ABSTRACT
The rationale for this research is that in sociology few studies have rigorously looked at individual-level suicide data. More common are aggregate analyses with units of analysis such as countries, states, counties, cities, etc. (e.g. Stack 1982, 1985; Wasserman 1983, 1984; Breault 1986; Girard 1988; Pescosolido and Georgiana 1989; Lester 1993). Moreover, one of the enduring themes of this literature concerns methodology, much of which is related to the ‘ecological fallacy’ (e.g. Moksony 1994; Pescosolido 1994). In one recent paper, Breault (1994) presents data to suggest that aggregate level findings at the national level, which are generally supportive of Durkheim’s theory, are ambiguous because results at lower levels of aggregation are substantially different. In the same paper, Breault suggests that the lower level results are inadequate for other methodological reasons, i.e. small sample size and limitations on the number of variables.
