If LLMs are increasingly used for data analysis, should we worry that they might encourage p-hacking?
A recent preprint suggests that the answer may be yes, at least in relatively complex analytical pipelines. I share some personal experience here 👇
www.linkedin.com/posts/stefan...
www.linkedin.com
If LLMs are increasingly used for data analysis, should we worry that they might encourage p-hacking?
A recent preprint, also featured in Science Magazine, suggests that the answer may be yes, at lea...
