Proving the Null Hypothesis

In most situations in which an experimental treatment or procedure is to be compared with a standard treatment or procedure, the null hypothesis to be tested states that the two procedures are equally effective. The alternative hypothesis is likely to be one-sided and to state that the experimental procedure is better than the standard. Advocates of the new procedure would hope that trials would be large enough to provide sufficient power to detect an assumed difference between the procedures, and thus that the null hypothesis would be rejected. If so, the probability of type I error (falsely rejecting a true null hypothesis) would be noted, and the new or experimental procedure could be accepted comfortably by those considering its implementation.