technology
AI Tools Drive Surge in Papers Analyzing Public Health Datasets
Researchers have traced a surge in low-value studies generated with AI assistance that cite the Global Burden of Disease study and NHANES survey. Journal editors report doubled submissions and growing difficulty distinguishing human from machine-generated work. OpenAI released its Prism paper-writing assistant earlier this year.
May 15, 9:58 AM·3m read1 source
The VergeAudio version
Tap play to generate a narrated version.
Developing·Limited corroboration so far. This page will refresh as more sources emerge.
A 2017 paper that assessed the accuracy of a particular type of statistical analysis on epidemiological data received a few dozen citations over the years before experiencing a sudden increase last summer. Peter Degen, a postdoctoral researcher at the University of Zurich Center for Reproducible Science and Research Synthesis, was asked by his postdoctoral supervisor to investigate why the paper was suddenly being cited every few days, hundreds of times.
Degen found that the citing papers all analyzed the Global Burden of Disease study compiled by the Institute for Health Metrics and Evaluation at the University of Washington.
The papers churned out predictions on topics ranging from stroke likelihood among adults over 20 to falls among elderly people in China. Searching on GitHub for related code, Degen discovered a Guangzhou-based company on Bilibili touting tutorials on producing publishable research in under two hours using its software tools and AI writing assistance.
Researchers who analyzed a subset of studies about headaches found they were rife with errors and misrepresentations.
The papers were not as obviously flawed as earlier AI-generated work, making them harder to filter out. “It’s a huge burden on the peer-review system, which is already at the limit,” Degen said. ” Matt Spick, a lecturer in health and biomedical data analytics at the University of Surrey and an associate editor at Scientific Reports, received three strikingly similar papers analyzing the US National Health and Nutrition Examination Survey (NHANES).
There was a sudden explosion in papers citing NHANES that followed a similar formula purporting to discover associations such as eating walnuts and cognitive function or drinking skim milk and depression. “If you’ve got enough computing power, you go through and you measure every single pairwise association, and eventually you find some that haven’t been written on before and you just publish,” Spick said.
