In a recent groundbreaking study, The New York Times collected one hundred cases over the past fifty years of "rampage killers" in America.  They separated out the nineteen teenage killers to make a study of them.  What they found is that while adults tended to act alone, kids often acted with the support of their peers.  In some instances, those kids who did the killing were helped along by other kids who drove them to school, showed them how to use a gun, helped them get a firearm, or simply came to watch.  There were times when these students were actually goaded into doing it.  Quite often the killers boasted about what they were planning and even encouraged friends to be a witness.

In 40 cases of school violence in the past twenty years, The Secret Service's National Threat Assessment found that teenagers often told someone before they did the deed.  Most of these kids are white and they prefer (and somehow acquire) semiautomatics.  Almost half had shown some evidence of mental disturbance, including delusions and hallucinations.  There's little doubt that at least some of them would score high on a psychopathy scale.

Common traits in the background of psychopathic children include:

• a mother exposed to deprivation or abuse as a child

• a mother who shows a tendency toward isolation

• a transient father or the family migrates frequently

• a mother who cannot maintain stable emotional connection with a child

• low birth weight or birth complications

• hypersensitivity or hyposensitivity to pain

• hyperactivity

• failure to make eye contact when touched

• absence of fear of strangers

• low frustration tolerance

• transient psychotic episodes

• sense of omnipotence

• easily distracted

• transient relationship

• cruelty toward others

Such children should be assessed for future dangerousness.

The idea of “dangerousness” has been a paramount issue in the legal/mental health arena, yet establishing an empirical body of data from which to make accurate predictions has been difficult.  Such research must meet seven criteria:

1. “Dangerousness” must be segregated into component parts: risk factors, harm, and likelihood of occurrence
2. a rich array of risk factors must be assessed from multiple domains
3. harm must be scaled in terms of seriousness and assessed with multiple measures
4. the probability estimate of risk must be acknowledged to change over time and context
5. priority must be given to actuarial research
6. the research must be done in large and broadly representative samples
7. the goal must be management as well as assessment.

In other words, risk assessment is a complicated business.

Expert John Monahan claims that all of the above criteria are met in the in-depth MacArthur Risk Assessment Study, which examined the relationship between mental disorder and violent behavior directed against others.  The researchers devised a comprehensive list of risk factors that 1) have been associated with violence in prior research, 2) are believed by experienced clinicians to be associated with violence, and 3) are hypothesized to be associated with violence by existing theories of violence or mental disorder.  This included factors not previously studied, such as social support, impulsiveness, anger control, psychopathy, and delusions.  Experts in these fields developed risk assessment instruments to assist with measurement and prediction. 

Of the assessment devices, Hare's PCL-R (the Psychopathy Checklist revised to include fewer items) appears to be the most effective.  That is, individuals at high risk to commit crimes can be reliably diagnosed with the PCL-R. 

Hare and some of his colleagues went on to develop the Psychopathy Screening Device (PSD) for children as a 20-item, 0-2 rating scale similar to the PCL-R.  It has a similar two-factor structure (Callus/Unemotional and Impulsive/Conduct problems), and was completed by teachers who interviewed each child assessed.  Researchers Fisher and Blair used the PSD in the context of reinforcement sensitivity with 39 children aged 9-16.  They found that poor performance on a card-playing task and on the moral/conventional distinction tasks were significantly correlated with behavioral disturbances.  That is, those children who played cards badly and also made little distinction between moral and conventional transgressions (like cheating) had higher ratings on the PSD.  Since adult psychopaths had similar results, this indicates that the PSD may be a reliable device for prediction of adult psychopathy.   

It seems advantageous to identify disorders like psychopathy along dimensional lines (traits and behaviors) rather than with the lists found in the DSM-IV offers.  Since the Hare PCL-R has been shown to be a more accurate predictor of dangerousness than the DSM-IV, it seems logical to conclude that the approach grounding the best instrument for detection and prediction is preferable. 

John McHoskey and his colleagues developed a similar instrument, the Kiddie-MACH scale.  MACH is a psychological concept that derives from the sixteenth-century writings of Niccolo Machiavelli, The Prince and The Discourses.  It was worked into a quantifiable personality construct that appeared to be involved in strategies that people used to gain and maintain interpersonal power.  The successful manipulator would have the following traits: a) lack of interpersonal affect, b) lack of concern with conventional morality, c) lack of gross psychopathology, and d) low ideological commitments.  These characteristics are central to the definition of psychopathy.  The Mach-IV contained twenty statements rated on a scale, and the Kiddie-MACH was used with children.  McHoskey studied MACH in the general population.  They integrated the psychopathy constructs found on the PCL-R with Machiavellianism and concluded that the Mach-IV is an accurate global measure of psychopathy.

More work must be done to determine which of the childhood psychopathy assessment devices is most effective, and this will probably require extensive longitudinal studies.  Yet the fact that researchers with specialized knowledge in psychopathy are moving away from the multiple conduct disorder diagnoses toward a more cohesive predictive construct should facilitate better agreement on the findings.  At this point, however, it should be clear that the limited intervention resources available ought to be directed toward children who exhibit traits common to a constellation of disorders: hyperactive, impulsive, attention-deficit and conduct problems, because children who manifest all of these seem to be the most strongly correlated with adult manifestations of psychopathy.  Classifications that focus primarily on behavior to the exclusion of personality characteristics will continue to fall short of the predictive value needed for locating the “fledging psychopath”—the person most likely to become a serious and chronic antisocial adult.