# Statistical Advisor, Hypotheses are Predictions

Hypotheses are predictions that one has prior to collecting the data concerning the pattern of results. For example, one may predict that males are more likely to endorse a questionnaire item than females, that age is related to respondents' agreement with a questionnaire item, etc.

We may have hypotheses about the specific shape of distributions; for example, many variables can be expected to follow the normal distribution. We may have hypotheses about differences between groups or samples, for example, gender differences. We may have hypotheses about differences between variables; for example, we may predict that students like art courses more than math courses (where the liking for arts courses would be recorded in one variable, and liking for math courses in the other; note that in this example each student would rate both courses). We may have hypotheses about relationships between variables; for example, we may predict that age is related to (correlated with) a respondents' expressed liking of a product. Finally, we may have hypotheses about differences in the relationships between variables in different groups. For example, we may predict that, while age is related to liking of a product in males, it is unrelated in females.