Jonathan Brauer
@RealJonBrauer
Associate professor at Indiana University. Study social influences on behavior & wellbeing in U.S. & abroad. Views are my own.
Check out our website! @Jake_Day4 and I created a site to publicly share course materials and musings about scientific theory and methods. reluctantcriminologists.com
#Statistics thought of the day: hbiostat.org/bbr/htest.html…
IMO this post is spot-on. Open science? First & foremost, let's make science comprehensible. As a PhD advisor, this quote hits home: "...what good is it to train students to do better science when better science won’t be rewarded? Are we just setting them up to exit academia?"
How can we reform science? I have some ideas. But I am not sure you’ll like them, because they don’t promise much. elevanth.org/blog/2025/07/0…
Y’all, I wrote a book! The Science of Second Chances is based on my decades of work studying crime and criminal behavior, first as an economics professor and now as a policy expert at Arnold Ventures. It shares what we know from cutting-edge research about how to intervene at…
Project codename: SKYNET
OpenAI is officially a military contractor. The generative AI powerhouse joins an army of Big Tech players making big bets on the Department of Defense. And one imagines that OpenAI's newly announced $200 million contract with the Pentagon is just the beginning...
New Content Alert: In a new blog post, Jon & Jake offer a pluralist response to Andrew Wheeler’s call to “build stuff” and discuss their recent preprints, in which they suggest how to move beyond Monopoly money publications to genuine scientific currency reluctantcriminologists.com/blog-posts/%5B…
Science funder explains why they are “no longer funding journal publications.” asterainstitute.substack.com/p/scientific-p…
We’re hiring! 🥳 The Criminal Justice team at @Arnold_Ventures is searching for a Policy Manager to join our team. Location: Houston, DC, or NYC Salary range: $100-130k Apply ASAP if you’re interested - we move quickly! Please spread the word. 🙏
#statstab #362 When does a significant p-value indicate a true effect? Understanding the Positive Predictive Value (PPV) of a p-value Thoughts: Cool illustration of how assumptions impact our beliefs about the next study #shiny #education #ppv #pvalue shinyapps.org/showapp.php?ap…?
E.g., We model ordinal Likert-type data using linear regression bc "conventional" & to "ease" interpretation. More appropriate alt models generated "identical" results (i.e., beta sign & stat significant). Trust us - a mean diff of 3.7 vs. 3.3 on non-metric measure is meaningful!
There are individual-group conflicts in all contexts. But academia is the one I personally experience in which people routinely justify doing the wrong thing because the right thing is hard or personally costly. We do the right thing because it is right, not because it is easy!
All journals should adopt this policy. Strict formatting requirements on initial submission are silly.
Note, however, that "explanation" also rejects CAUSAL prediction (ATE) as a proxy. The firing squad exemplifies an event whose ATE is close to one, yet its explanatory power (i.e., PN) is zero. Formally, ATE=P(y_x - y_x'), PN=P(y'_x'| x,y)
Predictive models guess outcomes based on observed associations, while causal explanations state what would have happened if a specific factor had been altered. A statistical model can excel at exploiting spurious correlations that have no causal connection to the outcome. A…