numbas, to math
@numbas@mathstodon.xyz avatar

Next month we're running a couple of training sessions and we're also trying a new drop-in session for anyone who wants to talk about Numbas.

The sessions are free and open to anyone with an interest in Numbas.

There's more information and links to register on the blog: https://www.numbas.org.uk/blog/2024/05/numbas-training-sessions-and-drop-in-hours-june-2024/

#Numbas #eAssessment #assessment #math #maths

ryanrandall, to TodayILearned
@ryanrandall@hcommons.social avatar

Idaho State University is hiring a Director of , , & : https://isu.csod.com/ux/ats/careersite/5/home/requisition/2180?c=isu

If you're a good fit, consider becoming my (boss's boss's) boss!

ByrdNick, to psychology
@ByrdNick@nerdculture.de avatar

We know that the task demands of cognitive tests most scores: if one version of a problem requires more work (e.g., gratuitously verbose or unclear wording, open response rather than multiple choice), people will perform worse.

Now we have observed as much in Large Language Models: https://doi.org/10.48550/arXiv.2404.02418

The tests included analogical reasoning, reflective reasoning, word prediction, and grammaticality judgments.

image/jpeg
image/jpeg
image/jpeg

ErikJonker, to ai
@ErikJonker@mastodon.social avatar

Tommorow, "Regulatory Models for Algorithmic Assessment: Robust Delegation or Kicking The Can?", digital seats available.
https://www.ucl.ac.uk/laws/events/2024/apr/regulatory-models-algorithmic-assessment-robust-delegation-or-kicking-can

EricCarroll, to random
@EricCarroll@zeroes.ca avatar

WHO document drops on transmission, so I opened this up as if it was a tactical nuke needing defusing.

But holy cow...

WHO Airborne Risk Indoor Assessment (ARIA) Technical Advisory Group with LIDIA MORAWSKA as the CO-CHAIR!?!?

CERN?!?!?

WHO Environment and Engineering Control Expert Advisory Panel (ECAP) with Cathy Noakes on it and ASHRAE liason?

An actual model?

I do see the dreaded Dr. IPAC Droplets on the committee... hmm.

<now reading intensely, this is going to take a few minutes>

Indoor in the context of : description of airborne transmission mechanism and method to develop a new standardized model for risk assessment

https://iris.who.int/handle/10665/376346

violetmadder,
@violetmadder@kolektiva.social avatar

@EricCarroll @IPEdmonton @BE

I wish we could see what's happening, who is on the side of public health and how they're teaming up to push against this nightmare, and who specifically is trying to stop them.

violetmadder,
@violetmadder@kolektiva.social avatar

@EricCarroll @hannu_ikonen @BE

I think a big part of our problem is that it's very hard for most people to imagine that ANYONE would actually do things so counterproductive and depraved as what's going on around us right now.

It's like that line in Fear and Loathing, "The only hope now, I felt, was the possibility that we’d gone to such excess, with our gig, that nobody in a position to bring the hammer down on us could possibly believe it."

petersuber, (edited ) to random
@petersuber@fediscience.org avatar

Kudos to the European Research Council (, @ERC_Research):

"To keep up with the times, the focus on journal publication should be dropped, in favour of achievements that give a more rounded picture…In line with @coarassessment's recommendations, the ERC’s new rules shifted away from the traditional listing of research posts held and papers published, to a four-page narrative curriculum vitae."
https://sciencebusiness.net/news/european-research-council/european-research-council-has-changed-how-it-evaluates-applicants

christianschwaegerl,
@christianschwaegerl@mastodon.social avatar

@petersuber @ERC_Research Should have been dropped decades ago. Perhaps that would have saved us from algorithmic social media, which is clearly modelled after science's publication metric system...

petersuber, to random
@petersuber@fediscience.org avatar

2023 was a year of significant progress in reforming in the direction of the principles. Thanks to @DORAssessment for the summary and large veins of the underlying activism.
https://sfdora.org/2024/02/12/2023-in-review-list-of-new-developments-in-research-assessment/

petersuber, (edited ) to random
@petersuber@fediscience.org avatar

Another reason not to rest research or journal/publisher/university on counts.
https://arxiv.org/abs/2402.04607

"Intrigued by a citation-boosting service that we unravelled during our investigation, we contacted the service while undercover as a fictional author, and managed to purchase 50 citations. These findings provide conclusive evidence that citations can be bought in bulk, and highlight the need to look beyond citation counts."

petersuber,
@petersuber@fediscience.org avatar

Update. " 'The capacity to purchase citations in bulk is a new and worrying development,' says Jennifer Byrne, a cancer researcher at the University of Sydney who has studied problematic publications in the biomedical literature…A researcher’s h-index and the number of citations they’ve garnered are often used for hiring and promotion decisions."
https://www.science.org/content/article/vendor-offering-citations-purchase-latest-bad-actor-scholarly-publishing

18+ Frances_Larina,
@Frances_Larina@sfba.social avatar

@petersuber

Gee, maybe - just maybe - the entire publish-or-perish and highly profitable journals system itself is the root of the problem. We were taught peer review would suss out all problems, but that was a smokescreen as the only mechanism allowed it is to publish a rebuttal in the same highly siloed journal.

boilingsteam, to linux
@boilingsteam@mastodon.cloud avatar
boilingsteam,
@boilingsteam@mastodon.cloud avatar

@dpwiz perfect calibration based on what? predictions that are not real?

dpwiz,
@dpwiz@qoto.org avatar

@boilingsteam on being right X% of the time when placing X% confidence

petersuber, to random
@petersuber@fediscience.org avatar

Kudos to the Volkswagen Foundation for experimenting with a new kind of for evaluating proposals.
https://researchonresearch.org/volkswagen-foundation-introduces-experimental-distributed-peer-review/

I have no opinion (right now) on the new method. I just want to applaud the experiment. The foundation is using its old and new methods side by side for a time and plans to compare the results.

More orgs that evaluate research (, , , agencies) should open up to this kind of experiment.

gepasi,
@gepasi@fediscience.org avatar

@petersuber it certainly will be interesting to see how this will go. Thinking of how competitive scientists are, I suspect reviews will be devastating... on the other hand it would be interesting if this worked well.

(and I also applaud)

petersuber, to random
@petersuber@fediscience.org avatar

and committees should also reward and data sharing.
https://commonplace.knowledgefutures.org/pub/5cttjyuh/release/1

"There have been many calls to expand P&T to include more types of nontraditional scholarship…and to formally acknowledge that there are more ways to share knowledge than the written word. We’re calling for an even more expansive view of scholarship: one that invites, accepts, and celebrates the entirety of research, including datasets."


🧵

petersuber,
@petersuber@fediscience.org avatar
daniel_huppmann, to climate German
@daniel_huppmann@mastodon.social avatar

The Austrian 🇦🇹 climate-science community is writing a national Report () following the process & structure of the - and at @uniinnsbruck, we're hiring an researcher to support the co-chairs & authors!

In collaboration with @ccca
👉 https://geographie.uibk.ac.at/unicms/chls/pdf/20240109_research_scholar_final_aar2.pdf

raptor, to security

is the acronym for . BSAM is an open and collaborative methodology developed to standardize the security evaluation of devices using Bluetooth technology.

https://www.tarlogic.com/bsam/

petersuber, to random
@petersuber@fediscience.org avatar

"The number of issued for research articles in 2023 has passed 10,000 — smashing annual records…Retractions are rising at a rate that outstrips the growth of scientific papers."
https://www.nature.com/articles/d41586-023-03974-8

PS: We can all list some of the factors at work here, and we should. My start: Paying more attention to where a work is published, and the fact of publication itself, than to the quality of what is published.

petersuber,
@petersuber@fediscience.org avatar

@timelfen
Yes, that's right. Sorry if it wasn't clear.

On we/they: Universities, funders, and researchers could all do better in paying more attention to the quality of research than where it's published.

The goals of research assessment create incentives to game the system. The goals should be high-quality research, not publication in high-prestige or high-impact journals, and not high scores on simplistic metrics, esp. journal-level metrics.

kellogh,
@kellogh@hachyderm.io avatar

@petersuber if this were a development team, i would treat the problem as the authors not being held responsible for their quality (or lack thereof). i’d make some policy changes that make the authors feel the impact more. but who knows, maybe research is different

another reaction is to sharpen reviews. the trouble is reviewers don’t have as much context, so they miss things that are more hidden. also, tackling it in review tends to slow down progress

petersuber, to Japan
@petersuber@fediscience.org avatar

"The University of Tokyo became the first Japanese university to sign the San Francisco Declaration on Research Assessment (, @DORAssessment) on 1 December 2023."

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