#statstab #545 Dynamic Meta-analysis: When Transparency Meets Multiplicity Thoughts: Seems like hard work but makes perfect sense. Combine this with live meta-analyses. #metascience #metaanalysis #evidence #multiverse drmattg.github.io/Uncertain_Ec...

#statstab #544 {Bambi} plot predictions Thoughts: Python package that works similarly to {marginaleffects} #python #stats #modelling #prediction #marginaleffects #effects #marginalia #reporting bambinos.github.io/bambi/notebo...

Sadly, #UCL school of Pharmacy has a symposium on herbal medicines today. My view (as published in the FT): "Herbal medicine: giving patients an unknown dose of an ill-defined drug, of unknown effectiveness and unknown safety." www.dcscience.net/magic-medici...

Modern problems require ancient solutions

#statstab #543 How hazard ratios can mislead and why it matters in practice Thoughts: Another *insert effect size measure* has issues paper. If you use HRs you better study up. #observational #hazardratio #effectsize #misconceptions #causalinference link.springer.com/article/10.1...

Woke up and chose #stats violence: I'm analysing a study, and to answer the research questions I need to conduct *both* an ANCOVA (baseline corrected) and an RM-ANOVA! Let the comments begin

Statistics aka Likelihoodmaxxing

#statstab #542 Improving the utility of non-significant results for educational research Thoughts: Another paper that tries to clarify some issue with NHST. #nhst #stats #pvalues #education #frequentist #misconceptions www.sciencedirect.com/science/arti...

{easystats} is not only highly functional and powerful, but their help is also a goldmine of accessible #statistics advice. Here explaining how to address high VIF values (hint: generally not by removing predictors, as often done) easystats.github.io/performance/... Thanks @easystats.github.io

Psychologists doing clustering be like: #statsmeme

I really wish researchers in my field would use ordinal models more. Trying to shove things into a linear (gaussian) model because "it's easier to interpret" is nonsense. You are not doing yourself or the field any favours.

15 new articles on statistical workflow! statmodeling.stat.columbia.edu/2026/05/29/1...

#statstab #541 Titles & Captions: Adding Context to Your Plots Thoughts: A good plot can showcase your entire paper. Normalise plotting effects of interest! #ggplot #dataviz #figures #plots #inference #r #legibility www.ggplot2-uncharted.com/module3/titl...

#statstab #540 Measures of Central Tendency for an Asymmetric Distribution, and Confidence Intervals Thoughts: Pseudomedian, Central Limit Theory, and more. #CLT #pseudomedian #confidenceintervals #distributions #signedranktest www.fharrell.com/post/aci/

Call for Contributions to Workshop: “Do experiments replicate? Philosophical Reflections on the Use and Misuse of Statistics and Econometrics” More info: xphi.net/2026/05/26/c... #PhilSci #stats

#statstab #539 Sufficiently Important Difference Thoughts: "This article 1) argues that MID should be replaced with “sufficiently important difference” (SID), a concept that includes balancing of benefits against harms" #mcid #effectsize #sesoi journals.sagepub.com/doi/full/10....

#statstab #539 Sufficiently Important Difference Thoughts: "This article 1) argues that MID should be replaced with “sufficiently important difference” (SID), a concept that includes balancing of benefits against harms" #mcid #effectsize #sesoi journals.sagepub.com/doi/full/10....

#statstab #538 A Tutorial on Conducting and Interpreting a Bayesian Independent T-Test Using Open-Source Software Thoughts: OK as a beginner guide, but not more. #bayesian #bayesfactor #priors #jasp #ttest #tutorial #guide onlinelibrary.wiley.com/doi/epdf/10....

I made a #Shiny app about post hoc power and effect-size uncertainty. If you accept an observed effect, you also need to accept its uncertainty. The app shows how power wobbles when you consider the CI, not just the estimate. Embrace the #wobble #samplesize open.substack.com/pub/mzlotean...

Land cover of Romania. #rayshader adventures, an #rstats tale

#statstab #537 {hdbayes} An R Package for Bayesian Analysis of Generalized Linear Models Using Historical Data Thoughts: An interesting approach to priors. I'm not v familiar w this so curious what others think. #bayes #bayesian #priors #historicaldata arxiv.org/html/2506.20...

#statstab #536 The Role of Item Distributions on Reliability Estimation: The Case of Cronbach’s Coefficient Alpha Thoughts: Part of a series of papers by Zumbo arguing for an ordinal alpha (that others contest) #ordinal #cronbachalpha #reliability journals.sagepub.com/doi/10.1177/...

'Can engaging in the arts slow ageing, as a recent study suggested?' I'm excited to share my new piece in @theconversation.com! Please share! Prepare for a drive-by lesson in reverse causality, residual confounding, and the importance of asking "compared to what?"

"If I hated to be the bearer of bad news, I wouldn’t have specialized in research methods." - pure comedy 😂 I need to put that quote somewhere asap.

#statstab #535 The Benefits Of Reporting Critical Effect Size Values Thoughts: You ran a study w/o a power analysis? Report the lowest value your test could have detected at 80% power and 5% alpha. #poweranalysis #effectsize #samplesize #posthoc osf.io/preprints/ps...

#statstab #534 Model-averaged Bayesian t tests Thoughts: I find this testing ensemble approach very amenable to theory building. Check a few models that include/exclude assumptions you may/not care about. #robust #ttest #bayes #bma #jasp link.springer.com/article/10.3...

#statstab #533 Analysis of Covariance and Standardization as Instances of Prediction Thoughts: An interesting read from some of the best. ANCOVA seems so simple yet it is quite complex. #ancova #analysis #standardization #prediction #marginaleffects www.jstor.org/stable/25300...

This is a very real element of (some) academia. People that have essentially cargo cult mentality of what being a researcher is, leading to bad practices.

#statstab #532 Fractional Bayes Factors for Model Comparison Free - O'Hagan (1995) Thoughts: Use a fraction of the data to convert an improper prior into a minimally informative prior. #bayesian #bayesfactor #priors #bain #hypothesis #testing doi.org/10.1111/j.25...

Klaus Fiedler passed away two days ago. joachimfunke.de/2026/05/13/n... In 2017 we did a fun interview for In Mind on the replication crisis. www.in-mind.org/blog/post/we... We didn't agree on everything, but I greatly respect he took on the role of Devil's Advocate as a scientific principle.

#stats #jasp #bayes Question: In a 2x2 rm design. I run a Bayesian RM-ANOVA, I then ask for post hoc comparisons for a main effect (a-b), I get BF=3. I also manually compute the contrast as [(x_a+y_a)/2 - (x_b+y_b)/2], the BF=3.1. Is the difference due to JASP standardising the data for t-tests?

#statstab #531 Effect Size Calculator [Campbell] Thoughts: A nice place for quick formulas for variance and effect sizes of various designs and data types. #metaanalysis #effectsize #CohenD #calculator #Variance #Eta #effects www.campbellcollaboration.org/calculator/e...

#statstab #530 On the Need to Revitalize Descriptive Epidemiology Thoughts: Paper that shows how to properly conduct a "descriptive" piece of research. More ppl should do this. #descriptive #research #guide #example #methods #epistemology #epistemology academic.oup.com/aje/article/...

Maybe the view should be 1. Epistemic accountability (ie. Duty to disclose; all research is made available in some form), and 2. Formally publish only what is "worthy" (triage, not suppression)

#statstab #529 How Do I Know What My Theory Predicts? Thoughts: I'd like to see more researchers adopt Dienes' framework and way of thinking about research. #bayesian #bayesfactor #evidence #epistemology #research #tutorial journals.sagepub.com/doi/10.1177/...

Speaking of crazy d values, this reminds me of this feature I was thinking about adding to the forest() function in metafor. #Rstats #MetaAnalysis

#statstab #528 {afex_plot} Publication Ready Plots for Experimental Designs Thoughts: Plotting your results should be a norm for all articles. #anova #dataviz #plots #plotting #r #rstats #afex singmann.github.io/afex_plot_in...

#statstab #527 How to interpret “confidence intervals” in observational studies Thoughts: A great example of a conversation that goes nowhere but is interesting to read #debate #unhelpful #discussion #forum #confidenceintervals #observational #inference discourse.datamethods.org/t/how-to-int...

I’m thrilled to have joined the Editorial Board for The Police Journal: Theory, Practice and Principles! #Policing #CriminalJustice #LawEnforcement

#statstab #526 Splines, B-splines, P-splines, and a disapproving kitten Thoughts: Nice R tutorial on splines with some explanations and illustrations. #splines #r #rstats #tutorial #guide #polynomial #nonlinear blog.djnavarro.net/posts/2025-0...

#statstab #525 Falsificationism and clinical trials Thoughts: A strongly worded paper on the role of Poperrian induction on inference. #Popper #falsification #induction #inference #philosophy #evidence #clinical onlinelibrary.wiley.com/doi/10.1002/...

#statstab #524 {simglm}: Tidy simulation and power analyses Thoughts: As your design becomes more complex, simulation is the only way to go. #simulation #r #rstats #tidy #tidyverse #poweranalysis #design #tutorial simglm.brandonlebeau.org

#statstab #523 Pre/Post design: the fallacy of comparing difference scores Thoughts: Another easy read on pre-post design and some concerns about change scores. #education #guide #prepost #ancova #change #ttest #design garstats.wordpress.com/2026/02/20/p...

#statstab #531 Using Extant Data to Improve Estimation of the Standardized Mean Difference Thoughts: I wonder if it's a good idea. Any edu/clinical people to chime in? Seems like using historic data. #smd #estimation #effectsize #variance #extantdata journals.sagepub.com/doi/10.3102/...

#statstab #530 Fisher vs Stouffer Thoughts: Combining p-values (say in a meta-analysis) is not a straightforward thing. Here are two methods. #fisher #stouffer #pvalues #metaanalysis #zscores #evidence #falsepositive #r #simulation timydaley.github.io/FisherVsStou...

#statstab #529 Understanding and misunderstanding randomized controlled trials Thoughts: Randomization is close to magic, but we still must be critical of our methods. #randomization #rct #design #critique #covariates #causalinference #ATE #confounder #bias doi.org/10.1016/j.so...

#statstab #528 Survey Bot Detection Workshop Thoughts: "Pick a cutoff value for detecting random responders" can you spot the bots? #OpenScience #bots #detection #sample #human #participants #deception #shinyapp #education #tutorial falkcarl.shinyapps.io/BotApp1/

Is a 55% replication rate too low, too high, or just right? Some thoughts on Tyner et al.’s (2026) recent study. #MetaSci #PhilSci