DRMN

where [[lambda]rmn] is an individual's hazard of re-arrest number r at age m; [[lambda].sup.0.sub.rmn] is an arbitrary non-negative value representing our prior (non-sample) belief about this hazard rate; and drmn is a flag indicating whether the nth individual was at risk of the rth arrest at the mth age.

(57) Hence, following Ryu, if we define the term [[lambda].sup.*.sub.rmn] = [[summation].sub.m]drmn[lambda]rmn for some appropriately redefined drmn, then we see that