“…As a result, most data sets included an array of risk factors that could represent a wide complement of theories and propositions. For instance, researchers interested in crime and deviance have used the NLSY data sets to investigate the general theory of crime (e.g., Turner and Piquero 2002), social learning and differential association (e.g., Sullivan 2006), deterrence (e.g., Lochner 2007), general strain (e.g., Katz 2000), social control (Harper and McLanahan 2004), Moffit's dual taxonomy (Turner, Hartman, and Bishop 2007), social support theory (Wright and Cullen 2001), as well as the relationship between offending and economic circumstances (e.g., Paternoster et al 2003), IQ (Cullen et al 1997;Herrnstein and Murray 1994), marriage (e.g., Bersani and Doherty 2013), immigrant status (e.g., Bersani In press), and a host of other factors. Clearly, the field has become adept at identifying the correlates of offending and crime, but whether and how those factors work directly, indirectly, additively, and interdependently as causes of crime is less clear (Agnew 2005;.…”