This study, using a certain simulation strategy, for the exact maximum likelihood estimator of θ from the MA(1), estimates appropriate percentiles, together with their standard errors, offering a new set of critical values for testing in finite samples H0: θ = -1, against H1: θ-1. In this...