| Popis: |
All of us have been guilty of favoring statistical procedures that support our hypotheses. This is fine if our hypotheses are correct. But what if they are false? No matter. By use of appropriate statistical procedures, applied in just the right way, even the saggiest of hypotheses can be propped up. This paper explains how. As an example, the usefulness of 14 useless acoustic features in the differentiation and recognition of lovebird calls will be shown statistically in two different ways. |
