Higher rates of physical symptoms among homeless women do not appear to be due to reporting bias: a comment on Ritchey et al. (1991)