.

 

   

Arguing Future Dangerousness
New Techniques for Assessing the Risk of Violence

By Brett C. Trowbridge and Charles H. Williams

This article supported by The Trowbridge Foundation

Originally published in Washington Criminal Defense, February 2000, Vol 14, No. 1

 

“[W]e are calling on clinicians to do risk appraisal in a new way--a way that is different from that in which most of us were trained. What we are advising is not the addition of actuarial methods to existing practice, but rather the complete replacement of existing practice with actuarial methods...Actuarial methods are too good and clinical judgment too poor to risk contaminating the former with the latter.” 

Vernon L. Quinsey, Grant T. Harris, Marnie E. Rice, and Catherine A. Cormier (1998).1 “As a professional enterprise, forensic psychological services to our criminal courts have gone through two phases. The first was a phase in which we sought to try out psychology in the courts, perhaps before we were actually ready to do so. The second phase, which began in the late 1980s, was marked by...a number of ambitious research initiatives to build a scientific base for better practice.

“Now we are beginning a third phase--the introduction of the new information and technology that this research has produced. We cannot enter this phase with a ‘build it and they will come’ mentality. Although the products of our recent efforts are likely to be used, all information and technology is susceptible to corruption when first put to use. We must try to introduce our new methods to courts, attorneys and clinicians in ways that minimize that threat and maximize their promise. That is an effort that is most likely to be successful if lawyers and psychologists accomplish it together.” Thomas Grisso (1999) 2

Science has a way of sometimes presenting us with a double-edged sword, whose explanations and discoveries cut both ways through legal controversies. Take, for example, the issue of future dangerousness that criminal law practitioners argue about incessantly in bail hearings, plea-bargaining, and sentencing--as well as in other proceedings. As long as the discourse was limited to vague generalizations about the predictive value of an individual’s criminal history, it remained possible to argue the issue either way--and without the testimony of experts.

Indeed, at one time early in the authors’ careers, it was an unshakeable article of faith that mental health professionals were no better at predicting future dangerousness than anyone else.

Yet, even when directly confronted with research showing the relatively low accuracy of their predictions--for example, in the 20-35% range for forensic specialists3--courts repeatedly sought the opinions of such professionals. In upholding the constitutionality of using psychiatric predictions of violence for the purpose of determining whom to execute, the U. S. Supreme Court in Barefoot v. Estelle (1983)4 said:

[I]f it is not impossible for even a lay person sensibly to arrive at that conclusion, it makes little sense, if any, to object that psychiatrists, out of the entire universe of persons whom might have an opinion on the issue, would know so little about the subject that they should not be permitted to testify.5

Once the constitutional issues were settled, the question shifted from whether violence could “sensibly” be predicted to how such predictions could be improved. This became the Holy Grail of one discipline in particular-- forensic psychology. 

In this article, we would like to describe significant recent advances in this field relating to the issue of future dangerousness--particularly as applied to identifying psychopaths and predicting recidivism in spousal assault, crimes of violence generally, and sex offenses. Although we are still a few years from seeing the general use of forensic prediction techniques in court, we believe that these instruments will change the nature of the discourse, because they will drastically reduce everyone’s uncertainty about future dangerousness.

Clinical Assessment

Historically, in North America, offenders suffering from severe mental illness were often committed to maximum-security mental institutions instead of being sent to prison. But usually only those offenders considered to be the least dangerous were subject to eventual release. Because those released did not constitute a random sample reflective of the population at large, it was not possible to obtain an unbiased estimate of the rate at which mentally ill offenders might commit future violence.

But researchers had an opportunity to conduct a natural experiment when the U. S. Supreme Court in Baxstrom v. Herold (1966),6 set in motion the release of some 967 offenders in a maximum security hospital in New York State who had been detained longer than their maximum sentence.

These offenders--none of whom had been acquitted by reason of insanity--were released to local hospitals and, for the most part, eventually to the streets. The resulting follow-up studies demonstrated a very low base-rate of post-release violent behavior. After 4.5 years, only two men had been reincarcerated for committing a violent crime.7 Yet all of the released men had been retained past their release dates because hospital staff thought they were still dangerous.

As of 1970 there were no psychological tests or instruments that had been shown to be useful in the prediction of violence.8 When estimating future dangerousness, forensic specialists traditionally used a clinical interview, supplemented, perhaps, by psychological testing such as the MMPI or the Rorschach, to identify factors that might increase or decrease the likelihood of future violent behavior. No other standardized procedures or established criteria existed for exercising clinical judgment in this situation.

One thing was certain, however: forensic specialists tended to over-predict future incidents of violence.

This occurred in part because people overestimate the rate at which violence occurs in the population at large. In one study conducted in 1980, members of a maximum-security hospital staff were asked, “If 100 men were randomly chosen from this maximum security hospital and released, how many would commit an offense causing bodily harm to another person within one year?” All made estimates that were much too high. Medically trained staff actually made higher estimates than non-medically trained staff.9

Excessive professional confidence masked the fallibility of clinical judgments. In another study conducted during the same era to determine the expertise of forensic psychiatrists, psychiatrists and high-school teachers were given file information about offenders who had been released some three years earlier. They were asked to rate the likelihood of a property offense, the likelihood of an assaultive offense, and the seriousness of an assault should one occur.

It turned out that psychiatrists disagreed among themselves more than the teachers did. The psychiatrists were, in fact, unable to demonstrate prediction expertise superior to lay persons. In addition, the professionals gave no more weight to testing than the laity did. The study concluded by questioning the usefulness of psychiatric examinations in predicting dangerousness.10

Confidence in the accuracy of clinical judgment was severely shaken as a result of these and other studies. Today, forensic specialists familiar with the research will readily acknowledge that clinical judgments, without more, must be regarded with considerable caution.11

Structured and Actuarial Assessment

Crime statistics show that violent behavior is actually a low-frequency event in the general population. So accurate prediction is a difficult, if not impossible, feat. For example, if on average only one of 100 persons were actually later violent, the prediction that none of them will be violent in the future would still be 99% accurate. There would only be one “false negative”--that is, only one person falsely identified as non-violent who later became violent.

As it happens, in the real world, far fewer than one in 100 men randomly chosen from the street will actually commit violent crimes in the future.

Currently, there is no prediction method powerful or sensitive enough to predict violence accurately among individuals in the general population. In fact, one can predict violence in the population better than chance by simply predicting that no one will be violent.

The situation is different, however, among populations with higher rates of future violence, such as those already incarcerated in correctional institutions or maximum-security hospitals. Here accurate predictions can be developed if the base-rate for violence is known and factored into the prediction. 

Even so, studies in which offenders are followed for five to ten years after release usually show base-rates of violent recidivism considerably lower than most people would venture -- generally, under 20%.12

By the late 1990s, two types of assessment devices had been developed by forensic psychologists to overcome the fallibility of clinical judgment. The first type takes the form of a structured checklist to guide the clinical assessment. These checklists are designed to ensure consideration of all relevant factors needed to decide an individual’s risk of future violence. The factors either have been demonstrated to correlate with recidivism in empirical studies or are so intuitively significant that they cannot be ignored.

The best known of these structured assessment devices--the HCR-20 for assessing general violence recidivism,13 the SARA for assessing risk of spousal abuse,14 and the SVR-20 for assessing sexual violence15 -- are the products of research at the Mental Health, Law and Policy Institute of Simon Fraser University in British Columbia.

The SARA (Spousal Assault Risk Assessment) allows evaluators to weigh 20 factors as they see fit to arrive at a summary rating of imminent risk of violence toward a domestic partner and others.16 Figure A lists the factors considered.

The structured assessment devices are criticized, however, because they employ factors that are, to some extent, subjective and fail to present probabilistic estimates of recidivism. The second type of risk assessment device is designed to overcome these problems. These devices--such as the VRAG discussed below--were developed and validated by the use of statistical or actuarial techniques.

Their scientific acceptance rests upon their ability to yield accurate estimates of recidivism, which is ensured in at least two ways. First, these devices are designed to be reliable in the sense that they are consistently easy to score. Repeated measurements of the same subject result in the same scores even when applied by different evaluators.

Second, these devices have been validated--that is, they have been shown to measure what they are supposed to measure. Actual recidivism, for example, more or less matches predicted recidivism.

In our opinion, any expert witness asserting that an individual is likely to be dangerous in the future should be challenged to demonstrate whether he or she is offering a structured or actuarial assessment or merely a clinical judgment. Obviously, if the witness cannot specify the particular assessment instrument used or cite the research demonstrating how the factors identified correlate with future violence, he or she is employing clinical judgment.

Psychopathy

Perhaps the earliest and most versatile of the validated risk assessment devices is the Psychopathy Checklist -- Revised (PCL-R) developed by Robert Hare of the University of British Columbia.17 This actuarial method of identifying psychopaths adds significantly to the best possible predictions based solely on criminal history.

Indeed, most knowledgeable forensic psychologists would concede the PCL-R’s general acceptance in their discipline.

Hare's theory--which has been corroborated by research--is that psychopaths commit a disproportionately large number of crimes. Although they make up less than 25% of the criminal population, psychopaths commit half of the serious crime and violence in North America. Studies show that they have a recidivism rate twice that of other offenders and a violent recidivism rate of about three times more.18

The PCL-R is the foundation of several current techniques used in assessing risk, including the VRAG and the SORAG discussed below. The factors scored are set forth in Figure B.

To use the PCL-R evaluators must rate subjects on each of the 20 variables on the basis of an interview and a review of their personal histories. To achieve a high degree of inter-rater reliability (agreement), evaluators need training in scoring the PCL-R. They also need to gather information from sources independent of subjects to avoid possible manipulation and deception by subjects. Scoring this checklist is neither time-consuming nor difficult as long as independent information is available.

The PCL-R has been validated on both male prison inmates and male forensic psychiatric patients.19 It can be used with adult males who have no criminal history. Psychopathy emerges early in life,20 but the use of the PCL-R on juvenile delinquents has been limited to research purposes.21 Even so, evaluators who make predictive judgments without using this device should be questioned why they did not.

Violent Recidivism

The Violence Risk Appraisal Guide (VRAG) is actuarial device for assessing the risk of violent recidivism. It was developed in the early1990s from research conducted at a maximum security hospital in Ontario, Canada. It is currently the best scientific instrument available for predicting violence among populations of individuals who have already committed at least one violent offense (an “index offense”).

The VRAG has an accuracy in predicting violent recidivism in the neighborhood of 75%.22

The VRAG incorporates the PCL-R but adds a list of other factors shown to increase the accuracy of prediction.23 The factors are listed in Figure C.

Weights are assigned to all of the variables, which are positively related to the probability of future violence with the exception, interestingly, of four variables--age at time of index offense, diagnosis of schizophrenia, victim injury in index offense, and female victim in index offense--all negatively related. Thus subjects who are older, who are diagnosed schizophrenic, who injure a victim in the index offense, or who choose a female victim for the index offense, are significantly less likely to be violent recidivists than other subjects.24

The PCL-R score is the most heavily weighed (up to +12). The elementary school maladjustment score is the next most heavily weighed, with the highest score (+5) following from frequent disruptive behavior and/or attendance or behavior resulting in expulsion or serious suspensions. The factors of personality disorder, separation from parent, non-violent criminal history, and failure on conditional release are each given equal weight (up to +3).25

Two other devices currently under development for estimating violent recidivism are considered especially promising, because their predictive power is potentially greater than the VRAG. These devices--the MacArthur Violence Risk Assessment device26 and the Classification and Regression Trees algorithm (CART)27--are not presently recommended, however, for clinical or risk assessment uses.

Again, because the data required for use of these validated instruments are relatively easy to gather and score, any mental health professional not using one of them when making predictions of violence can legitimately be asked why not. Indeed, at this point in history, the VRAG “is so far superior to anything previously available that not to consider its use . . . would be a difficult choice to justify.”28

Although the VRAG’s admissibility under the Frye standard29 has yet to be tested in Washington, the authors would be surprised if the issue has not been examined elsewhere. The VRAG can certainly be said to be generally accepted in the community of forensic psychology. Because it is a validated testing device, we believe that informed forensic specialists will agree with this point.

Sex Offender Recidivism

One of the basic premises of the Sexually Violent Predator Act codified in chapter 71.09 RCW is that forensic specialists can predict sexually violent recidivism. But while the constitutional issues may have been settled in decisions such as In re Young (Wash.1993)30 and Kansas v. Hendricks (1997),31 many technical risk-assessment issues have not.32

In response to the criticism that the VRAG is not applicable to sex offender populations, this device has been modified and developed into the Sex Offender Risk Appraisal Guide (SORAG). The SORAG, however, is only “marginally” better than the VRAG in predicting recidivism among sexual offenders.33

The SORAG differs from the VRAG principally in the addition of scoring for violent and sex offenses, sex offenses against males, and phallometrically determined sexual deviance. It turns out that the risk of sexual reoffending is lowest among father-daughter incest offenders. But among child molesters, offenders with boy victims have roughly twice the rate of sexual recidivism as offenders with girl victims. Rapists of adult women have intermediate rates of sexual recidivism but are more criminally versatile in that they are more likely to commit non-sexual offenses.34

Phallometric assessments not only discriminate child molesters from non-molesters and rapists from non-rapists but also predict sexual recidivism alone and interactively with the PCL-R.35

In 1996, Canadian researchers Karl Hanson and Monique Bussiere published a meta-analysis of factors shown to be predictive of sex offender recidivism.36 Among their findings was the fact that only 13.4 percent of 23,393 sex offenders reoffended in a four to five year period. Their study does not, however, present an organized scheme for considering and weighing the factors to make an actuarial prediction.

Figure D lists the factors of the Hanson-Bussiere study in the order in which they correlate with recidivism. Note that items 13, 16 and 19 are negative correlations.

Hanson recently developed the Rapid Risk Assessment for Sexual Offense Recidivism (RRASOR), a prediction instrument based on only four items easily scored from administrative records: prior sexual offenses, age less than 25, extra-familial victims in the index offense, and male victims in the index offense.37 This screening device uses weighted factors but shows only modest predictive accuracy.

Final Comments

Of course, further studies -- to refine these devices and broaden their validation bases--are underway as we speak. New tools are being developed to predict recidivism among juveniles.38 Not only is further research being conducted in universities and mental hospitals, but also correctional agencies are getting into the act--see, e.g., the Minnesota Sex Offender Screening Tool (MnSOST) developed by the Minnesota Department of Corrections. The pressure to assess the risk posed by offenders is clearly being felt among forensic specialists--particularly, in sexual-predator commitment cases.

The good news is that these new techniques of risk assessment will help most criminal defendants argue releases, pleas, and sentences. Most defendants will benefit as judges and prosecutors become more familiar and informed about this technology. Why? Because most defendants are not psychopaths and simply don’t present the same level of risk.

On the other hand, we can expect not only the courts but also the legislature and various law enforcement and correctional agencies to target psychopathic offenders. The bad news is that the pressure to identify such individuals will inevitably result in “false positives”--persons falsely identified as psychopathic or falsely classified as likely to reoffend.

It should not escape notice that these actuarial prediction techniques are yet another example where the technology precedes its explanation. We know the VRAG works but we don’t know exactly why--not unlike aspirin (until quite recently). There is no theory that convincingly explains the causes of criminal behavior.39 Until we have such an explanatory theory, we should not become complaisant about the prediction of recidivism.

We close with an observation by the inventors of the VRAG:

...[A]ctuarial measures of risk are best suited for ensuring that more intense treatments and more rigorous supervisory techniques are used with the offenders who need them most, those at high risk. The difficulty at present is that the characteristics of offenders that makes them high risk are also those that are related to poor response to intervention. Moreover, no treatments that reduce the recidivism of psychopathic offenders have yet been identified. If ever there was a problem that cried out for a solution, it is this one.40

Olympia lawyer Charlie Williams is the president-elect of WACDL. Olympia psychologist Brett Trowbridge, Ph.D., J.D., is the Executive Director of the Trowbridge Foundation, devoted to the advancement of forensic psychology; his work on this article was supported by the Trowbridge Foundation.

 Notes

1. V. L. Quinsey et al., Violent Offenders: Appraising and Managing Risk 171 (1998)

2. Thomas Grisso, “Improving Psychologists’ Assistance to Criminal Courts,” Address before the Psychological Expertise and Criminal Justice Joint Conference of the American Psychological Association and the American Bar Association, October 15, 1999, Washington, D.C.

3. See, e.g., John Monahan’s findings that for institutionalized mentally ill offenders who had a high base-rate for violence, “psychiatrists and psychologists are accurate in no more than one of three predictions of violent behavior over a several year period.” J. Monahan, The Clinical Prediction of Violent Behavior 47 (1981).

4. Barefoot v. Estelle, 463 U.S. 880, 77 L.Ed.2d 1090, 103 S.Ct. 3383 (1983).

5. Barefoot v. Estelle, 463 U.S. at 897; see, e.g., Schall v. Martin, 467 U.S. 253, 81 L.Ed.2d 207, 104 S.Ct. 2403 (1984) (upholding the constitutionality of a New York statute authorizing pretrial detention of juveniles based on finding of ‘serious risk’ that juvenile might commit crime).

6. Backstrom v. Herold, 383 U.S. 107, 15 L.Ed.2d 620, 86 S.Ct. 760 (1966).

7. See H. J. Steadman & J. J. Cocozza, Careers of the Criminally Insane: Excessive Social Control of Deviance (1974).

8. E. I. Megargee, “The Prediction of Violence with Psychological Tests,” 2 Current Topics in Clinical and Community Psychology 98 (C. D. Spielberger ed. 1970).

9. V. L. Quinsey, “The Long Term Management of the Mentally Disordered Offender,” Mental Disorder and Criminal Responsibility 137 (S. J. Hucker et al. eds. 1981).

10. V. L. Quinsey & R. Ambtman, “Variables Affecting Psychiatrists’ and Teachers’ Assessments of the Dangerousness of Mentally Ill Offenders,” 47 J. Consulting & Clinical Psychol. 353 (1979).

11. See, e.g., W. M. Grove & P. E. Meehl, “Comparative Efficiency of Informal (Subjective, Impressionistic) and Formal (Mechanical, Algorithmic) Prediction Procedures: The Clinical-Statistical Controversy,” 2 Psychol., Pub. Pol’y, & L. 293 (1996); Thomas Grisso & P. S. Appelbaum, “Structuring the Debate about Ethical Predictions of Future Violence,” 17 Law & Human Behavior 482 (1993).

12. V. L. Quinsey et al., Violent Offenders: Appraising and Managing Risk 40 (1998)

13. C. D. Webster et al., HCR-20: Assessing Risk for Violence, Ver. 2 (1997).

14. P. Kropp et al., Manual for the Spousal Assault Risk Assessment Guide (3d ed. 1999).

15. D. P. Boer et al., “Assessing Risk for Sexual Violence: Guidelines for Clinical Practice,” Impulsivity: Theory, Assessment and Treatment 326 (C. D. Webster et al. eds. 1997).

16. See also Donald G. Dutton, The Abusive Personality: Violence and Control in Intimate Relationships (1998).

17. Robert D. Hare, Manual for the Hare Psychopathy Checklist -- Revised (1991).

18. Robert D. Hare, Without Conscience: The Disturbing World of the Psychopaths Among Us 96 (1999).

19. G. T. Harris et al., “Violent Recidivism of Mentally Disordered Offenders: The Development of a Statistical Prediction Instrument,” 20(4) Crim. Just. & Behavior 315 (1993).

20. G. T. Harris et al., “Psychopathy and Violent Recidivism,” 15 Law & Human Behavior 625 (1991).

21. See A. E. Forth et al., “Assessment of Psychopathy in Male Young Offenders,” 2 Psychol. Assessment: J. Consulting & Clinical Psychol. 342 (1990); and R. Loeber, “Development and Risk Factors of Juvenile Antisocial Behavior and Delinquency,” 10 Clinical Psychol. R. 1 (1990).

22. Randy Borum, “Advances in Assessment of Dangerousness and Risk,” 2 Psychological Expertise and Criminal Justice 465 (1999).

23. C. D. Webster et al., The Violence Prediction Scheme: Assessing Dangerousness in High Risk Men (1994).

24. Why these somewhat counter-intuitive negative correlations to victim injury and female victims? We don’t know, and the authors don’t really explain.

25. The highest score for non-violent criminal history goes to robbery, arson, threatening with a weapon, grand larceny, malicious mischief, and fraud. Failure on prior conditional release includes parole and probation violation and revocation, failure to comply, bail violation, and any new arrests while on conditional release.

26. Henry Steadman et al., “A Classification Tree Approach to the Development of Risk Assessment Tools,” Law & Human Behavior (in Press). The MacArthur device is said to yield a predictive accuracy of 82% -- R. Borum, “Advances in Assessment of Dangerousness and Risk,” 2 Psychological Expertise and Criminal Justice 473 (1999).

27. W. Gardner et al., “A Comparison of Actuarial Methods for Identifying Repetitively Violent Patients with Mental Illness,” 20 Law & Human Behavior 35 (1996).

28. J. Monahan, “Review, The Violence Prediction Scheme,” 22 Crim. Just. & Behavior 446, 447 (1995).

29. Under Washington law, it is not enough that an expert vouch for the validity of a given explanatory theory or technique. There is an additional requirement derived from a federal case, Frye v. United States, 293 F. 1013 (D.C. Cir. 1923), that the theory or technique be generally accepted in the relevant scientific community. State v. Swan, 114 Wn.2d 613, 790 P.2d 610 (1990). The issue will probably be that of identifying the relevant scientific community -- forensic psychologists and/or other groups. There are also other admissibility issues under the Rules of Evidence (for example, probative value under ER 401, prejudicial impact under ER 403, specialized knowledge under ER 702, and reasonable reliance under ER 703). But the additional reliability requirements for admission under Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993) -- namely, testability, error rate and peer review -- have not yet been applied to state criminal proceedings.

30. In re Young, 122 Wn.2d 1, 857 P.2d 989 (1993).

31. Kansas v. Hendricks, 521 U.S. 346, 138 L.Ed.2d 501, 117 S.Ct. 2072 (1997).

32. See, e.g., the special theme issue on “Sex Offenders: Scientific, Legal and Policy Perspectives” found in 4 Psychol., Pub. Pol’y, & L. (1998). This issue is an invaluable resource for practitioners who defend “sexual predator” cases.

33. V. L. Quinsey et al., Violent Offenders: Appraising and Managing Risk 156 (1998).

34. R. K. Hanson & M. Bussiere, “Predicting Relapse: A Meta-analysis of Sexual Offender Recidivism Studies,” 66 J. Consulting & Clinical Psychol. 348.

35. M. L. Lalumiere & V. L. Quinsey, “The Discriminability of Rapists from Non-Sex Offenders Using Phallometric Measures: A Meta-analysis,” 21 Crim. Just. & Behavior 150 (1994); V. L. Quinsey & M. Lalumiere, Assessment of Sexual Offenders against Children (1996); M. E. Rice & G. T. Harris, “Cross-validation and Extension of the VRAG for Child Molesters and Rapists,” 21 Law & Human Behavior 231 (1997).

36. Reprinted in R. K. Hanson & M. T. Bussiere, “Predicting Relapse: A Meta-analysis of Sexual Offender Recidivism,” 66 J. Consulting & Clinical Psychol. 348 (1998).

37. R. K. Hanson, The Development of a Brief Actuarial Scale for Sexual Offense Recidivism (1997).

38. L. Augimeri et al., Early Assessment Risk List for Boys: EARL-20B, Version 1 -- Consultation Edition (1998); see R. Borum, “Assessing Violence Risk among Youth,” J. Clinical Psychol. (in Press).

39. Some authors have, however, done a convincing job of describing the dynamics of such behavior -- see, e.g., Aaron T. Beck, Prisoners of Hate: The Cognitive Basis of Anger, Hostility, and Violence (1999).

40. V. L. Quinsey et al., Violent Offenders: Appraising and Managing Risk 222-23 (1998) (emphasis supplied).

Figure A: SARA Factors

Criminal history

  1. Past assault of family members
  2. Past assault of strangers or acquaintances
  3. Past violation of conditional release or community supervision

Psychosocial adjustment

  1. Recent relationship problems
  2. Recent employment problems
  3. Victim of and/or witness to family violence as child or adolescent
  4. Recent substances abuse/dependence
  5. Recent suicidal or homicidal ideation/intent
  6. Recent psychotic and/or manic symptoms
  7. Personality disorder with anger, impulsivity or behavioral instability

Spousal assault history

  1. Past physical assault
  2. Past sexual assault/sexual jealousy
  3. Past use of weapons and/or credible threats of death
  4. Recent escalation in frequency or severity of assault
  5. Past violation of no-contact orders
  6. Extreme minimization or denial of spousal assault history
  7. Attitudes that support or condone spousal assault

Alleged (current) offense

  1. Severe and/or sexual assault
  2. Use of weapons and/or credible threats of death
  3. Violation of no-contact order

Back

Figure B: Hare PCL-R

  1. Glibness/superficial charm
  2. Grandiose sense of self-worth
  3. Need for stimulation/proneness to boredom
  4. Pathological lying
  5. Conning/manipulative
  6. Lack of remorse or guilt
  7. Shallow affect
  8. Callous/lack of empathy
  9. Parasitic life-style
  10. Poor behavioral controls
  11. Promiscuous sexual behavior
  12. Early behavior problems
  13. Lack of realistic, long-term goals
  14. Impulsivity
  15. Irresponsibility
  16. Failure to accept responsibility for one's actions
  17. Many short-term marital relationships
  18. Juvenile delinquency
  19. Revocation of conditional release
  20. Criminal versatility

Back

Figure C: VRAG Variables

  1. PCL-R score (psychopathy)
  2. Elementary school maladjustment score
  3. Diagnosis of any personality disorder
  4. Age at time of index offense (negatively related)
  5. Separation from either parent (except death) under age 16
  6. Failure on prior conditional release
  7. Nonviolent offense history score
  8. Never married or equivalent
  9. Diagnosis of schizophrenia (negatively related)
  10. Most serious victim injury (from index offense) (negatively related)
  11. Alcohol abuse score
  12. Female victim in index offense (negatively related)

Back

Figure D: Risk Factors of Sex Offense Recidivism

  1. Plethysmograph preference for children
  2. Scale 5 (Masculinity/Femininity) of MMPI
  3. Severely disordered
  4. Deviant sexual preferences (pretreatment)
  5. Prior sexual offenses
  6. Any personality disorder
  7. Negative relationship with mother
  8. Scale 6 (Paranoia) of MMPI
  9. Low motivation for treatment
  10. Victim stranger
  11. Antisocial personality disorder
  12. Plethysmograph preference for boys
  13. Victim female child (negatively related)
  14. Prior offenses (any non-sexual)
  15. Anger problems
  16. Age (negatively related)
  17. Early onset of sexual offending
  18. Prior offenses
  19. Victim related child (negatively related)
  20. Single (never married)
  21. Diverse sex crimes

Back

 

 

Home | Links | About the Trowbridge Foundation | Newsletter | Forensic Cases
Noteworthy Forensic Conferences | Abstracts of Available Articles & Presentations