Scientific Research Is Unreliable, Unreliable Scientists Report

[Moros]

A friend of mine is a bio-chemical researcher and she told me they have to play with the results every time; scrapping the (outcomes of) experiments they do not like, limiting results to the ones that fit,... It's institutionalised really.

I also had a friend (he already held a degree in engineering) who when writing his phd (in psychology) always had to correct his professor's mathematical suggestions, ideas and all. It was clear the man didn't really understand the program he was working with, not the basic principles of the maths involved with that kind of research. He called it embarrassing. When offered a spot to work with research team of the professor afterwards, he obviously politely declined.

Fragony already mentioned the two cases of clear fraud used in Dutch research? Well, it doesn't paint a pretty picture does it? To me it's clear that the sciences who rely to much on statistics, especially as those fields aren't rich in people who are actually strong at maths. That and the mere fact that the first rule of statistics is, correlation doesn't imply causality. When people find a correlation they end up using their creativity to explain it. When people want to explain something they'll find a correlation. And even worse but apparently common practice, people will even manipulate results to achieve correlation or a different form of mathematical backing.