t stats

i noticed the two formulas for calculating the t stat. one is [r*(n-2) ^ .5] / [1-r^2]^ .5 the other is the one thats like b0/std error. I am having trouble understanding the difference here. Is the top one used only for testing significance of the correlation coefficient? when do you use each one?

that equation is the test for correlation coefficient, you remember, how strongly two variables interact with each other from -1 to 1 the other equation to test regression coefficients in a regression is (you should of learned this on level 1). (H0 - Coefficient)/standard error of coefficient There are other tests out there, Chi-squared test, autocorrection test for autoregressive models, F-tests, Bruesh-Pagen, etc. These are the big ones, you need to learn them all.

word. yeah i havent visited this section in more than a month so im all hazy. thats the primary thing i think i need to remember is that the one with r in it is for the correlation coefficient and the other is for the regression coefficients in a regression (which i forgot from level 1).

I wouldn’t worry about it too much. I think they did to test correlation coefficient last year (the equation you posted). Quant last year was like 10% of exam, they asked question mostly related to mutli-regressions, problems with the standard error (autocorrelation, mutlicollinearity, heterodaskity), and seasonality. You can do poorly on quant and still pass so you’ll have to figure out your strategy.

thanks for the advice and help BiPoBoy. much appreciated