Post

Was this study performed and written by ChatGPT itself? Just poking at it for 20 minutes: 1) Table 3, the entry for Bašić et al. (2023a) says n=12. That paper used n=9 2) Almohesh (2024) used the "Chat GBT" (?), different learning methods, gamification, should have been excluded

nature.comnature.com

Nicole Hennig@nic221.bsky.social

The effect of ChatGPT on students’ learning performance, learning perception, and higher-order thinking: insights from a meta-analysis - Humanities and Social Sciences Communications https://www.nature.com/articles/s41599-025-04787-y #AI #education

More: 3) Same table 3, there are 2 entries for Chen and Chang, apparently 2 different studies, but they point to the same paper

10/5/25, 4:32 pm0 reposts2 likes

4) Same table, Hsu (2024), entry says CG=20/EG=20. That paper says 20/40

5) Hu (2024), entry says 67/68. It's 35/35.

I've seen enough. The paper in question is this one: www.nature.com/articles/s41...

nature.comThe effect of ChatGPT on students’ learning performance, learning perception, and higher-order thinking: insights from a meta-analysis - Humanities and Social Sciences CommunicationsHumanities and Social Sciences Communications - The effect of ChatGPT on students’ learning performance, learning perception, and higher-order thinking: insights from a meta-analysis

oh, there's even an entry on that table that doesn't have a paper. lovely.