G*****t 发帖数: 57 | 1 Hi, all,
I have a question about calculations for two independent sample...
Suppose two independent sample (A, B), only CV for each is given (CV_A, CV_B
), df also given (dfA, dfB). And we want the CV for the total population...
Here's what I thought:
Var(A+B) = VAR(A)+VAR(B)
Assume mean_A = mean_B,
CV(A+B) = sqrt(CV(A)^2+CV(B)^2)
Is there any problem for this calculation?
Shall I consider df values too as a weight? Say,
CV(A+B) = sqrt((dfA*CV(A)^2+dfB*CV(B)^2)/(dfA+dfB))
Thanks a lot!! | G*****t 发帖数: 57 | | s*********y 发帖数: 34 | 3 Var(A+B)=Var(A)+Var(B)+Cov(A,B)=Var(A)+Var(B) since A is independent of B.
_B
【在 G*****t 的大作中提到】 : Hi, all, : I have a question about calculations for two independent sample... : Suppose two independent sample (A, B), only CV for each is given (CV_A, CV_B : ), df also given (dfA, dfB). And we want the CV for the total population... : Here's what I thought: : Var(A+B) = VAR(A)+VAR(B) : Assume mean_A = mean_B, : CV(A+B) = sqrt(CV(A)^2+CV(B)^2) : Is there any problem for this calculation? : Shall I consider df values too as a weight? Say,
|
|