Blackwell Reference Online

Extract

Hypotheses are predictions that specify the relationships among the variables. The role of hypotheses in scientific research is to provide explanations for certain phenomena and to guide the investigation of related others. The development of scientific knowledge hinges ultimately upon the results from hypothesis testing. Formalized hypotheses consist of two types of variables: the independent and dependent variables. The former is the cause and the latter is the outcome. A good and well-worded hypothesis should (1) indicate the specific relationship between the dependent and independent variables to be examined; (2) suggest the nature of the relationship; and (3) imply the nature of the research design ( Cone & Foster 1993 ). Hypotheses, which are derived directly from a theory or theories, have to be testable. The hypothesis-testing process generally involves three steps. The first step is to formulate two hypothesis statements: a null hypothesis (often symbolized as H 0 ) that predicts no relationship between the variables in the population (e.g., H 0 : Social class is unrelated to deviant behavior) and an alternative hypothesis (H 1 ) that predicts a relationship between the variables (e.g., H 1 : Social class is related to deviant behavior). The null hypothesis should be mutually exclusive of the alternative hypothesis, meaning that there is no overlap between the two hypotheses. ... log in or subscribe to read full text