For simple random sampling,
I have calculated somemean square errors for ratio-type estimators such as Isaki estimator, and Prasad Singh estimator.
But, Mean Square Errors i obtained are negative. Does MSE eqaul to a negative value? I there a mistake ?
You have a mistake. MSE can never be negative, because its an expected value of a non-negative random variable (and thus is non-negative itself).