Aidan Wright
banner
aidangcw.bsky.social
Aidan Wright
@aidangcw.bsky.social
Psychopathology | Personality | Quant Methods

Professor of Psychology and Psychiatry | Eisenberg Family Depression Center | University of Michigan

Editor | Journal of Psychopathology and Clinical Science

Founder and Instructor | www.smart-workshops.com
At a *cheeky* stationary/novelty store and it seems like they’ve got stuff just for me
October 26, 2025 at 6:22 PM
Hey @dieworkwear.bsky.social , care to comment?
October 24, 2025 at 3:51 PM
Just noticed the classroom I teach in has this
October 8, 2025 at 3:56 PM
Spotted in Ann Arbor
October 1, 2025 at 11:00 PM
I forgot you’re an Arby’s guy
September 21, 2025 at 12:12 AM
Why whatever I owe you is late. She really hates when I work…
September 5, 2025 at 7:53 PM
I terminated my @zenbusiness.bsky.social account today, which they make difficult to do. I expressed my frustration during the process.

Imagine my surprise when I got an email saying I had written a positive review about them on trustpilot.com. I did not write that. This is fraud. Stay away.
August 26, 2025 at 8:41 PM
How I wish I wrote abstracts vs how I write abstracts…
August 21, 2025 at 3:25 AM
Cocoa does not want me to prep for tomorrow…
August 12, 2025 at 11:11 PM
The timing on this post from Twitter…
August 3, 2025 at 1:08 AM
This is peak lakehouse
August 3, 2025 at 12:12 AM
I don't have a great sense for the exact minimum. No doubt it will depend on eliciting prompt, construct, and model. But we do see that absolute error is modestly associated with length and at lower number of diaries/minutes (2 samples) we see things stabilize at some point.
July 27, 2025 at 4:50 PM
We compared them to other natural language processing methods, like LIWC and bespoke personality machine learning models, and they excelled compared to those, too.
July 26, 2025 at 8:02 PM
We also examined whether the LLM scored traits could predict "non global self-reported things," and they could. With important outcomes, like diagnoses and being in psychiatric treatment.

They don't beat self-report, except for treatment, which is the least contaminated by self-report method var.
July 26, 2025 at 8:02 PM
We then correlated both self and LLM scored traits with a whole bunch of other personality scales and looked at the profile correlations to see if the LLM traits had similar nomological nets.

They did. See this table. Note that LLM profiles are cross-method, and self report are mono method.
July 26, 2025 at 8:02 PM
We used two datasets, one with 30min of spontaneous through narratives and one with ~2 weeks of brief daily video diaries.

We had a bunch of LLMs score them for the big-5 then correlated with participant self-reported traits. This figure has our main results.
July 26, 2025 at 8:02 PM
Who says post publication peer review can't work. The first post pub review is in on our paper. Thanks, Joann!
July 18, 2025 at 2:51 AM
I'll be asking Joann to review the next paper I edit. Succinct. I like that.
July 18, 2025 at 2:07 AM
You into puppy feet? She recently had her nails done 💅 (the 9yo took my joking suggestion seriously when the babysitter came over)
July 9, 2025 at 12:49 AM
July 9, 2025 at 12:39 AM
In theory this gets to just regular fetch, but I’ve not seen it with mine
July 8, 2025 at 11:56 PM
Huge line for free smoothies this AM, but I must be getting old. My only thought was “I’d rather pay than take the time to wait”
June 28, 2025 at 12:52 PM
Lab road trip to Association for Research in Personality conference is off!
June 26, 2025 at 1:51 PM
Tell me you’re out of touch with how faculty work without telling me you’re out of touch with how faculty work
June 6, 2025 at 12:09 AM
Got this from a conference today. I don’t think that’s what the meant to say…
June 4, 2025 at 1:24 PM