when I was watching lv2 schweser. It said using n-2 degrees of freedom when using T-distribution. But When I was watching FRM schweser and lv1 text. It said using n-1 degrees of freedom. So confusing!! Any one could help me on this as I have no statistic background…
It depends on where you are using it. Basically the number of df = n - # parameters you are estimating. For example: a) Usual one-sample test of mean: estimate mu with X-bar => df = n - 1 b) Estimate linear regression: estimate intercept and slop => df = n - 2 c) Estimate quadratic regression: estimate intercept, slope, and curvature = > df = n -3 and so on…
is this level III material ?