ABSTRACT

Let and ?2 denote the multiple correlation coefficients of X1 with X(1) and with

respectively. Denote by We consider the following testing problems:

a. to test H10: ? 2=0 against the alternatives

b. to test H20: ? 2=0 against the alternatives

Let b[i] denote the i-vector consisting of the first i components of a vector b and C[i] denote the i×i upper-left submatrix of a matrix C. Partition S and S as

where S(22), S(22) are each of dimension p1×p1; S(33), S(33) are each of dimension p2×p2. Then