Erhao Ge
@erhaoworld.bsky.social
PhD (Human Biology and Evolution) of Evolutionary anthropology at UCL
“ Methods training is too often just something one is expected to soak up from colleagues and the published literature. PhD students have almost zero statistical training, and even less training in how to connect statistical analysis to theory.”
How can we reform science? I have some ideas. But I am not sure you’ll like them, because they don’t promise much. elevanth.org/blog/2025/07...
Which Kind of Science Reform
What hope is there for science reform, if we can't agree on what to reform? Right now, principles are more important than practices.
elevanth.org
November 10, 2025 at 9:19 AM
“ Methods training is too often just something one is expected to soak up from colleagues and the published literature. PhD students have almost zero statistical training, and even less training in how to connect statistical analysis to theory.”
Reposted by Erhao Ge
🚨 The Economist has been telling you for years that polygamy causes civil war by locking men out of marriage. A new article with @rebeccasear.bsky.social and @anthrolog.bsky.social explains that the demography of marriage markets doesn't actually work that way. 🧵
www.pnas.org/doi/10.1073/...
www.pnas.org/doi/10.1073/...
High rates of polygyny do not lock large proportions of men out of the marriage market | PNAS
There is a widespread belief, in both the scholarly literature and the popular press,
that polygyny prevents large numbers of men from marrying by ...
www.pnas.org
October 6, 2025 at 12:45 PM
🚨 The Economist has been telling you for years that polygamy causes civil war by locking men out of marriage. A new article with @rebeccasear.bsky.social and @anthrolog.bsky.social explains that the demography of marriage markets doesn't actually work that way. 🧵
www.pnas.org/doi/10.1073/...
www.pnas.org/doi/10.1073/...
Reposted by Erhao Ge
🚨𝗔𝘁𝘁𝗲𝗻𝘁𝗶𝗼𝗻 𝗲𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻 𝗲𝗻𝘁𝗵𝘂𝘀𝗶𝗮𝘀𝘁𝘀!
www.ehbea2026.com is now live for our 2026 European Human Behaviour & Evolution Association (EHBEA) conference in Leiden (NL🇳🇱) Check for 𝗮𝗯𝘀𝘁𝗿𝗮𝗰𝘁 𝘀𝘂𝗯𝗺𝗶𝘀𝘀𝗶𝗼𝗻, first 𝗽𝗹𝗲𝗻𝗮𝗿𝗶𝗲𝘀 & the [𝗔𝗜]𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻 𝘁𝗵𝗲𝗺𝗲 𝘁𝗿𝗮𝗰𝗸!
🗓️ 14–17 Apr 2026 | CBEN pre-conf 14 Apr
📍Pesthuis
#EHBEA2026
www.ehbea2026.com is now live for our 2026 European Human Behaviour & Evolution Association (EHBEA) conference in Leiden (NL🇳🇱) Check for 𝗮𝗯𝘀𝘁𝗿𝗮𝗰𝘁 𝘀𝘂𝗯𝗺𝗶𝘀𝘀𝗶𝗼𝗻, first 𝗽𝗹𝗲𝗻𝗮𝗿𝗶𝗲𝘀 & the [𝗔𝗜]𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻 𝘁𝗵𝗲𝗺𝗲 𝘁𝗿𝗮𝗰𝗸!
🗓️ 14–17 Apr 2026 | CBEN pre-conf 14 Apr
📍Pesthuis
#EHBEA2026
Overview | EHBEA2026
www.ehbea2026.com
October 10, 2025 at 10:27 AM
🚨𝗔𝘁𝘁𝗲𝗻𝘁𝗶𝗼𝗻 𝗲𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻 𝗲𝗻𝘁𝗵𝘂𝘀𝗶𝗮𝘀𝘁𝘀!
www.ehbea2026.com is now live for our 2026 European Human Behaviour & Evolution Association (EHBEA) conference in Leiden (NL🇳🇱) Check for 𝗮𝗯𝘀𝘁𝗿𝗮𝗰𝘁 𝘀𝘂𝗯𝗺𝗶𝘀𝘀𝗶𝗼𝗻, first 𝗽𝗹𝗲𝗻𝗮𝗿𝗶𝗲𝘀 & the [𝗔𝗜]𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻 𝘁𝗵𝗲𝗺𝗲 𝘁𝗿𝗮𝗰𝗸!
🗓️ 14–17 Apr 2026 | CBEN pre-conf 14 Apr
📍Pesthuis
#EHBEA2026
www.ehbea2026.com is now live for our 2026 European Human Behaviour & Evolution Association (EHBEA) conference in Leiden (NL🇳🇱) Check for 𝗮𝗯𝘀𝘁𝗿𝗮𝗰𝘁 𝘀𝘂𝗯𝗺𝗶𝘀𝘀𝗶𝗼𝗻, first 𝗽𝗹𝗲𝗻𝗮𝗿𝗶𝗲𝘀 & the [𝗔𝗜]𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻 𝘁𝗵𝗲𝗺𝗲 𝘁𝗿𝗮𝗰𝗸!
🗓️ 14–17 Apr 2026 | CBEN pre-conf 14 Apr
📍Pesthuis
#EHBEA2026
Reposted by Erhao Ge
Me and @watarutoyokawa.bsky.social want to thank everyone who attended, presented at, and helped organize #COSMOS2025! If you weren't able to make it, all teaching materials are already online and we'll be uploading videos of all talks soon cosmossummerschool.github.io/materials/
October 6, 2025 at 4:15 AM
Me and @watarutoyokawa.bsky.social want to thank everyone who attended, presented at, and helped organize #COSMOS2025! If you weren't able to make it, all teaching materials are already online and we'll be uploading videos of all talks soon cosmossummerschool.github.io/materials/
Kicked off #COSMOS2025 brilliant summer school on social learning and collective behaviour. First day absolutely packed with ideas ✅
September 30, 2025 at 10:46 AM
Kicked off #COSMOS2025 brilliant summer school on social learning and collective behaviour. First day absolutely packed with ideas ✅
Reposted by Erhao Ge
Key dimensions of social learning problems. @watarutoyokawa.bsky.social giving the first tutorial of #COSMOS2025 As always, course materials are opening available on our website cosmossummerschool.github.io/materials/
September 30, 2025 at 2:19 AM
Key dimensions of social learning problems. @watarutoyokawa.bsky.social giving the first tutorial of #COSMOS2025 As always, course materials are opening available on our website cosmossummerschool.github.io/materials/
Reposted by Erhao Ge
Can large language models stand in for human participants?
Many social scientists seem to think so, and are already using "silicon samples" in research.
One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.
THREAD 🧵
Many social scientists seem to think so, and are already using "silicon samples" in research.
One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.
THREAD 🧵
The threat of analytic flexibility in using large language models to simulate human data: A call to attention
Social scientists are now using large language models to create "silicon samples" - synthetic datasets intended to stand in for human respondents, aimed at revolutionising human subjects research. How...
arxiv.org
September 18, 2025 at 7:56 AM
Can large language models stand in for human participants?
Many social scientists seem to think so, and are already using "silicon samples" in research.
One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.
THREAD 🧵
Many social scientists seem to think so, and are already using "silicon samples" in research.
One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.
THREAD 🧵
Reposted by Erhao Ge
📣 Hot off the press at #ProcB!
🔍 Our study checked data/code-sharing policies in 275 eco/evo journals and compliance in Proc B (n=2,340) & Ecology Letters (n=571). Policies exist, but clarity and strictness vary, affecting reproducibility.
🔗 doi.org/10.1098/rspb...
🔍 Our study checked data/code-sharing policies in 275 eco/evo journals and compliance in Proc B (n=2,340) & Ecology Letters (n=571). Policies exist, but clarity and strictness vary, affecting reproducibility.
🔗 doi.org/10.1098/rspb...
September 17, 2025 at 6:30 AM
📣 Hot off the press at #ProcB!
🔍 Our study checked data/code-sharing policies in 275 eco/evo journals and compliance in Proc B (n=2,340) & Ecology Letters (n=571). Policies exist, but clarity and strictness vary, affecting reproducibility.
🔗 doi.org/10.1098/rspb...
🔍 Our study checked data/code-sharing policies in 275 eco/evo journals and compliance in Proc B (n=2,340) & Ecology Letters (n=571). Policies exist, but clarity and strictness vary, affecting reproducibility.
🔗 doi.org/10.1098/rspb...
Reposted by Erhao Ge
Nature research paper: One mother for two species via obligate cross-species cloning in ants
go.nature.com/45PYV6d
go.nature.com/45PYV6d
One mother for two species via obligate cross-species cloning in ants - Nature
In a case of obligate cross-species cloning, female ants of Messor ibericus need to clone males of Messor structor to obtain sperm for producing the worker caste, resulting in males from the same mother having distinct genomes and morphologies.
go.nature.com
September 5, 2025 at 10:42 AM
Nature research paper: One mother for two species via obligate cross-species cloning in ants
go.nature.com/45PYV6d
go.nature.com/45PYV6d