A Test of Three Different Explanations for Low Stimulus Response Discrimination in Phallometric Testing

ABSTRACT

Phallometric assessment is used to assess men’s sexual interest in children and to assist in risk assessment and treatment planning. A common response pattern, especially when the assessment is conducted in a forensic context, is an indiscriminate pattern of penile responses: No sexual stimulus seems to produce a substantially higher response than another. This indiscriminate response profile could be the result of (1) faking good (in particular, reducing the responses to child stimuli); (2) floor or ceiling effects caused by low or high arousability, or (3) non-exclusivity (the individual is similarly sexually interested in both children and adults). In this study of 2,858 adult male patients who underwent volumetric phallometric assessment for sexual interest in children between 1995 and 2011, we tested these three possible explanations. Results showed support for each of the explanations, but the variance accounted for in response discrimination was quite small when considering each explanation (separately or when considered together). We discuss avenues for future research to better discern the causes of indiscriminate responding in phallometric assessment.

Acknowledgments

Thank you to Dr James Cantor for providing access to the study data.

Disclosure Statement

No potential conflict of interest was reported by the authors.

Notes

¹ We also conducted the analyses with individuals with <1 cc of a response removed (n = 128 men) and the pattern of results remained largely unchanged.

² We also conducted the regression by entering the linear arousability term in step 1 and the quadratic term in step 2. The linear term explained 8% (95% CI .06 – .10) of the variance and the quadratic term explained an additional 5% (95% CI .11 – .15) of the variance in response discrimination.

³ We also conducted the regression by entering the linear age term in step 1 and the quadratic term in step 2. The linear term explained 4% (95% CI .03 – .05) of the variance (negative association with response discrimination) in response discrimination and the quadratic term did not explain any additional variance.

⁴ We conducted supplementary analyses where we just focused on victim age polymorphism for those with victims under 15 and 17+ (as in Michaud & Proulx, 2009) as well as a more restricted polymorphism variable of those with victims under 11 and 17+ (this categorization would ensure that those individuals who were classified as polymorphic were offending against victims that were very distinct in appearance). Less variance was accounted for in the analyses (η² = .001; 90% CI .000 – .005). These patterns were generally consistent in that those who were polymorphic had the lowest response discrimination score.