std error coefficient

will always be given to you for regs? no chance to calc? thx

You may have to back into it to complete an ANOVA table - that just seems like a gimme though. I doubt they’d deal with any other calcs on this other than asking you to interpret std error. I think you should know that the std error for an autocorrelation is 1/T^0.5

what like they give you the tstat and say whats the std error?

Most of the problems I’ve seen don’t list it and require you to back into it. Like if they you give you the regression coefficient, the t-stat, and the number of observations and then tell you to calculate the 95% confidence interval. I think we’ll have to be able to do that.

dirtydirty Wrote: ------------------------------------------------------- > what like they give you the tstat and say whats > the std error? Yeah, pretty sure there was a question on this on the 2006 CFAI exam where you had to get it to find the confidence interval. There was some backing into things involved.

In the free sample its tested twice.

on a couple of question you had to derive it from the t stat given the regression coefficient if you really want to know how it’s calculated (like I was): http://www.analystforum.com/phorums/read.php?12,736010,736010#msg-736010