ABSTRACT

Treating social support or sub-categories of it as homogeneous entities on the basis of empirical co-variance patterns, however, makes little sense. The suicide attempters arid the controls were not matched on marital status and, since marital status is often identified as an important determinant of social support, this could be expected to bias the results and had to be analytically controlled. One would have expected self-esteem and, indirectly, social support as a contributor to it, to differ substantially and independently between suicide attempters and controls. The pattern of results produced by this study underscores the danger of sidestepping conceptual problems and of relying too much on either intuitive concepts of social support or on empirical data analysis for the definition of relevant support categories or dimensions. The issue is the difference between empirical co-variation and functional equivalence; the former is a necessary condition for the latter, but cannot be derived from it.