ABSTRACT
Statistics courses and books—this one included—tend to resemble horoscopes. There are two senses to this resemblance. First, in order to remain plausibly correct, they must remain tremendously vague. Second, there are strong incentives for both astrologers and statisticians to exaggerate the power and importance of their advice. Statistical inference is indeed critically important. But only as much as every other part of research. The chapter focuses on the statistical analysis, perhaps because it is the only piece for which we have formulas and theorems. But every piece deserves attention and improvement. Statistical analysis can neither be uniquely credited with science’s success, nor can it be uniquely blamed for its failures and follies.
