1 2 3 4 Chapter Seven 5 6 7 Phallometric Assessment of 8 9 Sexual Arousal 10 11 12 Hannah L. Merdian 13 Department of Psychology, The University of Waikato 14 and 15 16 David T. Jones 17 Te Piriti Special Treatment Unit, Auckland 18 19 20 21 22 23 Introduction 24 25 The penile plethysmograph is a device which measures male sexual arousal by 26 means of a transducer around the subjects’ penis while various stimuli are 27 presented. Although this seems to be a simple enough premise, half a century 28 of research has failed to demonstrate that the assessment is either reliable or valid. 29 It does, however, seem to be a significant predictor of risk. 30 This chapter reviews the literature around the reliability and validity of 31 phallometric assessment, the limitations to its use, its contribution to risk 32 assessment, and alternatives to its use, and concludes with remarks on the future 33 of the phallometry. 34 35 36 The penile plethysmograph and its procedures 37 38 Originally developed by Kurt Freund to assess sexual orientation in men, the 39 penile plethysmograph (PPG) was later adapted to assess deviant sexual arousal 40 in male offender populations by Vernon Quinsey (Marshall, 1996). The prin- 41 42 International Perspectives on the Assessment and Treatment of Sexual Offenders: Theory, Practice 43 and Research, Edited by Douglas P. Boer, Reinhard Eher, Michael H. Miner, Friedemann Pfa¨fflin, 44 and Leam A. Craig. Ó 2011 John Wiley & Sons Ltd. Published 2011 by John Wiley & Sons, Ltd. 142 Assessment and Treatment of Sexual Offenders 1 ciple behind the instrument could not be much simpler. One attaches a device to 2 the penis of a subject, and measures what happens to it when the person is 3 exposed to a variety of possibly arousing stimuli, either visual or auditory. In a 4 typical phallometric assessment, the subject is seated privately in a comfortable 5 chair where they can attend to visual stimuli and auditory stimuli. Assuming that 6 penile arousal indicates sexual interest, a person’s arousal pattern in response to 7 various stimuli can be measured from the gauge around his penis. Often, 8 nonintrusive physiological measures such as galvanic skin response (GSR), 9 respiration and pulse rate are monitored in an attempt to detect suppression 10 or deliberate increases of arousal. 11 Freund’s initial device was based on a volumetric measure; an airtight glass 12 cylinder is fitted around the subject’s penis and the volume of air displaced in the 13 chamber is used as a measure of penile changes (Kalmus & Beech, 2005). While 14 sensitive and accurate, this technique has some drawbacks. Chief among them is 15 the fact that volumetric devices must be fitted by the technician, which is highly 16 unpalatable to many assessors. Circumferential gauges, on the other hand, as first 17 used by Fisher, Gross, and Zuch (1965), can be fitted by the client himself. There 18 are two types of circumferential gauges, both of which measure changes to the 19 diameter of the penis, usually about halfway up the shaft. Barlow gauges are thin 20 metal strips curved into an open circle, while rubber strain gauges are thin rubber 21 loops filled with mercury or indium-gallium. Both are commonly used in 22 correctional settings. Withthesetypes,changesin thecircumferenceofasubject’s 23 penis can be measured from changes in the electrical resistance of the gauge. 24 Overall, volumetric devices are superior to circumferential gauges, as they can 25 register changes in both length and diameter (Marshall, 2006). As noted by 26 Kalmus and Beech (2005), the initial stages of arousal in some men may result in 27 no change to, or even a decrease in, circumference in some men. (To understand 28 this, one might imagine filling the finger of a rubber glove with water; the end 29 may fill first, contracting the middle before the pressure balances and the middle 30 expands.) Kuban, Barbaree, and Blanchard (1999) compared the two gauge 31 types and found them equivalent for high responders, but volumetric gauges 32 appeared to be superior for low responders with a maximum penile increase 33 below 10% full erection. Nonetheless, circumferential PPGs are more wide- 34 spread due to their easier application and commercial availability, and it is 35 unlikely that anyone would use volumetric devices in a widespread correctional 36 application. 37 There continues to be a great deal of controversy about the use of phallometry 38 in correctional assessments. As Marshall and Fernandez (2000) point out, the 39 main problem is the lack of a sound empirical basis. Although the Association for 40 the Treatment of Sexual Abusers (ATSA) recommends that the use of phallo- 41 metric assessment should be used only to confirm a client’s self report of sexual 42 preferences (Howes, 2003), many treatment programs use phallometric assess- 43 ment to detect deviant sexual interests, determine treatment needs, and inform 44 risk assessments (Marshall, 1996; Marshall & Fernandez, 2003). It is also used Phallometric Assessment of Sexual Arousal 143 1 for behavioral treatment, either as a measure of success or for direct feedback to 2 the client in techniques such as covert desensitization (Adler, 1994), for 3 determining treatment progress (Blanchette, 1996) and for confronting an 4 offender’s denial of deviant arousal (Kercber, 1993). 5 Although early researchers were enthusiastic about the value of phallometry as 6 a fairly objective measure of male sexual arousal (see Marshall & 7 Fernandez, 2003b; Zuckerman, 1971), there is controversy about what exactly 8 the PPG assesses. Of course, few would argue the fact that sexual arousal in men 9 leads to swelling of the penis as a consequence of increased blood flow into the 10 genital area. However, as Singer (1984) points out, sexual arousal is a trichotomy 11 of an aesthetic feeling, an approach reaction, and a genital response. While the 12 penile plethysmograph seems an obvious measure for the latter, it says nothing 13 about the first two qualities. Gaither (2000) notes that the PPG only measures 14 one form of sexual arousal while sexual preference is a more holistic construct. 15 Whereas some studies have demonstrated that men’s subjective reports of their 16 sexual excitement correlate well with physiological measures, this was not true 17 for low levels of genital response (Singer, 1984). On the other hand, high 18 correlations have been demonstrated between phallometrically assessed and self- 19 reported sexual orientation in control populations (Lee-Evans, Graham, Har- 20 bison, McAllister & Quinn, 1975; Quackenbush, 1996) and more deviant 21 populations (Haywood, Grossman, & Cavanaugh, 1990), but both controls 22 and offenders reported subjective arousal that was not phallometrically indicated 23 and vice versa in this latter study. 24 It is also questionable whether physical arousal as measured by the PPG is a 25 sufficient measure to draw conclusions about behavior. Sexual offences might be 26 motivated by nonsexual reasons (Marshall & Fernandez, 2003) or some in- 27 dividuals might experience sexual arousal to deviant stimuli but would never act 28 on it. Also, even if phallometry is an accurate measure of arousal, it is not known 29 whether sexual preferences are an enduring trait that should be detectable in a 30 laboratory setting, or whether they are influenced by environmental factors to 31 the extent that the assessment situation would preclude accurate assessment 32 (Marshall & Fernandez, 2003a). 33 Controversies aside, it nevertheless seems unreasonable to dispute a link 34 between sexual arousal to deviant stimuli and inappropriate sexual behavior. As 35 Byrne (2001) concluded, despite the theoretical uncertainties regarding phallo- 36 metric assessment, sexual arousal is a large part of the genesis of sexual offending, 37 and the PPG is a useful measure of this arousal. 38 39 40 Psychometric properties 41 42 Aside from the fundamental concerns noted earlier, there are many issues 43 concerning the psychometric properties of the penile plethysmograph, including 44 a wide variety of methodological and demographic variables considered below. 144 Assessment and Treatment of Sexual Offenders 1 Unfortunately, many studies lack detailed descriptions of these factors, which 2 further aggravates a comprehensive review (Marshall, 2006). 3 4 5 Unstandardized assessments 6 7 Among other variations, phallometric research has used different stimulus 8 materials, different stimulus modalities, different presentation orders and times, 9 different gauges, and different hardware. Even relatively minor issues such as the 10 experimental instructions given can create considerable variability in the out- 11 comes. However, despite many attempts, none of these factors have been 12 standardized. 13 14 15 Stimulus variables 16 Obviously, if one is going to measure arousal that occurs in response to sexual 17 stimuli, the choice of stimulus materials is likely to have significant effects on the 18 results. Not surprisingly, there is significant variation among types of stimuli 19 used in the literature, roughly paralleling the development of the technology 20 used to create and present them. Earlier studies tended to use audiotapes, written 21 text or instructions to fantasize, or slides for visual materials; later studies mainly 22 use videotapes and current practice prefers audio and visual stimuli presented 23 from digital files on the recording computer. Visual materials may differ in 24 brightness, color, number of depicted persons, presence or absence of back- 25 ground, and in erotic or neutral content. They may also be either still visuals or 26 live video, although the latter is rarely used. Audio materials will vary in the voice 27 and dialect used, the nature of sexual activities described and the degree of 28 explicit description. As most phallometric assessments are intended to identify 29 subjects’ age and gender preferences, variability within the age categories 30 presented may have serious implications.