Alex Weber @weberam2@qoto.org

A paper to cite in response letters for why we choose to use GLM instead of “more conventional tests.” Great title too 🔥 #neuroscience #statistics https://www.sciencedirect.com/science/article/pii/S089662732100845X

Whenever I explain my #research at Google into mobile text editing, I'm usually met with blank stares or a slightly hostile "Everyone can edit text on their phones, right? What's the problem?"

Text editing on mobile isn't ok. It's actually much worse than you think, an invisible problem no one appreciates. I wrote this post so you can understand why it's so important.
https://jenson.org/text
#UXDesign #UX

Am I misunderstanding something?

This appears to be a stunningly irresponsible story in Science, claiming that up to 30% of the scientific literature is fake.

https://www.science.org/content/article/fake-scientific-papers-are-alarmingly-common

Below, the first two paragraphs of the story.

h/t @Hoch

I recorded a tutorial-type video on a Python Data Analysis project using Pandas, Numpy, Matplotlib, and Seaborn, and uploaded it to YouTube

https://www.youtube.com/watch?v=wQ9wMv6y9qc

Discussions: https://discu.eu/q/https://www.youtube.com/watch?v=wQ9wMv6y9qc

#programming #python

What's the best way to evaluate someone's publications if you don't have time to read them? 😬
e.g.,

-Characterizing dept research output. Say 7 active faculty and 3-5 papers each. (I don't have time to carefully read 21 papers.)

-Communicating publications to a Dean who doesn't know the field

-Evaluating a candidate or grant outside core expertise (so I don't really know the field)

Struggling a bit because

1. Journal impact factor is commonly used and also a poor indicator of quality

2. Gut feelings about journal "reputation" seem potentially arbitrary and prone to bias

Would love to hear your thoughts.

Takeaway from @cpsievert 's {bslib} #rstats 📦 presentation at #PositConf2023: If you are working in R Shiny, you REALLY ought to be using bslib to get lots more cool design capabilities!!
That's because basic Shiny doesn't have newer Bootstrap capabilities, since adding those would break existing code.
Package into: https://rstudio.github.io/bslib/
@rstats #RShiny

I love how the White House definition of open science includes "processes" next to "product"

"The [OSTP] defines Open Science as the principle and practice of making research products and processes available to all..."

https://www.whitehouse.gov/ostp/news-updates/2023/01/11/fact-sheet-biden-harris-administration-announces-new-actions-to-advance-open-and-equitable-research/

Open Science is often overly focused on the end result. I think it will be much more impactful if we make the *process that creates knowledge* open and accessible to all from beginning to end.

The blog of the UK Society for the Study of Labour History has published this piece by me on building a list of a radical online archives and the importance of digitised open access left-wing historical documents for researchers and activists. https://sslh.org.uk/2023/09/19/dont-mourn-digitise-building-a-list-of-radical-online-archives/

“The open access movement is fighting to get rid of paywalls altogether, and we are making progress. But in the meantime, thanks to Alexandra’s courage and creativity, researchers around the world have SciHub. It is my honor and privilege to recognize her with a 2023 EFF Award.”

How does it all work together?

Are there proposals of how multiple functional-connectional systems work together? Not just individual systems, but jointly.

Eg of individual systems: basal ganglia loops, cortico-thalamic loops, amygdala bidirectional connections with cortex etc.

#neuroscience
@cogneurophys

Synaptic transmission is noisy, presumably because the brain tries to keep energy costs down.

When we modelled this energy cost in an ANN, it spent its energy cleverly: it made synapses that were important for the task more precise, and synapses that were less task-important more noisy.

Kind of shockingly, this is also what the brain should do if it were being Bayesian.

Top work by PhD student James Malkin, with @conorjh and @laurence_ai

https://arxiv.org/abs/2309.03194

@PessoaBrain @manlius @DrYohanJohn A phase of matter is defined by the average behavior of individual particles. E.g. a magnetized phase is where the magnetic moment of most particles are pointing in the same direction, a solid phase is when the motion of individual particles is confined to a very narrow range, etc.

So if we observe the behavior of individual particles for sufficiently long, we can tell in which phase the system is.

A phase transition requires a control parameter — e.g. temperature. So, if we vary this, and observe the behavior of individual particles, we can tell when a phase transition happened.

Now, *predicting* a phase transition means, I suppose, knowing the precise value of the control parameter where it happens.

To perform this computation we need to know, in general, how the particles interact.

But to define what a phase is, and to measure it, we do not.

This is a great little game for learning git! Love it 😃 https://ohmygit.org

Provocative

As a young psychologist, this chills me to my bones. Apparently is possible to reach the stratosphere of scientific achievement, to publish over and over again in “high impact” journals, to rack up tens of thousands of citations, and for none of it to matter. Every marker of success, the things that are supposed to tell you that you're on the right track, that you're making a real contribution to science—they might mean nothing.

(I agree: time to rethink that any one of us is how this works).

I’m so sorry for psychology’s loss, whatever it is

https://www.experimental-history.com/p/im-so-sorry-for-psychologys-loss

A good friend works in a department run almost exclusively by social psychologists. He was told to expect to publish (not submit... publish) 4 papers a year, and make sure at least a subset of them appear in high impact journals like PNAS, Science, Psych Science, etc. if he wanted to get tenure.

The field cannot then act shocked when many of its superstars are found to have committed fraud or (best case scenario) sloppy research practices. The system practically begs for it to occur in order to succeed.

https://www.theatlantic.com/science/archive/2023/08/gino-ariely-data-fraud-allegations/674891/