Code sharing and reproducibility in survey-based social research: evidence from a large-scale audit
https://royalsocietypublishing.org/rsos/article/13/3/251997/480886/Code-sharing-and-reproducibility-in-survey-based
Code availability is bad with 35%. Reads like most code was not publicly available (period of analysis is 2015-2020). 1/
@openscience

The new functions in {dplyr} such as filter_out() look really useful. https://tidyverse.org/blog/2026/02/dplyr-1-2-0/ The biggest challenge will be me remembering them when needed instead of writing the code I have always written. Which would also imply not to use LLM assistance because it can hardly propose functions it has not been trained on, can it?

Welcome to #OpenScience Land
https://expedition-open-science.org/ This is a lightweight, very useful and accessible intro to key elements of Open Science (online and as a PDF). Seems particularly useful for introducing the topic to undergraduates @openscience @ZBW_MediaTalk

Okay, this data availability statement is correct because one can find the reproduction material for the article on the Dataverse. However, one has to search for it using the article title. Added value would have been to include the Dataverse DOI here.

@dfg_public erlaubt nun den Einsatz von LLMs in der Begutachtung unter Bedingungen.
https://www.dfg.de/de/aktuelles/neuigkeiten-themen/info-wissenschaft/2025/ifw-25-102 Die Bedingungen sind so gefasst, dass man lokal ein LLM das Gutachten schreiben lassen kann, den LLM-Einsatz ausweist und das Ganze einmal prüft.

{ggskewboxplots}: Enhanced Boxplots for Skewed Data in #Rstats
https://arxiv.org/abs/2511.17091v1 #dataviz Never thought about it, but asymmetric whiskers are plausible to me when the distribution is skewed. Plus I didn't know there are so many variants of boxplots. Shouldn't be surprised, though, because there are usually many takes on the some issue in stats

Dealing with 1 color #dataviz
https://www.matplotlib-journey.com/module2/dealing-with-one-color Very useful intro to the different ways of picking a color. Didn't know so far what RGB stands for and how HEX codes work. Plus there is the HSV method combining color, saturation and brightness.

Defense Against Dishonest Charts
https://flowingdata.com/projects/dishonest-charts/ #dataviz A highly instructive interactive guide on how charts can be deceiving and what to watch out for to avoid misinterpretation.

More than 30% of this century’s science Nobel prizewinners immigrated
https://www.nature.com/articles/d41586-025-03247-6 Largest beneficiary is, so far, the United States. Absolute numbers are a bit misleading because US higher ed & research system is much larger than any other system. Everything else equal, which it isn't, absolute numbers should be higher for the US

Does a new generation of social scientists have to publish more to achieve less? (from 2019)
https://blogs.lse.ac.uk/impactofsocialsciences/2019/04/02/publishing-and-perishing-does-a-new-generation-of-social-scientists-have-to-publish-more-to-achieve-less/ Intuition says "yes", and so says data for sociology.