Screencapped this from another board. I'm also going to post this to /lit/
Opinions on the state of the social sciences in the classroom setting, or within academia in general?
I think part of the reason you don't see smart people in social studies and liberal arts degrees is that they see the stuff this guy was talking about - professors selling their opinions and students unable to separate the objective from the subjective - and they get turned off to it up front.
The problem with changing this is that you need a lot of people to break the ice, if nobody ever calls the professors out for doing that kind of thing it will never change. The best possible solution, I think, is starting some kind of movement or group dedicated to fixing this, and collecting tons of members. I'm sure there a lots of people who agree about this, but because they're few and far between they'll just give in to the majority, assuming they were alone.
>>9164128
The fundamental problem with that is - on what platform? As much as a movement sounds nice, politics is very messy, and any new policy as such would be subordinated to larger ideological platforms which deal with broader issues with this particular problem with such a specific solution.
There's also the fact that students are primarily there to learn, which implies that they don't really know the material to begin with. Attacking your teachers on their lack of thoroughness just leads to students being punished for imprudence.
>>9164113
The social sciences went to shit when we got rid of the mathematics requirements training them to reason logically. If you look at the social scientists that aren't retarded, they all have heavily studied math and statistics.
>>9165494
This
This is actually the reason I went to econ from my previous major - I've seen through a lot of bullshit in the social sciences since then
>>9164113
Any meaningful question in the social sciences can be methodologically cast as some problem in data science and/or machine learning. It's underappreciated, but the greatest "social scientists" in existence right now are Facebook and Google. The proplem is that, frequently, they are focused on performance of their production models, rather than why they work so well to the exclusion of others.
Tools to understand the representations and statistical dependencies between variables learned by things like ANNs are just beginning to be a thing. To the extent that those in academia can capitalize on this, social science will advance.
Traditional hypothesis testing does not work for social science questions where both the vocabulary to discuss the problem and the problem itself are so ill-defined. First, you have to know the structure of the variables at play. This is where ML comes in which, unfortunately for humanities "researchers", involves math.