A note on testing the mems of normal distributions with known coefficient of variations

We consider two problems in sequential hypothesis testing. The first one deals with testing the mean of a normal distribution with known coefficient of variation. The standard sequential probability ratio test (SPRl) and the fixed-sample test are discussed and compared. The second problem, deals with a version of the Behrens-Fisher problem, for comparing the means of two normal-distributions with unknown variances,, Sequential and fixed-sample tests are developed when either (i) the coefficient of variations are known or (ii) the two distributions have the same unknown coefficient of variation. The fixed-sample test is compared with the" proposed sequential procedure.