The Abel Assessment of Sexual Interest (AASI)is an objective measure of a person’s sexual interest in various potential sexual partners, including males and females, of preschool, grade school, adolescent, and adult ages. The test is widely used in criminal matters, children’s court and family court cases where there have been allegations of inappropriate sexual conduct. There are four separate versions of the test for adult and adolescent males and females.
While no test can establish whether or not someone either did or did not commit a specific act, the test can establish a person’s patterns of sexual interest. That information typically can be presented to a jury. This test has largely replaced the Penile Plethysmograph (PPG) in court proceedings.
In the AASI, the client views slides depicting those gender and age categories. Outside the clients’ awareness, their viewing time of each slide is also being measured. The reliability and validity of viewing time or Visual Reaction Time (VRT) has been established as a reasonably accurate measure of the client’s sexual interests (the mental component), which in turn correlate strongly with sexual arousal (the physical reaction).
In addition to viewing pictures, the client completes a questionnaire that includes over 600 items. The questionnaire surveys interest or behaviors concerning normative sexual behavior, deviant but legal behavior, and criminal sexual acts. The questionnaire measures if the client is attempting to present himself in an overly positive manner. The test results include a classification as to how similar the client is in comparison to a reference group of “guilty deniers.”
AASI AND THE COURTS: The AASI has been accepted 15 times as meeting Daubert; denied for specific flaws 4 times; excluded as not relevant once; and reversed on appeal twice because the lower court record was insufficient for a full review. I will address here the criticisms from cases that often are cited. The information provided here is from published articles, peer-reviewed presentations, and personal communication with Dr. Abel regarding results of pertinent unpublished studies.
Whitehorse 2001, denied, citing:
- Insufficient Native Americans in the samples: Was not true at the time, Abel reported about 2% of the original sample was NA. However, the data base at the time was small, limiting examination of NA versus other ethnicities. The data base now has >3,000 NA subjects. Abel reports no significant difference across ethnicity.
- False Negative rate of 25% “wouldn’t assist the jury:” A debatable point, as the Daubert criteria cite “a known error rate,” not a minimum acceptable level. Regardless, the judge was citing a 1994 study by Abel, which was in regards to the earlier Abel Screen, not the current test.
- The test did not meet general acceptance: The record cites the ATSA guidelines (2001) as listing the PPG as a psychophysiological measure to use in evaluations. That portion of the guidelines was later revised to include PPG, AASI and polygraph as appropriate measures.
- The test excluded incest cases: That exclusion was true at the time, but no longer is true. After much internal research, Abel found no significant difference between incest and non-incest protocols.
Birdsbill 2003, denied, citing:
- AASI is intended for treatment, not diagnosis: I don’t know where this misunderstanding came from, the test is widely used in evaluations to assist the courts.
- The test doesn’t indicate a defendant’s tendency to abuse children: True, and there never was a claim it did. It measures sexual interest, not future behavior.
- There was no adequate control group: This is incorrect. Studies available at the time showed a known accuracy rate in discriminating offenders from non-offenders.
- There was no adequate peer review: The test is scored by Abel Screening, not by the clinician who administers the test. There seemed to be concern over this approach and the unknown algorithms to achieve the test results. Many studies have submitted data for blind interpretation and used the results in research independent of Abel, corroborating much of his original findings. That is a version of peer review.
- The error rate is poor: It is a known error rate, similar to tests and procedures accepted by the courts.
- The scientific community does not accept the AASI for the diagnosis of pedophilia: Nor do they accept any test as solely sufficient to render such a diagnosis. Apparent sexual interest in children on the VRT, or responses on the questionnaire indicating sexual urges and attraction to children, are relevant to rendering a diagnosis.
(This same court accepted the AASI in a later case.)
Ready V. MA, 2002, 2006, denied, citing:
- Inadequate testing of the technique: That was an issue with the earlier Abel Screen, but there had been testing done by the time of this case on both the VRT and the Questionnaire portions of the AASI. “Adequate” is a judgment call, apparently insufficient for this court.
- No studies on the questionnaire: Most studies focus on the VRT portion, but that is not true in all cases. Abel presented studies on the questionnaire as early as 2001.
- No studies on test-retest reliability: True at the time. Research since then has shown acceptable reliability.
- The formula was not subject to ‘rigorous scrutiny’: At the time, there was little published, and much (not all) came from Abel. That is no longer true. Since the algorithms for scoring are unknown, one can test the ‘formula’ only by inference from the results. Independent studies have replicated Abel’s results.
- No studies ‘outside the close-knit world of ATSA’: ATSA is the largest specialty organization focusing on sex offenders. It is no surprise most of the early work came through it. Studies have now been published in diverse journals outside ATSA. As to ATSA being ‘close-knit,’ the members are willing to disagree amongst themselves, sometimes quite vociferously.
- There were no pictures in the VRT of boys the specific age of the victims: The test does not claim to measure interest in a specific age, it measures interest in children in defined age and gender groups. It would seem the Court could rule the test therefore irrelevant. That is a different issue that whether the test meets the Daubert criteria as to what the test claims to measure.
- The susceptibility of the test to manipulation was unknown: True at the time. Concern has been expressed that, if a client knows about the test in advance (easy to find on the Net), he could ‘fake’ the results. Abel has since presented studies showing that subjects cannot alter their pattern of interest, even when told exactly how the VRT works.
SUMMARY: Many of the criticisms of the AASI were leveled against an earlier instrument, the Abel Screen. Other cases examined the AASI in the first version, early in the use of the instrument. Some criticism were valid, while other criticisms appear to reflect insufficient knowledge on the part of the testifying expert, and the Court consequently was not fully informed. The test is now in the third version. Ongoing research continually adds to our knowledge of the test. In my opinion, all the criticisms leveled have been addressed, and the test does meet the Daubert criteria.