Kevin Griffith

Stats enthusiasts: I like to weave narratives using insights cross-tabbed from the latest poll

Stats experts: I only estimate conditional expectations and I keep a gun nearby so I can shoot them if they're from a distribution I don't recognize

‘I Felt Like a Factory Worker.’ Why VA Psychologists Are Burning Out

Register for our upcoming webinar with @joannespetz.bsky.social!

healtheconomics.org/events

Fused LASSO doesn't shrink coefficients towards zero; the penalties cause coefficients with similar magnitudes to move towards one another. So you end up with fewer groupings, but not too few your predictions suffer

Two reasons

1) Random effects require much stronger assumptions that are unlikely to be met

2) We aren't trying to shrink coefficients towards zero. We are trying to empirically bin coefficients with similar effect sizes in a way that maintains predictive accuracy and removes our subjectivity

Has anyone implemented fused LASSO in Stata to group or bin coefficients for dummy variables?

pmc.ncbi.nlm.nih.gov/articles/PMC...

One week from today!

I've had so many conversations with people who are worried about the direction of #healthequity research -- the lack of rigor, the lack of theory, and overstatement of findings.

Worse yet is the defiant attitude by some that any effort to push on nuance or causality undermines the project.

Advancing the Science and Scholarship of Health Equity Health equity research has advanced substantially during the past 2 decades and has emerged as a foundational element of population health science and scholarship.1 JAMA Health Forum has embraced thi...

"When researchers cannot specify and be transparent about what equity means in their work, it risks becoming an aesthetic rather than a commitment."

Sharp piece by Choi et al outling several issues with contemporary research that have increasingly troubled me.

jamanetwork.com/journals/jam...

Assuming our history isn't sanitized, this period will be remembered as one of national shame.

www.yahoo.com/news/article...

Five classes I took in university (that I realize most people could use right now):

Research methodology
Statistics
Logic
Epistomolgy
Abnormal psychology

How it started / How it's going

They're trying so hard to justify a murder. And every American should know: if they kill you, this is what they'll do to you.

Central bank independence in practice The Peterson Institute for International Economics held a high-level conference in October 2025 in Washington, DC, to analyze the serious issues concerning central bank independence in practice. The g...

Good time to brush up on what the experts say on the topic: Central bank independence in practice www.piie.com/publications...

45 minutes until markets open

Now would be a really great time for Congress to reassert itself as a co-equal branch of government

It's now multiple pages

You know you're in the south when the walk home from #NashvilleNYE is a cacophony of (hopefully) celebratory gunfire 🎉

Actual 2025 tariff rates are, to date, far less than statutory rates, but the costs of the tariffs are mostly borne by the US – estimated tariff pass-through to US prices in 2025 is 94%. Tariffs on imports act like a 1 to 2pp tax on US manufacturing. brentneiman.com/research/GN2...

THIS THIS THIS. ALL OF THIS

THIS is why faculty resist technological strategies for teaching. There is no engaging with Edtech without this context

Tell that to Vanderbilt University Medical Center, they've effectively barred researchers from using your product for this reason!

Dr Russell D. Moore is a theologian, pastor, and the editor in chief of Christianity Today.

thank you!

thank you!

I was expecting my Google scholar and was pleasantly surprised 😄

Looking for examples of published health policy studies with clear methodological flaws. Preferably from a med or health policy journal. Send me your favorites!

Apparently, Vanderbilt no longer allows the use of Rev for transcription because the data are sent outside the U.S.

Qual researchers, what services for human transcription (not AI) are you using and like?

MAGAZINER: How many veterans have you deported?

NOEM: We haven't deported veterans

MAGAZINER: We are now joined on Zoom by a combat veteran you deported to Korea

Taken together, our results demonstrate consistently high veteran satisfaction regardless of the credentials of their anesthesia providers. FIN