Hypothesis testing

Abstract

You will recall (Chapter 9) that the refuting or supporting of a hypothesis is a major element in scientific methodology. A hypothesis is a reasonable conjecture about a particular state of nature. Two different ways of stating a hypothesis exist to match the two major approaches in research design, those of, (i) looking for differences between groups, and (ii) looking for relationships between groups. So we have hypotheses which make statements about differences or which make statements about relationships. For example, a hypothesis which states that there is a difference between the performance of primary-school boys and girls in reading skills is obviously a different hypothesis and requires different statistical measures from a hypothesis that states that there is a relationship between self-esteem and academic performance in secondary school pupils.