niraj_a Wrote: ------------------------------------------------------- > swaptiongamma Wrote: > -------------------------------------------------- > ----- > > No way we are dividing here by SEE. Ans has to > be > > C by the very fact that -1.5/0.45 = -3.333 > > > why do you divide the slope by the std error of > the coefficient? shouldn’t you divide the > coefficient by its std error? We’re trying to determine the significance of the slope coefficient only. SEE is for the entire regression.
swaptiongamma Wrote: ------------------------------------------------------- > niraj_a Wrote: > -------------------------------------------------- > ----- > > so .78 is not the coefficient? > > > > duly noted from now on: when they say > coefficient > > they mean the slope. i hate my ugrad stats prof > > for teaching us otherwise. > > > 0.78 is the intercept. right. she used to call it the coefficient term or the constant. C is easy to get now that that has been cleared up.
remember its a two tailed test and a significantly negative tstat implies the slope lies on the left hand side of the bell curve and that its significant = > hence significantly negative!
i don’t see how it can be anything other than C, unless there is more information that hasn’t been posted. coefficient t-statistic = coefficient/std. err. we got schwesed
oh n/m it is c
AudreyMwala Wrote: ------------------------------------------------------- > no its not c - wait for CP when he comes he will > solve it but keep trying Was this a decoy to do cpk in!!!