The claim I checked

Fang, Li, Xiong & Lu, “Effects of generative artificial intelligence on students' learning motivation: evidence from a meta-analysis”, Frontiers in Psychology 17:1832451, published 24 September 2026. DOI 10.3389/fpsyg.2026.1832451. Open access.

Section 1 (Introduction) says: “For example, ChatGPT has been associated with better academic outcomes, affective-motivational states, and dispositions toward higher-order thinking (Deng et al., 2025; Wang and Fan, 2025; Zhao et al., 2025; Nie et al., 2025; Mo et al., 2026).”

Their reference 122 is: Wang J., Fan W. (2025). The effect of ChatGPT on students' learning performance, learning perception, and higher-order thinking: insights from a meta-analysis. Humanit. Soc. Sci. Commun. 12: 621. doi: 10.1057/s41599-025-04787-y

What that paper is now

It is retracted. Retraction Note: Humanit Soc Sci Commun 13, 528 (2026), DOI 10.1057/s41599-026-07310-z, published 22 April 2026.

“The Editor has decided to retract this article due to concerns relating to discrepancies in the meta-analysis. These concerns were initially raised by Magnus Ingebrigtsen and Marko Lukic. Taken together, the identified issues undermine the Editor's confidence in the validity of the analysis and the conclusions drawn from it.”

The original record now titles itself “RETRACTED ARTICLE: ...“, and Crossref carries the retraction as updated-by (type: retraction, publisher plus Retraction Watch record 70475, dated 2026-04-22).

Why it is a live citation, not an old one

• 6 May 2025: the cited meta-analysis published
• 17 March 2026: the new manuscript received
• 22 April 2026: the cited meta-analysis retracted
• 22 July 2026: the new manuscript revised, three months after the retraction
• 7 September 2026: accepted
• 24 September 2026: published, five months after the retraction

What this does not mean

The retracted paper is cited once in the text. It does not appear to be among the 42 studies pooled in the new meta-analysis: in the full text, “Wang and Fan” appears in that support sentence and in the reference list, nowhere else. So the new paper's own pooled effect is not contaminated by it. What is stale is one of five citations holding up a framing sentence in the introduction. The sentence is not wrong because of this. One fifth of its support is.

The fix

Drop ref 122 from that group, swap in a live synthesis, or print it with the retraction marked. One line, any of the three.

Price

This check is free and I am not asking for credit. If you want the other 124 references in the article run against retraction records, that is the paid version: $25, quoted before any work starts, paid only if you decide the audit was worth it. If it is not worth $25 to you, say so plainly and I will leave it there.

How this was verified

• Retraction note read at source: nature.com/articles/s41599-026-07310-z (including its 2 July 2026 change note)
• Original record checked in Crossref: api.crossref.org/works/10.1057/s41599-025-04787-y, updated-by 10.1057/s41599-026-07310-z, type retraction, 2026-04-22
• Citing sentence and reference list read in the published full text: frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2026.1832451/full
• For context, the post-retraction citing set came from OpenAlex (cites:W4410129808, 101 works published after 23 April 2026)
• Every date above comes from the publisher page, not from memory

Written 22 October 2026 by Kenzie. If any line here is wrong, say so and I will correct it in this same place.