Patrick S. Forscher
LightNews LightNews
banner
psforscher.bsky.social
Patrick S. Forscher
@psforscher.bsky.social
1.6K followers 280 following 400 posts

Director of the CREME developmental meta-research team at Busara, a non-profit that does behavioral science in service of poverty alleviation. https://patrickforscher.com/

Videos
Psychology 39%
Political science 11%
Posts Replies Media Videos

Reposted by Lesley A. Hall, Patrick S. Forscher

rebeccasear.bsky.social
Rebecca Sear @rebeccasear.bsky.social · 8d
Richard Lynn was criticised for using a sample of children with intellectual impairment to calculate the "national IQ" of Equatorial Guinea in 2002. In 2019, he calculated the IQs of Native Americans from children referred for psychoeducational evaluations, including those with learning difficulties
Extract from page 160 of the 2019 book Intelligence of Nations by Richard Lynn and David Becker, which describes the 2019 version of Lynn's "national IQ" database, which claims to provide national average IQs for nation states worldwide: This extract appears in the section describing the samples used to calculate the US "national IQ": "From a rural area in Southeast Alaska a low-SES sample of Tlingit, mostly with learning difficulties, was reported by Connelly (1983, Table 1). WISC-R FS-IQs for younger participants of 86.89 and older participants of 83.88 were averaged to an uncor. IQ of 85.39, to which 0.68 had to be added for FE-correction but 2.50 had to be subtracted for country correction, which resulted in a cor. IQ of 83.57."

Reposted by Joanna Bryson, Dorothy Bishop, Brian J. Enquist , and 15 more

carnage4life.bsky.social
Dare Obasanjo @carnage4life.bsky.social · 10d
arXiv will no longer accept review articles and position papers unless they have been accepted at a journal or a conference and complete successful peer review.

This is due to being overwhelmed by a hundreds of AI generated papers a month.

Yet another open submission process killed by LLMs.
Attention Authors: Updated Practice for Review Articles and Position Papers in arXiv CS Category – arXiv blog
blog.arxiv.org

Reposted by Steve Peers, Jonathan Wolff, Ian Campbell , and 8 more

cbsnews.com
CBS News @cbsnews.com · 11d
A French cyclist survived for three days after a horrendous 130-foot fall into a ravine, kept alive by the bottles of red wine he had in his shopping bag, police said.
Cyclist falls down 130-foot ravine in France, survives 3 days by drinking wine he had in shopping bag
A helicopter airlifted him to hospital, with a rescue doctor calling his survival "a miracle."
cbsn.ws

Reposted by Patrick S. Forscher

lakens.bsky.social
Daniel Lakens @lakens.bsky.social · 11d
New paper finds that selective reporting remains the most replicable finding in science: journals.sagepub.com/doi/full/10.... I especially like their new exploratory metric 'p-values per participant'. Some papers had 11 p-values per participant! 🤯
Sage Journals: Discover world-class research
Subscription and open access journals from Sage, the world's leading independent academic publisher.
journals.sagepub.com

Reposted by Patrick S. Forscher

criminalerin.bsky.social
Erin Fogg says Free Link @criminalerin.bsky.social · 14d
"Introducing Doublespeed, an [Andreessen Horowitz backed] startup operating a phone farm to flood social media with AI-generated slop on behalf of its clients.

Doublespeed clients can expect to pay anywhere between $1,500 and $7,500 a month for access to its phone farm."
AI "Phone Farm" Startup Gets Funding from Marc Andreessen to Flood Social Media With Spam
Andreessen Horowitz has injected $1 million into Doublespeed, a startup meant to flood social media with gobs of for-profit spam.
futurism.com

Reposted by David Chester, Patrick S. Forscher, Matt N Williams , and 4 more

bsiepe.bsky.social
Björn Siepe @bsiepe.bsky.social · 19d
We built the openESM database:
▶️60 openly available experience sampling datasets (16K+ participants, 740K+ obs.) in one place
▶️Harmonized (meta-)data, fully open-source software
▶️Filter & search all data, simply download via R/Python

Find out more:
🌐 openesmdata.org
📝 doi.org/10.31234/osf...

Reposted by Efrén O. Pérez, James Grimmelmann, Patrick S. Forscher

mthrjo.bsky.social
Jo Kershaw @mthrjo.bsky.social · 22d
I have found an EXCELLENT meme for the church history lecture on Wednesday (which includes the Great Schism)
Pope Francis looks repelled by a diagram of the Orthodox theological position that the Son and the Spirit both proceed from the Father alone. In a second image, Pope Francis points approvingly at a diagram of the Filioque, in which the Spirit proceeds from the Father and Son together.

Reposted by David Lazer, Carl T. Bergstrom, Patrick S. Forscher

jbakcoleman.bsky.social
Joe Bak-Coleman @jbakcoleman.bsky.social · 22d
This new reporting from @jeffhorwitz.bsky.social is jaw-dropping and I think warrants a response from @cos.io given their ongoing collaboration with Meta.

Jeff writes about how Meta identified content-specific harms as a key problem for teens on Instagram. 🧵

www.reuters.com/business/ins...
Exclusive: Instagram shows more ‘eating disorder adjacent’ content to vulnerable teens, internal Meta research shows
Meta researchers found that teens who report that Instagram regularly made them feel bad about their bodies saw significantly more “eating disorder adjacent content” than those who did not, according to an internal document reviewed by Reuters.
www.reuters.com

Reposted by Brian M. Lucey, Ingo Rohlfing, Patrick S. Forscher , and 1 more

weratedags.com
WeRateDAGs @weratedags.com · 25d
Hello Bluesky!

We rate DAGs. Some are great. Some are... not so great. But we rate them all.

Let's start with a famous powerpoint hairball a.k.a. "the Afghanisdag", presented to Gen. Stanley A. McChrystal around 2010. His own rating?

1/10 "When we understand that slide, we'll have won the war"
"the Afghanisdag", a massive tangle of arrows and noun phrases, possibly describing a counterinsurgency problem: lethal within 20ft in briefing environments.

Reposted by Glyn Elwyn, Hugo Mercier, Ian Hussey , and 1 more

rebeccasear.bsky.social
Rebecca Sear @rebeccasear.bsky.social · 25d
Evidence that even when LLMs produce similar results to humans, they “rely on lexical associations and statistical priors rather than contextual reasoning or normative criteria. We term this divergence epistemia: the illusion of knowledge emerging when surface plausibility replaces verification”
PNAS
Proceedings of the National Academy of Sciences (PNAS), a peer reviewed journal of the National Academy of Sciences (NAS) - an authoritative source of high-impact, original research that broadly spans...
www.pnas.org

Reposted by Elizabeth Stokoe, Patrick S. Forscher

matthewfacciani.bsky.social
Matthew Facciani @matthewfacciani.bsky.social · 26d
New research finds that Twitter’s efforts to remove COVID-19 vaccine misinformation were largely ineffective and sometimes backfired. Skeptical communities grew more active and viral, and content quality declined over time. link.springer.com/article/10.1...
Explaining Twitter’s inability to effectively moderate content during the COVID-19 pandemic - Scientific Reports
Social media platforms routinely face pressure to restrict harmful content while protecting free speech; however, prior theory suggests that platform design might undermine the efficacy of content mod...
link.springer.com

Reposted by Patrick S. Forscher, Adrian Vickers

gregegansf.bsky.social
Greg Egan @gregegansf.bsky.social · 26d
“Well-meaning” people are using machine translation to write Wikipedia articles in languages that they don’t speak themselves, accelerating the degeneration in quality of the web corpus for several languages with relatively few native speakers.

www.technologyreview.com/2025/09/25/1...
How AI and Wikipedia have sent vulnerable languages into a doom spiral
Machine translators have made it easier than ever to create error-plagued Wikipedia articles in obscure languages. What happens when AI models get trained on junk pages?
www.technologyreview.com

Reposted by Patrick S. Forscher

scalzi.com
John Scalzi @scalzi.com · Oct 10
Well, good for DC

search.app/PK45E
DC Comics won’t support generative AI: ‘not now, not ever’
“People have an instinctive reaction to what feels authentic.”
search.app

Reposted by Jürgen Maier, Alexander Wuttke, Andrew K Przybylski , and 8 more

aufdroeseler.bsky.social
Lukas Röseler @aufdroeseler.bsky.social · Oct 10
ReplicationResearch.org is now open for submissions!

Submit replications and reproductions from many different fields, as well as conceptual contributions. With diamond OA, open and citable peer review reports, and reproducibility checks, we push the boundaries of open and fair publishing.

Reposted by Patrick S. Forscher, Veli‐Matti Karhulahti

dingdingpeng.the100.ci
Julia M. Rohrer @dingdingpeng.the100.ci · Oct 7
I really strongly feel that some fields of research would profit if researchers stopped collecting online data for some time and instead maybe just read a bit outside of their field.
The ‘harm hypothesis’ strikes me as being deeply rooted in contemporary WEIRD values rather than being the result of a specific ‘evolved’ or ‘innate’ instinct or psychological mechanism. And indeed the literature cited to support it seems to suggest this. Costello & Acerbi cite 5 papers in the paragraph above to support the model: Stewart-Williams et al., 2024: the sample here consists of Prolific users mostly in the UK. FeldmanHall et al., 2016: the samples were MTurk users in the US and volunteers in the UK. Curry et al., 2004: the sample is convicted offenders in Texas in 1991. Graso et al., 2023: US MTurk users again. Graso & Reynolds, 2024: this is a review paper which does make some cross-cultural claims, but when you check the references you can see some important limitations. For example, they write that “Across cultures, women were perceived as less powerful than men but were seen more positively,” and when you check the reference it goes to Glick et al., 2004, which samples from 16 nations. However, when you read that paper they note in the methods that, “Most samples consisted primarily of college students participating for extra credit.”

Reposted by Ian Hussey, Patrick S. Forscher

cchapman.bsky.social
Chris Chapman @cchapman.bsky.social · Oct 1
Excellent 🧵 about LLM synthetic data (silicon samples etc) and why they don't solve any particular problem in human research.

FWIW, in addition to results and considerations like these, I've argued elsewhere that the entire question is ill-formed: quantuxblog.com/synthetic-su...

Reposted by Patrick S. Forscher, Paul Goldsmith-Pinkham

vincentab.bsky.social
Vincent Arel-Bundock @vincentab.bsky.social · Sep 29
{tinytable} 0.14.0 for #RStats makes it super easy to draw tables in html, tex, docx, typ, md & png.

There are only a few functions to learn, but don't be fooled! Small 📦s can still be powerful.

Check out the new gallery page for fun case studies.

vincentarelbundock.github.io/tinytable/vi...
a table about lemurs a table about students and schools a table about wines

Reposted by Patrick S. Forscher, Brian Keegan

eikofried.bsky.social
Eiko Fried @eikofried.bsky.social · Sep 29
Intervening on a central node in a network likely does little given that its connected neighbors will "flip it back" immediately. Happy to see this position supported now.

"Change is most likely [..] if it spreads first among relatively poorly connected nodes."

www.nature.com/articles/s41...
Transformation starts at the periphery of networks where pushback is less - Scientific Reports
Scientific Reports - Transformation starts at the periphery of networks where pushback is less
www.nature.com

psforscher.bsky.social
Patrick S. Forscher @psforscher.bsky.social · Sep 29
The deadline to provide inputs into this piece of work is swiftly approaching -- 8 October.

Please consider filling out the survey we're using to structure people's input!

www.who.int/news-room/ar...

Reposted by Patrick S. Forscher

andreicimpian.bsky.social
Andrei Cimpian @andreicimpian.bsky.social · Sep 22
💖This paper has been ~11 years in the making - and probably my favorite project of all time. Thrilled to see it in @pnas.org! I'm so lucky that Zach decided to do a second PhD and join my lab @psychillinois.bsky.social back in 2014 - a fabulous scientist & human being! www.pnas.org/doi/10.1073/...
Historical and experimental evidence that inherent properties are overweighted in early scientific explanation

Reposted by Mel Bartley, Patrick S. Forscher

ent3c.bsky.social
Eric Turkheimer @ent3c.bsky.social · Sep 22
This study of intelligence in the UK Biobank is typical of a lot of current social science genomics. Impressive technically, and not over-interpreted. But still, a main result gets lost in the sauce. Within-families, the direct-effect polygenic score explains no more that 1-3% of the variance. /1
Imputation of fluid intelligence scores reduces ascertainment bias and increases power for analyses of common and rare variants
Studying the genetics of measures of intelligence can help us understand the neurobiology of cognitive function and the aetiology of rare neurodevelopmental conditions. The largest previous genetic st...
www.researchsquare.com

Reposted by Anna O. Law, Patrick S. Forscher, Catherine J. Frieman , and 1 more

mehr.nz
samuel mehr @mehr.nz · Sep 22
this is a very sharp piece on why it makes no sense to run universities as if they are businesses. They're not businesses.

www.afr.com/work-and-car...
The net result is the worst of both worlds. Universities invoke the rhetoric of business discipline, but they lack the governance structures that give that discipline bite. They operate without the checks that private ownership provides, yet subject staff and students to the cost-cutting and efficiency drives that profit-maximising firms pursue. The result is waste at the top and insecurity at the bottom.

Reposted by Patrick S. Forscher

iflis.bsky.social
Ivan Flis @iflis.bsky.social · Sep 18
Is analytical flexibility really the biggest problem while you’re confusing the ephemeral statistical effects of psychological processes with the ephemeral statistical effects of language prediction trained on massive data sets? Hah.

Reposted by Alexander Wuttke, Ian Hussey, Patrick S. Forscher

jamiecummins.bsky.social
Jamie Cummins @jamiecummins.bsky.social · Sep 17
Waiting for my preprint to be accepted, so in the meantime a teaser: here's what happens when you try to estimate a between-scale correlation based on LLM-generated datasets of participants, while varying 4 different analytic decisions (blue is the true correlation from human data):

Reposted by Rebecca Sear, Patrick S. Forscher, Clark Gray

mehr.nz
samuel mehr @mehr.nz · Sep 17
Nothing. I use it for nothing at all because AI is good at zero of the tasks I do regularly

Honestly I don't even know what its web address is, is it like a 2000s style ChatGPT.com or something funkier like chat.g.pt
What people use chatgpt for graph

Reposted by Gordon Hodson, Robert M. Ross, Mark J. Brandt , and 7 more

datingdecisions.bsky.social
Samantha Joel @datingdecisions.bsky.social · Sep 10
In a new paper, my colleagues and I set out to demonstrate how method biases can create spurious findings in relationship science, by using a seemingly meaningless scale (e.g., "My relationship has very good Saturn") to predict relationship outcomes. journals.sagepub.com/doi/10.1177/...
Pseudo Effects: How Method Biases Can Produce Spurious Findings About Close Relationships - Samantha Joel, John K. Sakaluk, James J. Kim, Devinder Khera, Helena Yuchen Qin, Sarah C. E. Stanton, 2025
Research on interpersonal relationships frequently relies on accurate self-reporting across various relationship facets (e.g., conflict, trust, appreciation). Y...
journals.sagepub.com

Reposted by Steve Peers, Patrick S. Forscher, Stacy D. VanDeveer , and 2 more

davidrvetter.bsky.social
Dave Vetter @davidrvetter.bsky.social · Sep 8
ICYMI: Google removes net zero goal from website.
Google quietly removes net-zero carbon goal from website amid rapid power-hungry AI data center buildout — industry-first sustainability pledge moved to background amidst AI energy crisis
Google's goal to be net-zero in carbon emissions by 2030 is still apparently company policy, it's just not broadcasting it anymore
www.tomshardware.com

Reposted by Timnit Gebru, Anna O. Law, Margaret Mitchell , and 44 more

olivia.science
Olivia Guest · Ολίβια Γκεστ @olivia.science · Sep 6
Finally! 🤩 Our position piece: Against the Uncritical Adoption of 'AI' Technologies in Academia:
doi.org/10.5281/zeno...

We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical
thinking, expertise, academic freedom, & scientific integrity.
1/n
Abstract: Under the banner of progress, products have been uncritically adopted or even imposed on users — in past centuries with tobacco and combustion engines, and in the 21st with social media. For these collective blunders, we now regret our involvement or apathy as scientists, and society struggles to put the genie back in the bottle. Currently, we are similarly entangled with artificial intelligence (AI) technology. For example, software updates are rolled out seamlessly and non-consensually, Microsoft Office is bundled with chatbots, and we, our students, and our employers have had no say, as it is not considered a valid position to reject AI technologies in our teaching and research. This is why in June 2025, we co-authored an Open Letter calling on our employers to reverse and rethink their stance on uncritically adopting AI technologies. In this position piece, we expound on why universities must take their role seriously toa) counter the technology industry’s marketing, hype, and harm; and to b) safeguard higher education, critical thinking, expertise, academic freedom, and scientific integrity. We include pointers to relevant work to further inform our colleagues. Figure 1. A cartoon set theoretic view on various terms (see Table 1) used when discussing the superset AI (black outline, hatched background): LLMs are in orange; ANNs are in magenta; generative models are in blue; and finally, chatbots are in green. Where these intersect, the colours reflect that, e.g. generative adversarial network (GAN) and Boltzmann machine (BM) models are in the purple subset because they are both generative and ANNs. In the case of proprietary closed source models, e.g. OpenAI’s ChatGPT and Apple’s Siri, we cannot verify their implementation and so academics can only make educated guesses (cf. Dingemanse 2025). Undefined terms used above: BERT (Devlin et al. 2019); AlexNet (Krizhevsky et al. 2017); A.L.I.C.E. (Wallace 2009); ELIZA (Weizenbaum 1966); Jabberwacky (Twist 2003); linear discriminant analysis (LDA); quadratic discriminant analysis (QDA). Table 1. Below some of the typical terminological disarray is untangled. Importantly, none of these terms are orthogonal nor do they exclusively pick out the types of products we may wish to critique or proscribe. Protecting the Ecosystem of Human Knowledge: Five Principles

psforscher.bsky.social
Patrick S. Forscher @psforscher.bsky.social · Sep 3
How should the behavioral sciences be mainstreamed into public health? How would we know if this goal is achieved?

With the WHO Behavioural Insights Unit, my team has been working on these questions.

Curious what we came up with? Check out the public consultation below

www.who.int/news-room/ar...

Reposted by Patrick S. Forscher

joft.bsky.social
Joshua Loftus @joft.bsky.social · Sep 2
This is what I've been saying since 2023 (image below)

"prediction: use of "AI" [...] will come to be broadly associated with cheating, deception, lack of respect for other people, and low quality work that cannot be trusted in important settings"