You do not understand my point or the point of statistics in general, it’s not the question of analyzing a single piece of data or the life of one person in such depth that you can predict what that person or post will do or why that person or post is doing that.
It is a matter of having very simple metrics that you can use for statistical analysis, let me give you an example, I as a programmer can tell my program to count swear words in each post and assign a vulgarity value on a scale of 1-5. The machine has absolutely no need to be sentient or posses cognitive abilities. Then I can sift through 1mil posts, record their likes and come up with a value that will tell me how many of highly vulgar posts are liked, comparing this to how many of other posts are liked, I have learned what my chances of getting liked are given that I write a vulgar post.
As you can see, it’s a methodological approach, it’s a mathematical science and the metrics can be obtained with very little need for understanding.
/edit
Let me put this in simple terms, the post in my hypothetical example were liked because they were vulgar, period, because that is what I tested for.
OMG, that’s all I’ve been saying the whole time! What exactly are you trying to argue about? I’ve only been talking about why people hit the like button and how statistics won’t provide that.
You do realize that swear words can vary from region to region? A swear word to one person is not a swear word to another, even in the same language. Your numerical methods may be methodical and flawless but you initiated it with a cognitive assumption.
Wait wait wait…now you’re contradicting your first point by claiming to know why people liked your hypothetical example. Just because that’s your stated metric that you’re testing for, doesn’t mean that your conclusion is true. You must go to great lengths to control all the variables at hand for that statistic to have any actual meaning. Simply counting swear words in a sentence and then determining popularity of a post because of its vulgarity ignores all other aspects of a post. That’s like trying to determine if peanut butter gives you cancer while ignoring the fact that your entire control group smokes 10 packs of cigarettes a day.
If there was little need for understanding when you set up the test, then why would you expect any meaning derived from it to be valid?
Here let me highlight the point that you made and should really just stick to:
Your statistical analysis showed you that you have an X% chance for words that you have deemed vulgar to get liked should you post them in any of the locales that you have sampled.
However, saying that they were liked BECAUSE they were vulgar, period, without having controlled your test sufficiently to rule out the other variables is an inherently flawed conclusion.
You do see the difference, right? One states probabilities of a post with vulgar content getting liked for a defined set of circumstances, while the other establishes a blanket reason, a conclusion, a “Why” for liking the posts.
You know what, maybe I should deny your existence and your argument until you can prove to me WHY you exist and of course I’m not going to be swayed from your non-existence by mere scientific proof, since of course a picture of you or your birth certificate or argument of evolution doesn’t prove WHY you exist.
I was told if I could get enough likes from Ark I could have sex. Now Ark does this. Finding out it was just a joke has ended any prospect of that, THEN he takes it back. Too late for me. Thanks for nothing
