ABSTRACT
The univariate and bivariate statistics described in the preceding chapters rarely can provide convincing tests of hypotheses or the theories from which they were derived. Many of the statistical tools already discussed can be employed in multivariate analysis. A highly simplified example can illustrate the way in which crosstabs and bivariate statistics can be adapted in order to conduct a multivariate analysis. A variety of statistical procedures are designed specifically for multivariate analyses that can be used in a wide range of situations and that provide easily interpretable results. Statisticians have devised ways for us to adapt multiple regressions in order to compensate for some of these problems. Coefficients may fail to attain statistical significance even where there is a substantial relationship, leading us to falsely identify bivariate relationships as spurious.
