THE DISCRIMINABILITY OF RAPISTS FROM NON-SEX OFFENDERS USING PHALLOMETRIC MEASURES - A METAANALYSIS

The authors examined how well identified rapists could be discriminated from non-sex offenders using phallometric assessments, what variables might moderate this discrimination, and whether rapists respond more to descriptions of rape than to consenting sex. Eleven primary and five secondary phallometric studies involving 415 rapists and 192 non-sex offenders were examined using meta-analytic techniques. Study effect sizes averaged 0.82 (95% confidence interval 0. 16 to 1.49). Only stimulus set was a statistically significant moderator of effect size: Stimulus sets that contained more graphic rape descriptions produced better discrimination between rapists and non-sex offenders. There was a trend for stimulus sets that contained more exemplars of rape descriptions to achieve better discrimination. Also, rapists responded more to rape than to consenting sex cues in 9 of the 16 data sets and in all 8 of those using the more effective stimulus sets.