General Discussion
In reply to the discussion: Good science is good. Poorly designed science is bad. [View all]LostOne4Ever
(9,755 posts)In fact, there are quite a few lines. Falsifiability comes to mind almost instantly. Starting with a conclusion and searching for evidence being another. Here they are in detail:
http://rationalwiki.org/wiki/Pseudoscience
http://en.wikipedia.org/wiki/Pseudoscience
http://www.ithaca.edu/beins/methods/materials/char-pseudo.htm
I have not looked through but about 10 or so of the 88 pages of the notebook from your article but there are scientific reasons to exclude data beyond "confirmation bias." For instance, if you discover some sort of system error in the experiment and can show that it was influencing your data you USUALLY will want to drop that data and try again. This has nothing to do with confirmation bias.
Generally though when one refers to confirmation bias one is referring to people only counting evidence "confirms" their pre-existing conclusions and ignoring evidence to the contrary. Which is a bit different from thinking that there might be a problem with ones data because the variance from the average is greater than 2 standard deviations or because your precision is piss poor. Either way, science at least tries to reproduce said experiments to prove their verifiability and employs a variety of requirements to minimize this problem.
IOW, at least science does try to minimize confirmation bias as opposed to openly embracing it like pseudoscience does.