30k medical studies may be invalid due to contaminated cell cultures

Re:30k medical studies may be invalid due to contaminated ce... by arkh on Thursday October 19, 2017 @03:21AM (#15506056)

> we found a very interesting biological pattern in the data

Isn't it related to p-hacking? Get lot of data, search for a pattern then publish some theory the pattern would "prove". Instead of having a theory first, setting up experiments which can disprove it then analyzing the data from those experiments.

[ Parent | on HN ]

Re:30k medical studies may be invalid due to contaminated ce... by biomcgary on Thursday October 19, 2017 @04:16PM (#15511198)

The observed pattern exactly fit into a previously established hypothesis and was totally unexpected. I did not slice and dice the data with multiple statistical tests. The only way it could have been related to p-hacking is that I am familiar with a bunch of biological theories and could recognize data patterns that are compatible with any of them. So, maybe a Bonferroni-type correction of 30-70x (i.e., number of theories my internal neural net has been trained on)?

[ Parent | on HN ]

Re:30k medical studies may be invalid due to contaminated ce... by mercer on Thursday October 19, 2017 @08:21AM (#15507241)

I'd say if they were explicit about using the data to construct a theory, it's okay, no?

[ Parent | on HN ]