Up to $80k to design and run a research Study on Duolingo English learners’
real-life communication skills

Deadline: May 31st 2024

#CogSci

PaulGrahamRaven, to random
@PaulGrahamRaven@assemblag.es avatar

This is probably a bit of a long shot here, but nonetheless: has anyone caught a New Model Army live show recently, and if so, was it good? They're playing round the corner from me next week, but I don't want to drop >40euro on discovering that they're a pale shadow of the band I saw in 1995...

tomstafford,
@tomstafford@mastodon.online avatar

@PaulGrahamRaven not seem them for a few years, but more recently than 1995 (maybe 2019??) and they were great

tomstafford, to random
@tomstafford@mastodon.online avatar

What nonsense is this? I'm looking at a journal article and I get an institution specific DOI, one that other people can't use to find the article?!

https://doi-org.sheffield.idm.oclc.org/10.1093/qje/qjv030

tomstafford,
@tomstafford@mastodon.online avatar

If I find the article without institutional login I'm presented with the universal, generic, DOI I really want

https://doi.org/10.1093/qje/qjv030

tomstafford, to random
@tomstafford@mastodon.online avatar

Trump supporters target black voters with faked AI images

https://www.bbc.co.uk/news/world-us-canada-68440150

"Donald Trump supporters have been creating and sharing AI-generated fake images of black voters to encourage African Americans to vote Republican."

1/

#GenAI #Disinformation

tomstafford, to psychology
@tomstafford@mastodon.online avatar

NEW NEWSLETTER

How persuasive is AI-generated propaganda? https://tomstafford.substack.com/p/how-persuasive-is-ai-generated-propaganda

Reasonable People #51: Bullet review of a new paper suggesting LLMs can create highly persuasive text and will supercharge covert propaganda campaigns.

I'm not persuaded

#persuasion #disinformation #psychology #psyops #propaganda #ai

tomstafford, to random
@tomstafford@mastodon.online avatar

Saturday morning technical question:

Is there a way to spoof my location (IP address) so online I appear to be at a different place within the UK?

I don't want a VPN so I can appear to be in the UK when I'm not, I want to be in Sheffield and appear to be in another UK location, so I can receive any geo-targetted adverts for that place

tomstafford, to de
@tomstafford@mastodon.online avatar

We've had one of these De'longhi espresso machines for ~10 years, and it has reliably made at least 2 coffees a day, nearly every day.

Also, I bought it secondhand for £10

That's ~7000 coffees, or approx 0.1p/coffee.

Is there any way I can identify similarly well engineered, long-lived and great value products? Ideally before purchase!

https://www.delonghi.com/en-gb/ec230-bk-stilosa-manual-espresso-maker/p/EC230.BK

#De'longhi #Coffee #Espresso

tomstafford, to random
@tomstafford@mastodon.online avatar

We need to have a frank conversation about higher education. These confected rows benefit nobody

https://www.theguardian.com/commentisfree/2024/feb/06/woke-foreign-students-uk-universities-starved-cash

fact check: true

tomstafford, to random
@tomstafford@mastodon.online avatar

A partially randomised approach to internal funding allocation: A pilot by the University of Leeds’ Research Culture team

https://eprints.whiterose.ac.uk/208005/

lakens, to random
@lakens@mastodon.social avatar

You can now sign up for the first full-day workshop of the Paul Meehl Graduate School for Metascience on Meta-Analysis and Bias Detection, co-taught by Robbie van Aert and myself: https://paulmeehlschool.github.io/metaanalysis/ Workshops are free and open for anyone.

tomstafford,
@tomstafford@mastodon.online avatar

@lakens nice. Feel free to send a message to metaserv@jiscmail.ac.uk to promote any events via the list

tomstafford, to random
@tomstafford@mastodon.online avatar

This series reboot is really taking Chronicles of Narnia in a new direction

lakens, to random Dutch
@lakens@mastodon.social avatar

I am writing a paper on the benefits of preregistration. In one section I will address what most people find the strongest criticisms on preregistration. So, which points do you feel I should discuss?

tomstafford,
@tomstafford@mastodon.online avatar

@lakens

(not that i endorse these but)

  • that it is an unreliable signal of rigour (because reviewers don't check prereg)
  • that it is inappropriate for observational data sets / exploratory data analysis
  • that it crowds out/ overshadows serendipitous discovery
  • that it is only appropriate for a very narrow range of disciplines / subdisciplines and inappropriately demanded of researchers outside the core of quantative disciplines where it was developed

#preregistration

tomstafford,
@tomstafford@mastodon.online avatar

@renebekkers @lakens didn't these people get the "it's a plan, not a prison" memo ;-)

tomstafford,
@tomstafford@mastodon.online avatar

@renebekkers @lakens i am slowly becoming reconciled to the idea that a lot of successful advocacy is repeating yourself again and again for every new audience

  • All
  • Subscribed
  • Moderated
  • Favorites
  • JUstTest
  • mdbf
  • ngwrru68w68
  • modclub
  • magazineikmin
  • thenastyranch
  • rosin
  • khanakhh
  • InstantRegret
  • Youngstown
  • slotface
  • Durango
  • kavyap
  • DreamBathrooms
  • provamag3
  • GTA5RPClips
  • tacticalgear
  • normalnudes
  • tester
  • osvaldo12
  • everett
  • cubers
  • ethstaker
  • Leos
  • anitta
  • cisconetworking
  • megavids
  • lostlight
  • All magazines