>>12365845>go ahead and point it outAlready found two in engineering. I did contact the publisher, but got no reply.
I know this is what science is supposed to do, point out errors, correct them, move on. This is not reality in our publish or perish culture anymore. Nobody wants to admit mistakes, as it can cost you a career.
>what is made up in any of these studies.You can post those links by getting rid of the dashes in some way.
Most studies im talking about have conductor bias, low samples rates, or biased population samples.
Then there are also biology papers. I found two where the data was published alongside the actual research and quickly computed the mean and standard deviations for a few tests they performed (both were in stem cell research, geared towards angel disease). The mean of their datasets agreed with theoretical values, but their standard deviations were around 80 - 120%. In other words, their data was meaningless.
This happens all the time. Check out p-hacking or generally the reproducibility crisis. I work in the cross-section of machine learning and engineering. Machine learning papers are often a huge load of horse shit and not reproducible.
