An academic study claiming positive effects of OpenAI-developed ChatGPT on student achievement has been retracted nearly a year after its publication. The paper, published by Springer Nature, was withdrawn due to inconsistencies in its analysis and lack of confidence in its conclusions.

The study in question had wide repercussions on social media and was accepted by many as gold standard data proving the benefits of ChatGPT in education. Ben Williamson from the University of Edinburgh stated that the article presented very striking claims and was quickly accepted in academic circles.

Methodological errors and doubts in the study

The retracted article analyzed the results of 51 different studies to measure the impact of ChatGPT on student performance, perception of learning, and higher-order thinking skills. The researchers claimed that they calculated the difference between groups using ChatGPT and control groups that did not use the meta-analysis method.

The results of the study suggested that ChatGPT greatly improved learning performance and positively affected the perception of learning. However, these findings faced serious criticism after they were published in the journal Humanities & Social Sciences Communications on May 6, 2025.

Ben Williamson stated that the study synthesized very low quality research and brought together studies with different methods incomparably. Williamson emphasized that it is not possible for so much quality work to be completed and published just two and a half years after ChatGPT’s launch in November 2022.

Residual effect after withdrawal

Although the article was retracted, it was cited 262 times in peer-reviewed journals within Springer Nature during its publication. The study, which collected citations from 504 different sources in total, reached approximately half a million readers and entered the 99 percentile among articles in terms of attention score.

This reveals how quickly flawed or poorly based research can spread in the academic world and have a long-lasting impact. In order to maintain credibility in academic publishing, such retractions are considered a critical step in cleaning up the scientific literature.

How do you think the fact that such mistakes are noticed so late in academic studies affects the education world?