Observation, in which the scientist observes what is happening, collects information, and studies facts relevant to the problem. In this stage, statistics suggests what can most advantageously be observed and how data might be collected. Hypothesis, in which the scientist puts forth educated hunches or explanations for observed findings and facts. In this stage, the statistician helps format observations in a form that is comprehensible and understandable. Prediction, in which the anticipatory deductions based on hypotheses are put forward in testable ways. Statistics can help only a little at this stage of analysis, for predictive insights are often intuitive and creative rather than numerical. Verification, in which data are collected to test predictions. In judging the extent to which predictions are borne out by observation, we recognize that data and predictions almost never agree exactly, even when theories are correct.
