Sex offender management using the polygraph: A critical review (original) (raw)
Abstract
Reducing recidivism is a central goal in the treatment of sex offenders. In Europe, there is an increased interest in using the polygraph (“lie detector”) as a tool in the treatment and risk assessment of convicted sex offenders. This interest originated from optimistic reports by American clinicians who argued that polygraph testing in the treatment of sex offenders is akin to urine analysis in the treatment of drug addiction. In this article, we critically examine the validity and utility of post-conviction sex offender polygraph testing. Our review shows that the available evidence for the claims about the clinical potential of polygraph tests is weak, if not absent. We conclude that portraying post-conviction polygraph testing as analogous to urine analysis is inaccurate, misleading, and ultimately, risky.
Introduction
Recidivism among sex offenders poses a serious problem. Data from the United States Bureau of Justice Statistics (2003), show that 5% of about 10,000 sex offenders who were released from prison in 1994 were rearrested for a sex crime within 3 years. At 43%, rearrests for any type of crime were substantially higher. A review of 61 studies (n
=
23.393) demonstrated that sexual recidivism was 13% after a mean follow-up of 4–5 years (Hanson & Bussiere, 1998). Recidivism for any reoffense was 36%. Moreover, these statistics are an underestimation of true recidivism rates. Follow-up periods during which recidivisms are monitored are by definition restricted, and not all crimes lead to rearrest or reconvictions (Kilpatrick, Edmunds, & Seymour, 1992). Unfortunately, highly effective treatment programs do not yet exist. Meta-analyses on the efficacy of sex offender treatment programs have shown that cognitive-behavioral therapy can decrease recidivism, but that this decrease is only modest (Hall, 1995, Hanson et al., 2002). More recent and more carefully designed studies failed to find any treatment effects at all (Marques et al., 2005, Schweitzer and Dwyer, 2003).
Sex offenders often deny the seriousness and impact of their crimes. Embracing a more truthful and complete report of their crimes is considered an important aspect of treatment. Risk assessment would also benefit from offenders' truthful self-report about their criminal history and pathology. Contemporary risk assessment is based on actuarial risk predictors, which are risk factors derived from empirical research (Hanson, 2003). Hanson and Bussiere's meta-analysis showed that sexual offense recidivism was best predicted by measures of sexual deviancy (e.g., deviant sexual preferences, prior sexual offenses) and, to a lesser extent, by general criminological factors (e.g., age, total prior offenses). Although actuarial risk assessment has improved the prediction of recidivism compared to clinical judgement (Dawes et al., 1989, Grove and Meehl, 1996, Grove et al., 2000, Quinsey et al., 2006), this prediction is not faultless (Hanson, 2003, Harris et al., 2003). Langton et al. (2007) assessed the predictive ability of 6 contemporary actuarial instruments in 468 sex offenders who were followed for an average of 5.9 years. Predictive accuracy was expressed in terms of the area under the Receiver Operating Curve (ROC a). This statistic can be interpreted as the probability that a randomly chosen sex offender who reoffended obtained a higher score than a sex offender who did not. Values range from 0 to 1, with .5 reflecting chance level. The results showed that for all instruments, accuracy was significantly above chance. With ROC a values ranging from .57 to .72 for any crime, and from .61 to .71 for sex crimes, accuracy was, however, only modest.
Offender's records are used to score actuarial risk assessment instruments, and, although objective, this information is often incomplete. As an example, the number of prior sex offenses present in the records is probably an underestimation of the true number of sex offenses. Relying exclusively on file information, thus, poses a serious problem, and good practice requires corroborating offender's self-report. This, in turn may prove problematic as offenders have a vested interest in concealing their true behavior, as good behavior is often rewarded with such benefits as parole. In this context, some authors have referred to offenders' fake good tendencies as “supernormality” (Cima et al., 2003). This supernormality is a major obstacle for accurate risk assessment. All together, both treatment and risk assessment in sex offenders could benefit from more honest self-report.
In this review, we critically evaluate a method for verifying the veracity of offenders' self-reports; post-conviction sex offender polygraph testing (PCSOT). This method has steadily gained popularity among clinicians over the last decade, and was used in 70% of community sexual abuser programs for adult males in 2002 in the United States (McGrath, Cumming, & Burchard, 2003). Proponents argue that this method contributes to a more accurate and more complete sexual history taking, more accurate offense descriptions, superior assessment of high-risk behaviors, and better compliance with probation conditions. The popularity of the method among practitioners is illustrated in the opinion expressed by the former chair of the American Polygraph Association: “Those of us who conduct post-conviction sex offender testing know that it has been the missing link in preventing recidivism (Consigli, 2002, p. 239)”. American practitioners have presented the role of the polygraph in the treatment of sex offenders as similar to that of urine analysis in the treatment of drug addicts (English, Jones, Patrick, & Pasini-Hill, 2003, p. 414). Both are viewed as objective and accurate methods to verify whether the offender's reports are truthful. These positive claims received much attention, and PCSOT was recently introduced in Europe. In the United Kingdom and the Netherlands, highly secure treatment hospitals have started PCSOT studies. European users appear to share Consigli's enthusiasm, promoting widespread use. At the time of writing, in the United Kingdom, the Management of Offenders and Sentencing Bill was in progress, extending the use of polygraph tests for sex offenders on license.
In this review, we discuss whether this enthusiasm is justified. We will first describe the general principles of polygraph testing and how it is used in PCSOT. We will then critically examine the results from PCSOT studies. Finally, we will discuss whether there is any evidence for the notion that PCSOT reduces recidivism, and whether PCSOT is, indeed, the urine test of sex offender treatment.
Section snippets
The polygraph and the detection of deception
The type of polygraph test used in PCSOT is a version of the Comparison Question Technique (Grubin, Madsen, Parsons, Sosnowski, & Warberg, 2004). This type of test is also known as the Control Question Test (CQT; National Research Council, 2003, Reid, 1947), and is best known from its use by law enforcement agencies in criminal investigations. During a CQT, relevant and comparison questions are asked. The relevant questions are directly related to the incident under investigation (e.g., “On the
Polygraph validity
Ever since its introduction, the CQT has been the object of an intense debate, and remains controversial (Ben Shakhar, 2002, Ben Shakhar and Furedy, 1990, Fiedler et al., 2002, Furedy, 1996a, Furedy, 1996b, Honts et al., 1995, Lykken, 1998, Raskin and Honts, 2002). The debate focuses on the assumptions underlying the CQT, thus calling into question the test's accuracy. Many critics have argued that it is impossible to determine the origins of the physiological responses. Stronger responses to
The polygraph and sex offenders
An important reason for lower accuracy in screening applications than in specific-incident testing is that the former involves considerably more ambiguity for the examinee (National Research Council, 2003). This becomes clear when looking at the function of the CQT in PCSOT more closely. PCSOT intends to verify self-reported information about past or current behavior. Firstly, it entails verifying the accuracy and completeness of the sexual history information provided by the offender. This
Disclosure through polygraph testing
Some have argued that the polygraph has its utility, regardless of its accuracy. “I am less concerned about the reliability of polygraph. The fact is that every time I refer a client for polygraph testing, I end up with significantly more information than I started with (Veeder, 1995, in Hagler, 1995)”. Several reports have recently flourished, suggesting that the polygraph is highly successful in obtaining previously undisclosed information (e.g., Ahlmeyer et al., 2000, Emerick and Dutton, 1993
Bogus pipeline
There is a long tradition in social psychology to develop methods that aim to reveal a person's true opinion on sensitive topics. Standard paper-and-pencil tests are believed to be unreliable, because they are subject to response biases such as social desirability (Crowne & Marlowe, 1964). One approach social psychologists have used to circumvent social desirability is attaching the person to an impressive but useless machine. People respond more truthfully when they are led to believe that
Reducing recidivism
The most critical question is not to what extent PCSOT leads to the disclosure of information, but whether it reduces recidivism. Such reduction could be either direct, due to the deterrent effect of an upcoming polygraph on behavior, or indirect through increased disclosure improving risk assessment and/or treatment outcome. We will argue that there is evidence against the former, and no evidence for the latter.
PCSOT has been used since the 1960s, and has been described as the missing link in
Concluding remarks
Treatment of sex offenders is a serious issue. Not only does recidivism occur frequently, the public in many countries is highly sensitive to the topic (Lösel, 2000). With this in mind, it is understandable that practitioners resort to unorthodox techniques such as polygraph testing. At present, however, there is no evidence supporting the accuracy of the CQT in PCSOT. Relevant theory and research suggests that this accuracy will be lower than that obtained in criminal investigations. Users
References (57)
Child Abuse and Neglect
(1997)
- FuredyJ.J.
The North American polygraph and psychophysiology: Disinterested, uninterested, and interested perspectives
International Journal of Psychophysiology
(1996)
- FuredyJ.J.
Some elementary distinctions among, and comments concerning, the ‘control’ question ‘test’ (CQT) polygrapher's many problems: A reply to Honts, Kircher and Raskin
International Journal of Psychophysiology
(1996)
- GannonT.A. et al.
Cognitive distortions in child molesters: A re-examination of key theories and research
Clinical Psychology Review
(2006)
- HontsC.R. et al.
Polygrapher's dilemma or psychologist's chimaera: A reply to Furedy's logico-ethical considerations for psychophysiological practitioners and researchers
International Journal of Psychophysiology
(1995)
- WardT. et al.
Cognitive distortions in sex offenders: An integrative review
Clinical Psychology Review
(1997)
- AbramsS. et al.
Polygraph testing of the pedophile
(1993)
- AhlmeyerS. et al.
The impact of polygraphy on admissions of victims and offenses in adult sexual offenders
Sexual Abuse: A Journal of Research and Treatment
(2000)
- AndreassiJ.L.
Psychophysiology: Human behavior and physiological response
(2000)
- Ben ShakharG.
A critical review of the Control Questions Test (CQT)
- Ben ShakharG. et al.
Theories and applications in the detection of deception: A psychophysiological and international perspective
(1990)
- BranamanT.F. et al.
Polygraph testing in sex offender treatment: A review of limitations
American Journal of Forensic Psychology
(2005)
- BumbyK.M.
Assessing the cognitive distortions in of child molesters and rapists: Developments and validation of the MOLEST and RAPE scales
Sexual Abuse: A Journal of Research and Treatment
(1996)
- Bureau of Justice Statistics
Recidivism of sex offenders released from prison in 1994
(2003)
- CimaM. et al.
The other side of malingering: Supernormality
Clinical Neuropsychology
(2003)
- ConsigliC.J.E.
Post-conviction sex offender testing and the American Polygraph Association
- CrossT.P. et al.
A critique of the validity of polygraph testing in child abuse cases
Journal of Child Sexual Abuse
(1992)
- CrossT.P. et al.
Polygraph testing and sexual abuse: The lure of the magic lasso
Child Maltreatment
(2001)
- CrowneD.P. et al.
The approval motive
(1964)
- DawesR.M. et al.
Clinical versus actuarial judgment
Science
(1989)
- EmerickR.L. et al.
The effect of polygraphy on the self report of adolescent sex offender: Implications for risk assessment
Annals of Sex Research
(1993)
- EnglishK. et al.
Sexual offender containment: Use of the postconviction polygraph
Annuals New York Academy of Sciences
(2003)
- FiedlerK. et al.
What is the current truth about polygraph lie detection
Basic and Applied Social Psychology
(2002)
- GannonT.A. et al.
Increasing honest responding on cognitive distortions in child molesters: The bogus pipeline revisited
Sex Abuse
(2007)
- GroveW. et al.
Comparitive efficiency of informal (subjective, impressionistic) and formal (mechanical, algorithmic) prediction procedures: The clinical–statistical controversy
Psychology, Public Policy, and Law
(1996)
- GroveW.M. et al.
Clinical versus mechanical prediction: A meta-analysis
Psychological Assessment
(2000)
- GrubinD. et al.
Accuracy and utility of post-conviction polygraph testing of sex offenders
British Journal of Psychiatry
(2006)
- GrubinD. et al.
A prospective study of the impact of polygraphy on high-risk behaviors in adult sex offenders
Sexual Abuse: A Journal of Research and Treatment
(2004)
Cited by (0)
The first two authors contributed equally to the manuscript and are listed in alphabetical order. Bruno Verschuere is a post-doctoral fellow of the Scientific Research Foundation – Flanders, Belgium (FWO).
Copyright © 2008 Elsevier Inc. All rights reserved.