Roy Levy
@RoyLevy11
Followers
224
Following
286
Statuses
57
Quantitative methodologist & psychometrician, professor at Arizona State University, https://t.co/jARQr06p6F
Joined February 2018
RT @dmcneish18: For anyone who might be interested in applying to PhD programs in quant methods this year, nearly all of the faculty in the…
0
40
0
@shanetutwiler @dmcneish18 Have you hacked into my to-do list? 🙂 Plans for extending our work to approach to situations with discrete latent variables are in the works!
0
0
2
@dmcneish18 @JonathanDowlin Seconding this. If you wish to proceed as a frequentist, I'd suggest Rosseel's Structural-after-Measurement (SAM) approach, rather than modifying our (Bayesian) MUPPET approach. Early evidence suggests they perform similarly along certain criteria
1
0
2
@quantitudepod Teaching this topic now, so I can’t resist commenting with one of my favorite quotes (Jackman, 2009, p. xxviii) “Adopting an explicitly Bayesian approach would resolve a recurring source of confusion for these researchers, letting them say what they mean and mean what they say.”
1
0
4
RT @EDMSatUMD: Bayesian Statistical Modeling, a 3-day online First Course followed by a 2-day online Second Course. With Roy Levy (@RoyLev…
0
9
0
@lishinskialex @quantitudepod They think their shots at me go unnoticed, but I'm playing the long game of persuasion. It started with an interview, and now they frequent(ist)ly (and accurately!) impute what I would say. We're well on our way to Season 12 episode 5: "Roy was Right All Along"
1
0
4
RT @kareem_carr: As a statistician, it is extremely frustrating to me to see an account called “World of Statistics” with over 1.5M followe…
0
5K
0
RT @WomenMeasure: We are now accepting applications for the 2023 Women in Measurement ResearcHER program! This program is designed to suppo…
0
6
0
RT @dmcneish18: Dynamic fit index cutoffs now support 1) categorical data 2) non-normal continuous data 3) missing data 4) second-order m…
0
41
0
The interplay among scoring, reliability, validity, and modeling vexes many applied researchers and methodologists alike. A new paper aims to bring new insights (and some old wisdom) to further our understanding.
New paper out now -- This is my response to the recent comment on the sum score paper co-authored by @MelissaGWolf . (short thread)
0
2
14
@DerekCBriggs Those who ignore measurement are doomed to rediscover it, be it in innocent, comical, or tragic ways.
0
0
2
@quantitudepod Me: [listening to this episode while washing dishes] 10 yr old: What are you listening to? Me: A podcast about stats [getting excited talk to my kid about stats] 10 yr old: There sure are a lot of beeps in there... So we had a conversation, but not the one I hoped for
2
0
7
RT @dmcneish18: New 2-year postdoc opening in my lab starting in the fall/winter to work on method development in latent variable modeling…
0
44
0
I don't disagree, Fortune Cookie, but I was taken a little aback by this.
0
1
5
@quantitudepod You're sick, home from school. You pick your favorite of Daddy's mugs for hot cocoa. He says it's from a podcast. It opens with music and then 2 guys talking, making references to things you've never heard of. And then they say something that validates your entire worldview.
1
0
4
@quantitudepod It’s disturbing that Patrick thinks digging up a corpse would make me happy. It’s more disturbing that he's right.
1
0
6
@jeffreymgirard @quantitudepod Yep, one of the earliest treatments of Bayes FA was motivated by concerns for Heywood cases. Just specify a prior that says variances can't be negative. Problem solved/ precluded. A nice example where we have knowledge and Bayes allows us to easily build that into our model.
1
0
1