134 5 Court and the Legal System—Adult Foremicsutilizes statistical predictors (Melton ct al., 1997). Milner and Campbell (1995) sug-
gest that a combination of actuarial and clinical methods will provide the most
accurate risk assessment.
While the literature remains controversial, there are certain factors that have con-
sistently been shown to be significantly related to future violent behavior. Among
such factors are the individual's score on the Hare Psychopathy Checklist, a his-
tory of criminal behavior, a history of substance abuse, and the age of the offender
(G. Harris, Rice, & Quinsey, 1993). Other studies have attempted to develop pre-
dictive models for specific types of offenders. Blanchette (1996) states that there are
special considerations in conducting risk assessments for sexual offenders, such as
assessing the individual's cognitive processes, their general lifestyle, and their history
of sexual deviance. Moreover, as noted by Quinsey, Lalumiere, Rice, and Harris
(1995), risk assessments vary considerably depending on the specific type of sexual
offender. For this reason, mental health professionals who are not trained specifi-
cally in the assessment of sexual offenders are likely to draw erroneous conclusions
regarding their risk. Another common population for whom mental health profes-
sionals tend to conduct inaccurate risk assessments is the mentally ill. It has been
suggested that this may be due to the illusory correlation between mental illness
and violence or the belief that an individual is more dangerous simply because he
or she is mentally ill (Melton ct al., 1997). Therefore, it is crucial for mental health
professionals to truly have expertise with the specific population on whom they
purport the ability to conduct risk assessments.
Melton et al. (1997) provides guidelines for the most appropriate ways for mental
health professionals to communicate the results of their risk assessments to the courts.
These authors stress the importance of refraining from using language which suggests
that their opinion is absolute. They also suggest that experts present information
to the court regarding the factors which have been empirically shown to enhance
an individual's risk for violent behavior. Finally, these researchers encourage mental
health professionals to provide the court with a statement as to the poor validity of
violence prediction. These suggestions are in stark contrast to the method employed
by the psychiatrist who testified in the Barefoot v. Estelle (1983) case.
In California, even those mental health professionals who are not accustomed to
their work entering the legal system have been faced with the issue of conducting
risk assessments. In 1976, the landmark case of Tarasoffv. Regents of the University of
California delivered a decision that requires therapists to take preventative measures
if any reasonable therapist would believe that their client is likely to harm an iden-
tifiable victim in the near future. Currently, most jurisdictions have a statute similar
to that of Tarasoff in California (Melton et al., 1997). Thus, this statute brought the
issue of risk assessment into the lives of all therapists. In so doing, controversy exists
among mental health professionals concerning the damage that this form of risk
assessment has on the therapeutic process and the ethical principle of confidentiality
between a therapist and his or her client.