iiinova — where your digital self keeps evolving

Tom Carpenter, PhD @tcarpenter.bsky.social · 28/12/2025

“Instead of” there is doing a lot of work If I were still teaching, I would be talking about how to use AI to make us better writers… E.g. giving us feedback or critique… But the mental muscles I went to train have to to be the students’

250

Tom Carpenter, PhD @tcarpenter.bsky.social · 06/12/2025

Ask copilot or chat gpt. It can output a csv no sweat.

060

Tom Carpenter, PhD @tcarpenter.bsky.social · 28/11/2025

Possibly. We might also admit some things are not very knowable— without other “ways of knowing”

070

Tom Carpenter, PhD @tcarpenter.bsky.social · 28/11/2025

I often hear researchers feeling bad about this. But it’s called prioritization. The key is to do it intentionally and strategically. You have finite time and can’t do everything.

191

Tom Carpenter, PhD @tcarpenter.bsky.social · 27/11/2025

On the one hand, people need better education about AI On the other hand, do we even understand what it means to be conscious in the first place?

150

Tom Carpenter, PhD @tcarpenter.bsky.social · 19/11/2025

Ah, but read what I said. I’m not saying it’s impossible to do valid online research. I am saying that volume will need to shrink. It’s going to be harder to stay on top of this. Some folks will give up or mess it up. There will be an arms race. Reviewers will start raising flags. Etc.

100

Tom Carpenter, PhD @tcarpenter.bsky.social · 19/11/2025

The volume of social psych papers that ran on prolific and mturk is going to need to shrink soon, and that will have a big impact on a lot of small labs

162

Tom Carpenter, PhD @tcarpenter.bsky.social · 14/11/2025

A huge pain, but doable with undergrad participants essentially for free (or at least historically so). I’m not saying what is right… I’m saying what is driving behavior

020

Tom Carpenter, PhD @tcarpenter.bsky.social · 13/11/2025

If we want this to change, then we need to make it feasible TMP factor: Time, money, pain Whether a study gets done depends a lot on how difficult it is. The question is one of ROI. I suspect many researchers are thinking to themselves, “that’s a great thing, but it’s not something I can do”

290

Reposted by Tom Carpenter, PhD

Jess Calarco @jessicacalarco.com · 01/11/2025

We have progressed from data collection to data analysis.

My 11-year-old sitting with her pile of Halloween candy, sorting it into a bar graph

970344664087

Reposted by Tom Carpenter, PhD

Menotti Minutillo @menotti.bsky.social · 30/10/2025

ladies and gentlemen...we got him

171181784003

Reposted by Tom Carpenter, PhD

Greg Pak @gregpak.net · 10/10/2025

* hoppening

121000146

Tom Carpenter, PhD @tcarpenter.bsky.social · 04/09/2025

But where does 1a say anything about personhood? I’m reading it and it seems clear that you can’t restrict speech—and that would obviously apply to one or more people organized under a LLC. IMO the bigger question is “when is money free speech, vs when is it corruption”

210

Reposted by Tom Carpenter, PhD

Cameron Patrick @cameronpat.bsky.social · 18/08/2025

this slide is from a colleague's introductory stats course, I think it fits many statisticians' experiences

Slide titled: "Assumptions of the model and model checking"
with a scatterplot with axes how much people should worry vs how much people do worry.

89224

Reposted by Tom Carpenter, PhD

Adam L @adam-lg.bsky.social · 28/07/2025

45. Academia doesn't reward building useful tools nearly as much as it should

2344

Reposted by Tom Carpenter, PhD

Adam L @adam-lg.bsky.social · 27/07/2025

14. We mostly evaluate latent variable models with the equivalent of Rorschach tests

1111

Reposted by Tom Carpenter, PhD

Adam L @adam-lg.bsky.social · 27/07/2025

5. You should use a precision-recall curve for a binary classifier, not an ROC curve

1232

Reposted by Tom Carpenter, PhD

JLRay @jlray.bsky.social · 25/07/2025

Wow. Scientists have edited mosquito DNA to prevent the spread of malaria to humans "while supporting essential physiological functions... and negligible fitness costs" to the mosquito population. Potentially ending the mosquito-born spread of malaria to humans. www.nature.com/articles/s41...

361080310

Tom Carpenter, PhD @tcarpenter.bsky.social · 23/07/2025

… set of paths consistently supported by the data. Even getting that down is a trick. And making sense of it is fraught and doesn’t get you much further than one would get from regression. But at least then we would have some confidence we understand the correlational relationships!

000

Tom Carpenter, PhD @tcarpenter.bsky.social · 23/07/2025

… the model is correct and then gives you what the path would be under that specification. There’s nothing different when we go to SEM other than your ability to p-hack goes up exponentially. IMO this would be a great place to use machine learning approaches to train / tune models to find …

100

Tom Carpenter, PhD @tcarpenter.bsky.social · 23/07/2025

… all those hypotheses together (in the same way that ANOVA contest many multiple comparisons at once). There’s nothing different between this and running a bunch of regressions and claiming the results support the way you specified those models. In reality, it’s the reverse. Regression assumes …

100

Tom Carpenter, PhD @tcarpenter.bsky.social · 23/07/2025

Yes and see this a lot in social too. Proper use of SEM implies a particular philosophy of hypothesis testing in regression contexts. An omitted path is hypothesizing that path is exactly 0. A non-omitted path hypothesizing it is non-zero. Model fit is effectively the joint set of …

100

Tom Carpenter, PhD @tcarpenter.bsky.social · 23/07/2025

… SEM for causal discovery. However, if you have a good read on the causal process, it can be great for estimating parameters such as factor, loadings or paths with latent variables

050

Tom Carpenter, PhD @tcarpenter.bsky.social · 23/07/2025

This is probably not anything you don’t already know …. But I did a lot of SEM work and will repeat it anyway. The model assumes you know the causal structure. Fit indices will confirm that the model is a fit to the data, but many incorrect models can fit the data. So I would not use …

291

Tom Carpenter, PhD @tcarpenter.bsky.social · 19/07/2025

Curious how this compares to the cost of living per state

020

Reposted by Tom Carpenter, PhD

Julia M. Rohrer @dingdingpeng.the100.ci · 08/07/2025

Let's say you want to include age as a predictor in your model. How do you do that? Here's an illustration of three options -- it's for a paper I'm working on (so if you feel like anything could be tweaked...).

Plot that depicts the average importance people in my data assign to their friendships (y-axis, on a scale from 1 to 5, depicted with 95% confidence intervals) by their age (x-axis, from 18 to 60).

Depicted are 3 different ways to model importance of friends as a function of age.
Using age as a linear predictor: this imposes a linear trajectory which comes with very tight confidence intervals (i.e., uncertainty is low).
Using age as a categorical predictor: this imposes no trajectory whatsoever but instead simply reproduces the means by age. The confidence intervals are very wide, in particular for those ages not well represented in the data (i.e., uncertainty is high).
Age splines: This results in a smooth trajectory that follows some of the bumps in the data, but not all of them. The confidence intervals are somewhere between the linear and the categorical case (i.e., uncertainty is medium)

3215629

Reposted by Tom Carpenter, PhD

respectful huff @alexqarbuckle.bsky.social · 05/07/2025

There should be a corner at Home Depot where a guy with a table saw will slice you off custom lengths of hot dog from an infinite hot dog coming out of the wall

844101610

Reposted by Tom Carpenter, PhD

Ian Carlos @ianxcarlos.bsky.social · 25/06/2025

There were two girls at Wawa just now talking about funny movies and one said, “Have you ever seen the movie Office Space? It’s an old people movie but it’s funny”

4076802365

Tom Carpenter, PhD @tcarpenter.bsky.social · 24/06/2025

Counterpoint: the ability to chat with an article or literature and find patterns in our own work that perhaps we missed I think has a lot of potential to augment our scientific work

100

Reposted by Tom Carpenter, PhD

Emilio Ferrara @emilioferrara.bsky.social · 23/06/2025

🤖Thrilled to share our latest work☄️ Have you ever wondered what LLMs know but they are not saying? We built an auditing framework to study information suppression in LLMs, and demonstrated it to quantify and characterize censorship in DeepSeek. Read more: arxiv.org/abs/2506.12349

arxiv.org

Information Suppression in Large Language Models: Auditing, Quantifying, and Characterizing Censorship in DeepSeek

This study examines information suppression mechanisms in DeepSeek, an open-source large language model (LLM) developed in China. We propose an auditing framework and use it to analyze the model's res...

172

Reposted by Tom Carpenter, PhD

Ida Momennejad @neuroai.bsky.social · 20/06/2025

Pleased to share our ICML Spotlight with @eberleoliver.bsky.social, Thomas McGee, Hamza Giaffar, @taylorwwebb.bsky.social. Position: We Need An Algorithmic Understanding of Generative AI What algorithms do LLMs actually learn and use to solve problems?🧵1/n openreview.net/forum?id=eax...

316537

Tom Carpenter, PhD @tcarpenter.bsky.social · 18/06/2025

building intuition around systems of matrix data and how we manipulate them should be right after (or right before) basic calc (integrals, derivatives, partial derivatives)

210

Tom Carpenter, PhD @tcarpenter.bsky.social · 18/06/2025

Someone please explain why linear algebra isn’t taught more in high school in the United States? Seems like maybe if you’re lucky you get a few lectures on matrices and that’s it.

260

Reposted by Tom Carpenter, PhD

John Schwartz @jswatz.bsky.social · 08/06/2025

Under the Trump agenda, energy will cost more. And when energy costs more, everything costs more. www.nytimes.com/2025/06/04/c...

nytimes.com

Electricity Prices Are Surging. The G.O.P. Megabill Could Push Them Higher.

12312

Tom Carpenter, PhD @tcarpenter.bsky.social · 05/06/2025

Many academics are taught “do more” instead of “prioritize”. They are taught the academic superhero myth, that “truly great/smart” scholars can handle it. So they sabotage their own success in service to the cult of personality

072

Tom Carpenter, PhD @tcarpenter.bsky.social · 05/06/2025

Galaxy brain: what does anything measure?

110

Tom Carpenter, PhD @tcarpenter.bsky.social · 04/06/2025

I love this technique because it gives a cool way to isolate change and stability components of within-person measurement using latent variables. The stable portion of IATs is far more predictive of other individual-difference measures than one would think given traditional scoring/analyses

271

Tom Carpenter, PhD @tcarpenter.bsky.social · 04/06/2025

I love this analysis technique because it gives a cool way to isolate change and stability components of within-person measurement using latent variables. The stable portion of IATs is far more predictive of other individual-difference measures than one would think given traditional scoring/anakysus

031

Tom Carpenter, PhD @tcarpenter.bsky.social · 04/06/2025

That moment when you had completely forgotten about a cool data app you’d made, only to find others posting about it ;) Thanks @calvinklai.bsky.social!

061

Reposted by Tom Carpenter, PhD

Calvin Lai @calvinklai.bsky.social · 04/06/2025

Paper w/ @tcarpenter.bsky.social & @alexgoedderz.bsky.social at PSPB!🚨 The standard IAT is only 5 min long. We found that making the IAT longer by taking it multiple times greatly improves predictive validity. 🧵below, with practical advice about how to run IATs! LINK: osf.io/preprints/ps...

45016

Reposted by Tom Carpenter, PhD

Calvin Lai @calvinklai.bsky.social · 04/06/2025

BOTTOM LINE: Aggregating 2 IATs together and using CFA can lead to much higher predictive validity. Finally, I wanted to thank Tom and Alex! This paper came out while they were moving out of academia & I was going on parental leave, so we hadn't publicized it. Better late than never! 😉

051

Reposted by Tom Carpenter, PhD

Calvin Lai @calvinklai.bsky.social · 04/06/2025

FAQ Q1: 4 IATs is a lot!? A: You only need 2 IATs to reap the benefits of aggregating. Q2: It's a lot of work to set up these models. A: Here's a shiny app to make it easy: tcarpenter.shinyapps.io/trait_iat/ Q3: I don't want to learn CFA. A: Averaging IATs together gets you some of the benefits.

151

Tom Carpenter, PhD @tcarpenter.bsky.social · 01/06/2025

What should universities look like 10 to 15 years from now? Lots of intersecting headwinds… dramatic change is likely. Biggest IMO is that teaching universities become cheap again by reverting to “room with chairs and a prof”. Think community college, but with amazing faculty.

010