Lightnews — Scholar-powered news

Lisa DeBruine @debruine.bsky.social · 7h

We're working on a web app version that doesn't require R expertise, and it is not aimed specifically at coders, so we don't want to use an R-typical name.

1

Lisa DeBruine @debruine.bsky.social · 7h

CheckMate was one favourite possibility, but seems very difficult to google and is the name of a bunch of other apps

1

Lisa DeBruine @debruine.bsky.social · 8h

It's in the same universe as StatCheck and RegCheck (and can actually contain them) so MetaCheck seems like a natural name, but there is already a project in the same domain with that name (although it might be inactive, the last update was 3 years ago).

subugoe.github.io/metacheck/

Crossref Metadata Compliance

Automatically check metadata compliance for hybrid open access (OA). Includes a webapp to send compliance reports by email.

subugoe.github.io

2 2

Lisa DeBruine @debruine.bsky.social · 8h

Please help us, #MetaScience community!

It's time to decide on a forever name for papercheck (scienceverse.github.io/papercheck/). We don't want it to be confused with papercheck.ai, and we plan to check other research artifacts like repo contents, data, code, and prereg. Any suggestions?

Check Scientific Papers for Best Practices

A modular, extendable system for automatically checking scientific papers for best practices using text search, R code, and/or (optional) LLM queries.

scienceverse.github.io

2 16 11

Lisa DeBruine @debruine.bsky.social · 1d

Yeah, that's what I ended up doing. I looked into adding them as non-exported internal data, but those are always lazyloaded, regardless of settings, so it didn't help.

r-pkgs.org/data.html#se...

Lisa DeBruine @debruine.bsky.social · 1d

I love the OSF, but I'm going to have to stop using it for student projects if it continues to take 30 seconds every time I load a new page.

8

Reposted by Lisa DeBruine

Elizabeth Gadd @lizziegadd.bsky.social · 7d

Adding my voice to the chorus battling to retain an increased focus on people, culture, and environment in the next REF... wonkhe.com/blogs/the-ba...

The battle for people, culture and environment

Lizzie Gadd makes the case for embracing People, Culture and Environment and sets out why reverting to REF 2021 would be a mistake

wonkhe.com

8 16

Lisa DeBruine @debruine.bsky.social · 7d

devtools::install_github("debruine/reprex.lazyload")

reprex.lazyload::mydata # works
reprex.lazyload::myfunc() # fails

library(reprex.lazyload)
myfunc() # works

2

Lisa DeBruine @debruine.bsky.social · 7d

I'm having an #rdev problem using data in an r package inside a function when calling the function without loading the package. I made a reprex to demonstrate the problem (code in the next post) but I think it has something to do with lazyloading data.

mydata is a data table
myfunc() returns mydata

1

Lisa DeBruine @debruine.bsky.social · 22d

We're meeting tomorrow before the first class on Friday, so we'll check out some different options and see what they prefer. I'm pretty comfortable with the command line, and can see how that could be a more straightforward interface, so we'll definitely try that one out.

Lisa DeBruine @debruine.bsky.social · 22d

I need #rstats meme help. I've just updated my MSc book and all of the chapters start with a (probably outdated) meme except chapter 10 (next steps). Any ideas?

psyteachr.github.io/reprores-v5

Data Skills for Reproducible Research

This book provides an overview of skills needed for reproducible and open research using the statistical programming language R and tidyverse packages. It covers reproducible workflows, data visualisa...

psyteachr.github.io

2 9 12

Lisa DeBruine @debruine.bsky.social · 24d

Thanks! Luckily, the RStudio cheat sheet got html versions recently.

Do you know if VScode is enough better than RStudio to merit using a different IDE than the rest of the class and book instructions?

1

Lisa DeBruine @debruine.bsky.social · 25d

Excited for today’s seminar by @juliehulme-psyc.bsky.social about the experiences of disabled and neurodivergent students.

2

Lisa DeBruine @debruine.bsky.social · 28d

Fab! I wish I'd learned about this before I spent all day yesterday updating the alt text for my dataviz chapter. I'd thought my figure captions were good enough, but they were more high-level about the purpose of the plots, and less detail about what the plots specifically convey.

2

Lisa DeBruine @debruine.bsky.social · 28d

Thank you so much! These are fantastic resources

1

Reposted by Lisa DeBruine

Sajedeh Rasti @sajedehra.bsky.social · Sep 10

Time is running out to register for our pre-symposium workshop and Meta Research Symposium event at Paul Meehl Graduate School
Check the information and registration links below:

paulmeehlschool.github.io/workshops/

6 7

Lisa DeBruine @debruine.bsky.social · 28d

This fall I will have a blind student in my coding class for the first time. Do any other instructors or visually impaired coders have advice beyond making sure my book has useful alt-text for the images? #rstats #accessibility

Class book: psyteachr.github.io/reprores-v5/

6 36 26

Reposted by Lisa DeBruine

James Bartlett @bartlettje.bsky.social · Sep 8

Our contribution to an upcoming book on teaching open science was "how to teach reproducible research". There are plenty of lesson plans or individual courses out there but less on how you can build skills across a whole degree.

osf.io/preprints/ps...

OSF

osf.io

1 7 17

Lisa DeBruine @debruine.bsky.social · Aug 22

I have discovered the most delicious oatmeal topping ever. Stir in a spoon of biscoff spread and top with flaky sea salt (measure with your heart, it really depends how you're feeling abut sweet vs salty on a given day).

2 9

Reposted by Lisa DeBruine

ricvolpe @ricvolpe.bsky.social · Aug 15

Happy to share the first article of my PhD journey at @uofglasgow.bsky.social and @socialai.bsky.social

@debruine.bsky.social, Prof Stumpf and I reviewed ~150 articles from 10 disciplines about trust in AI and synthesised how it forms in affective human-AI relationships: doi.org/10.1016/j.ch...

Our categorisation of concepts relevant to trusting emotional support from generative AI, underpinned by four categories: propensity to trust, motivations to depend on (convenience, expertise, empathy, positive regard, and personalisation), perceived risks that produce vulnerabilities (privacy, harm, and organisational control) and assurances that support an unquestioning attitude (credibility, congruence, and accountability). Additionally, attachment and social connectedness sit on a spectrum between motivations and risks as they present positive and negative effects on trust.

1 4

Lisa DeBruine @debruine.bsky.social · Aug 19

I'll probably add this to Papercheck after I test it a bit more (it only works for articles just now) as it will demonstrate some lower-level functions we've been meaning to add to assess the accuracy of references and citations.

What sort of related reference functions would you find useful?

2

Lisa DeBruine @debruine.bsky.social · Aug 19

What was I supposed to do this morning?

Make comments on a draft of the Papercheck paper.

What did I actually do?

Write an R function to add DOIs from OpenAlex to a .bib file (the bibtex citations you get from google scholar bizarrely don't have DOIs)

1 18

Lisa DeBruine @debruine.bsky.social · Aug 8

I want to highlight a bunch of different original data sources to show the variety of ways data needs to be processed, so I'm on purpose not just getting a bunch of data that is already structured the same way from the SDG Data Portal (which is fab)

unstats.un.org/sdgs/datapor...

1

Lisa DeBruine @debruine.bsky.social · Aug 8

I've started 6 chapters in the past 3 days. It's very draft content, but I'd love feedback on whether this is the sort of thing people might use and what you'd want out of it.

What metadata about each version would you want to be able to decide if a version is good for a lesson you're planning?

A plot showing water access by development level from 2000 to 2022. High-income countries have nearly 100% safe access, while access steadily worsens for lower income levels, but steadily improves over time.

1 3

Lisa DeBruine @debruine.bsky.social · Aug 8

Awesome, I didn't know that!

I'm finding tons of data sources, and it's really emphasising the need for good data skills. So many ways to label countries! Most provide data in such a non-tidy format, like with text notes in cells with numbers and unmarked note rows before or after tabular data.

1