A testing strategy is proposed to avoid the ambiguity of a "negative result." It distinguishes important effects from those in a buffer area. If a result falls into the buffer area, the effect can be declared positive with a probability, or declared as pending. There may be clinical uses. For example, a new blood pressure lowering drug is tested with mean 145 mm Hg, sd 13 mm Hg for the study group. A buffer is set for reduction between 4 and 10 mm. Sample size can be calculated for conventional testing, but a test with a buffer area yields a markedly reduced sample size.