Nicola Romanò @nicolaromano@qoto.org

@nicolaromano
I agree, a bad idea to push novelty scores. Making perceived novelty a review/selection/acceptance criterion does disservice to the scientific community and their stakeholders. I firmly believe that scientific quality, study inegrity, ethics and replicability should form the basis of a decision to accept and publish a manuscript, and NOT the perception of flashiness or novelty. For that same reason I also abhor the weight and importance given to a journal's calculated or perceived impact score (or whatever scoring factor is used), whether by submitting authors, tenure review committees, readers and the audience (lay press), etc...

@ColinTheMathmo When I was younger it was all just arithmetic practice drills (which I knew I could've been doing with a calculator...) and then when we moved on to "higher math" there was just never any purpose grounded in reality.

Algebra? I still do not actually know what the point of solving for X is. No one ever told me. They were just like "do this a bunch." I did. A purpose never occurred to me. It is now decades later and I don't know why I had to do that.

@ColinTheMathmo

"Fun" is a kind of engagement. It is also subjective. Learning can be enjoyable; that's not identical to "fun".

Learning is better when not unpleasant, let alone miserable. But when you ask a person (of any age) to learn something, for no obviously good reason—other than to avoid punishment—they will find it unpleasant.

Curiosity is specific. Utility a bit more general. But neither work on kids who expect to be celebrities due to magical thinking.

@ColinTheMathmo @rakhichawla I think that the "fun" stuff can be important for getting people over the initial activation barrier. I.e. those who say they're "just not good at maths" without ever seriously engaging with it. But I agree that fun is not enough in the long term.

Why probability probably doesn...

This seems spot on.

Introducing {oomph} an #RStats pkg technical demonstration of 500x faster named subsetting of vectors and lists

https://github.com/coolbutuseless/oomph

Given a static named list/vector, `oomph` subsets 100 elements from n=200k list 500x faster than R's standard method, & 1000x less memory allocation

Notes:
* Uses an order preserving minimal perfect hash
* Suited to static objects only - hashing object would need to be recalculated for every addition/removal
* A dynamic minimal perfect hash would be welcomed

Hey are you a recent PhD with interests in #rstats and #psychology? Fancy joining the amazing stats teaching team in Psychology at Edinburgh University? Closing date for applications: 27 January https://edin.ac/3OVNTTE

BBC's Visual and Data Journalism Cookbook for R Graphics and their {bbplot} #RStats 📦are useful resources for making publication-quality graphics in
#R

https://bbc.github.io/rcookbook/

#ggplot2 #DataViz

Hey, every fucking company these days, when you send me that email telling me that I have a new document to view, that I have to access by clicking on a link, then entering my credentials, then waiting for you to text me a code that I have to type in before having to hit through your terrible menu to find where you hid the messages... then looking again because that messages page wasn't the type of message the email was talking about...

Could you, I dunno... maybe mention what the fuck the document is fucking about in the god damned mother fucking email!!?!!??!

Hello World :)

I am pleased to present you a small project that I have been working on these last weeks:

MARL (Mastodon Archive Reader Lite) is a small web app that allows you to explore in detail the content of your Mastodon posts archive, including attached files (images, videos, sounds), and with different search options.

https://github.com/s427/MARL

🙏 Boosts welcome! 🙏

(More info in the post below 👇)

#Mastodon #archive

The problem of fake science papers when trying to write a systematic review https://www.science.org/content/article/systematic-reviews-aim-extract-broad-conclusions-many-studies-are-peril

#surveycomparison #representationbias
New #R-package out now!
"sampcompR" provides functions to easily compare surveys against benchmark surveys (e.g. for bias estimation) on a univariate, bivariate, and multivariate level.

By Björn Rohr & Barbara Felderer

https://bjoernrohr.github.io/sampcompR/

Confidence intervals and p-values can be either calculated with normal, parametric methods or as bootstrap confidence intervals and p-values and additionally adjusted for multiple testing.

@tehstu I don't know how to get Microsoft to fix that, but I *do* know my solution to that whole fiasco last year... was to load Linux and never looking back since.

A few weeks ago I asked everyone for some examples of fictional maths teachers 👩‍🏫✖️➗

Earlier this week we released a @sci_burst episode all about maths anxiety & maths teachers in #popculture, touching on some of your suggestions!

✨ Link in replies ✨

#maths #science #scicomm #podcast