ABSTRACT
The univariate and bivariate statistics described in the preceding chapters often help us to understand the subjects we are researching. However, univariate and bivariate analyses rarely can provide convincing tests of hypotheses or the theories from which they were derived. To test a hypothesis convincingly, we must be able to rule out major alternative rival hypotheses. Although a sound research design can sometimes allow us to dismiss alternative rival hypotheses, social scientists commonly find that they must rely on data analysis rather than research design to examine the validity of rival hypotheses. This requires the use of multivariate analysis, which is analysis of the simultaneous relationship among three or more variables.
