Models as Prediction Machines: How to Convert Confusing Coefficients Into Clear Quantities - Julia M. Rohrer, Vincent Arel-Bundock, 2026 Psychological researchers usually make sense of regression models by interpreting coefficient estimates directly. This works well enough for simple linear model...

Good news everyone 🥳 Our (w @vincentab.bsky.social) primer on models as prediction machines (with the marginaleffects package) is finally officially published!>

journals.sagepub.com/doi/10.1177/...

Nick Offerman Visits Criterion Closet To Rebuild Shelves NEW YORK—In an eight-hour-long video posted Tuesday to the Criterion Collection’s YouTube channel, actor Nick Offerman can be seen paying a visit to the Criterion Closet to rebuild the film library’s ...

“Don’t tell me you’ve been using particle board in here,” said the 55-year-old woodworker

These are essential questions to be asking right now. Ignoring the power of these models (which are only getting smarter by the month) is a recipe for professional obsolescence for one's self and their students.

Excited to share that @russpoldrack.org will be speaking at @princeton.edu this Tuesday, March 24!

Talk: "The Promise and Perils of AI-Assisted Coding in Science" 12:00 PM ET

Workshop: "Testing and Validating Code Developed Using AI" 11:00 AM ET

The Zoom is open to all (see flyer for details)!

[134] Figuring Out Figure 1 - Data Colada A few years ago our Journal Club discussed an interesting methods paper entitled, “Putting Psychology to the Test: Rethinking Model Evaluation Through Benchmarking and Prediction” (.htm). This post describes my attempt to understand what’s happening in Figure 1 of that paper, which shows that extremely simple experiments can generate extremely negative R2s. I learned a...

New post on Data Colada. I spent a very long time trying to figure out a figure in a methods paper that is not in my area of expertise. It shows extremely *negative* R-squared values from replications of a true effect. The question on my mind was: WHAT?? I learned some stuff. datacolada.org/134

Data Organization in Spreadsheets Karl W. Broman & Kara H. Woo Pages 2-10 | Received 01 Jun 2017, Accepted author version posted online: 29 Sep 2017, Published online: 24 Apr 2018 1. Introduction 2. Be Consistent 3. Choose Good Names for Things 4. Write Dates as YYYY-MM-DD 5. No Empty Cells 6. Put Just One Thing in a Cell 7. Make it a Rectangle 8. Create a Data Dictionary 9. No Calculations in the Raw Data Files 10. Do Not Use Font Color or Highlighting as Data 11. Make Backups 12. Use Data Validation to Avoid Errors 13. Save the Data in Plain Text Files ABSTRACT Spreadsheets are widely used software tools for data entry, storage, analysis, and visualization. Focusing on the data entry and storage aspects, this article offers practical recommendations for organizing spreadsheet data to reduce errors and ease later analyses. The basic principles are: be consistent, write dates like YYYY-MM-DD, do not leave any cells empty, put just one thing in a cell, organize the data as a single rectangle (with subjects as rows and variables as columns, and with a single header row), create a data dictionary, do not include calculations in the raw data files, do not use font color or highlighting as data, choose good names for things, make backups, use data validation to avoid data entry errors, and save the data in plain text files.

Every day is a good day for sharing one of the most useful papers about research data ever written. PLEASE get your people to understand and follow this advice.

www.tandfonline.com/doi/full/10....

A new Department of Cognitive Science is being created at Bocconi University in Milan, Italy.

Here is the call for a cluster hire (for around 10 faculty) in all areas of cognitive science, at both junior and senior levels:

www.unibocconi.it/en/faculty-a...

Deadline: May 4th, 2026

[133] Heterofriendly: The Intuition for Why You Always Need Robust Standard Errors - Data Colada When I taught my first PhD-level methods course, I invited students to submit questions about any topic in statistics or methodology. Six out of 10 students asked about the same topic: robust & cluste...

Answering the most popular question in a PhD methods course
datacolada.org/133

Willie Colón, salsa pioneer, has died at 75 The South Bronx bandleader took the Latin genre to new heights while recording for Fania Records.

The South Bronx bandleader took the Latin genre to new heights while recording for Fania Records. n.pr/4kLTCKV

late to the party, but congrats, Felix!

PIano being dropped on car in car testing facility

Would p-curve work if you dropped a piano on it?
datacolada.org/129

Completely agree. Averages can hide weird patterns. We make a related point about stimuli in psych experiments and propose a visualization here: psycnet.apa.org/record/2026-... (or datacolada.org/126 for a summary)

Colada[128] The Best Audit Study and its interesting shortcoming
datacolada.org/128

The Effect: An Introduction to Research Design and Causality The Effect: An Introduction to Research Design and Causality, Second edition is an excellent teaching text about research design, specifically concerning research that uses observational data to make ...

The second edition of The Effect has been delayed a bit... because there were so many orders they had to switch to a bigger print run! Thank you everyone for your support, and check out the second edition here: www.routledge.com/The-Effect-A...

Thank you!!

Completely agree!

In those cases, it can feel like the kind thing to do is to either not mention them at all or to soften them so much that they become obscured. And that’s where I think there is an inevitable tension between kindness and truth-seeking
…

I largely agree with this, and I think it’s something that some critically minded people tend to underweigh. The biggest challenge in practice, I think, is that certain criticisms-no matter how carefully you frame them-will still be perceived as unkind
…

Don't assume, plot
datacolada.org/126

Same (except with base r). I find it quite enjoyable, and it has made me more fastidious with figures.

It could work. I'm particularly interested in confounds and causal identification, because so called 'conservative confounds' could create other unintended problems. But my concern might apply to this as well.

Interesting, I was thinking about this in the context of causal identification rather than statistical significance. I was curious about counterarguments like "what seems like a 'conservative confound' that goes against the observed effect may actually change how the other mechanisms work."

thanks!

Does anybody know of a paper or blog discussing arguments of the form "we found our results despite this confound that works against our effect"? Essentially, how to think about confounds that go against the hypothesized or observed effect.

Thinking Clearly with Data An engaging introduction to data science that emphasizes critical thinking over statistical techniques

Parts of this book would probably work: press.princeton.edu/books/paperb...

Really proud of this new work out @psychscience.bsky.social. Led by the amazing but bluesky-less Amanda Geiser and with @deborahsmall.bsky.social.

We show that when comparing moral wrongs, people are (much) more willing to “scale up” than to “scale down” condemnation and punishment…

Measurement error, Nobel Prize winning research, friendly disagreements, and more....
datacolada.org/124

🏆 Research integrity consultant and image forensics expert @elisabethbik.bsky.social has uncovered fraudulent data in over 7,600 scientific papers and exposed the practices of ‘paper mills’ that produce counterfeit scientific articles. She is honoured with the €200K Individual Award. Congrats!

Roy Ayers, whose 'Everybody Loves The Sunshine' charmed generations, dies at 84 The vibraphonist, composer and jazz-funk pioneer helped inspire the neo-soul movement, and his best-known song was sampled over 100 times.

The vibraphonist, composer and jazz-funk pioneer helped inspire the neo-soul movement, and his best-known song was sampled over 100 times.