If you have watched the strange, sad decline of psychology over the last decade, you may have started to wonder if any conclusions are legitimate. With evolutionary psychology claiming voting Republican is an adaptive function and social psychologists claiming people can predict the future, there is a lot of woo getting published.
While scientists are sort of gracious about it - psychology departments are usually in the humanities buildings, not the science ones - the downside science is only starting to realize is that it damages the credibility of all science to have nonsense lumped in.
Motivated reasoning, selectively excluding subjects, amending the experimental procedure after designing the study, all those things are biasing studies to get more positive findings - and no one cares about replicating results, since they are too busy doing their own surveys of undergraduate students to bother.
The good news is, the fraud is starting to get caught and that is primarily due to young researchers who thought their field was trying to be science but found their elders are more like carnival hucksters. Sure, there is still a lot of political science 'your voting is in your genes' tripe making it through psychology peer review but that is a fad brought on by the wide-open possibilities of epigenetics. 20 years from now those people will be laughed at, and rightfully so.
Tia Ghose at LiveScience has the takedown of psychology and also highlights some of the positive ways the ethical researchers are instituting accountability.
Psychology Literature Littered With Conclusions That Aren't Real