brms book update: I have added a new chapter on generalized linear multilevel models (GLMMs). See paulbuerkner.com/software/brm... for the latest draft of the book. I would love to hear your thoughts!
Paul Bürkner - The brms Book
paulbuerkner.com
October 30, 2024 at 3:36 PM
brms book update: I have added a new chapter on generalized linear multilevel models (GLMMs). See paulbuerkner.com/software/brm... for the latest draft of the book. I would love to hear your thoughts!
That feeling when you spend six months struggling to implement Bayesian GLMMs with imputed data only to find that the observed vs. imputed data analysis results are identical.
Yes, I am a moron.
Yes, I am a moron.
October 8, 2025 at 11:34 PM
That feeling when you spend six months struggling to implement Bayesian GLMMs with imputed data only to find that the observed vs. imputed data analysis results are identical.
Yes, I am a moron.
Yes, I am a moron.
New preprint by @semihaktepe.bsky.social 🎉
We compare ANOVA/SDT/GLMM for binary judgments in 20 datasets of the truth effect. #lme4
Main conclusion:
"GLMMs are a theoretically sound and practically robust method and thus superior for analyzing binary judgments in social and cognitive psychology.”
We compare ANOVA/SDT/GLMM for binary judgments in 20 datasets of the truth effect. #lme4
Main conclusion:
"GLMMs are a theoretically sound and practically robust method and thus superior for analyzing binary judgments in social and cognitive psychology.”
January 28, 2026 at 8:10 AM
New preprint by @semihaktepe.bsky.social 🎉
We compare ANOVA/SDT/GLMM for binary judgments in 20 datasets of the truth effect. #lme4
Main conclusion:
"GLMMs are a theoretically sound and practically robust method and thus superior for analyzing binary judgments in social and cognitive psychology.”
We compare ANOVA/SDT/GLMM for binary judgments in 20 datasets of the truth effect. #lme4
Main conclusion:
"GLMMs are a theoretically sound and practically robust method and thus superior for analyzing binary judgments in social and cognitive psychology.”
November 24, 2025 at 8:54 AM
not uncommon for frequentist mixed models (esp. GLMMs) to give weird/implausible estimates for random effects parameters or convergence issues, and going Bayesian with even mild priors usually fixes it
March 3, 2026 at 10:36 AM
not uncommon for frequentist mixed models (esp. GLMMs) to give weird/implausible estimates for random effects parameters or convergence issues, and going Bayesian with even mild priors usually fixes it
The world if those phighting glmms didnt exist:
April 2, 2025 at 10:05 AM
The world if those phighting glmms didnt exist:
“Overall, GLMMs are a theoretically sound and practically robust method and thus superior for analyzing binary judgments in social and cognitive psychology.”
nice work, @semihaktepe.bsky.social @danielheck.bsky.social
nice work, @semihaktepe.bsky.social @danielheck.bsky.social
Analyzing Binary Judgments: A Comparison of ANOVA, Signal Detection Theory, and Generalized Linear Mixed Models in the Context of the Illusory Truth Effect: https://osf.io/xn397
January 28, 2026 at 2:10 AM
“Overall, GLMMs are a theoretically sound and practically robust method and thus superior for analyzing binary judgments in social and cognitive psychology.”
nice work, @semihaktepe.bsky.social @danielheck.bsky.social
nice work, @semihaktepe.bsky.social @danielheck.bsky.social
A fully funded team of field assistants, multiple bio markers (preferable using new techniques no one else has published on), and analyzed with a dozen GLMMs, even though they don’t even understand GLMMs.
September 14, 2023 at 12:21 PM
A fully funded team of field assistants, multiple bio markers (preferable using new techniques no one else has published on), and analyzed with a dozen GLMMs, even though they don’t even understand GLMMs.
@epigenci.bsky.social makes a pretty convincing case for pseudobulk (as opposed to GLMMs) in multi-sample single cell studies here: doi.org/10.1093/bioi... . #bioinformatics
Pseudobulk with proper offsets has the same statistical properties as generalized linear mixed models in single-cell case-control studies
AbstractMotivation. Generalized linear mixed models (GLMMs), such as the negative-binomial or Poisson linear mixed model, are widely applied to single-cell
doi.org
January 31, 2025 at 10:55 PM
@epigenci.bsky.social makes a pretty convincing case for pseudobulk (as opposed to GLMMs) in multi-sample single cell studies here: doi.org/10.1093/bioi... . #bioinformatics
I was coming to give these examples (GLMMs and logistic regression)
March 3, 2026 at 11:32 AM
I was coming to give these examples (GLMMs and logistic regression)
4/7
However, once you mathematically correct for this massive surveillance bias using Bayesian phylogenetic dyadic GLMMs, the noise dissipates. Reservoir competence isn't just random; it emerges as a predictable biological trait.
However, once you mathematically correct for this massive surveillance bias using Bayesian phylogenetic dyadic GLMMs, the noise dissipates. Reservoir competence isn't just random; it emerges as a predictable biological trait.
June 4, 2026 at 4:21 PM
4/7
However, once you mathematically correct for this massive surveillance bias using Bayesian phylogenetic dyadic GLMMs, the noise dissipates. Reservoir competence isn't just random; it emerges as a predictable biological trait.
However, once you mathematically correct for this massive surveillance bias using Bayesian phylogenetic dyadic GLMMs, the noise dissipates. Reservoir competence isn't just random; it emerges as a predictable biological trait.
Since you mentioned it, I also can't help but recommend @andrew.heiss.phd's blog post: www.andrewheiss.com/blog/2022/11...
Marginal and conditional effects for GLMMs with {marginaleffects} | Andrew Heiss
Use the {marginaleffects} package to calculate tricky and nuanced marginal and conditional effects in generalized linear mixed models
www.andrewheiss.com
August 7, 2025 at 1:29 AM
Since you mentioned it, I also can't help but recommend @andrew.heiss.phd's blog post: www.andrewheiss.com/blog/2022/11...
Doing SDT via GLMMs is probably one of my favorite types of modelling - it was the first, relatively simple, example of doing generative stuff and inferential stuff in the same model, and not the classic two step "get estimates and -then- throw them in an ANOVA".
New preprint by @semihaktepe.bsky.social 🎉
We compare ANOVA/SDT/GLMM for binary judgments in 20 datasets of the truth effect. #lme4
Main conclusion:
"GLMMs are a theoretically sound and practically robust method and thus superior for analyzing binary judgments in social and cognitive psychology.”
We compare ANOVA/SDT/GLMM for binary judgments in 20 datasets of the truth effect. #lme4
Main conclusion:
"GLMMs are a theoretically sound and practically robust method and thus superior for analyzing binary judgments in social and cognitive psychology.”
January 28, 2026 at 2:56 PM
Doing SDT via GLMMs is probably one of my favorite types of modelling - it was the first, relatively simple, example of doing generative stuff and inferential stuff in the same model, and not the classic two step "get estimates and -then- throw them in an ANOVA".
In wonderful Liverpool for @britishecolsoc.bsky.social annual meeting, with some stuff from my PhD that hasn't been published yet!
Come and see me to find out about nightjar lunar ecology using hidden markov models, and glmms! Poster session 1 TONIGHT @ 18:00!
Come and see me to find out about nightjar lunar ecology using hidden markov models, and glmms! Poster session 1 TONIGHT @ 18:00!
December 11, 2024 at 4:22 PM
In wonderful Liverpool for @britishecolsoc.bsky.social annual meeting, with some stuff from my PhD that hasn't been published yet!
Come and see me to find out about nightjar lunar ecology using hidden markov models, and glmms! Poster session 1 TONIGHT @ 18:00!
Come and see me to find out about nightjar lunar ecology using hidden markov models, and glmms! Poster session 1 TONIGHT @ 18:00!
that our guide and tutorial will help to demistify the methods behind GLMMs used to model nominal-scale data. Figure below from the preprint, an illustration of probit and logit nominal models.
February 2, 2025 at 5:32 PM
that our guide and tutorial will help to demistify the methods behind GLMMs used to model nominal-scale data. Figure below from the preprint, an illustration of probit and logit nominal models.
REAL ONES KNOW IT WAS GLMMS
June 8, 2024 at 3:01 AM
REAL ONES KNOW IT WAS GLMMS
NEW METHODS ARTICLE: Phylogenetic GLMMs open doors to study evolution of discrete traits. We show how binary models extend to ordinal & nominal traits, using bird data, and provide tutorials to make these methods accessible to evolutionary biologists:
doi.org/10.1093/jeb/...
Mizuno et al.
doi.org/10.1093/jeb/...
Mizuno et al.
Promoting the use of phylogenetic multinomial generalised mixed-effects model to understand the evolution of discrete traits
Abstract. Phylogenetic comparative methods (PCMs) are fundamental tools for understanding trait evolution across species. While linear models are widely us
doi.org
December 11, 2025 at 2:32 PM
NEW METHODS ARTICLE: Phylogenetic GLMMs open doors to study evolution of discrete traits. We show how binary models extend to ordinal & nominal traits, using bird data, and provide tutorials to make these methods accessible to evolutionary biologists:
doi.org/10.1093/jeb/...
Mizuno et al.
doi.org/10.1093/jeb/...
Mizuno et al.
HE'D MAKE GLMMS BUT HAVE ALL THE HOMESTUCK CHARACTERS
December 18, 2024 at 10:41 AM
HE'D MAKE GLMMS BUT HAVE ALL THE HOMESTUCK CHARACTERS
Closing the Generalised Linear Models in R course, with a look at powerful extensions of GLMs such as GAMs and GLMMs.
A big thank you to @vdveenb.bsky.social for delivering another outstanding edition of the GLMs in R course!
www.physalia-courses.org/courses-work...
A big thank you to @vdveenb.bsky.social for delivering another outstanding edition of the GLMs in R course!
www.physalia-courses.org/courses-work...
May 29, 2026 at 2:41 PM
Closing the Generalised Linear Models in R course, with a look at powerful extensions of GLMs such as GAMs and GLMMs.
A big thank you to @vdveenb.bsky.social for delivering another outstanding edition of the GLMs in R course!
www.physalia-courses.org/courses-work...
A big thank you to @vdveenb.bsky.social for delivering another outstanding edition of the GLMs in R course!
www.physalia-courses.org/courses-work...
Only 1 seat left!
#GLMMs in #Rstats course with @florianhartig.bsky.social
www.physalia-courses.org/courses-work...
#GLMMs in #Rstats course with @florianhartig.bsky.social
www.physalia-courses.org/courses-work...
September 14, 2026 at 12:47 PM
Only 1 seat left!
#GLMMs in #Rstats course with @florianhartig.bsky.social
www.physalia-courses.org/courses-work...
#GLMMs in #Rstats course with @florianhartig.bsky.social
www.physalia-courses.org/courses-work...
get away from that gacha husband in those glmms 😭😭😭
January 11, 2025 at 4:20 AM
get away from that gacha husband in those glmms 😭😭😭
Spent my Friday night finally getting close to a working solution on average marginal effects for GLMMs with varying coefficients directly in Stan. It actually fucking works this time, so not a wasted evening for once!
March 2, 2024 at 5:42 AM
Spent my Friday night finally getting close to a working solution on average marginal effects for GLMMs with varying coefficients directly in Stan. It actually fucking works this time, so not a wasted evening for once!
This would also seamlessly make the connection to ordered data you mention in the discussion, but also connect it to SDT models of rating data, which also are GLMMs (link.springer.com/article/10.3...; vuorre.com/posts/sdt-re... this also has ROC details).
January 28, 2026 at 11:00 AM
This would also seamlessly make the connection to ordered data you mention in the discussion, but also connect it to SDT models of rating data, which also are GLMMs (link.springer.com/article/10.3...; vuorre.com/posts/sdt-re... this also has ROC details).
Just as a comment: you're right that F1 and F2 ANOVA used to be standard in psycholinguistics, but these have been replaced by GLMMs and LMMs in the past decade. Haven't seen an ANOVA in a psycholinguistics paper in a good few years!
January 28, 2026 at 6:24 PM
Just as a comment: you're right that F1 and F2 ANOVA used to be standard in psycholinguistics, but these have been replaced by GLMMs and LMMs in the past decade. Haven't seen an ANOVA in a psycholinguistics paper in a good few years!