So I'm taking Exam C for the 100th time because I'm dumb. And I am having a lot of trouble with biased/unbaised forms of variance. Can someone help? Here's how I'm thinking about it.
Method of Moments says Var(X) = EX^2 - (EX)^2 Now this is the same as Sum(x-mean)^2 / n, right? And that would be biased?
If we use Sum(x-mean)^2 / (n-1) then we get the unbaised variance. To get the unbaised from the Method of Moments, we could multiply Method of Moments version by n/(n-1).
Also variance of xbar is sigma^2/n. In this case, what is the sigma squared? Is it Sum(x-mean)^2 / (n-1)?
also, population is /n and sample is /n-1. but how can you tell if you should use sample of population?
Sorry if these are basic, stupid questions, but I've been through the manual and still continually screw this up. And if I fail this exam again I'm going to quit being an actuary and be homeless.
Method of Moments says Var(X) = EX^2 - (EX)^2 Now this is the same as Sum(x-mean)^2 / n, right? And that would be biased?
If we use Sum(x-mean)^2 / (n-1) then we get the unbaised variance. To get the unbaised from the Method of Moments, we could multiply Method of Moments version by n/(n-1).
Also variance of xbar is sigma^2/n. In this case, what is the sigma squared? Is it Sum(x-mean)^2 / (n-1)?
also, population is /n and sample is /n-1. but how can you tell if you should use sample of population?
Sorry if these are basic, stupid questions, but I've been through the manual and still continually screw this up. And if I fail this exam again I'm going to quit being an actuary and be homeless.
Variance in Exam C: bias and method of moments?