7 153 DISCUSSION is, the networks in this dissertation are estimated through pairwise Markov random fields (PMRF; Costantini et al. 2015, van Borkulo et al. 2014). Although this approach is currently mostly used in estimating networks regarding psychopathology on cross-sectional data, the causal interpretation of these models is limited because of their undirected edges (Briganti et al., 2022), which makes it impossible to determine whether the activation of a dynamic risk factor causes or is caused by the activation of another dynamic risk factor, or whether the causal impact is bidirectional. 7.4.2 LIMITATIONS OF THE NBM-RSR Some limitations of the NBM-RSR should be acknowledged. First, the network of dynamic risk factors in the NBM-RSR should contain variables which have been demonstrated to have predictive value for sexual reoffending. However, empirical studies on the predictive properties of dynamic risk factors generally do not consider interrelationships among these factors. This could lead to spurious associations between some of the dynamic risk factors and sexual reoffending. That is, a dynamic risk factor might falsely appear to have predictive value due to an undetected interrelationship with other dynamic risk factors (Andrews et al., 1990; Bonta & Andrews, 2017). The network analysis in Chapters 3 and 4 found that most but not all commonly known dynamic risk factors remain related to sexual reoffending after controlling for all other dynamic risk factors in the network. However, currently, there is neither sufficient empirical research nor a well-developed method to determine which dynamic risk factors are directly causally connected to sexual reoffending and which only gain predictive power through their connection with other dynamic risk factors (and, thus, should be part of the external field). Second, despite their observed causal influence, the position of protective factors (e.g., characteristics of offenders, their environment, or their situation, that reduce the risk of future criminal behavior; de Vogel et al., 2009) remains unclear in the current NBM-RSR. Third, Chapter 6 addresses the development of dynamic risk factors only to a limited extent, even though the NBM-RSR does allow for a comprehensive discussion of this topic from, for example, a developmental psychological approach. In particular, the causal influence of factors in the external field on dynamic risk factors and thus on the likelihood of sexual reoffending through lifespan should be included in a more thorough elaboration of the NBM-RSR. For example, more attention could be paid on how (developmental) changes in cognitive, motivational, psychophysiological, and social functioning will affect (the interrelationship of) dynamic risk factors. Fourth, despite its implications for tertiary prevention, to date the NBM-RSR provides little guidance for primary and secondary prevention of sexual offenses. Again, factors in the external field provide the opportunity to applicate the NBM-RSR more broadly. Hopefully, a future broader application of the NBM-RSR could provide guidance for increased effectiveness of primary and secondary prevention for sexual offenses by indicating for example which sociocultural factors and situational factors are expected
RkJQdWJsaXNoZXIy MjY0ODMw