We should dispense with use of the confusing term "null hypothesis" in educational research reports. To explain why the term should be dropped, the nature of, and relationship between, scientific and statistical hypothesis testing is clarified by explication of (a) the scientific reasoning used by Gregor Mendel in testing specific hypotheses derived from his general inheritance theory and (b) the statistical reasoning used in applying the chi-square statistic to his experimental data. The Mendel example is followed by application of the same pattern of scientific and statistical reasoning to educational examples. A better understanding of the related, but separate, processes of scientific and statistical hypothesis testing, including the role of scientific hypotheses (i.e., proposed explanations) and scientific predictions (i.e., expected test results), not only reveals why null statistical hypotheses and predictions need not be stated, but also reveals how we can improve the clarity of our research reports and improve the quality of the research reported by insuring that alternative scientific hypotheses and theories are in fact tested.