Three features distinguish crime science from criminology: it embraces the physical, computer and engineering sciences as well as the social; it focuses on crime rather than criminals, and it is single-minded about cutting crime, rather than studying it for its own sake. Crime science was conceived by the British broadcaster Nick Ross in the late 1990s in order to recruit scientific methods to crime prevention, with encouragement from the then Commissioner of the Metropolitan Police, Sir John Stevens and Professor Ken Pease.
Situations have not been entirely ignored in criminology. Birkbeck3 and LaFree (1993) and Sampson and Lauritsen (1994) have reviewed various situational approaches to crime, deviance, and violence, including the work of symbolic integrationists, opportunity and routine activity theorists, and those who advocate the study of criminal events. Most empirical research on situations, however, has involved the event as the unit of analysis, with no link to individual histories or individual patterns of behavior. To study the behavior situation relations advocated by Mischel and Shoda (1995)4, we would need to study the same individuals across multiple situations or contexts.
To proceed with such an approach, we must define situations. From the various situational perspectives on crime there has not as yet emerged a consensus as to how situations should be conceptualized. Let us turn back to the summer camp study (Shoda et al., 1994) for a possible model. In that study the researchers observed behavior occurring within different levels, which they referred to as ecological settings, nominal situations, and interpersonal situations. Table 1 provides examples of these different levels. The ecological settings represented are camp, school, and home. Within each ecological setting are nominal situations, which they describe as “dictated by the structure of the particular ecology (the setting)” (Shoda et al., 1994: 675). Thus, at