The dataset of "Diagnostic efficacy of Kimi and DeepSeek in detecting statistical results errors in biomedical journals and suggestions for the use"

The dataset records the 10 “GenAI model + prompt method” strategies mentioned in our paper titled "Diagnostic efficacy of Kimi and DeepSeek in detecting statistical results errors in biomedical journals and suggestions for the use", particularly the "Statistical Rules" we summarized, as well as the terminology prompting methods we applied to the 21 positive and negative samples, for the reference of our peers.

At the same time, we hope more colleagues will pay attention to the issue of statistical errors in biomedical journals and work together to improve the comprehensiveness and accuracy of the "Statistical Rules" and “terminology prompting”, thereby further enhancing GenAI's capability to assist in the review of statistical result errors.To prevent misuse, this dataset is available upon request.This version also provides information on the cases and controls used in the articles, including the types and quantities of statistical errors, as well as GenAI diagnostic findings.