Schrödinger's Straight

If we don't want a continuous index but labels/categories that catch more of this heterogeneity the four categories from RoW make sense to me.

Though using democracy with adjectives might be useful for people if they want to communicate specific deficits of a diminshed subtype of a regime.

I guess the problem is that there's considerable variation in how liberal or contested regimes are. Plotting V-Dem's Polyarchy score against binary measures (BMR/DD) shows massive overlap. Same 0/1 labels "hide" vastly different democratic quality.

Electoral autocracy is probably the term you are looking for? They have de jure multiparty elections but those elections are not necessarily free and fair. @vdeminstitute.bsky.social's Regimes of the World measure codes countries in that way based on conditions of their continuous indices

Not a bad day to head into the office

Extract from the Charlemagne column in the Economist, reading: "more working means more money, and more money means added clout in global affairs. Until the 1960s Europeans toiled longer hours than Americans, and mattered somewhat more in the world. That is no coincidence"

And the freestyle causal inference top Play of the Week goes to the @economist.com Charlemagne piece on Europeans needing to work more

just like buying a new empty notebook, switching todo apps will *definitely* make me more productive

Yes, we had a lot of partial dependence plots initially! I'll update the code and produce some today.

Asked ChatGPT how to expand a data frame with a categorical variable and count.

Thought it was hallucinating when it suggested `uncount()`, but nope! It's a legit #tidyverse function. Reverse a count with grouped `uncount()`.

Love it!

If you want to study 'TV news' (for which ideally you should get the images), here are the transcripts from archive.org

github.com/notnews/arch...

Happy 25th anniversary to R 1.0.0! 📊👩‍💻⌨️🎉🍾

Congratulations! So well deserved!

10 maps of Germany for each month from January to October 2024. 268 points on each map represent the DWD weather stations active since 1961. The points are colored by the deviation of the monthly average temperature in 2024 from the long-term monthly average between 1961 and 1990. There’s hardly any station which had an average in any month that was below the reference period. Highlights: February 6.2 °C above the reference period; March 4.0 °C, August 3.4 °C.

Yesterday brought 18°C in Cologne (DE), which is 3°C above the previous record temperature for a 24th of November. This made me revisit a map I made last year.
Most of 2024 has been warmer than the long-term average for Germany from 1961-1990.

#30DayMapChallenge Day 25 Heat #rstats #ClimateChange

These are pictures from my commute to campus. We've hired a lot of junior faculty recently, you'll be part of a young, dynamic, and growing department!

Feel free to reach out if you have any questions!

Assistant Professor of American Politics The Department of Political Science at Colorado State University invites applications for an entry-level tenure-track Assistant Professor position in American politics with a specialization in politic...

Hey #polisky, we are hiring in American Politics at Colorado State!

Great colleagues, wonderful state, 350 miles of bike lanes, and over 300 days of sunshine. We have a great group studying AP, behavior and CSS more generally. There also is an HPC available!

jobs.colostate.edu/postings/148...

#polisky Friends, I'm collecting job market resources for social science undergrads. I want to include examples of jobs that students can have with a social science BA, an intro stats course, and R coding skills.

What are successful stories or job ads that I should include?

now tallying libraries used in replication files of **34** social science journals + support for python

github.com/recite/softv...

if stata mavens know how to tally imports in stata, let me know. (ssc install is not used frequently)

w/ @danweitzel.net

Are you interested in democratic backsliding?

With @honoratam.bsky.social and Joep van Lit we are organizing a workshop on "Early Signs of Democratic Recession in Established and Backsliding Democracies"

Join us at the @ecpr.bsky.social joint sessions

polisky

ecpr.eu/Events/Event...

Polisky, check out this super cool resource: demscore.se

"An innovative e-infrastructure providing free access to harmonized data from several of the world’s largest datasets on Democracy, Environment, Migration, Social Policy, Conflict, and Representation"

Oh, very cool. Thanks for sharing this!

New Polisky working paper here: github.com/soodoku/part...

GitHub - soodoku/partisan-gaps: How do (biased) guessing encouraging features and guessing agnostic ... How do (biased) guessing encouraging features and guessing agnostic coding techniques affect the partisan gap? - GitHub - soodoku/partisan-gaps: How do (biased) guessing encouraging features and ...

You can find the most recent version of our working paper on GitHub: github.com/soodoku/part...

The most up to date PDF is here: github.com/soodoku/part...

Our results paint a mixed picture of democratic competence. Smaller partisan gaps are partly a consequence of the fact that the average respondent doesn’t know the facts. It is mostly partisan guessing masquerading as partisan gaps. 6/7

This working paper sheds light on the source of the partisan gap as well. Our results suggest that the gap is not due to partisan cheerleading of the knowledgeable but by motivated guessing of the politically ignorant. 5/7

Removing these inflationary features from survey questions produces knowledge estimates that are more strongly associated with established predictors of political knowledge, such as interest, participation and education. 4/7

We find that common features of commercial polls like not asking don’t know, inserting a partisan cue, and treating unconfident answers as knowledge inflate the partisan gaps by up to 50%. 3/7