iiinova — where your digital self keeps evolving

Reposted by Joseph Sexton

Tom Olino @olinotom.bsky.social · 23/09/2026

I will be reviewing applications this admissions cycle (pending funding). This time, I'll be looking for folks interested in research within our training clinic. We have implemented ROM, and want to build on that infrastructure. Other research is possible; the primary goal is clinical research.

1119

Reposted by Joseph Sexton

Ashley Watts @ashleylwatts.bsky.social · 23/09/2026

🌟 Vanderbilt Psychology has an open rank position available in our clinical area! Job ad here: apply.interfolio.com/194049 (Note that this position is separate from our other psych department's asst prof search: apply.interfolio.com/194058) RTs appreciated!!

01818

Joseph Sexton @josephfsexton.bsky.social · 13/09/2026

"'My job is gone because of ChatGPT, and I was being invited to train the model to do the worst version of it imaginable,' she says. The idea depressed her. But her financial situation was increasingly dire [...] so she turned on her webcam and said 'hello' to Melvin [an AI interviewer]."

000

Joseph Sexton @josephfsexton.bsky.social · 24/08/2026

Definitely! Let me get my semester schedule under control over the next week, then I'll reach out to touch base. Could be the beginning of some Philly measurement nerd meetups...

110

Joseph Sexton @josephfsexton.bsky.social · 24/08/2026

Thanks Griffin! Would love to catch up some time, hope you're doing well

000

Joseph Sexton @josephfsexton.bsky.social · 21/08/2026

The paper, code, and supplements are all on OSF: osf.io/ybnup/. Would love to hear thoughts on next steps and future directions!

osf.io

OSF

020

Joseph Sexton @josephfsexton.bsky.social · 21/08/2026

In closing, we consider the consequences of underfactoring. As Thurstone put it in 1947: "Too many factors can do no harm, but too few factors are sure to cause trouble." Unidimensional scales permit simpler measurement and theory -- but it's for naught if they're invalid.

240

Joseph Sexton @josephfsexton.bsky.social · 21/08/2026

The problem is that ruling out one multidimensional model does not rule out multidimensionality altogether. EFAs revealed meaningful 2-factor alternatives with theoretical and diagnostic implications.

120

Joseph Sexton @josephfsexton.bsky.social · 21/08/2026

Issue 3: Failure to explore multidimensional alternatives. There's a relative dearth of exploratory work. For SUD, the (faulty) logic was something like: (a) an established 2-factor model was bad, (b) a 1-factor model seems fine given fit indices and loadings, ergo (c) SUD is unidimensional.

120

Joseph Sexton @josephfsexton.bsky.social · 21/08/2026

(This finding pairs especially well with @ashleylwatts.bsky.social’s recent paper in Nature Mental Health. bsky.app/profile/ashl...)

140

Joseph Sexton @josephfsexton.bsky.social · 21/08/2026

Across 11 SUDs, when we dropped asymptomatic individuals, loadings decreased markedly, and the first-to-second eigenvalue ratio (of focal interest for gauging dimensionality) dropped by a median of 60%.

120

Joseph Sexton @josephfsexton.bsky.social · 21/08/2026

Issue 2: Asymptomatic cases drive unidimensionality. The use of highly skewed data (also common in psychopathology) yields upwardly biased tetra/polychoric correlations. This inflates 1-factor loadings + the first eigenvalue. In practice, anything can be unidimensional if you add enough zeros.

140

Joseph Sexton @josephfsexton.bsky.social · 21/08/2026

Indeed, we find nearly perfect fit indices for 1-factor models of 11 SUDs. But when we inspect model fit more rigorously -- using dynamic fit index cutoffs and recently developed robust fit indices -- the 1-factor models actually turn out to be quite poor.

130

Joseph Sexton @josephfsexton.bsky.social · 21/08/2026

Issue 1: Conventional fit indices and cutoffs are overly lenient. With categorical data and WLSMV estimation (common in psychopathology), even egregiously false models produce fit indices that meet common cutoffs. It's trivially easy to find support for a 1-factor model.

130

Joseph Sexton @josephfsexton.bsky.social · 21/08/2026

With undergrad Abbe Knutsson and @ashleylwatts.bsky.social, I situate my argument around three central issues. We take DSM's substance use disorder criterion set as a case example. Data are drawn from NSDUH (US population-based sample, N = 104k+).

120

Joseph Sexton @josephfsexton.bsky.social · 21/08/2026

📝 Excited to share an accepted first-authored paper at Clinical Psych Science! Why are so many mental disorders treated as unidimensional? We show support for 1-factor models is often just an artifact of conventional psychometric methods -- which weren't designed with psychopathology in mind.

4397

Joseph Sexton @josephfsexton.bsky.social · 25/07/2026

Saying this w your full chest to a **neuropsychologist** is so extremely embarrassing

000

Joseph Sexton @josephfsexton.bsky.social · 25/07/2026

Have they spoken to a psychologist ever

100

Reposted by Joseph Sexton

Ashley Watts @ashleylwatts.bsky.social · 09/07/2026

Excited to share a paper recently accepted at Nature Mental Health! 😄 For decades, we've assumed that covariance among psychopathology informs comorbidity. What if we’ve been wrong?

913461

Joseph Sexton @josephfsexton.bsky.social · 04/07/2026

When you share a scathing critique of current psychology and someone sends you a paper by Paul Meehl

090

Joseph Sexton @josephfsexton.bsky.social · 11/06/2026

The concept of a bisexual lighting mega Gaudi drone show at the Sagrada Família

Screenshot of Twitter post showing the new topping for the Sagrada Familia + a drone show in the background of Gaudi, drones shining blue and purple and pink

020

Joseph Sexton @josephfsexton.bsky.social · 27/05/2026

I’m not a particularly fashionable person, but nowhere do I feel so dismally unfashionable as I do in Copenhagen

010

Joseph Sexton @josephfsexton.bsky.social · 26/05/2026

I show how the seeming unidimensionality of SUD (and likely many other psychiatric constructs) is in fact driven by asymptomatic respondents -- individuals that experience none of the SUD criteria, but drive up the correlations between items. Let's talk about it :)

110

Joseph Sexton @josephfsexton.bsky.social · 26/05/2026

Very excited for #aps26bcn... the schedule is stacked! Any factor analysis nerds or skeptics wanna meet up at my poster? 🗣️ "Accounting for Zero Inflation Worsens the Structural Validity of the DSM-5 Substance Use Disorder Criterion Set" (w/ @ashleylwatts.bsky.social) 📍 Poster 122, 8-9pm Friday

250

Joseph Sexton @josephfsexton.bsky.social · 26/05/2026

Now also seems like a perfect time given -- in clinical science, anyways -- growing comfort with formal theoretical and nonlinear approaches

110

Joseph Sexton @josephfsexton.bsky.social · 26/05/2026

Last fall in a rage against polychoric correlations I fell upon several damning papers by Foldnes and Grønneburg (beginning ~2020) which exposed me to copula approaches and their importance. They've been in my mind since then, and I'd love to spend more time on them moving forward

110

Joseph Sexton @josephfsexton.bsky.social · 26/05/2026

This is a great resource that I hadn't paid enough attention to earlier, thanks for sharing!

000

Joseph Sexton @josephfsexton.bsky.social · 25/05/2026

Unfamiliar w utility of pseudorandom numbers but have been desperate for a better tutorial for copulas + have a strong hunch they'll be in vogue once someone translates those extremely dense psychometrika papers

210

Joseph Sexton @josephfsexton.bsky.social · 11/05/2026

My favorite thing about Love’s truck stops is their consistently niche selection of Pokémon plushies. Zero recent starters but ten Togetics from Pokémon Gold and Silver (1999)

040

Reposted by Joseph Sexton

Simon Hammann @simonhammann.bsky.social · 20/04/2026

Don't be shy to take on a little two-week side project. These five months will be the most precious three years of your academic journey.

191958537

Joseph Sexton @josephfsexton.bsky.social · 07/04/2026

I've long felt this observation is very understudied given how extremely inconvenient it is/was. as I recall it, 2020 was filled with assumptions that the suicide rate would skyrocket, and 2021 was filled with questionable post-hoc explanations

010

Joseph Sexton @josephfsexton.bsky.social · 07/04/2026

What are good papers squaring high suicidal ideation with low suicide rates during COVID?

110

Joseph Sexton @josephfsexton.bsky.social · 02/04/2026

all told, I'm opting to just not think about it yet ❤️😭

110

Joseph Sexton @josephfsexton.bsky.social · 02/04/2026

I seriously, seriously lucked out by having such amazing mentors over the years. Among many others, hugeeee thank you to @ashleylwatts.bsky.social, @kirstyclarkphd.bsky.social, and bsky-less Ken MacLeish, Steve Hollon, and Hao Wu. Very excited for (I hope) continued collaborations :)

030

Joseph Sexton @josephfsexton.bsky.social · 02/04/2026

Very stoked to share that I'm joining Temple University's Clinical Psych PhD program this fall! I'll be doing critically minded and methods-heavy work on measurement, depression, and suicide in adolescents with @olinotom.bsky.social. Feeling super grateful! If you're in Philly, say hi :)

11313

Reposted by Joseph Sexton

Nashville Banner @nashvillebanner.bsky.social · 31/03/2026

If Metro Nashville wants to build dedicated transit lanes on state-controlled corridors, it needs the sign-off of the TN General Assembly. This week, the legislature is nearing approval of the city’s plans for dedicated lanes downtown and along Gallatin and Nolensville pikes. @stelliott.bsky.social

nashvillebanner.com

Dedicated transit lanes in Nashville nearing state approval - Nashville Banner

The Tennessee General Assembly is close to approving dedicated transit lanes in Nashville, paving the way for transportation improvements.

083

Joseph Sexton @josephfsexton.bsky.social · 31/03/2026

“how precarious is research funding in the US these days?“ well there’s a subreddit where students are making ggplots to predict whether GRFP comes out on a Tuesday

150

Joseph Sexton @josephfsexton.bsky.social · 24/03/2026

great work, Philipp, and congrats on this milestone!

110

Joseph Sexton @josephfsexton.bsky.social · 19/03/2026

I'm telling you right now, do not trust that cutoff CFI > .95. he is not worth your time. he will give you trust issues and then some

010

Joseph Sexton @josephfsexton.bsky.social · 27/02/2026

NSF updated the GRFP website with FAQs, mostly to do with outcry about (many) applications being returned without review. "The number of applications RWR has not changed substantially" they say. Translated: the proportion RWR likely ~doubled.

030