main | contact | curriculum | bookstore | library | newsletter | personal |

Brent E. Turvey
March, 1996

"Dangerousness: Predicting Recidivism in Violent Sex Offenders"

Turvey, B., "Dangerousness: Predicting Recidivism in Violent Sex Offenders," Knowledge Solutions Library, March, 1996, Electronic Publication, URL:

Note: Brent E. Turvey, MS is a full partner of Knowledge Solutions, LLC.
He can be reached for comment or consultation by contacting:
Knowledge Solutions; 61535 S Hwy 97, #9-148; Bend, OR 97702
Phone 541-318-8293; Email:


When a violent sex offender has been tried and convicted of his crime(s), he may be proffered to a mental health facility of some variable nature for assessment and treatment. The mission statements and treatment goals of North American facilities vary widely. However, due to the fact that bed-space and funds are finite, all such institutions are compelled to make reviews and evaluations of patients for release back into society. All institutions treating violent sex offenders will eventually release a percentage of their sex offender population. Therefore all such institutions invariably tender themselves to the considerations of sex offender dangerousness. Definitions vary widely. The operant definition of dangerousness in this work will be: the likelihood that the sex offender will recidivate.

When a deciding body releases a convicted violent sex offender from a mental health institute back into society, it has potentially put the wolf back into the sheep pen. The reality is that the sex offender could do it again, possibly escalating. That is precisely why good tools for dangerousness assessment are necessary.

Case Example:

By the time Jerry Brudos was sixteen, in 1955, his fantasies had already developed into strong and complex motivations for his behavior. He had an extensive collection of lingerie that he had stolen from the girls in his neighborhood. He masturbated frequently, using the stolen garments to fuel and develop his fantasies.

Eventually Jerry asked one of the neighbor girls whom he had stolen lingerie from to come over to his home on a night that he knew the house would be empty. She was eighteen years old. He had promised to give her private belongings back. She was met in his house by a masked individual with a knife who forced her into Jerry's bedroom. He then forced her to remove her clothing and pose while he took photographs. Later, her statement to the police was specific in mentioning that she believed the masked individual was actually Jerry Brudos.

When the masked individual left, she put her clothes back on. As she was leaving, Jerry appeared and claimed he had been in the barn and that someone had locked him in. She left quickly, quite terrified, she later told the police.

Jerry developed the pictures he had taken of her and kept them.

In 1956, Jerry got a seventeen year-old girl into his car, offering her a short ride. He drove her to a secluded, abandoned farmhouse. He dragged her screaming from the car and then beat her face and breasts. He then ordered her to removed her clothing. A couple from a nearby farm happened by and interrupted the incident.

Jerry was turned over to the State Police, who proceeded with an investigation. A search was made of his room and several years worth of photographs of frightened young girls and stolen women's lingerie were discovered. Jerry maintained that none of this material belonged to him.

In April of 1956, Jerry Brudos was committed to the Oregon State Hospital in Salem. On April 16th, one Psychiatrist wrote of Jerry "There is no evidence of suicide, homicide, or destructive urges,". Jerry Brudos was released before the end of the year. His official diagnosis had been "Adjustment Reaction of Adolescence with sexual deviation, fetishism."

Jerry Brudos would later abduct, pose, mutilate, kill, then rape Linda Slawson, Jan Whitney, Karen Sprinker and Linda Salee before his second capture and incarceration. All of the deaths involved extreme acts of sexual sadism. Despite his confessions and subsequent denials, the events are recorded in detail by Jerry's own camera, as he photographed his victims moment by moment. He kept those photos hidden in a secret compartment in his workbench at home, along with nude photos of his wife. Some of the photos include Jerry himself, interacting with the victims(Ressler[14]).

The question begged by the Brudos case is very basic: Could Jerry Brudos's later crimes have been predicted by any known factors and then subsequently any recidivism prevented by continued incarceration?

The Goal

The objective of this work is to present the best tools and research available for predicting sex offender dangerousness. The goal behind these tools is to provide for prediction and prevention of sex offender recidivism; to effectively screen any offenders who are likely to recidivate from re-entering society. This would result in a reduction of those sex crimes specifically targeted by the dangerousness assessment techniques to be mentioned.

The techniques for predicting future dangerousness to be suggested in this paper are inappropriate age selection of victim by the offender, and offender pattern analysis. Hence, this paper will review the findings of Dr.Vernon Quinsey of Queens University, and the findings of the National Center for the Analysis of Violent Crime.

The Victims

Recent statistics generated by the BJS point to an alarming shift in victim population of choice in reported rapes since 1987. For the years 1973 through 1987, the average number of all rapes to females reported annually to the police in the U.S. was 154,000, with a 53% estimated reporting rate(Harlow[6]).

This data shows 54,400 rapes to females within the 12-19 year old range, representing roughly 35% of the total number of reported rape victims.

During 1992, only 109,000 rapes to females were reported to law enforcement agencies in the U.S. That's well below the 154,000 average rapes of females per year for the sample in Table 1. However, one half of the females raped in 1992 were juveniles under the age of eighteen years old(Half of Women...[4]). Furthermore, 17,000(16%) of the females raped in 1992 were under the age of twelve[4]. The average age of female victims of rape continues to fall.

This data represents a fundamental shift in the way law enforcement and mental health must view the population female rape victims. Subsequently, it must also represent a fundamental shift in the way law enforcement and mental health view sex offenders. The known female rape victim of choice, after a fashion, is white more than 50% of the time, and she is a juvenile more than 50% of the time[4].

While overall rape is down over the past ten years, the number of rapes to girls under the age of eighteen years old has almost doubled. That increase is significant to the relevance of the first method proposed for the prediction of dangerousness, as will be discussed in METHOD #1.


Inappropriate Age Selection

Dr. Vernon Quinsey has been doing clinical research on sex offender dangerousness for the past decade and a half. He currently holds the Psychology Chair at Queen's University in Kingston, Ontario. He does most of his research sampling out of the sex offender population from the Penetanguishene Mental Health Centre, a maximum security psychiatric hospital right across the street from the University.

One of his findings regarding dangerousness, in 1986, concerned a group of 360 such offenders. That study showed a lack of relation between clinical diagnosis of dangerousness and actual recidivism(Quinsey[13]). One factor he pointed to was the general disagreement between psychiatrists as to a method for predicting dangerousness.

Another factor Dr. Quinsey pointed to was objectivity. He published a quality control study in December of 1986 which data showed that, "Perceptions of treatability, however, may be more like moral judgments or judgments of likeability than they appear to be. If perceptions of treatability are moral judgments, then the negative correlation of dangerousness and treatability makes sense,"(Quinsey[10]).

In 1990, Quinsey published another study, providing some basic points of interest for those interested in assessing dangerousness. He collected recidivism and readmission data on an average of 46 month follow-up periods on 54 rapists. Sexual recidivism and violent recidivism were well predicted by phallometrically measured sexual interest in nonsexual violence and degree of psychopathy (Quinsey[11], Hare[5]).

Using the same method to gauge the range of the offenders' sexual preferences (measurement of penile tumescence before, during, and after audio descriptions of an array of sexual encounters, with a baseline tumescence recovered between each description), Quinsey published an earlier study with associated results. This time with a sample of child molesters, he demonstrated that " with children was the most highly related to victim damage...compared to others, subjects who had injured their victims engaged in violence or sadism during the rape, showed little sexual interest in adults and a lot of interest in coercive sex with children"(Quinsey[9]).

It is possible to list dozens more such published studies conducted by Dr. Quinsey and his associates, that indicate the following are excellent phallometric predictors of recidivism:

1) The preference for violence or sadism during rape

2) Degree of Psychopathy of an offender[5]

3) The inappropriate sexual age preferences as measured phallometrically

In 1991, Dr. Quinsey published a study which data supported, vehemently, a specific predicting element of a sample of child molesters (Quinsey[12]). The recidivism of 136 extrafamilial child molesters was determined on an average of a 6.3 year follow-up. 43% of those in the study committed a new violent or sexual offense. 58% were arrested for an offense of some kind or were returned to the institution. Inappropriate sexual age preferences as measured phallometrically were related to new convictions for sexual offenses. Inappropriate age choice is operationally defined by Quinsey [12] as either a juvenile the age of 15 or under, or as a juvenile at least 5 years younger than the offender. Also of note is the result that behavioral treatment was shown to have no effect on recidivism. Patient self reports were not related to recidivism at all.

So Vernon Quinsey's contribution to dangerousness in the last 15 or twenty years has been supportive of these results, consistently:

Pyschopathy is positively correlated to recidivism

Violence/sadism are positively correlated with recidivism

Inappropriate age choice of victim is a consistent predictor of recidivism

Patient self reports of sexual preferences have no bearing on recidivism

Psychiatrists tend not to be very objective about dangerousness assessment

The tools provided for by Vern Quinsey don't require a lot of guesswork. They are all excellent predictors with good data to back them up. They provide a solid foundation for any multivariate approach to dangerousness assessment.

The most objective assessment tool is of course the offenders inappropriate age choice of the victim. The juvenile population of female victims of rape accounts for more than half of all reported rapes. The relevancy of the victim age indicator of dangerousness for extrafamilial child molesters, in light of the growing number of victims in that age range, should not be ignored. Simply put, Quinsey provides a tool that could have applications to the offenders who are responsible for 50% or better of the female victim population. That is seriously significant.

A problem with this methodology is pointed out by Dr. William Marshall who shows clearly that sexual arousal, i.e., erectile responses, is not always a functional part of sexual offenses committed against women(Marshall[8]). He makes an extremely valid point. Sexual assaults take many forms. The lack of an erect penis on the part of a male offender is no true sign of a lack of male sexual arousal during such an incident. Nor does such an assault have to be the product of arousal at all. That's why you could never rely on just one indicator when assessing dangerousness.

Quinsey has given us three things to look at. Psychopathy, for which there is a checklist designed very carefully and thoughtfully by Dr. Robert Hare[5]. A preference of the sex offender for sadism and violence, which can both be estimated by victim damage and measured in phallometric studies. Lastly, inappropriate age choice, which has been measured by phallometric studies. Each by itself makes a fairly convincing argument for predicting recidivism. Together they are terribly convincing.


Offender Type

This model of thinking is not biological. It rates the offender's behavior from the victim's point of view and from the offenders point of view. On the subject of recidivism, Robert Ressler puts it very well indeed when he states, "I prefer to reason from the which a man whose violent fantasies had been developing since early childhood eventually realized those fantasies..."(Ressler[14]). There is a pattern to the violent sex offender's crime that makes it likely he will offend again. The first part is the fantasy, which develops and grows within the offender over time. The second part is an increase in violence or sadism, which agrees with Quinsey. The overall pattern is one of tandem fantasy/behavior escalation.

Groth[3] defines sadism as a single psychological experience that fuses sexuality and aggression. Violence is damage done to the victim without sexuality in the mind of the offender. So sadism is done for the pleasure of the offender; to feel sexual pleasure from knowing the pain he is giving the victim. Violence is just meant to hurt the victim, with the offender no feeling any specific sexual pleasure from it.

Case Example:

Let's think about Jerry Brudos when he was 16: He'd been collecting women's panties and shoes and underthings from a very early age. He had an enormous collection; looked through it all the time, felt the fabric on his arms and face. Smelling those private things all the time. Putting them on. Touching them to himself. Masturbating chronically.

Then the rush of going out at night and sneaking into neighbor's houses. Into rooms of girls he watched during the day. Older girls with developing bodies. Seeing them around town. Imagining what they wore underneath their dresses and skirts. Feeling powerful once he had their panties and their bras and shoes. Seeing them in the neighborhood later. Knowing he had their private things, with their private smells still on them. He had them. He knew the secret of the underthings against their skin, and against his own.

After a while, Jerry's fantasy gets more involved. He wants a participant. He wants to know up close the secret that the panties and the bras hide. He wants to see those private places that make those private scents that he has held for so long. So Jerry plans very carefully: invite an older girl over, one whose items he'd stolen. Promise to give her things back. Make sure it's an evening when the house will be empty. No interruptions. Get a mask, a knife, and a camera. The camera is the only way to preserve what he will see, so that he can see it again and again, whenever he wants. And she comes over, and he takes out the knife, and she does whatever he wants in his room. The knife keeps him in control. And she's afraid and he is very excited. And he burns a roll of film. When she's gone, he doesn't care. He's gotten the pictures; a frightened, naked girl. She is crying and begging to be let go. And only he can do that because he has the knife. He develops the pictures himself and keeps them with the lingerie, masturbating often, fantasizing about events over and over again. And he gets more victims, and takes more photos. It felt so good to have that control, and to keep reliving it.

But it doesn't last long enough. So his fantasy develops to make it last longer, with more interaction with the victim. Not just pictures. Eight months later he gets a girl into the car with him, but he doesn't have his knife. He drives her someplace secluded and uses his hands, beating her to make her feel his control. Later, after his release, his fantasy would escalate further. His behavior would come to include a pistol, pre-mortem bondage, and the forced posing of victims in lingerie he had stolen from other women, often after or during the moment of death.

Of note is the fact that Jerry Brudos strangled his victims. He did not kill them with the knife or the gun. Those props were just to gain and keep control. They would have made his victims too messy for potential lingerie photographs later, ruining the fantasy.

Jerry Brudos's clear escalation of fantasy and behavior, which included sadism, was a prime indicator of dangerousness. His fantasy was a long time in the brewing. His behavior was deliberate, fueled by the fantasy. He liked what the fantasy made him feel, and he wanted to feel more.

As Ressler points out[14], this kind of pattern is indicative of someone who is going to do it again when they are released. The sexual sadist is one who has established an enduring pattern of sexual arousal in response to sadistic imagery (Ressler[15]). And sadism is an inextricable part of the anger-excitation rapist classification. Quinsey would agree, saying of an offender with sadistic fantasies that behavioral treatment does not seem to affect recidivism[12].

According to the rapists classification scheme, designed by the National Center for the Analysis of Violent Crime(Hazelwood[7]), from the work done by Groth[3], and more thoroughly published in The Crime Classification Manual(Ressler et al[15]), there is one general type of rapist that will be more likely to repeat their crimes after release and is generally resistant to treatment. This is the anger-excitation rapist.

The classification of an anger-excitation rapist in this scheme involves both offender data and victim data. As a note, Hazelwood reports a more accurate classification rate from the victim's report of the rape(mid 90%)[7]. Essentially it involves extensive interviews with both offender and victim, and a significant time investment by the attending psychiatrist. Below are the things that the psychiatrist should be look for to discern an anger excitation rapist.

The likely profile of the anger-excitation rapist [15].

Uses a con to gain victim confidence

Sexually stimulated by victim's responses to the infliction of physical and emotional pain.

Seeks total fear and submission from victim

Well planned, executed methodically


Offender brings his own weapons, bindings, props

Offender is sexually and verbally selfish

Offense lasts for an extended period of time

Sadism may be muted with little physical injury to victim, but may include eroticized aggression, often focused on genital areas, as:

sexual bondage

likely record the event

The victim is susceptible to a con

The victim is almost always a stranger

These are the sex offenders who are most likely to do it again. They are offenders who can sustain violence and sexual arousal in tandem. They have an above average intelligence, they have a desire to cause the victim pain for their own sexual pleasure, they have a rich, deeply developed fantasy life[15], and they are most often resistant to treatment of any kind[12]. When they are released from their perspective institutions, as was Jerry Brudos at age 17, they are the most likely to do it again. By applying this classification to such individuals, it is possible to screen them from re-entry into society. The subsequent reduction of recidivism would represent a significant decrease in those kinds of crimes.


Self reporting tends to be a heavily relied upon staple for most psychiatrists when assessing offender progress. This method assumes that the offender wants to get well. That is rarely the case with a violent sex offender.

As Quinsey has already demonstrated, Dangerousness often becomes a likeability test for the offender. A moral judgment rather than an empirical one made by the psychiatrist. This is not a problem for the violent sex offender. He is a master of impression management [2], [5]. He knows just what to say to appear healthy and likable. "Virtuelessness paints a more convincing picture of virtue than virtue can of itself,"[5].

The most sadistic and prolific violent serial sex offenders fool their respective psychiatrists quite easily. Many such offenders have been under the care psychiatrists, who coincidentally believed them to be making excellent progress, and all the while they killing and raping human beings[2]. Well-known examples of this eventuality include Arthur Shawcross, Monte Rissel, and Edmund Kemper.

An extreme example of this is an instance related by S.A. John Douglas of the NCAVC[5]. In his most recent work, S.A. Douglas tells of an inmate named Thomas Vanda in Joliet Penitentiary who was convicted then incarcerated for the homicide of a young lady. S.A. Douglas approached the prison psychiatrist about Mr. Vanda, who stated that Mr.Vanda was responding very nicely to medication and treatment. Bible study was mentioned as having a positive effect on Vanda. The prison psychiatrist also stated with affirmity that Vanda was an excellent candidate for parole if his progress continued.

As it turned out, the prison psychiatrist made it policy not to learn the details of offenses committed by particular inmates. It is his feeling that such details unfairly influence his judgment against his clients, in this case, Mr.Vanda. But S.A. Douglas was not dissuaded and proceeded to recount the details of Mr. Vanda's crime.

Thomas Vanda had joined a church group. After everyone had gone home, he propositioned the young lady who had hosted the meeting. The young lady turned him down and Mr. Vanda became violent. He got a knife from her kitchen, and proceeded to inflict multiple sharp force injuries to her arms, hands, chest, and abdomen. Her wounds were fatal. As she was on the floor, hemorrhaging, Mr. Vanda inserted his penis into an open wound in her abdomen and masturbated until he ejaculated into that wound.

The psychiatrist became very upset at S.A. Douglas, and asked him to leave.

The arousal associated directly with the element violent sadism is key to the pattern of the recidivent violent sex offender. Mr. Vanda's behavior at that crime scene suggests that he will re-offend if released. Furthermore, his continued interest in bible study suggests a refinement of his previous con.

The violent sex offender is a manipulator. He uses self-reporting to his advantage. He knows how to get people to like him. He is a model inmate. He is a model patient. His specialty is impression management[5]; getting you to buy into his con. His words, therefore, cannot be believed. Self reporting is unreliable.

Only offender behavior can be assessed with any near sense of objectivity. The lesson from the Vanda case, and others much like it, is clear; without knowing the offender's history, and his previous patterns of behavior, current behavior can not accurately be assessed for progress in a treatment program of any kind.


The NCAVC and Dr.Vernon Quinsey have each provided the mental health community with some very timely and useful tools and guidelines for the prediction of violent sex offender dangerousness. Their research is sound and ongoing.

Dr. Quinsey's biological method of phallometric measurements is, as mentioned, not to be used without other criteria, and is specific only to sex offenders from a maximum security psychiatric facility. The NCAVC's classification method is more organic, based on a broader range of reported personality and behavioral markers. A combination of the two approaches would likely yield the best results.

Just a side note on the psychopathy checklist. It is an excellent tool, the only problem being that each individual psychiatrist has his or her own individual interpretation of it. That is its greatest weakness. More discussion on psychopathy is definitely warranted. On the question of recidivism of psychopaths, the author recalls a course as a Psychology undergraduate taught by a retired psychotherapist. In a lecture on the necessity of strong positive regard for the client at all times, he said, quite sternly, that if you ever come in contact with a psychopath, don't walk; run. And if they're a sadistic psychopath, see to it that they get locked up forever.

The ideal assessment regiment for sex offenders who are candidates for release from a psychiatric facility would include the following;

Degree of Psychopathy(Groth[3], Hare[5])

Inappropriate age choice of victim measurement(Quinsey[12])

Classification of rapist type(Ressler[14],[15])

With these tools of dangerousness assessment at their disposal, those who would unlock the doors and set free the likes of Jerry Brudos and Thomas Vanda may be better dissuaded. If those individuals who are most likely to recidivate can be more easily identified and screened by the mental health community, then there will be a reduction in the most violent and dangerous types of sex offenses in our society. We will subsequently have fewer juvenile girls who are dead or in therapy because they happened to be found by a monster that the mental health community failed to keep out of society.


1.Bernard, Thomas & Vold, George, Theoretical Criminology 3rd Ed., (New York: Oxford University Press, 1986)

2.Douglas, John, "Mindhunter" (New York: Plenum, 1995)

3.Groth, A. Nicholas, Men Who Rape: The Psychology of the Offender, (New York: Plenum Press, 1979)

4."Half of Women Raped During 1992 Were Younger than 18 Years Old," BJS Statistics, USDOJ, Advance for Release at 5P.M. EST, Wednesday, June 22, 1994

5. Hare, Robert, "Without Conscience" (New York: Pocket Books, 1993)

6. Harlow, Caroline, "Female Victims of Violent Crime," BJS Statistics, (USDOJ: January 1991)

7.Hazelwood, R. & Reboussin, R. & Warren, J. & Wright, J., "Prediction of Rapist Type and Violence from Verbal, Physical, and sexual scales," Journal of InterpersonalViolence, Vol. 6, No.1, March, 1991, pp.55-67

8. Marshall, William & Blader, Joseph, "Is Assessment of Sexual Arousal in Rapists Worthwhile? A Critique of Current Methods and the Development of a Response Compatibility Approach," Clinical Psychology Review, 1989, Vol.9 569-587

9.Quinsey, Vernon & Chaplin, Terry "Penile Responses of Child Molesters and Normals to Descriptions of Encounters with Children Involving Sex and Violence" Journal of Interpersonal Violence, Vol.3, No.3, September 1988 259-274

10.Quinsey, Vernon & Cyr, Mireille, "Perceived Dangerousness and Treatability of Offenders: The Effects of Internal Versus External Attributions of Crime Causality" Journal of Interpersonal Violence, Vol.1, No.4, December 1986 458-471

11.Quinsey, Vernon & Harris, Grant & Rice, Marnie, "A follow-Up of Rapists Assessed in a Maximum Security Psychiatric Facility," Journal of Inter-personal Violence,Vol.5, No.4, December 1990 435-448

12.Quinsey, Vernon & Harris, Grant & Rice, Marnie, "Sexual Recidivism Among Child Molesters Released From a Maximum Security Psychiatric Insti-tution", Journal of Consulting and Clinical Psychology, 1991, Vol. 59, No.3 ,381-386

13.Quinsey, Vernon & Maguire, Anne, "Maximum Security Psychiatric Patients: Actuarial and Clinical Prediction of Dangerousness," Journal of Inter-personal Violence, Vol.1, No.2, June 1986 143-171

14. Ressler, Robert, Whoever Fights Monsters, (New York: St. Martin's Press, 1992)

15. Ressler, Robert & Douglas, John & Burgess, Ann & Burgess, Allen, The Crime Classification Manual (New York: Lexington Books, 1992)

Updated 2/19/02
Knowledge Solutions LLC 1996 - 2002