Answering the most popular question in a PhD methods course
datacolada.org/133
02.03.2026 19:03
👍 23
🔁 10
💬 1
📌 2
Answering the most popular question in a PhD methods course
datacolada.org/133
I will give another talk about mediation analysis and I have finally figured out what I want to say about "design-based" solutions:
- longitudinal data can help rule out certain types of confounding
- if you sequentially intervene on X and M, you need assumptions to piece the estimates together
>
Research by Caviola et al in the US and Spain finds children are less likely than adults to prioritize humans over animals in moral dilemmas, regardless of cognitive capacity, suggesting species favouritism develops as we grow older:
buff.ly/DB4Swhc
Happy to share our latest study published in PNAS.
Using data from 274,316 French students, we find that lower-SES students are less likely to wait for better university offers, even when waiting would lead to more prestigious or better-fit programs.
www.pnas.org/doi/10.1073/...
Abstract Objective: To examine whether induction of labor and epidural analgesia are associated with breastfeeding initiation, maintenance, and breastfeeding problems. Design: Population-based longitudinal cohort study. Setting: The Norwegian Mother, Father and Child Cohort Study (MoBa), linked with data from the Medical Birth Registry of Norway. Sample: A total of 73,069 mothers with singleton infants and vaginal births. Methods: Information on induction of labor, method of induction, and epidural analgesia were obtained from registry data. Breastfeeding initiation, maintenance of breastfeeding at six months, and breastfeeding problems were reported by mothers six months postpartum. Logistic regression analyses estimated associations between birth interventions and breastfeeding outcomes, adjusting for several maternal sociodemographic and obstetric variables, and maternal and fetal health. Main outcome measures: Initiation of breastfeeding within the first week, maintenance of breastfeeding to six months postpartum, and reported breastfeeding problems. Results: Induction of labor was associated with reduced likelihood of initiating breastfeeding (adjusted OR 0.65, 95% CI 0.56 - 0.77) and maintaining breastfeeding at six months (adjusted OR 0.86, 95% CI 0.81 – 0.91), and with increased likelihood of breastfeeding problems (adjusted OR 1.11, 95% CI 1.01 – 1.22). Epidural analgesia was also associated with lower initiation (adjusted OR 0.68, 95% CI 0.59 – 0.77) and maintenance (adjusted OR 0.79, 95% CI 0.76 – 0.83), and higher rates of breastfeeding problems (adjusted OR 1.12, 95% CI 1.05 – 1.21). Conclusion: Induction of labor and epidural analgesia may be a risk factor for reduced likelihood of breastfeeding. Awareness can help clinicians target individualized breastfeeding support postpartum.
New preprint! 🎉 Using a sample of over 73k births across Norway, we found that induction of labor (with either oxytocin, prostaglandin, or amniotomy) and epidural analgesia are associated with lower likelihood of initiating breastfeeding doi.org/10.22541/au....
Lien vers la newsletter : psychopapers.kessel.media/posts/pst_02...
6. Roue des privilèges
➡️ Une roue des privilèges pour visualiser les positions dominantes en science
7. Autisme et microbiote (partie abonnés)
➡️ Le lien entre microbiote et autisme repose surtout sur petites études, modèles animaux douteux et essais cliniques négatifs.
4. IA et base de données NHANES
➡️ Une explosion d’articles médiocres générés avec l’IA à partir des données du NHANES, avec des fausses découvertes.
5. Menace et extrême droite
➡️ Trois expérimentations indiquent que se sentir menacé n’augmente pas le conservatisme politique.
2. Revues prédatrices en médecine respiratoire
➡️ Le nombre de journaux a fortement augmenté, sans hausse comparable des articles dans les revues majeures.
3. Entretien international du trauma
➡️ L’ITI, un entretien clinique pour diagnostiquer le trauma, vient d’être validé dans plusieurs pays.
Nouvelle Newsletter ! Au programme :
1. Prédire le suicide
➡️ Des médecins aidés par un algorithme de machine learning prédisent les tentatives de suicide 3 à 10 fois mieux que les médecins seuls.
psychopapers.kessel.media/posts/pst_02...
A team of young people talking and working together, sharing and writing down ideas, using laptops.
📣 @i4replication.bsky.social and CERDI are organizing a Replication Game on May 29, 2026 in #ClermontFerrand.
🧠💡 Teams of researchers, PhD, and master's students will reproduce the results of an academic article.
Register now: www.surveymonkey.ca/r/F6XWX9F
It's 2025 and you can still claim an effect with results like this at JAMA Psychiatry. cc @urisohn.bsky.social @lakens.bsky.social @heinonmatti.bsky.social
jamanetwork.com/journals/jam...
I don't disagree but historically governments have done so only in high stakes areas (e.g. medicine). I'm afraid the social sciences will be left alone because their success is (deemed) irrelevant
Encore une preuve que Batman n’a pas besoin de construire des écoles pour servir la JUSTICE
Very interesting set of studies about how agreement with punished acts undermines the perceived legitimacy of the punisher. Reminds me a bit of the Mullen & Nadler (2008) study that found that people were more likely to steal a borrowed pen from the researcher after being exposed to an unjust law.
Many have asked for the LLM Survey paper. The release was bungled a bit by PNAS, but it is live now: www.pnas.org/doi/10.1073/...
Help me out quant psych hivemind. I have two competing SEM models fitted to a dataset. They are fairly unconstrained and I established the factor structure in a separate data fold. As a result, both have excellent fit (non-sig chi-sq, CFI/TLI > .99, RMSEA = .02).
This has been many years in the making. We wanted to do it right—not simply get it done.
Our goal was to contribute, however modestly, to an academia that genuinely strives for inclusivity.
May the AWoP—small as the step may be—help move academia toward greater inclusion. 🤝🫂🌈🍉🏳️⚧️✊⚖️♿🦮🌐
Update your syllabus and stay on the frontier - it will increase your students’ wages. Epic work by my colleagues @barbarabiasi.com and @profsongma.bsky.social #linkoftheday
www.barbarabiasi.com/uploads/1/0/...
QUI AURAIT PU PRÉDIRE ?!
new paper by Sean Westwood:
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
We can no longer trust that survey responses are coming from real people.”
Fascinating (and important to many fields of study).
Link to PNAS paper is in 2/2.
1/2
Calibrating scientific skepticism www.wiringthebrain.com/2018/07/cali... - I wrote this a few years ago in relation to claims of transgenerational epigenetic inheritance in humans, but the issues relate equally to the kind of microbiome studies we assess in the paper linked below...
Nouvel épisode sur le podcast Nota Bene sur Lovecraft, son histoire, et son univers. On me voit géant, vert, en train de sourire, avec des tentacules à la place des mains. Je suis entouré de fumée et d'éclairs, et on peut voir face à moi de dos une silhouette humaine en noir. On peut lire en bas "Cthulhu".
Oyez, oyez, fans de science-fiction, de fantastique, et de fantasy !
Dès maintenant, et ce sur toutes les plateformes de podcast, vous pouvez aller écouter mon épisode dédié à Lovecraft et son histoire !
Bonne écoute !
https://podcasts.apple.com/fr/podcast/nota-bene/id1573434643
Transparent and comprehensive statistical reporting is critical for ensuring the credibility, reproducibility, and interpretability of psychological research. This paper offers a structured set of guidelines for reporting statistical analyses in quantitative psychology, emphasizing clarity at both the planning and results stages. Drawing on established recommendations and emerging best practices, we outline key decisions related to hypothesis formulation, sample size justification, preregistration, outlier and missing data handling, statistical model specification, and the interpretation of inferential outcomes. We address considerations across frequentist and Bayesian frameworks and fixed as well as sequential research designs, including guidance on effect size reporting, equivalence testing, and the appropriate treatment of null results. To facilitate implementation of these recommendations, we provide the Transparent Statistical Reporting in Psychology (TSRP) Checklist that researchers can use to systematically evaluate and improve their statistical reporting practices (https://osf.io/t2zpq/). In addition, we provide a curated list of freely available tools, packages, and functions that researchers can use to implement transparent reporting practices in their own analyses to bridge the gap between theory and practice. To illustrate the practical application of these principles, we provide a side-by-side comparison of insufficient versus best-practice reporting using a hypothetical cognitive psychology study. By adopting transparent reporting standards, researchers can improve the robustness of individual studies and facilitate cumulative scientific progress through more reliable meta-analyses and research syntheses.
Our paper on improving statistical reporting in psychology is now online 🎉
As a part of this paper, we also created the Transparent Statistical Reporting in Psychology checklist, which researchers can use to improve their statistical reporting practices
www.nature.com/articles/s44...
J’avais écrit ça sur cette question uneheuredepeine.blogspot.com/2019/02/le-m... (prière de lire jusqu’à la fin avant de m’engueuler ok ?)
Are methodological and causal inference errors creating a false impression that the gut microbiome causes autism? In this strong analysis, Mitchell, Dahly, and Bishop question the evidence.
They show that triangulation in science requires multiple robust lines of research.
A quick (1000 words) read to enjoy with your morning coffee or afternoon tea:
"Psychology wants to stay WEIRD, not go WILD"
Why hasn't psychology diversified it samples, methods, theories, etc.? Because it doesn't want to. osf.io/preprints/ps...
