Debate continues between quantitative versus qualitative conceptualization of psychopathology in general and suicidal behavior or drug abuse in particular. Are suicidal and substance abuse behaviors dimensional and continuously distributed, with quantitative (but not qualitative) differences based on levels of severity or between continua of severity, with “pathological” and “non-pathological” existing at opposite ends of the distribution but without clear cut points? In contrast, are there actual cut points that demarcate discrete taxa or diagnostic subgroups (Hill, 2002; Hinshaw et al. 2002; Leboyer et al. 2005)? Arguably, psychopathology has properties that are amenable to dimensional and categorical assessment, and therefore both types of assessment should be used. Recent findings show that a dimensional analysis of suicidal behavior and related risk factors yield dimensions that are opposite and independent, such as under- versus over-engagement and rejection-turmoil (Hyde et al. 2005). Furthermore, it may be the case that suicidal behavior is itself an independent clinical outcome with both dimensional and categorical properties (Leboyer et al. 2005; Windle, 2004).
