Misuse of Psychophysiological Arousal Measurement Data

Jack S. Annon*

ABSTRACT: Psychophysiological arousal measurement data (plethysmograph) is being misused in some state treatment programs, probation and parole departments, and courts.  The high percentage of some control subjects with no history of deviant behavior who respond with deviant arousal patterns means that at the present time arousal measurement data cannot be used by itself to diagnose someone as a sexual deviant or to predict past or future behavior.

I am gravely concerned about what I see as a growing misuse of psychophysiological arousal measurement data in some state treatment programs, probation and parole departments and courts.  Based on my training, research, and experience in this area, I respectfully offer the following information for your serious consideration.

Throughout the l970s and through most of the l980s I was the only practitioner in the state of Hawaii advocating the use of, and using, psychophysiological measurement of arousal patterns as one component in a comprehensive psychosexual assessment of males and females accused, or convicted, of sexual offenses.  For years my use of the plethysmograph to suggest appropriate treatment targets in males and females was dismissed as unreliable and without much merit.

Unfortunately, at the present time the pendulum has now swung to the opposite extreme position, not only here in the state of Hawaii, but across the United States as well.  Out-of-context arousal measurement data are more and more being misused to determine the dangerousness of an individual relative to sentencing purposes, to assess dangerousness relative to parole and conditional release purposes, and to dictate probation requirements, such as whether or not a person may travel.  This is despite the fact that the only research-based supportive function of such measurement is in providing information for specific treatment purposes.

Psychophysiological Screenings Versus Comprehensive Assessments

I believe there are two primary reasons for this misunderstanding in the use of arousal measurement data.  First, there is a failure to discriminate the difference between a psychophysiological screening from a comprehensive assessment, and second, there is a lack of awareness of the clinical and research data underlying the use of such arousal measurements.

In regard to the first area, a screening usually consists of one, or two at most, stimuli in a given category (e.g., one or two slides depicting rape, one or two slides depicting a given sex and age, or one or two tapes describing sexual activity with a child, etc.).  Such a screening is usually carried out in a relatively brief amount of time (e.g., from a half hour to an hour or so), and covers a wide range of possible behaviors (e.g., exhibitionism, voyeurism, rape, child molestation, mutually consenting adult interactions, etc.).  Generally instructions are to look at, or listen to, the stimuli, then "allow yourself to respond in whatever way you wish."  The advantage of these screenings are that you can cover a wide range of possible behaviors in a relatively brief time.

The purpose of such a screen is not to predict past or future behavior, nor to suggest treatment targets — its primary purpose is only to select areas that suggest where a more comprehensive assessment is appropriate.  Such screenings can be used in an institutional system for screening a large number of individuals in a relatively short period of time, in order to determine which individuals would be appropriate for a more comprehensive assessment in a given area.  For example, a particular individual may respond with arousal to listening to a rape tape description equal to, or higher than, his response to a taped mutually consenting adult interaction; but does not respond with much arousal to cues relating to children, or to telephone calls, or peeping.  The next step would be a comprehensive assessment of that area dealing with rape, which would cover issues of physical aggression, verbal coercion, and humiliation, as contrasted with mutually consenting sexual interactions.

Unfortunately, some practitioners with a limited scientific background and knowledge call such screenings a "comprehensive assessment" and then make diagnoses and recommendations — not only in terms of treatment targets, but as to the degree of "dangerousness" of the person as well.  This is a serious misuse of screening data.  As Dr. William Pithers, the immediate past president of the Association for the Treatment of Sexual Abusers, stated in a 1988 deposition concerning this very issue: "... studies that had failed to differentiate populations typically used single stimuli per category; and studies that appeared to differentiate populations with a reasonable degree of reliability and validity used multiple instances of stimuli."

By contrast, a comprehensive assessment would involve:

1. gauge calibration prior to the beginning of every assessment and if there are any questions, directly after the assessment;
2. the use of some form of signal detection task for visual stimuli, with ideally a method for visually monitoring the client's direction of gaze, such as with a closed circuit' camera that is focused on the client's face;
3. a request for the client to summarize or describe the previously presented stimuli that he or she has either heard or seen;
4. a minimum of four stimuli in each stimulus category;
5. instructions randomly given that for half of the stimuli in a given category he or she is to look at it or listen to it, and "experience it as clearly and vividly as possible"; and the other half of the instructions to "suppress your response by any mental means."
6. in contrast to a short presentation in the "screening" process, a stimulus presentation of generally two minutes in length or longer in each stimulus category;
7. a number of subsets of stimuli in a given category (for example, if on the "screening" the individual showed a significant response to descriptions of sexual interactions with male children, then one might want to assess the following child conditions in a comprehensive assessment: a) the child initiates; b) the child is related; c) mutually consenting child interactions; d) verbal coercion e) physical coercion; f) sexual physical brutality; and, g) physical brutality without any sexual interactions).

There are three ways of analyzing the responses from the data that are collected.  One method is to look at the overall percentage of arousal to the different cues to see which significant differences in percentage might indicate appropriate therapeutic targets.  A second method is to review responses to specific behavioral descriptive cues that either bring on a systematic increase or decrease in arousal.  A third method is to compute a comparative index (comparing appropriate versus inappropriate responses) for diagnostic purposes.  This index is computed by using the generally agreed upon formula by which the client's arousal responses to the different offense behaviors are divided by his or her arousal responses to the mutual adult situations.

Use of Psychophysiological Arousal Measurement Data

Finally, I return to the second area of concern which is the appropriate use of the data collected.  Some practitioners with limited training will take any response, regardless of the amount of response, to a deviant stimuli as an indication that the person is a "pedophile," "rapist," "exhibitionist," etc.  Unfortunately, some practitioners will go even further and make statements that unless the individual receives treatment, the person is a danger to the community and has a high probability of acting out his or her behavior.  They then make suggestions as to whether or not to incarcerate the individual.  This is an unethical and dangerous misuse of the data.  Dr. Pithers nicely sums it up in the deposition mentioned previously:

... I know of no psychometric procedure of psychophysiological procedures that can be used to demonstrate with psychological certainty that a person has committed a legal offense or engaged in child sexual abuse or is likely to do so in the future.  That is the province of sorcerers and witches, not of a psychologist.  It clearly asserts that the practitioner has special powers beyond which most psychologists would assert themselves to have; and, therefore, I believe it is a highly inappropriate response and potentially one for consideration by an ethical board.

What is the research base for prediction from arousal patterns?  There is a very limited base.  Dr. Vernon Quinsey and his colleagues in 1980 reported on 30 child molesters who were released from incarceration, who had received therapy, and who were followed up for an average of 29 months.  The pre-release arousal measurement data showed a small but significant relationship with whether the child molesters were convicted of a new child offense.  However, when this study was enlarged by adding new subjects to a total of 132, and the follow up time extended to an average of 34 months, no relationship between the post treatment arousal patterns and recidivism was found.  The only predictive finding found was that the sexual arousal data taken from the initial testing of 100 treated and untreated child molesters was significantly related to recidivism.  One interpretation of these data is that arousal patterns do not always persist through time.  In sum, our research-based data is extremely limited in this area and is too preliminary to be used as a basis for any prediction in general.

Another related aspect is the percentage of arousal to deviant stimuli.  As mentioned before, some practitioners with limited training and experience state that any arousal to deviant stimuli indicates a deviancy and should be treated.  Dr. Pithers points out, what many of us in the field know, that almost every male will find some kind of arousal to a deviant stimuli.  We look for the ratio to non-deviant stimuli, as well as significant elevations.  Percentages are deceiving.  A practitioner may diagnose an individual based on data indicating less than 10% of a full erection and categorize an individual and recommend treatment.  Or a probation officer may want to know in which "deviant" categories does an offender show 20% or more of an erection in order to help formulate probation monitoring.

Dr. Richard Laws, the current president of the Association for the Treatment of Sexual Abusers, and one who has set up a number of behavioral laboratories for the assessment and treatment of sex offenders, considers anything below 20% of maximum erection to be "no arousal."  He also believes that 20% to 40% of erection is considered "low arousal," and does not justify treatment.  It is his belief that arousal from 40% to 60% suggests where judgments can be clearly made.  (His chapter outlining this use of arousal data is part of the training manual for the clinical training project of the Department of Public Safety of the state of Hawaii).

Arousal Patterns in Subjects Who Are Not Sex Offenders

Another most important reason not to make diagnoses or predictions based solely on arousal measurement data is that it appears that arousal responses to deviant stimuli are not limited to sex offenders.  As Dr. Pithers states "... there appear to be people in society who do have disordered arousal patterns who, to the best of my knowledge, have never sexually offended."

Empirical research support for this statement was done right here in our own state by Dr. Gary Farkas in his doctoral dissertation research in 1979.  As part of his research he assessed the arousal patterns of 42 male university students ranging in ages from 18 to 38, including a diverse mixture of ethnicities including Americans of Caucasian, Oriental, and Polynesian ancestry.  He found that his data from the normal college students were more comparable to patterns found for rapists' arousal than to past studies of normals.  Although he assumed that the laboratory sample had more in common with fellow college students than with convicted rapists, he observed that the subjects in his study evidenced substantial arousal to descriptions of coercive and violent sexual behavior and showed patterns much like those of rapists evaluated by similar procedures by Dr. Gene Abel and others.

Even stronger support for this position comes from the research conducted by Dr. William Farrall (designer and manufacturer of the plethysmograph used exclusively in the state of Hawaii) for his doctoral dissertation.  Dr. Farrall developed a stimulus set for assessing the arousal patterns of sex offenders using a video format with audio stories and still photographs.  Results of his assessment of sex offenders eventually indicated that he could correctly determine 87.7% of those tested as having deviant arousal patterns.  Furthermore, even the highest, or second highest arousal correctly determined the age and gender preference in 66.6% of the cases.

As with Dr. Farkas, I served on Dr. Farrall's dissertation committee and in light of my own research as well as being aware of the literature, I advised Dr. Farrall to go further and use his stimulus set for assessing "non-sex offenders" for control purposes.  He then recruited volunteers from newspaper advertisements seeking "normals."  They were selected and interviewed and asked not to volunteer if they were ever involved in sexual deviant activities such as pedophilia, incest, or rape.  They were also cautioned not to participate if they ever had fantasies or thoughts regarding sex with children or any other paraphilias.  This control group, after the initial screening, comprised 24 individuals ranging from 20-59 years old.

To Dr. Farrall's surprise, 53% of his control group produced deviant profiles.  Unfortunately, because of the anonymity of this control group, it was impossible to reach them for follow-up interviews or retesting.

Dr. Farrall then searched for a second control group.  He contacted Dr. Molinder, who is the co-author of the Multiphasic Sex Inventory (MSI) to see how they found people for their control group for the MSI.  Dr. Molinder confirmed that they had extreme difficulty as well, and said they had used people from service groups who did not meet criteria for severe character disorder on the Minnesota Multiphasic Personality Inventory (MMPI).

Dr. Farrall then set out to find his "squeaky clean" control subjects with the use of preset criteria on the MMPI and the MSI.  If any sign of deviancy, faking, or admission of deviant acts were found in the MSI, or if the MMPI indicated a severe character disorder, the subjects were excluded from his study.  Subjects were taken from service clubs in the community that consisted mainly of family men who appeared to be stable in their careers.  Even out of this group only 50% tested were free of deviant responses to the MMPI and MSI.  Dr. Farrall then extended his efforts in Utah and Grand Island, Nebraska to find other members.  He also contacted church groups, and of the 18 volunteers here, only 10 met the control group criteria.  After considerable screening he finally ended up with a control group of 18 people who stated they had no deviant fantasies or had ever been involved as an adult in any illegal sexual activities, and who passed the MMPI and MSI criteria.

Again, to his surprise, 16.6% of his "squeaky clean" control group responded to some deviant stimuli.  Dr. Gene Abel, as well as others that I have talked with, also report similar experience in assessing "normals."  In referring to his "squeaky clean" control group Dr. FarralI sums it up: "... If it is true that nearly 17% of the men in our population have deviant arousal, it should be of considerable concern and there is probably a need to address the issue nationally. It would be of considerable interest to study why these men do not offend."

Observations and Suggestions

In sum, this research indicates that it is both unethical and a serious misuse of arousal measurement data to diagnose someone, or predict someone's past or future behavior, solely on the out-of-context use of arousal measurement data.  The only valid purpose of arousal measurement data is to select appropriate targets for treatment and, later, to see whether the treatment was effective in order to assess the probability of the deviant behavior occurring in the future.

Furthermore, the use of such data by non-treatment providers who deal with sex offenders, and who make judgments concerning sentencing, parole, conditional release, and probation considerations, can also seriously misuse the data.  This, obviously, has unnecessary negative ramifications for a given individual.

In consideration of all of the above, as some of you are aware, my reports to the courts, parole, probation, treatment providers, and to attorneys will not provide detailed specific information pertaining to the raw data that has been collected during arousal measurement.  I will make a statement as to the individual's arousal responses in general, and indicate specific treatment targets.  However, I will not be reporting indices, or percentages, or idiosyncratic cues that could possibly be misused by a non-treatment provider (such as the case where a probation or parole officer mandates that an individual in the sex offender program should perform "satiation").  Releasing such raw data to a non-treatment specialist is not only a violation of the ethical code of the American Psychological Association, but a violation of Hawaii state law as well.

On the other hand, if I receive an appropriate signed release by any individual that I assess and/or treat, I would be more than happy to release the more detailed results of the psychological and psychophysiological testing to any sex offender treatment specialist who has the necessary education, specialized training, and other professional credentials required to validly interpret psychological test results and other assessment data as related to sex offenders.

Unless it has been changed recently, such a qualified sex offender treatment specialist, as specified by the Hawaii Sex Offender Treatment Program:

1. must hold a doctoral degree in psychology granted by an accredited institution of education (or hold a masters' degree and provide services under the direct supervision of a licensed psychologist);
2. must have competence in the diagnosis of psychological disorders and in psychotherapy, as demonstrated by holding a valid Hawaii license to practice psychology;
3. must have specialized competence in sex therapy and human sexuality, as demonstrated by documented education, training, and supervised clinical supervision field;
4. must have specialized competence in the behavioral and/or clinical assessment and/or treatment of paraphilias in sex offending behaviors as documented by specialized education, training, and supervised clinical experience in the field; and
5. must have at least 1000 hours of diagnostic or general psychotherapy with sex offenders or similar populations.

As I would certainly not feel qualified to recommend sentencing terms, or to dictate parole or probation-related legal requirements to an individual, I do not feel it appropriate that a parole or probation officer should dictate treatment procedures, or provide co-therapy leadership in sex offender treatment groups.  All of us have as our major goal the protection of the public.  Each of us has a fairly well-defined role and responsibility for such protection.  However, in addition, I also have the responsibility for providing hopefully effective treatment for a patient so as to prevent any future victims of sexual aggression.

As a licensed clinical psychologist and a board certified forensic psychologist, I take full responsibility for selecting the most helpful therapeutic procedures available for each of my patients, in addition to relapse prevention group therapy, and maintenance group therapy.  I do my hest to keep up with all relevant clinical and research-based procedures in this area, as well as continue to share my findings and procedures with sex offender treatment specialists throughout the United States and Canada.

While some may believe that the opinions that I have expressed here are isolated and mine alone, on the contrary. it is my belief that these views are shared by the majority of the people working in this area.  In support of this belief I will make a copy of this memorandum available to my professional colleagues, along with an invitation to telephone or write me in response to anything that I have said.

I hope that this information has been of some assistance to you, and I stand ready to provide any further information in response to any questions that any of you may have on any particular point.

* Jack S. Annon is a clinical and forensic psychologist at 10088 Bishop Street, Suite 506, Honolulu, Hawaii, 96813.  This is from a letter that Dr. Annon sent to several judges, attorneys, and treatment providers in Hawaii and elsewhere.  [Back]


[Back to Volume 5, Number 1]  [Other Articles by this Author]

Copyright 1989-2014 by the Institute for Psychological Therapies.
This website last revised on April 15, 2014.
Found a non-working link?  Please notify the Webmaster.