Home
Creators
Search Recent Random
Posts
Search DMs Popular Hash Lookup Tags Random
Importer
Import FAQ
Register Favorites Keys Review DMs Logout
Community
Telegram ThePornDude
Home Creators Posts Import Register Favorites Logout

The SHOCKING TRUTH about Causal Inference with Julia Rohrer (YouTube) (Patreon)

Published:
2026-02-21 04:51:00
Edited:
2026-02-22 21:58:58
Imported:
2026-02
Tags:

Content

We’re joined by psychologist Julia Rohrer (Leipzig University) to talk about causal inference, open science, and what psychology has actually learned after the replication crisis.

We discuss why researchers can’t avoid causal claims, post-treatment bias, estimands vs models, incentives around uncertainty, and whether reforms improved science or just paperwork.

One of the rare episodes where we recommend the guest’s work without reservation.

Links

-Julia Rohrer website

https://www.juliarohrer.com/

- The 100% CI blog

http://www.the100.ci/

- Causal inference for psychologists who think causal inference is not for them (2024)

https://doi.org/10.1111/spc3.12948

- Loss-of-confidence project (2021)

https://doi.org/10.1177/17456916211035069

- Birth order and personality (2015)

https://doi.org/10.1073/pnas.1506451112

- BEMC May 2024 talk — Causal confusions correlate with casual conclusions

https://www.youtube.com/watch?v=7Yx0Kz3VQwI

- Less casual causal inference — research talk by Julia Rohrer

https://www.youtube.com/watch?v=7WgH0gE7pZ4

00:00 Welcome to Decoding the Gurus + Introducing Julia Rohrer

01:18 Optimism vs. pessimism: where psychology reform stands now

04:59 Tech & training catch up: OSF, coding skills, Git, and AI help

07:18 How we teach methods: from stats recipes to a causal perspective

18:46 The RCT obsession: when “correlation ≠ causation” becomes a trap

24:28 Beyond experiments: mechanisms, hard-to-explain associations, and toolkits

29:00 Fixing causal talk in papers—and why the public debate makes it harder

47:21 Storytelling vs reality: why gurus sell hope to desperate audiences

53:10 Birth order research recap: the one small IQ effect (and lots of nothing)

57:22 Why “sexy stories” win: media narratives, debunking, and what birth order really means

01:01:10 Open science’s growing pains: badges, preregistration deviations, and “open-washing”

01:08:42 Incentives, prestige journals, and shifting norms: can academia reward rigor?

01:19:45 Closing reflections: causal diagrams, meaningful effect sizes, and optimistic reforms

Files

The SHOCKING TRUTH about Causal Inference with Julia Rohrer

We’re joined by psychologist Julia Rohrer (Leipzig University) to talk about causal inference, open science, and what psychology has actually learned after the replication crisis. We discuss why researchers can’t avoid causal claims, post-treatment bias, estimands vs models, incentives around uncertainty, and whether reforms improved science or just paperwork. One of the rare episodes where we recommend the guest’s work without reservation. Support the Show If you enjoy the show and want to support us, you can find us on Patreon, where we post bonus episodes and extra content: ► Patreon: https://www.patreon.com/decodingthegurus Find Us Elsewhere ► Twitter/X: https://twitter.com/GurusPod ► Podcast (all platforms): https://decoding-the-gurus.captivate.fm/ ► Reddit: https://www.reddit.com/r/decodingthegurus/ Links -Julia Rohrer website https://www.juliarohrer.com/ - The 100% CI blog https://100ci.github.io/ - Causal inference for psychologists who think causal inference is not for them (2024) https://doi.org/10.1111/spc3.12948 - Loss-of-confidence project (2021) https://doi.org/10.1177/17456916211035069 - Birth order and personality (2015) https://doi.org/10.1073/pnas.1506451112 - BEMC May 2024 talk — Causal confusions correlate with casual conclusions https://www.youtube.com/watch?v=7Yx0Kz3VQwI - Less casual causal inference — research talk by Julia Rohrer https://www.youtube.com/watch?v=7WgH0gE7pZ4 00:00 Welcome to Decoding the Gurus + Introducing Julia Rohrer 01:18 Optimism vs. pessimism: where psychology reform stands now 04:59 Tech & training catch up: OSF, coding skills, Git, and AI help 07:18 How we teach methods: from stats recipes to a causal perspective 18:46 The RCT obsession: when “correlation ≠ causation” becomes a trap 24:28 Beyond experiments: mechanisms, hard-to-explain associations, and toolkits 29:00 Fixing causal talk in papers—and why the public debate makes it harder 47:21 Storytelling vs reality: why gurus sell hope to desperate audiences 53:10 Birth order research recap: the one small IQ effect (and lots of nothing) 57:22 Why “sexy stories” win: media narratives, debunking, and what birth order really means 01:01:10 Open science’s growing pains: badges, preregistration deviations, and “open-washing” 01:08:42 Incentives, prestige journals, and shifting norms: can academia reward rigor? 01:19:45 Closing reflections: causal diagrams, meaningful effect sizes, and optimistic reforms

Comments

Andreas Singer

I love the thumbnail. It's shocking!

CrystalJade

Super interesting discussion. Julia was such a great guest. Good choice, as these concepts aren’t well understood in general. It’s definitely a theme exploited by the gurusphere.

jpops

Title feels a bit on the click bait side

Cole Rees

Just a heads up, when I click play, I get "sign in to prove you're not a bot". The sign in link doesn't respond to clicking. The YouTube icon does redirect me to YouTube on my phone browser, which I am logged in to. I can't seem to copy the link to the video out of the patreon app. Presumably I could get the link if I was viewing patreon in the mobile browser. I'm on a pixel, no VPN. But I could find your video by googling it :)

thefigmaster

CI site doesn't seem to work for me

Jeanine Marie

I really enjoyed this. Thank you!!

Aimée Bower

Found this conversation really interesting. Thank you!

Password1234!

Here's my two cents on the discussion 20mins in: I agree whole heartedly with the base of the arguments around arbitrary rules and checklist approaches toward testing models and their assumptions. Much more nuance is needed with arguments made for vs against inference with any design used. BUT, a large portion of our students WANT simple, easy to understand rules around building an argument for/vs against some treatment, and (arguably) they will not use these skills often enough to justify extensive understanding. The number of students studying a social science who are willing to contend with questions like "but what are the assumptions of this test actually telling me?" and "do my findings add/deduct strength from inference a/b/c, and if so, by how much?" are extremely limited. I don't think that means the focus around open science and properly understanding causal inference isn't very important, but I do think that means that we should adjust our expectations around what we consider good progress for understanding these concepts. Nonetheless, I would love to see more critical thinking around methods and statistics used within the broad category of social science and will continue to advocate for the open science framework. I'm content with the slow roll, so long as people are willing to listen to me explain why it needs to be faster!

driver3

Speaking as a non-academic, much of this went over my head. However, I re-listened to some portions, and I liked it. It's a rich topic that many of us think about for our entire lives in one way or another ("what really causes this or that... and am I sure?"), and it was good to hear from an expert. There was also a good summary offered at some point, as to the sloppy way that gurus approach causal inference (if I am not misphrasing). I have a kind of "related request": I'm active in campaigning to lower GHG emissions (and to clean up the pollution that has already been emitted, as possible) in order to address the climate crisis. While the opponents of action on climate change tend to rely on a ready-made extensive grab-bag of many different arguments, s.... ome of those arguments relying on mis- and dis-information, .... in order to try to reduce and delay action, one of the big tactics they use, if not the biggest one, goes something like this (relying on mis- or dis-information). "It's not an emergency, very few (if any) are dead directly from climate change, and one could even say that the net number of lives has been saved by warming. Scientists have not yet proven conclusively that increased mortality results from human-caused climate change and so please stop your gratuitous liberal hysteria, the adults in the room have actual important business to talk about." Over the last few years I have become convinced that it is important to break through such arguments and counter them with whatever work that modern scientists are doing so that we non-academics can say clearly and unapologetically something like: "Peer-reviewed science shows that an estimated range of about x-y extra number of people dies every year, beyond what would be expected, and this number is attributable, through credible scientific methods, to climate change." [I may be getting some loaded words wrong, but you get the idea.] For example, I see this number: "....What is already known on this topic. Climate change is one of the most pressing public health threats of the 21st century, contributing to more than 250 000 deaths each year..." from here: https://pmc.ncbi.nlm.nih.gov/articles/PMC10862342/ BMJ Med . 2024 Feb 10;3(1):e000627. doi: 10.1136/bmjmed-2023-000627 Published research on the human health implications of climate change between 2012 and 2021: cross sectional study Victoria L Bartlett 1, Harry Doernberg 1, Maryam Mooghali 2, Ravi Gupta 3, Joshua D Wallach 4, Kate Nyhan 4,5, Kai Chen 4, Joseph S Ross 2,6,7,✉ PMCID: PMC10862342 PMID: 38352020 Abstract Objective To better understand the state of research on the effects of climate change on human health, including exposures, health conditions, populations, areas of the world studied, funding sources, and publication characteristics, with a focus on topics that are relevant for populations at risk. --------- This is arguably a poor example, because the number does not seem to come out of the study (it seems to be just assumed from others), and because I'm not sure what this means ("....Provenance and peer review: Not commissioned; externally peer reviewed.") However, I like that link because it gives a sense of the increasing amount of attention that scientists are bringing to this important topic. I think Chris brought up the topic of smoking. As I understand it, there were difficult (epidemiological and others, I guess) science issues that came up in the effort to research smoking and its harms, and (if I understand) this delayed taking life-saving policy actions. I think much the same is happening with climate change. I think scientists such as are occasionally on the DTG show could help us all understand better the progress that has been made in science of making credible estimates of how many extra deaths are occurring every day "attributable to" (or insert whatever wording is appropriate here) climate change.