#statstab #642 PublicationBiasBenchmark
Thoughts: R package for benchmarking publication bias correction methods through simulation studies.
#publicationbias #rstats #r #methods #bias #simulation #petpeese #rmse #robma #3psm #pcurve
fbartos.github.io/PublicationB...
Thoughts: R package for benchmarking publication bias correction methods through simulation studies.
#publicationbias #rstats #r #methods #bias #simulation #petpeese #rmse #robma #3psm #pcurve
fbartos.github.io/PublicationB...
Benchmark for Publication Bias Correction Methods
Implements a unified interface for benchmarking meta-analytic publication bias correction methods through simulation studies (see Bartoš et al., 2025, <doi:10.48550/arXiv.2510.19489>). It provides 1) ...
fbartos.github.io
October 20, 2026 at 4:31 PM
#statstab #642 PublicationBiasBenchmark
Thoughts: R package for benchmarking publication bias correction methods through simulation studies.
#publicationbias #rstats #r #methods #bias #simulation #petpeese #rmse #robma #3psm #pcurve
fbartos.github.io/PublicationB...
Thoughts: R package for benchmarking publication bias correction methods through simulation studies.
#publicationbias #rstats #r #methods #bias #simulation #petpeese #rmse #robma #3psm #pcurve
fbartos.github.io/PublicationB...
I conducted another robustness report on the second meta-analysis I found on genAI impacts on writing. Like my last robustness report, the effect was much lower after correction, around g = .319[−.026, .956]. Caution is needed for genAI ed research.
www.researchgate.net/publication/...
www.researchgate.net/publication/...
(PDF) Can GenAI Boost Writing Scores? Another Robust Bayesian Meta-Analysis (RoBMA) of Writing Ability
PDF | A new meta-analysis (Teng, 2026) has shown that genAI has a strong impact on writing with similar effects as another recent meta-analysis (Dong,... | Find, read and cite all the research you nee...
www.researchgate.net
September 19, 2026 at 9:53 AM
I conducted another robustness report on the second meta-analysis I found on genAI impacts on writing. Like my last robustness report, the effect was much lower after correction, around g = .319[−.026, .956]. Caution is needed for genAI ed research.
www.researchgate.net/publication/...
www.researchgate.net/publication/...
Using multiple publication-bias approaches (RoBMA @fbartos.bsky.social), we found that the evidence becomes considerably less conclusive.
This does not mean Montessori is ineffective.
It means the current evidence base is more uncertain than average effect sizes alone suggest
This does not mean Montessori is ineffective.
It means the current evidence base is more uncertain than average effect sizes alone suggest
July 7, 2026 at 2:51 PM
Using multiple publication-bias approaches (RoBMA @fbartos.bsky.social), we found that the evidence becomes considerably less conclusive.
This does not mean Montessori is ineffective.
It means the current evidence base is more uncertain than average effect sizes alone suggest
This does not mean Montessori is ineffective.
It means the current evidence base is more uncertain than average effect sizes alone suggest
@fbartos.bsky.social @maxmaier.bsky.social & EJ Wagenmakers have extended RoBMA with 3-level publication bias-adjusted model-averaged meta-regression models in R along with the usual code-free and intuitive @jaspstats.bsky.social implementation. Nerdy details are osf.io/preprints/ps...
OSF
osf.io
July 28, 2025 at 4:19 PM
@fbartos.bsky.social @maxmaier.bsky.social & EJ Wagenmakers have extended RoBMA with 3-level publication bias-adjusted model-averaged meta-regression models in R along with the usual code-free and intuitive @jaspstats.bsky.social implementation. Nerdy details are osf.io/preprints/ps...
CRAN updates: BatchJobs bigrquery BiostatsUHNplus chopin duckdb itol.toolkit nhanesA parameters reddPrec RKorAPClient RoBMA snowquery vibass #rstats
September 10, 2025 at 4:02 PM
CRAN updates: BatchJobs bigrquery BiostatsUHNplus chopin duckdb itol.toolkit nhanesA parameters reddPrec RKorAPClient RoBMA snowquery vibass #rstats
Is GenAI Helpful for Writing? A Re-Analysis Using Robust Meta-Analysis (RoBMA): https://osf.io/uk3vf
June 30, 2026 at 9:47 PM
Is GenAI Helpful for Writing? A Re-Analysis Using Robust Meta-Analysis (RoBMA): https://osf.io/uk3vf
Our findings should not be interpreted as set in stone. The RoBMA framework used here can be updated as new evidence accumulates. Nevertheless, our reanalysis provides little support for reliable 2D:4D differences between groups defined by sexual orientation. 8/9
March 19, 2026 at 8:39 AM
Our findings should not be interpreted as set in stone. The RoBMA framework used here can be updated as new evidence accumulates. Nevertheless, our reanalysis provides little support for reliable 2D:4D differences between groups defined by sexual orientation. 8/9
We also fitted multilevel RoBMA meta-regressions to account for dependence across studies, hands, and group comparisons. Comparisons among sexual orientations provided weak to moderate evidence against differences in 2D:4D ratios, and no differences were observed between the right and left hand. 7/9
March 19, 2026 at 8:39 AM
We also fitted multilevel RoBMA meta-regressions to account for dependence across studies, hands, and group comparisons. Comparisons among sexual orientations provided weak to moderate evidence against differences in 2D:4D ratios, and no differences were observed between the right and left hand. 7/9
Updates on CRAN: lfebd3 (0.2.0), MKomics (0.8), MVar (2.2.9), MVar.pt (2.2.9), oolong (0.7.0), RoBMA (4.0.0), RoBSA (1.0.4), ronfig (0.0.10), summata (0.11.5), wru (3.1.0)
May 7, 2026 at 3:36 PM
Updates on CRAN: lfebd3 (0.2.0), MKomics (0.8), MVar (2.2.9), MVar.pt (2.2.9), oolong (0.7.0), RoBMA (4.0.0), RoBSA (1.0.4), ronfig (0.0.10), summata (0.11.5), wru (3.1.0)
Updates on CRAN: CepReg (0.1.3), extraSuperpower (1.6.2), FastRet (1.3.0), NUETON (0.2.0), OhdsiReportGenerator (2.0.2), revert (1.0.0), RoBMA (3.6.1), rpql (0.8.3), shinyfilters (0.2.0), VBphenoR (1.1.0), xrf (0.3.1)
December 18, 2025 at 2:50 AM
Updates on CRAN: CepReg (0.1.3), extraSuperpower (1.6.2), FastRet (1.3.0), NUETON (0.2.0), OhdsiReportGenerator (2.0.2), revert (1.0.0), RoBMA (3.6.1), rpql (0.8.3), shinyfilters (0.2.0), VBphenoR (1.1.0), xrf (0.3.1)
Updates on CRAN: BatchJobs (1.10), bigrquery (1.6.1), BiostatsUHNplus (1.0.3), chopin (0.9.9), duckdb (1.3.3), itol.toolkit (1.1.12), nhanesA (1.4), parameters (0.28.2), reddPrec (3.0.1), RKorAPClient (1.2.0), RoBMA (3.6.0), snowquery (1.3.0), tmap (4.2), vibass (1.0.3)
September 10, 2025 at 9:04 PM
Updates on CRAN: BatchJobs (1.10), bigrquery (1.6.1), BiostatsUHNplus (1.0.3), chopin (0.9.9), duckdb (1.3.3), itol.toolkit (1.1.12), nhanesA (1.4), parameters (0.28.2), reddPrec (3.0.1), RKorAPClient (1.2.0), RoBMA (3.6.0), snowquery (1.3.0), tmap (4.2), vibass (1.0.3)
Updates on CRAN: qtl2ggplot (1.2.6), R2OpenBUGS (3.2-4), REDCapTidieR (1.2.4), riskscores (1.2.3), RoBMA (3.5.1), ruler (0.3.1), sportyR (2.2.3), xegaGaGene (1.0.0.4)
July 28, 2025 at 9:10 PM
Updates on CRAN: qtl2ggplot (1.2.6), R2OpenBUGS (3.2-4), REDCapTidieR (1.2.4), riskscores (1.2.3), RoBMA (3.5.1), ruler (0.3.1), sportyR (2.2.3), xegaGaGene (1.0.0.4)
Updates on CRAN: LFM (0.3.1), PAMscapes (0.14.0), RoBMA (3.5.0), rquest (1.0.5), VertexWiseR (1.3.2)
June 11, 2025 at 9:28 AM
Updates on CRAN: LFM (0.3.1), PAMscapes (0.14.0), RoBMA (3.5.0), rquest (1.0.5), VertexWiseR (1.3.2)
Updates on CRAN: HVT (25.2.1), maptiles (0.9.0), netmeta (3.1-0), RoBMA (3.4.0), ssifs (1.0.4)
February 4, 2025 at 1:24 PM
Updates on CRAN: HVT (25.2.1), maptiles (0.9.0), netmeta (3.1-0), RoBMA (3.4.0), ssifs (1.0.4)
Updates on CRAN: admiral (1.2.0), card.pro (2.3.0), fucom (0.0.3), imuf (0.5.1), MuPETFlow (0.1.1), RobLox (1.2.3), RoBMA (3.3.0), visOmopResults (1.0.0)
January 15, 2025 at 9:16 PM
Updates on CRAN: admiral (1.2.0), card.pro (2.3.0), fucom (0.0.3), imuf (0.5.1), MuPETFlow (0.1.1), RobLox (1.2.3), RoBMA (3.3.0), visOmopResults (1.0.0)
Updates on CRAN: bcdata (0.5.0), gptoolsStan (0.2.0), hexfont (0.5.1), qs2 (0.1.4), RcppCCTZ (0.2.13), reproducible (2.1.2), RoBMA (3.2.0), VertexWiseR (1.2.0)
December 12, 2024 at 9:21 AM
Updates on CRAN: bcdata (0.5.0), gptoolsStan (0.2.0), hexfont (0.5.1), qs2 (0.1.4), RcppCCTZ (0.2.13), reproducible (2.1.2), RoBMA (3.2.0), VertexWiseR (1.2.0)
I don't disagree with any of what you said. I've even been more open to the notion that parsimony shouldn't be seen as a the "best or more natural" path. But I'm currently leaning heavily on the design school (like what Senn does) and also on relaxing assumptions (like what the ROBMA package does)
March 5, 2026 at 8:20 PM
I don't disagree with any of what you said. I've even been more open to the notion that parsimony shouldn't be seen as a the "best or more natural" path. But I'm currently leaning heavily on the design school (like what Senn does) and also on relaxing assumptions (like what the ROBMA package does)
This led me initially to the ROBMA and robust t-test approach, where you fit multiple models (with varying) assumptions and combine the evidence. But - as Julia mentioned before - this opens up the question of your Estimand. I guess doing science is just pick your poison ☠️
February 17, 2026 at 11:42 AM
This led me initially to the ROBMA and robust t-test approach, where you fit multiple models (with varying) assumptions and combine the evidence. But - as Julia mentioned before - this opens up the question of your Estimand. I guess doing science is just pick your poison ☠️