#statstab #596 Generalized Additive Latent and Mixed Models {galamm} Thoughts: Complex ideas require complex (but appropriate) modelling strategies. #glmm #gam #r #rstats #lavaan #stan #latent #bayesian #bayes docs.ropensci.org/galamm/

Generalized Additive Latent and Mixed Models

Estimates generalized additive latent and mixed models using maximum marginal likelihood, as defined in Sorensen et al. (2023) <doi:10.1007/s11336-023-09910-z>, which is an extension of Rabe-Hesketh a...

docs.ropensci.org

#statstab #595 Required Sample Size to Detect the Mediated Effect Thoughts: Do with this information what you will. #mediation #samplesize #poweranalysis #power #error #type2 #causalinference journals.sagepub.com/doi/full/10....

Sage Journals: Discover world-class research

Subscription and open access journals from Sage, the world's leading independent academic publisher.

journals.sagepub.com

These ratio outcome variables are a plague in the sciences. If your analysis treats 1/2 and 10/20 as the same evidence, then you have discarded uncertainty in the outcome. Model what you measure, not what you construct. Use the denominator as an exposure instead.

vincentholst.bsky.social@vincentholst.bsky.social · 5d ago

The study is based on the CD index, a citation-based measure of how much the successors of a work still rely on its predecessors, running from -1 (consolidating) to +1 (disruptive). Key detail: a paper with zero references automatically gets CD = +1 as soon as it is cited once.

#statstab #594 Measurement Theory: Why it is Possible to Measure Psychological Phenomena Thoughts: I need to engage with this properly, but seems cool. #psychometrics #mesurement #theory #validity #reliability #history elliedegb.quarto.pub/psychometric...

#statstab #593 Lord’s Paradox: The Power of Causal Thinking Thoughts: This is Judea Pearl's take on the paradox. Will this ever be solved in a way that all are happy? #confounding #paradox #lordsparadox #causalinference #ancova #DAG causality.cs.ucla.edu/blog/index.p...

#statstab #592 ANOVA: Mixed-Effects (F-Test) {pwrss} Thoughts: A slightly more comprehensive view of interaction testing in ANOVA. #poweranalysis #samplesize #rstats #r #package #guide #effectsize #interaction cran.r-project.org/web/packages...

#statstab #591 What is Incidence Rate Ratio? Thoughts: A simple overview, but can be a starting point for IRRs in count models. #irr #regression #stats #poisson #negativebinomial #effectsize www.statology.org/incidence-ra...

What is Incidence Rate Ratio? (Definition & Example)

This tutorial provides an explanation of incidence rate ratio, including a formal definition and an example.

statology.org

#statstab #590 Questions about statistical claims in paper from recent Nobel prize winners; some general challenges in trying understand nonlinear patterns using quadratic regression Thoughts: A pattern could just be properties of the model #quadratic statmodeling.stat.columbia.edu/2025/10/15/q...

#statstab #589 The Permutation Test: A Visual Explanation of Statistical Testing Thoughts: If you stay in a frequentist framework, this should be your default. #permutation #randomization #nhst #ttest #pvalue #tutorial #guide www.jwilber.me/permutationt...

#statstab #588 Test if two coefficients in a linear model are both significant Thoughts: The bigger p-value of the two would do it. #regression #equivalence #nhst #pvalues #discussion stats.stackexchange.com/questions/66...

Test if two coefficients in a linear model are both significant

Is there a single test one can perform that would be significant if two coefficients in a linear model are both different from 0, but would not be significant if only one of them differs from 0? Ye...

stats.stackexchange.com

#statstab #587 Some Ways to Make Regression Modeling More Helpful Than Misleading Thoughts: Part of a series of papers and commentary. Very useful reference. #regression #modelling #inference #guide #commentary onlinelibrary.wiley.com/doi/10.1002/...

Some Ways to Make Regression Modeling More Helpful Than Misleading

Click on the article title to read more.

onlinelibrary.wiley.com

New post: Diagnosticity and SDT's d′ & c can rank confession techniques differently, but because they answer different questions. Before declaring one metric “better,” check whether they were entered into the same competition. #statistics #SDT #estimands open.substack.com/pub/mzlotean...

#statstab #586 Uses of Uncertain Statistical Power: Designing Future Studies, Not Evaluating Completed Studies Thoughts: 3 uses of power: comparing the performance of diff. procedures, planning studies, and evaluating completed studies. #power #nhst psycnet.apa.org/fulltext/202...

📢 Excited about our (Louisa Kulke & Fatima Felisberti) new paper in Motivation & Emotion! It echoes a theme from my recent ISRE talk: rigorous methods, precise estimates, and theory-driven predictions are essential if we want cumulative emotion science. #emotion link.springer.com/article/10.1...

Judging emotional authenticity from dynamic facial expressions: a preregistered Bayesian analysis of individual differences - Motivation and Emotion

Accurately judging whether another person’s emotional display is genuine or deliberate is a core component of social interaction, yet prior research linking this ability to socio-cognitive traits has ...

link.springer.com

Nice project on Estimation Statistics plots for complex designs that I covered in #statstab #481. The right-middle plot separates p-fishers from scientists. If you instinctively say "but I was to plot all possible differences", not planned (theory informed) comparisons, you're not doing science.

Physalia-courses@ONLINE@physaliacourses.bsky.social · 3w ago

A great read on moving beyond traditional ANOVA and towards estimation-based approaches for multi-group comparisons. www.nature.com/articles/s41...

#statstab #585 Collapsibility: Odds Ratios versus Risk Ratios Thoughts: In light of recent social media discussions, here is the same conversation 13y ago I had saved. #oddsratios #riskratios #collapsibility #inference #debate #discussion #statistics stats.stackexchange.com/questions/48...

#statstab #584 How are the likelihood ratio, Wald, and Lagrange multiplier (score) tests different and/or similar? Thoughts: I don't recall why I saved this link. But might be useful. #likelihood #stata #wald #lagrange #inference stats.oarc.ucla.edu/other/mult-p...

#statstab #583 Statistical methods: Basic concepts, interpretations, and cautions Thoughts: A bit of history and a bit of theory. Useful overview of p-values. #pvalues #nhst #estimationstatistics #fisher #neyman #compatibility #history arxiv.org/pdf/2508.10168?

#statstab #582 Is Repeated Bayesian Interim Analysis Consequence-Free? Thoughts: It's good to have debates about papers that make bold claims. #discussion #debate #bayesian #interim #error #sequential #priors discourse.datamethods.org/t/is-repeate...

Is Repeated Bayesian Interim Analysis Consequence-Free?

Being still new to Bayesian analysis, and currently focusing more and more on platform trials and interim analyses, I thought this was interesting. Interested to hear others thoughts on the paper! Ab...

discourse.datamethods.org

#statstab #581 Marginal and conditional summary measures: transportability and compatibility across studies Thought: "Because marginal and conditional summary measures do not generally coincide, their naïve pooling in evidence synthesis can produce bias" #estimands arxiv.org/html/2507.21...

A few very good references for analysing pre-post data, and why change scores aren't the way.

Peter Tennant@pwgtennant.bsky.social · last mo.

Senn 2006: onlinelibrary.wiley.com/doi/10.1002/... Vickers & Altman 2001 www.bmj.com/content/323/...

#statstab #580 A comment on replication, p-values and evidence Thoughts: What does a p-value have to do with a replication probability? Maybe not much. #bayesian #pvalues #probability #replication #neynan #fisher #significance doi.org/10.1002/sim....

A comment on replication, <i>p</i>‐values and evidence S.N.Goodman, <i>Statistics in Medicine</i> 1992; <b>11</b>:875‐879

Click on the article title to read more.

doi.org

#statstab #579 {pmsampsize} Sample Size for Development of a Prediction Model Thoughts: Do you have enough data for your next Predictive Model? Hint: it's probably more than n=30 per cell. #prediction #samplesize #rstats #r #package #guide #r2 cran.r-project.org/web/packages...

#statstab #578 Accuracy and Consistency of Visual Analog Scales in Ecological Momentary Assessment and Digital Studies Thoughts: Interesting paper. Not my fav way to collect data of this type. #vas #visualanalogscale #psychometrics #scale #ordinal #reliability media.eiko-fried.com/publications...

#statstab #577 Revisiting the Table 2 fallacy: A motivating example examining preeclampsia and preterm birth Thoughts: Another useful citation for the Table 2 fallacy to convince your friends. #table2 #causalinference #regression #inference #error onlinelibrary.wiley.com/doi/abs/10.1...

Revisiting the Table 2 fallacy: A motivating example examining preeclampsia and preterm birth

Background A “Table Fallacy,” as coined by Westreich and Greenland, reports multiple adjusted effect estimates from a single model. This practice, which remains common in published literature, can ...

onlinelibrary.wiley.com

#statstab #576 Zero Truncated Negative Binomial Weighted Weibull Distribution and Its Application Thoughts: Lots of cool work with new distributions for specific data/scenarios. Let's move beyond Normal. #negativebinomial #weibull #distributions #MLE link.springer.com/article/10.1...

Zero Truncated Negative Binomial Weighted Weibull Distribution and Its Application - Lobachevskii Journal of Mathematics

Abstract In this article, we propose two new discrete distributions, the negative binomial-weighted Weibull and the zero truncated negative binomial-weighted Weibull distributions. Some statistical pr...

link.springer.com

#statstab #575 How to create a(n almost) fully Bayesian outcome model with inverse probability weights Thoughts: An interesting way to do IPTW within a Stan/bayesian framework. #brms #stan #iptw #causalinference #weighting #rstats #r #bayesian #tutorial www.andrewheiss.com/blog/2021/12...

How to create a(n almost) fully Bayesian outcome model with inverse probability weights | Andrew Heiss

Use a posterior distribution of inverse probability weights in a Bayesian outcome model to conduct (nearly) fully Bayesian causal inference with R, brms, and Stan

andrewheiss.com

#statstab #574 When the Mean is Misleading: a Guide to Ordered Regression for Meaningfully Modeling Ordinal Outcomes Thoughts: The "it's fine to analyse likert/ordinal data with t-test" camp is not doing themselves any favours #ordinal #likert #probit #regression link.springer.com/article/10.1...

When the Mean is Misleading: a Guide to Ordered Regression for Meaningfully Modeling Ordinal Outcomes - Journal of Quantitative Criminology

Objectives We demonstrate the utility of ordered regression models for analyzing ordinal outcomes frequently encountered in criminology, comparing their predictive accuracy and estimation against conv...

link.springer.com

#statstab #573 Means to valuable exploration II: How to explore data to modify existing claims and create new ones Thoughts: A guide on the many types of exploratory research. #exploratory #research #tutorial #guide #hypothesis open.lnu.se/index.php/me...

An excellent example of why the language of "controlling for x, y..." has led to so many mistakes in statistical modelling. (there are also like 4 other things wrong in that paragraph)

Saloni@scientificdiscovery.dev · 2mo ago

Oh no! Oh no!!