ABSTRACT
Statistics is an exact mathematical science. However, when applied in decision-making, there are many human choices that need to be made. The choices can make the statistical outcome go one way or the other – to reject a treatment or to accept the treatment. The choices could be made to support a person’s bias or preference – to justify a particular action or to condemn the action. This chapter is about the cautions needed when applying statistics to support or direct decisions. The choices include formulating the hypothesis, choosing the level of significance, choosing the number of data, ensuring that data are independent, and many other issues related to the context of the decision to be made. The statistical hypothesis is not the supposed cause-and-effect mechanism. The hypothesis and associated attribute are expected observable outcomes if the mechanism is as supposed.
