2007 年 44 巻 2 号 p. 23-38
Two main models, namely, Terri Moffitt's hypotheses and Travis Hirschi and Micheal Gottfredson's General Theory of Crime, have been proposed since around 1990, as an explanation of the consistency and change of individual criminality. There has been a heated debate over the relative validity of the two models. This research attempts to contribute to the solution of this debate. By defining criminality that gives rise to individual offending behavior as a probability process, the two models are expressed in a single framework. We examined whether the predictions of these two models better fit the actual observation obtained from the event history analysis of the recidivism of releasees from a Juvenile Classification Home. The result suggests that Moffitt's hypotheses are more congruent with the observation.