> It’s a very simple matter of doing the correct comparisons using the correct methods. Well .. if theres 25,000 lines of 200 different variables and 45 of them correlate, how do you pick which are meaningful? "People who are tall really like rice!". Meh.
Certain assumptions about the world are made.
>you very easily dismiss things that don’t sit within your realm of experience or competence or are not consistent with your belief system Sorry, I do understand statistics pretty well and frankly this is how statistics does work. You have the data and you apply your belief system to it. That's called "your hypothesis".
>Same with the intelligence review. >on the pay equity issue >You just have to know what you are doing You see? You basically admitted it - you just have to know which correlations to ignore.
To a certain extent when you want to look at a very large, multivariate data set you really have to apply real world observations (you can't build your hypothesis off nothing). I can tell you that in my real world experience, people who live their job - come in early - leave late - don't know anything except work - get paid the most.
Male OR female - if they have a picture of their kids on their work desk - if they talk about their kids a lot - they don't get paid as much. They don't live to work, they work to have money for their kids. People who get paid a lot, I never hear them say "my kid is sick I have to leave". You see?
And women tend towards having this type of "my kids are my entire life" attitude. (Which is pretty dang reasonable considering, I'm not being critical of the attitude.) If I was looking at that type of data that would be the type of observation that would drive how I put variables together.
This would be a good separate thread, although, contentious and one sure to draw the ire of the leftist hate mob here on HC.
