Quote Originally Posted by Disciple View Post
As someone without a stats degree I wonder about this. There is a lot of math I am not smart enough to understand, but what I do manage to follow becomes useful and illuminating rather than confusing. Naively it appears to me that statistics is fundamentally about understanding and describing what you are sampling and the biases of the way in which that sample is taken. How does this high school perspective differ from statistics at an advanced level?
Basic stats gives the foundation. Everything thing else is application at higher levels. How do you figure bias? How do you get a valid sample? How do you figure the Z score? Are there null sets? What's the p value and how many standard deviations are you away from the mean? How often are you dumping your data and running the analysis?

Usually the numbers are not the problem. The problem is in understanding the methodology and the techniques researchers use to come up with their data.

I have been published, never alone and always with a handful of other journal authors. One thing most of us do not do is statistics. We work with the research people and they will tell us the best methodology, and then we get the data and give it to them to run through their software. And the more frequently you run the numbers, the better the result tends to be. Just like predicting the weather.

But I understand how to read a research article, read the methodology, understand the application and how they reached the conclusions.