doinkboy
banner
doinkboy.bsky.social
doinkboy
@doinkboy.bsky.social
Social Science Data Analysis (especially Dyadic, Mediation, SEM; davidakenny.net) and My Twisted Take on Political & Cultural Topics; Reposts Are Not Necessarily Endorsements
Well deserved.
November 24, 2025 at 1:08 PM
One of my all-time favorite YouTube videos. Never fully appreciated Prince before viewing it.
November 23, 2025 at 2:01 AM
Reposted by doinkboy
November 18, 2025 at 12:20 PM
Reposted by doinkboy
* We should not rely on meta-analysis or Cochrane to take out bad papers, publishers should do this - do better screening before they get published.
* Some Cochrane reviews still contain works from authors in the top 10 of the Retraction Watch Leaderboard.
#IRICSydney
November 17, 2025 at 1:32 AM
Reposted by doinkboy
IRT often feels like one hot fever dream where everybody tries to hide the fact that it's just a nonlinear factor model. I had friends who weren't trained in it who essentially picked up that it had magic properties that factor models don't.
November 8, 2025 at 5:40 AM
Shouldn't people be impressed that you got the effect with an N of 34? Is it not a waste of resources, including participant resouces, to have unnecessarily large sample sizes?
November 7, 2025 at 4:26 PM
Would be informative to have a quantitative measure of agreement for I and Q, such as the intraclass correlation.
November 6, 2025 at 12:11 PM
My daughter loved the Cranberries and once made a Dolores O'Riordan doll and threw it onstage when she was performing. When You're Gone is one of my all time favorite sings. O'Riordan passing was a very sad day.
November 4, 2025 at 12:57 PM
p smacking?
November 4, 2025 at 12:32 PM