Psychophysiologic Assessment of Erectile Response and its Suppression as a Function of Stimulus Media and Previous Experience With Plethysmography.

Most clinicians specializing in the treatment of sex offenders would agree that an accurate assessment of an offender's pattern of sexual arousal (e.g., to deviant vs. nondeviant sexual stimuli) can be important to successful treatment. Despite its importance, such assessment is often difficult to obtain owing to the frequent efforts of many offenders to deny or minimize their deviant response patterns (Marshall, 1994).

Since the 1960s, penile plethysmography (PPG) has been used to provide a relatively objective measure of sexual arousal. Most commonly, PPG is used to evaluate a sex offender's relative sexual arousal, through treatment, to a variety of interactions (e.g., sex with children, rape of an adult, consensual adult sex), assessed via audio-taped descriptions of such activities. Used in this manner, PPG can help the clinician working with sex offenders to (a) recognize deviant arousal patterns and (b) evaluate the success of treatment aimed at the modification of these patterns. However, the evidence supporting the validity of the technique has been mixed (e.g., Simon & Schouten, 1991). This is particularly true when assessing convicted sex offenders who remain "nonadmitters" (Freund, Chan, & Coulthard, 1979). Further, while PPG has been found to have specificity as high as 95%, its sensitivity has been assessed as only near 50% (Freund & Blanchard, 1989).

One likely reason for PPG's mixed validity findings is the susceptibility of the technique to some degree of voluntary control (Freund, 1961, 1963, 1967; Henson & Rubin, 1971; Lalumiere & Earls, 1992; Laws & Holmen, 1978; Mahoney & Strassberg, 1991; McNaulty & Adams, 1991; Quinsey & Bergensen, 1976; Quinsey & Carrigan, 1978; Quinsey & Chaplin, 1988). Further, this body of literature makes clear that, in general, men are far more successful at voluntarily suppressing sexual arousal to a preferred stimulus than they are at generating arousal to a neutral or nonpreferred stimulus (e.g., Mahoney & Strassberg, 1991).

Several variables have been suggested as having a potentially important impact on a man's ability to dissimulate his sexual arousal while being assessed via PPG. One such variable is experience with this assessment procedure. Freund, Watson, and Rienzo (1988) demonstrated that those with greater experience with penile plethysmography were more effective at misrepresenting their pattern of arousal. However, it appears that no attempt was made in this study to ensure that participants were actually attending to the erotic stimulus materials.

In clinical usage, a potential experience effect is an important consideration. Sex offenders are typically assessed plethysmographically several times during the course of their treatment, often including feedback on the results of each testing session. If men are better able to suppress their arousal as a result of experience with plethysmography, this effect should be an important consideration in interpreting assessment results later in an individual's treatment.

Another variable that may mediate men's ability to suppress their arousal is the nature of the erotic stimulus. It has been demonstrated that, for most individuals, videotape stimuli produce higher levels of sexual arousal than do audiotapes (Abel, Barlow, Blanchard, & Mavissakalian, 1975; Julien & Over, 1988; Sakheim, Barlow, Beck, & Abrahmson, 1985). This stimulus effect could make it relatively difficult to suppress sexual arousal to video material. Further, Card and Farrall (1990) found that, independent of initial levels of sexual arousal, sex offenders were less able to suppress their erections when presented with sexually explicit videotapes than with sexually explicit audiotapes. This effect may be the result of the greater ecological validity of the video material (i.e., it coming closer to a real-life sexual experience; Davis & Bauserman, 1993).

Card and Farrall (1990) also informally observed that, as in lie detection, changes in galvanic skin response (GSR) seemed to be associated with attempts to fake during PPG assessment. Another measure used in lie detection is finger pulse amplitude (FPA), measured by a photoplethysmograph. Podlesney and Raskin (1978) found FPA to discriminantly decrease when participants attempted to deceive the experimenters. The value of FPA in identifying men's attempts to fake during penile plethysmography has yet to be evaluated in published research.

The Present Study

The clinical importance of PPG requires that we have the best information possible regarding the procedural possibilities for minimizing dissimulation and for detecting attempts to fake when using the procedure. The present study was designed to address these issues.

The following hypotheses were tested: When men's sexual arousal is assessed plethysmographically, they would be more successful at suppressing their arousal when (a) they were experienced with the assessment procedure and (b) the explicit stimulus material is audio tapes instead of video tapes with an audio component. It was also hypothesized that (c) using novel stimuli with experienced participants (as opposed to stimuli with which they were familiar) would reduce the experience effect. Finally, we expected that (d) attempts at suppression would be detectable through measurement of galvanic skin response (GSR) and finger pulse amplitude (FPA), independent of stimulus modality, previous experience, or novelty of the stimuli.

METHOD

Participants

Research participants were male college students recruited from an introductory psychology course subject pool, receiving extra credit for their participation. They completed a brief initial questionnaire attesting that they met the following inclusion criteria: (a) between 19 and 40 years of age, (b) taking no medications that might interfere with sexual arousal, (c) not suffering from medical conditions that might interfere with arousal, (d) having no previous experience with PPG, and (e) believing that they were likely to become sexually aroused by depictions of sexual activity between a man and a woman. No further information about participants was collected in order to help assure them of maximal anonymity. Therefore, demographic descriptive statistics are unknown and cannot be reported.

A total of 42 participants signed up for the study and arrived at the laboratory. Six of these failed to achieve detectable arousal during the initial test video, and were excluded. In addition, data from one participant were discarded due to a procedural error in data collection. Therefore, data from a total of 35 participants were included in the analyses.

Experimental Design

The study was a mixed factorial 2 x 2 x 2 (within-subjects) x 2 (between-subjects) design. The within-subjects factors were stimulus type (audio vs. audiovisual), attempted suppression of arousal, and experience with the procedure. The between-subjects factor was whether, during the last (i.e., 3rd) experimental trial (when they were experienced with the procedure), participants were exposed to the same stimulus to which they were previously exposed or to a novel stimulus. Participants were randomly assigned to the between-subjects conditions, while order of stimulus presentation was preassigned and counterbalanced.

Stimulus Materials

Four (two audio and two audiovisual) 5-min tape segments were constructed. The two audiotape segments described consensual heterosexual activity between strangers. The two audiovisual segments were chosen from commercially-produced full-length sexually explicit videos. The content and context of the audiovisual segments was similar to those of the audio segments.

The audio segments were recorded onto videotape, with a blank visual field showing on the screen, in order to standardize the presentation of all segments. All segments were carefully selected to present no violence or overt degradation, and only portrayed contact between one man and one woman.

Apparatus

A two-room arrangement was used, with participants seated in a private, locked room with no windows or observation devices to the experimenter's room. All physiological data were collected via a multi-channel computerized polygraph system (CPS). Penile plethysmographic (PPG) data were gathered using a mercury-in-rubber strain gauge placed halfway down the shaft of the penis, data from which were fed into the CPS. Skin conductance was recorded from Beckman 16mm Biopotential Ag-AgC1 electrodes. The sensors were taped to the palmar surface of the index and middle fingers of the participant's nondominant hand. A photoplethysmograph pickup strapped with Velcro over the tip of the palmar surface of the ring finger of the participant's nondominant hand collected finger pulse data.

A computer joystick box, specially constructed to allow only left-right movement, was used to encourage attendance to the stimuli. The box had a sign taped to it indicating what directional movement would signify. Data from the joystick were displayed on the experimenter's monitor. The experimenter watched the monitor for joystick movement, but these data were not collected for statistical analysis.

A television monitor was placed across from the participant's chair to afford him full view of the screen and adequate perception of the audio. A videocassette recorder placed in the experimenter's room provided the signal for the television monitor.

Procedure

Overview. The experimental procedure, depicted in Table 1, involved a baseline assessment followed by three trials, with each trial consisting of two segments. In each segment, the participant was presented with both an audio and an audiovisual stimulus period (in counterbalanced order across participants). Participants were instructed that, during the first half of each period, they were to respond "normally," while during the second half, they were to attempt to suppress their plethysmographically-assessed arousal.

Table 1. Depiction of the Experimental Design

Participant                  Trial 1
                  Segment 1          Segment 2

   P1          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S
   P2          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S
   P3          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S
   P4          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S

Participant                Trial 2
                  Segment 1           Segment 2

   P1          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S
   P2          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S
   P3          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S
   P4          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S

Participant                  Trial 3
                 Segment 1            Segment 2

   P1          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S
   P2          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S
   P3          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S
   P4          Audio   Video       Audio   Video
               N   S   N   S       N   S   N   S

Note. The four rows, P1 to P4, represent the four possible sequences to which any given participant may be randomly assigned. The cell entries N and S refer to experimental instructions either to respond Normally or to Suppress arousal. The italicized cells under Trial 3 refer to conditions in which participants received novel stimuli.

Procedural details. As each participant arrived at the laboratory, he was seated in a comfortable chair in a private room and given an informed-consent form explaining the procedure and the placement of the gauges. The experimenter then left the room, leaving the participant in complete privacy to remove his pants and underpants, place the strain gauge, and cover his lap with a clean hospital gown. When the participant signaled his readiness through the intercom, the experimenter returned, placed the GSR and FPA sensors, then left.

In his adjacent room, the experimenter adjusted the CPS software to provide optimal readout, then played a short (approximately 3 min) sexually explicit baseline videotape through the participant's monitor. This tape graphically displayed sexual activity between a man and a woman. The experimenter watched the PPG data trace during this presentation for signs of the participant's erectile response. If less than approximately 5% of estimated full erectile response was detected, and no equipment malfunction was apparent, the procedure was terminated. The participant was given full experimental credit and assured that, due to the nature of the laboratory environment, failure to become significantly aroused was common and did not signify a problem for the individual. Six potential participants were excluded in this way.

Participants allowed to continue were contacted over the intercom and asked to place the joystick box under his dominant hand, on the arm of his chair. He was reminded of the proper use of the joystick: tilting it to the left if the man depicted in the tape seemed to be receiving more pleasure than the woman, and to the right if the woman seemed to be receiving more pleasure. He was instructed to make these evaluations continuously.

Following the baseline assessment, Trial 1 began. The experimenter started the first experimental tape, consisting of either an exclusively audio component or video with audio. Approximately 150 seconds into this tape, the experimenter sent a tone through the intercom. Each participant had been previously instructed that, when he heard the tone, he was to try to make himself unaroused any way he could without touching his penis or the equipment attached to it. The tape was allowed to play to its end (for another approximately 150 seconds), after which the experimenter watched the PPG trace for detumescence. Throughout the study, the same criteria was used for "sufficient detumescence": At least 3 minutes had elapsed and the participant evidenced less than 5% full arousal.

After sufficient detumescence, the second segment of Trial 1 began. A new tape was started. If the previous tape was audio, this tape was audiovisual, and vice versa. Again, the participant had been instructed to respond normally until signaled, after which he was to again attempt to suppress his sexual arousal.

At the conclusion of this second segment, the experimenter examined the data charts for both segments of Trial 1 in order to derive an estimate of the participant's success at reducing his arousal. This estimate was an approximate percentage of reduction observed, from the previous maximum arousal achieved. This estimate was then shared with the participant via the intercom.

After sufficient detumescence, Trial 2 began. In this trial, the same procedure, using the same tapes presented in the same order as in Trial 1, was repeated in its entirety. This part of the procedure was designed only to provide a participant two more opportunities to practice arousal suppression and receive feedback on his success: Data from this trial were not analyzed.

After both segments of Trial 2, the procedure varied depending on whether or not the participant was assigned to the novel condition. If not, the same tapes were again presented in the same order as before; otherwise, similar but novel tapes were presented in the same order. That is, if the participant was previously exposed to an audio tape first in Trials 1 and 2, then in Trial 3, a novel audio tape was presented first, followed by a novel audiovisual tape. Since every participant had repeated the procedure (i.e., two tapes) and been given feedback twice prior to this trial, Trial 3 represented the experienced trial for all participants.

After Trial 3, the participant was asked to remove the equipment and get dressed. The experimenter then returned to the participant's room to administer a short debriefing and questionnaire, answer any questions, and then thank and dismiss him.

Data Reduction

The first 20 seconds of FPA and GSR data were discarded from each period (i.e., normal responding and suppression), as an orienting response was expected in reaction to the participant perceiving the start of the tape and the intercom tone. Data collected during the remaining 90 seconds of each period during Trials 1 and 3 were analyzed. As mentioned above, Trial 2 was considered a practice run, so data from this trial were not analyzed. Skin conductance (GSR) responses were scored as follows. First, the magnitude of each spike was calculated relative to the low point prior to the occurrence of the spike. For each period during Trials 1 and 3, a mean GSR response value was determined. Finger photoplethysmography (FPA) responses were scored by taking the amplitude of each pulse peak and computing the area under the curve.

For penile plethysmography (PPG) data, the first 80 seconds were discarded, and only the next 30 seconds of data during each period were analyzed. This procedure allowed each participant sufficient time to achieve arousal during the first part of the test period, when he was allowing himself to respond normally, and sufficient time to reduce his arousal during the second part of the test period when he was attempting to suppress. Again, only data from Trials 1 and 3 were subjected to analysis. The unit of measurement was the area under the curve (Abel, Blanchard, Murphy, Becker, & Djenderedjian, 1981) associated with erectile response during the 30-second arousal measurement period.

While GSR data appeared to meet requirements for the assumption of normal distribution, both PPG and FPA data were excessively skewed. Consequently, each of these data sets was subjected to a square-root transformation.

For each of the measures (PPG, GSR, and FPA), data from Trial 1 represented the inexperienced condition, while data from Trial 3 represented the experienced condition.

RESULTS

An analyses of variance (ANOVA) was performed on each of the three physiological measures, using suppression (no suppression attempt vs. attempts to suppress), stimulus type (audio vs. audiovisual), and experience (inexperienced vs. experienced, i.e., Trial 1 vs. Trial 3) as within-subjects factors, and novelty (familiar vs. novel stimuli in Trial 3) as a between-subjects factor. Since PPG responses were the primary data of interest, results for this measure are presented first.

Penile Plethysmograph

As expected, a strong and significant main effect for suppression was found, F(1,33) = 24.07, p [is less than] .001. As a group, participants were successful in suppressing their arousal (M = 1324 vs. M = 821). A significant main effect was also found for stimulus type, F(1,33) = 5.21, p [is less than] .05; participants showed more arousal overall to audiovisual stimuli than to audio-only stimuli. However, this stimulus effect was complicated by a significant stimulus x experience interaction, F(1,33) = 4.22, p [is less than] .05.

The four means of this interaction were compared using a Tukey procedure. Results revealed that, when in the experienced condition (i.e., Trial 3), participants were significantly less aroused to the audio stimuli (M = 809) than they were to the audiovisual stimuli (M = 1222). Further, their arousal during the audio-experienced condition (M = 809) was also significantly lower than it was to the audio (M = 1108) or the audiovisual (M = 1121) stimuli when participants were inexperienced (i.e., Trial 1). None of the other means were significantly different from each other. Figure 1 graphically depicts these findings. Further analysis demonstrated that, as revealed in Figure 2, the significantly (p [is less than] .05) greater arousal to video than audiovisual material with experienced participants (Trial 3) held in the suppress as well as the non-suppress condition.

[Figures 1-2 ILLUSTRATION OMITTED]

Contrary to expectation, no significant main effect or interaction involving arousal were found for novelty of the stimuli. As presented in Table 2, participants were not differentially able to suppress their arousal when exposed to novel versus familiar stimuli during Trial 3.

Table 2. Means and Standard Deviations for Penile Plethysmograph

                                      Inexperienced    Experienced
      Condition                        M      SD        M      SD

Not in novel stimulus
 condition (N = 18)
  Video
    Not suppressing                  1515    1085     1416     869
    Attempting suppression            742     507      853     722
  Audio
    Not suppressing                  1251     904      994     503
    Attempting suppression            929     629      591     444
In novel stimulus
 condition (N = 17)
   Video
     Not suppressing                 1256     976     1626    1411
     Attempting suppression           973     559     1004     542
   Audio
     Not suppressing                 1423    1035     1119     726
     Attempting suppression           831     542      658     384

Note. Figures represent the square root transformed area under the curve.

Galvanic Skin Response (GSR)

A significant main effect for stimulus type was found for GSR, F(1,33) = 4.43, p [is less than] .05; participants' mean GSR response amplitudes were greater during audio presentations (M = 1733) than during audiovisual presentations (M = 1542). Contrary to expectation, no other significant main effect or interaction was found.

Finger Pulse Amplitude (FPA)

The same ANOVA as for GSR was performed on FPA data. A significant main effect was found for experience, F(1,33) = 10.15, p [is less than] .01. Participants evidenced greater FPA response during their experienced run (M = 1664) than during their initial run (M = 1337). A significant main effect was also found for novelty of the stimuli, F(1,33) = 7.30, p [is less than] .05; participants exposed to familiar stimuli during Trial 3 had a greater FPA response (M = 1741) than those exposed to novel stimuli (M = 1246). No other significant main effect or interaction was found. There was, however, a nonsignificant trend observed for suppression, F(1,33) = 2.63, p = .11. Participants demonstrated a somewhat greater FPA response when attempting to suppress (M = 1558) than when they were not attempting to suppress (M = 1443).

Debriefing Questionnaire

The debriefing questionnaire, administered after all physiological testing was completed, asked about participants' experience during the study. Of primary interest was a question asking for a brief description of any techniques used while attempting to suppress sexual arousal. Unfortunately, over half the participants left this item blank. Of those responding to this item, virtually all described attempting to suppress using a cognitive technique. Such techniques included thinking of repugnant imagery while still attending to the tape, or focusing on some real or imagined undesirable quality of the actors.

DISCUSSION

Arousal as a Function of Media Type

It was hypothesized that men would be less sexually aroused to audio materials than to audiovisual materials. The results indicated a more complex picture than expected; an interaction between media type and experience with the procedure. When participants were inexperienced (Trial 1), only a nonsignificant difference was observed between their arousal to audio versus audiovisual stimuli. After experience (Trial 3), however, while their PPG-assessed arousal to audiovisual stimuli remained about the same, their arousal to the audio-only stimuli was significantly and substantially reduced (see Table 2). It is likely that the audiovisual stimulus, with its sights as well as sounds, remained more compelling (and therefore resulted in less satiation) by the third exposure.

Why our participants failed to demonstrate a significant effect for media type (e.g., Julien & Over, 1988) under the inexperienced condition is not immediately apparent. Perhaps the specific stimuli chosen was responsible (e.g., unusually highly arousing audio stimuli). The particular participants may have played a role. The sociopolitical climate from which our participants were drawn is rather conservative. It is likely that many participants had not had previous exposure to highly sexually explicit material (audio or video tapes). As a consequence, initial exposure to such material in this study may have had an unusually highly arousing effect, even under the audio-only condition.

Suppression of Sexual Arousal

As expected, and consistent with previous research (e.g., Mahoney & Strassberg, 1991), men in this study were, on average, successful at suppressing their PPG-assessed sexual arousal. This effect held, even though participants (a) were required to perform a semantic tracking task designed to encourage their attending to the sexually explicit stimuli, and (b) were offered no explicit reward for being able to suppress.

This finding has sobering clinical implications, as the PPG is commonly used to assess the sexual arousal shown by sex offenders and suspected offenders. Men who are able to significantly suppress their arousal to some (e.g., deviant) stimuli may be able to present an unrealistically normal picture of their sexual response pattern. This would be true whether one's arousal pattern was being assessed using a measure based on absolute or relative (i.e., a deviancy index) scores.

It is also interesting that, at least according to their self-reports, participants relied exclusively on psychological (i.e., cognitive) rather than physical techniques to suppress arousal. None mentioned using muscle contractions such as the "pumping" described by Freund et al. (1988). To the extent that these reports are valid, it suggests that techniques designed to detect deception via changes in muscle tension (e.g., Card & Farrall, 1990) may be of limited value. Further study into this issue is clearly warranted. It would also be of value to explore the relationship between thought content and suppression effectiveness. Identifying the most effective specific thoughts in arousal suppression could be of value in (a) better understanding the role of cognitive mechanisms in sexual dysfunction and (b) the development of better cognitive strategies for use in the treatment of compulsive sexual behavior.

Contrary to our predictions, men's ability to suppress PPG-assessed sexual arousal did not vary significantly as a function of stimulus type (audio vs. audiovisual) or experience with the procedure. However, their overall success at suppression, in combination with their already reduced sexual arousal when experienced with PPG and being exposed to audio tapes, resulted in a particularly low level of arousal in this condition (see Table 2). The clinical implications of this are explored later in this section.

Stimulus Novelty

The hypothesized effects of novelty of stimulus materials on sexual arousal or ability to suppress did not occur. On their third trial, participants demonstrated no greater arousal when presented with a novel stimulus (i.e., a new audio or audiovisual tape) than to a tape to which they had previously been exposed twice. They were also no less able to suppress their arousal in response to these novel stimuli than to the familiar ones.

These findings may have been a function of the particular stimulus materials used. It was decided to standardize the stimuli in the study, so they could be counterbalanced in order of presentation. Therefore, all stimuli presented similar activities (i.e., fellatio, cunnilingus, vaginal intercourse) between one man and one woman in a similar context. This standardization of the materials may have counteracted the intended novelty manipulation. That is, participants may have perceived the scenarios as so similar as to not be really novel, after twice previously being exposed to comparable stimuli.

GSR and FPA

Previous research had suggested that both GSR and FPA are effective in assessing general deception (e.g., Podlesney & Raskin, 1978) and might be useful in detecting attempts at misrepresentation during penile plethysmography (Card & Farrell, 1990). Neither measure, however, was significantly associated with attempts at arousal suppression in this study. Still, it may be premature to conclude that these indices are of little value in detecting deception during plethysmography.

In general lie detection, the person who is attempting to falsify his/her responses is also trying to avoid being detected in this deception. Such was not the case in the present investigation (i.e., there was nothing for participants to be caught at). In real life settings in which plethysmography occurs, however, fear of being detected at misrepresentation may render measures such as GSR or FPA more effective than they were here. Clearly, more research is needed to assess their value in this regard. The few significant effects associated with GSR and FPA were not predicted, nor do they lend themselves to obvious interpretation.

Limitations of the Present Study

There are several features of this study that limit the generalizability of our findings. The most obvious of these is the nature of our participants. Most were young (early 20s), all were college students, and many may never have been exposed to highly sexually explicit material. The extent to which their responses to our experimental manipulations can be generalized to the sex offender population is an open question.

Another limit to generalizability concerns the stimulus materials used. While we found some differences and some nondifferences associated with media type (audio vs. audiovisual), the extent to which these findings are a function of the specific audio and audiovisual stimuli used is unknown. Other examples of such media might produce a different pattern of results.

Finally, the experience manipulation in this study was rather different from the way in which experience is typically obtained in clinical assessments of sex offenders. Specifically, in clinical practice many stimulus segments are presented in a given session, and feedback, when provided, is rarely immediate. Further, months often elapse before the next assessment session. It is unclear how these procedural differences might impact the generalizability of our findings. Certainly, these results should be considered tentative until replications of this study specifically addressing these methodological limits can be undertaken.

Conclusions

The results of this study again demonstrate the need for effective measures to (a) reduce men's ability to suppress their arousal during plethysmography and (b) detect attempts at such suppression. While many of the predicted effects associated with stimulus type and experience were not evidenced here, our results still have potentially important implications for those relying on this assessment technique.

Many clinicians utilizing penile plethysmography rely on an arbitrary, a priori cutoff score (e.g., 10%, 15%, or 20% of estimated full arousal) as an index of significant sexual arousal, with scores below this level interpreted as clinically (as opposed to statistically) nonsignificant. Even when not attempting to suppress, experienced participants in our study responded with significantly less arousal to audio than to audiovisual stimuli. This already low arousal level dropped even further when the men were actively suppressing.

The resulting level of plethysmographically-assessed sexual arousal, under the audio-only/experienced/suppression condition, could, in many cases, fall below such predetermined cutoffs. This, in turn, could lead to the erroneous conclusion that the individual was not sexually interested in the target being portrayed (i.e., a false-negative error).

There are often legal and other obstacles to using highly explicit videos in assessing sex offenders. However, our findings argue strongly in favor of their use, versus audio-only stimuli, when assessing men plethysmographically. This seems particularly true when these men are experienced with the procedure and are motivated to misrepresent their pattern of sexual arousal.

REFERENCES

Abel, G. G., Barlow, D. H., Blanchard, E. B., & Mavissakalian, M. (1975). Measurement of sexual arousal in male homosexuals: The effects of instructions and stimulus modality. Archives of Sexual Behavior, 4, 623-629.

Abel, G. G., Blanchard, E. B., Murphy, W. D., Becker, J. V., & Djenderedjian, A. (1981). Two methods of measuring penile response. Behavior Therapy, 12, 320-328.

Card, R. D., & Farrall, B. S. (1990). Detecting faked penile responses to erotic stimuli: A comparison of stimuli conditions and response measures. Annals of Sex Research, 3, 381-396.

Davis, C. M., & Bauserman, R. (1993). Exposure to sexually explicit materials: An attitude change perspective. Annual Review of Sex Research, 4, 121-209.

Freund, K. (1961). A laboratory differential diagnosis of homo- and heterosexuality -- an experiment with faking. Review of Medicine, 7, 20-31.

Freund, K. (1963). A laboratory method for diagnosing homo- or hetero-erotic interest in the male. Behavior Research And Therapy, 1, 85-93.

Freund, K. (1967). Diagnosing homo- or heterosexuality and erotic age preference by means of a physiological test. Behavior Research And Therapy, 5, 209-228.

Freund, K., & Blanchard, R. (1989). Phallometric diagnosis of pedophilia. Journal of Consulting and Clinical Psychology, 57, 100-105.

Freund, K., Chan, S., & Coulthard, R. (1979). Phallometric diagnosis with "Nonadmitters." Behavior Research and Therapy, 17, 451-457.

Freund, K., Watson, R., & Rienzo, D. (1988). Signs of feigning in the phallometric test. Behavior Research and Therapy, 26, 105-112.

Henson, D. E., & Rubin, H. B. (1971). Voluntary control of eroticism. Journal of Applied Behavioral Analysis, 4, 37-44.

Julien, E., & Over, R. (1988). Male sexual arousal across five modes of erotic stimulation. Archives of Sexual Behavior, 17, 131-143.

Lalumiere, M. L., & Earls, C. M. (1992). Voluntary control of penile responses as a function of stimulus duration and instructions. Behavioral Assessment, 14, 121-132.

Laws, D. R., & Holmen, M. L. (1978). Sexual response faking by pedophiles. Criminal Justice and Behavior, 5, 343-356.

Mahoney, J. M., & Strassberg, D. S. (1991). Voluntary control of male sexual arousal. Archives of Sexual Behavior, 20, 1-16.

Marshall, W. L. (1994). Treatment effects on denial and minimization in incarcerated sex offenders. Behavior Research and Therapy, 32, 559-564.

McNaulty, R. D., & Adams, H. E. (1991). Voluntary control of penile tumescence: Effects of an incentive and a signal detection task. The Journal of Sex Research, 28, 557-577.

Podlesney, J. A., & Raskin, D. C. (1978). Effectiveness of techniques and physiological measures in the detection of deception. Psychophysiology, 15, 344-359.

Quinsey, V. L., & Bergensen, S. G. (1976). Instructional control of penile circumference in assessment of sexual preference. Behavior Therapy, 1, 213-221.

Quinsey, V. L., & Carrigan, W. F. (1978). Penile response to visual stimuli: Instructional control with and without auditory sexual fantasy correlates. Criminal Justice and Behavior, 5, 333-341.

Quinsey, V. L., & Chaplin, T. C. (1988). Penile responses of child molesters and normals to descriptions of encounters with children involving sex and violence. Journal of Interpersonal Violence, 3, 259-274.

Sakheim, D. K., Barlow, D. H., Beck, J. G., & Abrahamson, D. J. (1985). A comparison of male heterosexual and male homosexual patterns of sexual arousal. The Journal of Sex Research, 21, 183-198.

Simon, W., & Schouten, P. (1991). Plethysmography in the assessment and treatment of sexual deviance: An overview. Archives of Sexual Behavior, 20, 75-91.

Manuscript accepted September 27, 1999

Jeffrey A. Golde, Donald S. Strassberg, and Charles M. Turner University of Utah

The authors are indebted to John Kircher and Claire Jensen for the invaluable assistance in the conduct of this study, and to John Wincze and other reviewers for comments on an earlier draft of this paper.

Address correspondence to Donald S. Strassberg, Ph.D., Department of Psychology, 390 S. 1530 E., University of Utah, Salt Lake City, Utah 84112; email: [e-mail].