Sort:  

Well, I've also looked at some research about aging and how it affects mental faculties and so on and while it was quite interesting, they had a sample size in the tens, so different factors they were assessing like education and doing mental exercise like sudoku and crosswords regularly were represented by groups of literally a few people. I don't know how you write about your "conclusions" with a straight face when the variable you are looking at is represented by 4 people or something. And don't get me started on the we had 24 subjects, but we compiled our analysis on 21 because 3 were outliers (read including them would have made our study even more useless and our hypothesis even less substantiated) and we excluded them. But 3 people are more than 10% of your sampling...

I understand finding experiment subjects is hard, but working with sample sizes that are simply to small to be representative about anything simply ads no real value to your field and just obscures reality with a pile of papers to cite with claims and conclusions that have not actually been properly substantiated.

they get funding. that's how they write "conclusions" with a straight face.