Lecture 33 (Wednesday, April 3, 2013)
1. Douglas et al (2003) ▯violence risk for 100 patients released into community
Interrater reliability (kappa) = .61
AUC (physical violence) = .70
AUC (any violence) = .67
2. Douglas et al (2005) ▯compared HCR20, VRAG, PCLR, VORAS on prediction violent
HCR20 and VRAG most useful
PCLR and VORAS “just okay”
Not as strong as HCR20 and VRAG
PCLR is not as good by itself as it is in combinations with other tests
Level of Service InventoryRevised (LSIR)
Originally called the Level of Supervision Inventory before the revision
Like the HCR20, it is a structured interview not a selfreport inventory
Also developed in Canada in the late 1970s
Were not initially designed to predict violence, they were originally designed to assess the level of
supervision that would be required for any inmate that was about to get out of parole (how close of an
eye we should keep on them)
54 items grouped rationally into 10 scales
The grouping of the scales is rational rather than empirical ▯does the item look as though it is related
to the thing that we are trying to measure? (construct validity)
Scale 1 – Criminal history
Any prior convictions, escape from correctional institutions, punished for institutional misconduct
Scale 2 – Education/Employment
What is their highest level of education?
What is their employment history? Regularly employed? Employed at time of arrest?
Would ask these individuals in this interview, but also check the records
Scale 3 – Financial
1 Lecture 33 (Wednesday, April 3, 2013)
Financial problems? Large abnormal debts (gambling, etc.), problems paying debts
Reliance on social assistance
Scale 4 – Family/Marital status
Dissatisfaction with marital situation
Scale 5 – Accommodations
More than 2 address changes last year
High crime neighbourhood?
Scale 6 – Leisure/Recreation
Absence of recent participation in organized activity
Could make better use of time – are they getting enjoyment out of this activity? If not, should try
Scale 7 – Companions
A social isolate?
Some criminal acquaintances
Scale 8 – Alcohol/Drug problems
Alcohol problems ever?
Drug problems ever?
Both are factors associated with higher risk of reoffending
Scale 9 – Emotional/Personal
Severe interference, psychosis?
Mental heath treatment in the past?
Possible predictor of recidivism – usually a negative predictor of violent reoffending.
Most often, people with mental illness are going to be victims as opposed to perpetrators
Scale 10 – Attitudes/Orientation
Supportive of crime
Unfavourable toward convention
Because the LSI includes so many variables that are predictive of violent reoffending, it is used for
the purpose of assessing violence risk
2 Lecture 33 (Wednesday, April 3, 2013)
Room for clinical judgment as well
History of Psychopathy as a Construct
Date back to ancient Rome
Appear in Shakespeare, they appear everywhere
1. Emil Kraepelin (ca 1895)
Distinguished between two forms of madness: schizophrenia and manic depressive disorder (mood
Moral insanity, a third type of madness he described
Defective ability to restrain: “the reckless gratification of…immediate egotistical desires”
High in impulsivity, egocentricity, low in empathy or concern for others
2. Hervey Cleckley, prison psychologist
Someone who is involved in the first publicized cases of multiple personality disorder
Wrote “The Mask of Sanity” (19411988)
Psychopaths are not mentally ill in the sense that a schizophrenic or manic person is, they are
perfectly in touch with reality and they are absolutely ordinary people from every observable
standpoint until you start dealing with them