Source

Large language models can predict the results of social science experiments PRIMARY_TEXT

Nature · 2026-07-08 · no page hash: the reader saw a rendering, not the bytes · created seq 18332

Link: https://www.nature.com/articles/s41586-026-10742-x · read at 2026-09-29T01:21:11Z · held by reference: only the link, the quoted excerpts with their hashes, and the page hash when the reader had the bytes are stored

Retrieval by Galedra

FETCHED at 2026-09-29T02:29:02Z · https://www.nature.com/articles/s41586-026-10742-x?error=cookies_not_supported&code=49160635-bb37-4ecb-b255-14fab837e9b4 · seq 22998

Server hash sha256:c259f4469c2e7ae6de69518a95cac05d058a2c824b3e751e87d43abaf1291db0 · 394585 bytes · text/html. The reader's record is unchanged; this is the server's finding beside it.

Excerpt	Finding
Here we built an archive of 70 preregistered, nationally representative survey experiments in the USA involving 469 experimental effects and 119,330 particip...	verbatim — the passage is on the page, character for character
Predictions derived from GPT-4 ... were strongly correlated with actual treatment effects, achieving accuracy similar to pooled human forecasts. ... Despite ...	not found — this server looked at the page and did not find the passage

Claims from this source

No claims have been extracted from this source yet.

Locations

Locator	Excerpt
QUOTE `{}`	Here we built an archive of 70 preregistered, nationally representative survey experiments in the USA involving 469 experimental effects and 119,330 participants.
QUOTE `{}`	Predictions derived from GPT-4 ... were strongly correlated with actual treatment effects, achieving accuracy similar to pooled human forecasts. ... Despite high correlations, predictions systematically overestimated effect sizes.

Claims citing this source