>>11325765I encountered two studies that I attempted to replicate and both papers had fatal errors.
One made simple errors in specified data (which I caught) I explained this (misspelled DNA) as a typo, albeit a total failure (key sequence in question.) After inferring the correct sequence, I was able to replicate reliable results. Attempts to establish connection to the authors across the Great Firewall were fruitless. The subject being esoteric, I did not attempt to publish my results.
The other paper reported success with a technique daignostic test to differentiate 4 closely related species if insect. In this case, the test subject DNA I sourced from other places invalidated the diagnostic for these species. In this case, the study attempted to establish a diagnostic for a species, but failed as it was not diagnostic for my controls.
This author was awarded a graduate degree and won a prestigious institutional job, and to my knowledge never returned to the esoteric subject. I established contact and communication with one of the co-authors and an expert in the field who acknowledged the problems and with another grad student was sourcing a more diverse set of samples and trying again. My results, being negative, were never published, nor did I try.
So keep in mind, /Sci/-tards, that what you will find in the literature, are (mostly) positive results, a structural bias in the publish/peer review process, as well as FALSE positives which you cannot easily detect. While you can't assume a particular study isn't a false positive, you should be aware that there are many reasons an experiment isn't replicating having nothing to do with what you are doing and if nothing else is working, investigate further.
