HappySqurriel said:
The social sciences (for the most part) don't use statistics, they abuse it. Which is quite remarkable because statistics is an unfortunately necessary abomination of mathematics. When using statistics correctly you're always very open about the uncertainty of what you're saying because you can not prove anything is true with statistics; you can only demonstrate that they're likely to be true. The vast majority of papers published in the social sciences would never be published in a reputable scientific journal because they lack statistical rigour; and in many (probably most) cases this is on purpose because they would never be able to make the claims they're making if they had to approach the problem in a scientific way. |
You've had a go at social science statistics before, i remember. My response would be that much of the lack of scientific rigor simply comes from the difficulty of doing original research (e.g. hugely extensive polling that would necessary for some things). For instance, my paper relies heavily on Freedom House democracy indeces, and so is only as scientifically sound as their methods are, due to my lack of resources as an undergraduate.
Monster Hunter: pissing me off since 2010.