Re-thinking the analysis of intergenerational social mobility: a comment on John W. Fox's "Social class, mental illness, and social mobility".

The method of analyzing social mobility described by Fox (1990) is flawed in its adjustment for between-group differences in destination status when estimating the extent of the mentally ill's mobility as compared with the general population. Use of the recommended model with hypothetical data sets resulted in a significant finding when no overall upward or downward mobility occurred, and a non-significant result when the downward mobility of a psychotic group was contrived to be massive. An alternative model for the test of group differences in mobility is suggested within the framework of log-linear analysis commended by Fox (1990). This method indicated significantly more downward and less upward mobility in mentally ill groups when data from four studies were re-analyzed. We conclude that the weight of evidence from published studies supports the notion of social selection-drift, although this does not imply the inconsequence of social factors in the aetiology of schizophrenia (and other psychoses) or in its prognosis and occupational consequences.