A psychology journal has retracted a paper on how using generative AI can affect confidence in work tasks after its sole author refused to share data with the editors.
The April paper, published in Technology, Mind, and Behavior, which is published by the American Psychological Association (APA), purported to describe the results of a survey of 1,923 adults recruited online to perform AI-assisted tasks. The paper was featured in an APA press release and articles in TIME and Futurism.
As we wrote in June, experts in the field began flagging issues with the paper’s data on PubPeer shortly after it was published. Sandra Grinschgl, who studies human-technology interaction at the University of Bern in Switzerland, was the first to do so, and later enlisted the help of colleagues Ian Hussey and Malte Elson. Their concerns included inconsistencies in graphs and results, study design, “erroneous” references and a lack of ethics approval.
According to the Aug. 31 retraction notice, the journal requested data as part of its investigation, and the author of the paper, Sarah Baldeo, and the approving research ethics body, ID Quotient Advisory Group, declined. According to the notice, Baldeo and ID Quotient said “the requested transfer of both anonymized participant-level materials and analytic syntax was not authorized under the participant-consent and Canadian [Research Ethics Board] framework governing the study either publicly or under a data-use agreement.”
Baldeo is the founder and CEO of ID Quotient Advisory Group, which provides “neuroscience-enabled IT & AI consulting,” according to its website. A self-described AI neuroscientist, she is also a Ph.D. candidate in AI and neuroscience at Middlesex University in England.
Baldeo told Retraction Watch “the article should be removed from that specific journal record because the journal’s requested post-publication verification pathway was incompatible with the participant-consent and Canadian REB framework governing the study.” She also said she does “not agree with all characterizations in the retraction notice.”
Grinschgl told us she, Elson and Hussey have not heard from the journal, despite flagging their concerns directly via email in May.
“Empirically, not sharing data is unfortunately normal and under the Canadian consent/REB framework there might be genuine limits on releasing participant-level data,” Grinschgl said. “However, even where participant data can’t be shared, that doesn’t extend to demonstrating the platform the study was run on, or to sharing the statistical/analytic code, neither of which should contain participant data.”
“Withholding those is not normal, it is odd to us. We still doubt that the data collection platform ever existed,” Grinschgl said. “Further, it seems like a remarkable loophole to us that an author can set up a company to grant their own ethics approval.”
Hussey helped create the INSPECT-SR reproducibility checklist for research papers, and psychologist Elson has been involved in assessing the trustworthiness of published research for over a decade. His work previously led to the retraction of a high-profile paper about the psychological impact of first-person shooter video games.
Richard Landers, the editor-in-chief of Technology, Mind, and Behavior, and Christopher Ferguson, the article’s action editor, both declined our request for comment.
The notice also states Middlesex “formally indicated that the research was not approved, conducted, or supervised by the University, and requested withdrawal of its affiliation from the article.” In a now-deleted LinkedIn post archived here, the university wrote about Baldeo’s “landmark study” and said it was “proud to see #TeamMDX leading the conversation on how AI is shaping the way we think and work.”
In a Bluesky thread about the retraction, Hussey flagged Baldeo’s book, 100 Ways to Future-Proof Your Brain, for containing possible hallucinated references. The book’s publisher, Arc & Edge, was acquired by ID Quotient, Baldeo’s company, on September 8, as announced in a press release written by the group, “Canadian Bestselling Author Buys Her Publisher.” Her book appears to be the only one published by Arc & Edge; however, she denied she owned the publisher before her company acquired it.
When we asked about the alleged fake references, she said she was “not able to respond to generalized social-media allegations.”
She also said she requested a retraction of the paper on June 1 “if correction was not available because I could not authorize disclosure of participant-protected materials outside the participant-consent framework and Canadian TCPS 2 research-ethics principles – which were disclosed pre-publication in January and February 2026 to the APA journal.”
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on X or Bluesky, like us on Facebook, follow us on LinkedIn, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].
I am one of the people named in the article.
> When we asked about the alleged fake references, she said she was “not able to respond to generalized social-media allegations.”
I did not provide specific examples on Bluesky of problematic references in the book, but will do so here. These are verifiable claims, not allegations.
The following references appear in the book’s reference list, starting on page 212. I provide the references verbatim.
Dall’Orso, S., Cellini, N., & Sarasso, S. 2021). Sleep and the Glymphatic System: Emerging Investigations into Brain Clearance. Frontiers in Neurology, 12, 730526. https://doi.org/10.3389/ fneur.2021.730526
– DOI does not resolve. I do not find an article of this title under a different DOI.
Ling, D. M. (2012). Finding stillness within the shaken brain. CMAJ: Canadian Medical Association Journal, 184(17), 1935-1936. https://doi.org/10.1503/cmaj.121366
– DOI resolves to an article by the same title but different author, issue, pages.
Hadlington, L. (2016). Cognitive failures in daily life: Exploring the link with internet addiction and problematic mobile phone use. Frontiers in Psychology, 7, 861. https://doi.org/10.3389/fpsyg.2016.00861
– DOI resolves to an article by different authors and title on an unrelated subject. An aritcle by this author and title, with a different journal, year, and DOI does exist.
Iacoboni, M. (2014). Understanding others: Imitation, language, empathy. Frontiers in Human Neuroscience, 295. https://doi.org/10.3389/fnhum.2014.00295
– DOI resolves to an article by different authors and title on an unrelated subject. A 2003 book chapter with this title and author may exist.
Hope, H., et al. (2021). Building cognitive resilience through uncertainty tolerance. Scientific Reports, 11, 10392. https://doi.org/10.1038/s41598-021-89980-0
– DOI resolves to an article by different authors and title on an extremely unrelated subject. I do not find an article of this title under a different DOI.
Lambert, G. W., Reid, C., Kaye, D: M., Jennings, G. L., & Esler, M. D. (2016). The importance of sunlight and vitamin D on human health and brain function. International Journal of Epidemiology, 45(5), 1453-1462. https://doi.org/10.1093/ije/dyw033
– DOI resolves to an article by different authors and title on an unrelated subject. I do not find an article of this title under a different DOI.
Brown, S. L., Brown, R. M., & Preston, S. D. (2013). Neural activation in the caregiving system in late-life parents. Psychology and Aging 28(2), 352-363. https://doi.org/10.1037/a0031576
– DOI resolves to an article by different authors and title on an extremely unrelated subject. I do not find an article of this title under a different DOI.
Monteiro, T., Marques, T., & Paton, J.J. (2020). Using temperature to analyze the neural basis of a latent temporal decision. bioRxiv. https://doi.org/10.1101/2020.08.24.264580
– DOI resolves to an article by different authors and title on an extremely unrelated subject. An article by these authors and title does exist.
Sturm, V. E., Brown, J. A., Verstaen, A., et al. (2023). Awe as a psychosocial resource: Its role in stress reduction and health promotion. Frontiers in Psychology. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10018061/
– DOI resolves to an article by different authors and closely related title on the same subject. I cannot find an aritcle by the listed authors and title.
Thank you for sharing all this useful info!
This sort of thing is why I loathe LLMs and will never, ever use them for anything, especially not related to my work. It’s hard enough to track down citations in papers written by humans because of the dozens of different ways of formatting references that have been used and the inevitable typos and genuine mistakes.
A bit of software making up fake ones is, to put it mildly, infuriating.
I just tried to go through and click many of the links you referred to cause I wanted to read the articles and all of those links seem to be linked to the wrong articles. Can you re-upload with the right links? I do appreciate the links and article information but I just wish I could get to those articles. Thanks again
The point of my post is that these links, which are copied verbatim from the author’s book, do not exist or link to the wrong articles. I.e., I provide them as evidence of problematic references that are consistent with being AI hallucinations.
We really appreciate you doing so! By providing the references verbatim, you gave us much depth of understanding of the issue. Your post is so powerful, it’s doubtful that the author will respond.
God almighty, the whole point of listing those citations is that THEY ARE FAKE.
Some might notice a pattern.
You are dealing with a smart cookie here. Fancy setting up your own ethics committee and a company behind the scenes to get a single publication. Wheels within wheels. Thank goodness for sleuths and Retraction Watch!
I’m glad the paper retracted it but it never should have passed peer review. I’m a PhD student, nowhere near an expert in methods and the whole thing was immediately suspicious. A paper with that size N would have been a massive undertaking to clean/code and should have generated a lot more information. I’ve handed in class papers with more rigor. And the issue with the study platform was ridiculous – what kind of platform tracks that level of interaction? The only way to have done it would have been to create a wrapper that calls an API but it doesn’t seem like that’s what occurred.
Reading the PeerPub thread and her shockingly ridiculous claims on LinkedIn about the fMRI scans are enough to make me question if she should ever have anything published again, much less receive a PhD after what appears to be such blatant misconduct.