sharoz

@sharoz@scicomm.xyz

Perception, cognition, and data visualization scientist

(he/him)

#VisionScience #CognitiveScience #Neuroscience

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

sharoz, to random

Had a great discussion on with White House representatives today. I encouraged shifting to a mandatory model for sharing publicly funded research data, stimuli, analyses, code, and other replication and reproducibility materials.

A few key points:

  1. For research that cannot be shared (e.g., due to participant privacy concerns) explicitly require that the reason for not sharing the data be stated in the resulting articles.

...

debivort, to random
@debivort@drosophila.social avatar

It's pretty disappointing how many influential, personable scicom folks are still on twitter, as Musk is hocking anti-trans hit pieces.

I get some of their tweets here through bots, but I'm thinking of cutting that off.

What do you all think about this?

sharoz,

@debivort @NicoleCRust

  • Search, especially for my own threads, so I can use them to answer the same Q
  • Looking through who someone's following to find more people without needing to click multiple times
  • Being able to open and follow an account via an external link
  • Quote tweets
  • Giphy integration (no fun allowed)
  • Video upload (not links, uploads!)
  • No "pick your server" technobable
  • Go where the non-science followers are. Links posted to both get more clicks on Twitter.
sharoz,

@debivort @NicoleCRust

Another big one is the attitude of "you're using mastodon wrong" that a lot of users got during the big migration 6ish months ago. It nearly turned me off from mastodon completely early on. I had to move servers because I was told I have too many followers. That is antithetical to scicom.

sharoz,

@debivort @NicoleCRust

Don't get me wrong, I have a lot of positive things to say about mastodon. The edit feature alone is fantastic. But it's tough to attract the audience that science communicators want to communicate to with mastodon's cultural and technical hurdles.

sharoz,

@debivort @NicoleCRust

Oh and I forgot the biggest one: the algorithm. Purely chronological is an algorithm, not the absence thereof. It means never having context when you see a reply. It means not being able to filter to a good summary after you've been off the site for a day or two. It makes the most interesting longer back-and-forth discussions impossible to understand unless you've been in it from the beginning.

I think bluesky has the right approach of algorithms as tabs.

sharoz,

@NatureMC @debivort @NicoleCRust
These are effortful workarounds, not solutions. You're right that it's better than nothing, but the knowledge of which hashtags to follow is tough to acquire. Twitter had that "trending" pane of popular hashtags, which was great because it was also tuned to your network... before the whole mod team was fired.

sharoz,

@mrcompletely @debivort @NicoleCRust

I agree. But I think the features I've outlined are necessary to attract a general non-techy non-scientist audience. And failure to attract them will prevent the science communicators fully adopting mastodon as their primary platform.

It's the same debate as "Why do people use Mac/Windows when Linux is free?". A UX focus that prioritizes general users instead of techies will attract the most people irrespective of corporate malevolence.

sharoz,

@mrcompletely @debivort @NicoleCRust

Yeah, but this thread is about science communicators.

sharoz, to random

The fancier your file format, the shittier your data.

sharoz,

@odr_k4tana

I have seen so many instances of one file format bedded in another. Like CSVs with json or xml in one of the columns.

sharoz,

@odr_k4tana It's also disturbingly common

sharoz, to random

Is there a term for this fallacy?

Edit: Maybe https://en.wikipedia.org/wiki/Texas_sharpshooter_fallacy ?

sharoz,

@JoranJongerling @odr_k4tana

Yes, but I was looking for the broader fallacy. I think of HARKing as picking which statistical tests to perform. Whereas I'm trying to describe a case where an alpha threshold was chosen after the results were known.

sharoz,

@odr_k4tana @JoranJongerling

🤣

sharoz, to random

Helped someone debug some tidyverse data processing issues. It turns out "NA" was a legitimate code used in their data and readr by default interprets it as NA, not a string. Careful folks!

Edit: for anyone who doesn't know, read_csv() has an na parameter. The default is na = c("", "NA"). Setting it to na = "" fixed the issue.

sharoz,

@odr_k4tana It was an acronym. All the other cells in that column were also 2-letter acronyms. The person who recorded the data used Excel and SPSS, so they had no way of knowing it'd be a problem.

sharoz,

@odr_k4tana

Most data processing and analyses shouldn't require knowing how to program.

sharoz,

@odr_k4tana Statistical analysis and data processing should be point-and-clickable

sharoz,

@datamaps @odr_k4tana To be clear, I do think that rstats is the best option. But that's because GUI analysis tools are so badly made and closed source. But they don't have to be.

sharoz,

@odr_k4tana @datamaps
I've never been able to get JASP to work with data that has replicate trials (i.e. all data I've ever worked with). I generally recommend people avoid it because it just doesn't work for most cogsci-style data.

sharoz, to random

So much development on Rstudio, but updating R still isn't automated and easy.

sharoz,

@mccarthymg Using a command line is the exact antonym of automated and easy 😉

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

This week, Science published a stunningly irresponsible news story entitled "Fake scientific papers are alarmingly common" and claiming that upward of 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.

Headline and intro notwithstanding, the story itself later notes that the detector doesn't actually work and flags nearly half of real papers as fake. Does the reporter just not understand that?

h/t @Hoch

sharoz,

@ct_bergstrom

I always use a personal email address because institutional email addresses are inherently short-term. And I only have international co-authors on about a third of my papers.

According to this "detector", 2/3s of my papers are fake 🤡

sharoz,

@ct_bergstrom

Good example case for @cfiesler

SBMost, to random

Putting together a lecture on study tips that are backed by the science of memory, and I admit I may be a little too pleased with myself for this closing slide.

sharoz,

@SBMost That would be an amazing resource to share publicly outside of the course!

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