Assessing Risk

Card Set Information

Assessing Risk
2015-04-10 12:48:02
Psychology Criminology
Psychology,Forensic Psychology
Show Answers:

  1. How has the idea of dangerousness changed?
    • Started off as individuals who threatened political and social stability in the early 19th century 
    • This became repeat offenders who threatened the community 
    • Shift from property offences to violent offences in the 1930s
    • Psychology became increasingly involved around the 60s, assessing dangerousness and medicalising the phenomenon
  2. What was the act that focussed on dangerousness?
    • Criminal Justice act of 2003
    • Outlines specific provisions for dangerous offenders 
    • In cases where life cannot be given, courts can give an extended 8 year sentence if it looks like the offender is going to cause serious harm to the public in cases of violent or sexual offences of over 10 year sentences 
    • Serious harm constitutes death or serious personal psychological or physical injury
    • This may violate the principle that people are punished on what they have done, not what they might do
  3. What are the four possible prediction outcomes?
    • True positive: predict that they will reoffend and they do 
    • False positive: predict that they will reoffend and they don't 
    • True negative: predict that they wont reoffend and they don't 
    • False negative: predict that they won't reoffend and they do
  4. What is the selection ratio?
    • The proportion of predictions that behaviour will occur
    • False positives + True positives divided by the total
  5. What is the base rate?
    • The rate at which behaviour actually occurs in the relevant sample
    • True positives + False negatives divided by the total
  6. What is the prediction accuracy?
    • The proportion of all correct predictions
    • True positives + True negatives divided by the total
  7. What are the two crucial principles for predicting recidivism?
    • The greater the difference between selection ratio and base rate, the greater the potential for inaccuracy 
    • The rarer the behaviour (the lower the base rate) the more difficult it is to predict true positives
  8. How are predictions sometimes misleading?
    The number of correct predictions does not automatically mean the predictor is accurate, he could have a higher selection ratio and correctly predict 50 yet get 50 wrong whereas someone else could predict 50 and get them all right
  9. What are the two types of error in prediction?
    • High false positives 
    • Over-predict recidivism rates
    • There is poor specificity (fail to identify non offenders)
    • This could lead to a more oppressive police force
    • High false negatives 
    • Under prediction of recidivism 
    • Implications for community safety
  10. What are the trade-offs between false positives and false negatives?
    • As we increase sensitivity we decrease specificity meaning we recognise fewer non offenders  
    • Raising the cut off will reduce the number of false positives but increase the number of false negatives
    • Lowering the cut off will increase the number of true positives but also the number of false  positives
  11. What is the Receiver Operator Characteristic curve?
    • Used to decide on cut-offs
    • True positive ratio is plotted against the true negative ratio for each cut off
    • .90-1.0 accuracy is excellent and .50-.60 is a failure
  12. What is clinical prediction?
    • Decisions on risk based on a clinical judgment of the offender by a mental health professional
    • Poor accuracy- Sepejak et al (1984). Judged the accuracy of predicting recidivism 2 years down the line. Psychiatrists= .20, Psychologists =.17
    • Lack of cpncensus amongst professionals as to what constitutes 'dangerous' or 'violent' with over 250 definitions for aggression 
    • Improper psychometric tests are used to measure dangerousness- Hinton (1983). Prior to 1980 the majority of tests used contain questions that are unreliable or misleading, and they still aren't perfect
    • Conservative decisions are made to be safe
    • Base rates are overestimated 
    • Situational factors are ignored
  13. What is actuarial prediction?
    • Factors that are known to predict crime are combined with statistically derived weightings to form a composite measure 
    • These measures have been found to be around 10% more effective in predicting recidivism (Grove et al, 2000)
  14. What is the VRAG?
    • The violence risk appraisal guide -Quinsey et al (1998)
    • 12 item inventory used to assess someone's risk of violence following release within a specific time frame for violent, mentally disordered offenders 
    • Used to get an idea of their psycho social history 
    • Incorporates the Hare checklist for psychopathy
  15. What is the SORAG?
    • The Sex offender risk appraisal guide- Quinsey et al (1998)
    • Modified version of the VRAG used to assess risk of sexual violence in a specific timeframe following the release of violent and mentally disordered offenders 
    • 14 item checklist
  16. What is the PCL-R
    • Psychopathy checklist (revised)-Hare (1991)
    • Not intended as a recidivism measure but is often used as one 
    • 20 item scale
  17. What is the RRASOR
    • Rapid risk assessment for sex offence recidivism - Hanson 1997
    • Brief 4 item screening assessment for males who have already been convicted of a sexual offence 
    • Relies on information from files on the offender
  18. How effective are the various tests of recidivism?
    • Reasonably- Barbaree et al (2001)
    • Meta analysis measured the accuracy of predictions from the various measures 
    • Any reoffending: VRAG: .77 , SORAG: .76 , PCL-R: .71 , RRASOR: .60  
    • Serious reoffending: VRAG: .69 , SORAG: .73 , PCL-R:.65 , RRASOR:.65
    • Sexual reoffending: VRAG: .61 , SORAG: .71, PCL-R: .61, RRASOR: .77