by Tania L. Abramson and Paul R. Abramson
How did psychological tests of pedophilia, and the diagnosis of Parental Alienation Syndrome (PAS), gain admissibility as expert testimony in the courtroom? The United States Supreme Court, in Daubert v. Merrell Dow Pharmaceuticals, ruled that scientific knowledge was the principle that dictates the admissibility of evidence. To now claim that psychological tests of pedophilia and PAS represent scientific knowledge makes a mockery of science itself.
There’s a long history of notorious diagnostic causation theories involving women. Nervous females were hysterics; refrigerator mothers caused autism. Vengeful mothers, under the banner of Parental Alienation Syndrome, are now believed to alienate their children from fathers who have been accused of child sexual abuse.
Are mothers more likely to be alienators in custody battles? Perpetrators of child sexual abuse are overwhelmingly male. The U.S. Justice Department estimates that approximately 30% of perpetrators of child sexual abuse are also family members. The gender disparity in protecting kids from familial sexual abusers is a byproduct of the differential probability of molesting kids. The nature of the crime itself would also dictate strong emotional reactions from the protecting parent.
Richard Gardner, a psychiatrist, coined the phrase Parental Alienation Syndrome (PAS). Based upon his impressions, Dr. Gardner concluded that spiteful mothers invented child sexual abuse claims. Those claims, Dr. Gardner believed, were designed to insure the sole custody of children in protracted custody battles. To further facilitate their objectives, vindictive mothers would then alienate their children from their fathers, hence the name of the syndrome PAS.
Dr. Gardner also believed, according to his 1992 self-published book True and False Accusations of Child Sexual Abuse, that vengeful mothers were women who were rejected by their husbands, and are now exacting retribution by inventing claims of child sexual abuse. Inventing claims of child sexual abuse may also have resulted, Dr. Gardner further noted, from women fantasizing about adult-child sex. Dr. Gardner nonetheless considered all sexual paraphilias, including pedophilia, valuable mechanisms of species survival.
It’s easy to craft a psychologically enigmatic narrative in the absence of data. Narratives that are based upon nothing more than impressions however will never amount to science. If they were purported to be science, psychology and psychiatry would crumble under the weight of cryptic storylines.
To avoid this kind of disintegration, scientists elevate impressions, hunches, and theories, by objectifying them so that predictions from the theory can then be rigorously tested. The outcome of testing is in turn used for the continued evolution of the theory. Daubert in fact promotes falsification as a cardinal element of science. Theories are elevated to science only through rigorous attempts to falsify them through testing.
Eight criteria comprise PAS. Weak, frivolous, or absurd rationalizations for the deprecation, for example. None of these criteria however have been objectively defined, nor is PAS accompanied by objective diagnostic measures. That this syndrome only arises in fierce custody battles, where experts in PAS are the only ones who can discern it accurately, is hardly reassuring as well.
One of the few empirical studies on PAS, a doctoral dissertation by Carlos A. Rueda, which was then published in The American Journal of Family Therapy, examines inter-rater concordance. Dr. Rueda is a licensed social worker with a doctorate from the online Walden University. Dr. Rueda describes his dissertation as sanctioned by Dr. Richard Gardner.
The final statement of the abstract of this paper is a curious one. The data gathered from the completed surveys was sufficiently reliable to suggest a wider study for the purpose of classification in the next edition of the DSM. As the reader will soon discover, this study was anything but that, raising questions about the propriety of this project. Was it simply to campaign for the inclusion of PAS in the DSM?
A small sample of therapists, whose degrees are never mentioned, were given 5 cases to evaluate. Were they psychiatrists, psychologists, or marriage and family therapists? In either case, it would be very useful to know. The 5 cases themselves were previewed and evaluated by a panel of experts in the field of child custody and PAS. Was Dr. Gardner one of the experts? Were family law attorneys experts too?
Most striking however was the fact that three-quarters of the therapists (44 out of 58) refused participation. If the sample was small, and not representative of the range and expertise of therapists to begin with, three-quarter attrition is so dismal that one wonders how the research was published in the first place.
Dr. Rueda also conflates inter-rater concordance with reliability, both of which have no relevance to the validity of PAS.
Presume for a moment that we have created a syndrome called Hollywood Producer Delusions (HPD) to categorize people who delude themselves into believing that they are Hollywood producers. We then create 5 sample cases of what our experts (undefined) believe are deluders, and we then ask a small number of Hollywood types (undefined) to participate in a study to see if they can agree, based upon the criteria we claim to be definitive of deluders, that our 5 cases are indeed truly deluders.
Three-quarters of our sample however refuses to participate. Nevertheless, the remaining twenty-five percent is largely in agreement. Would that finding prove that HPD is a syndrome? Or is it simply that our small group was in agreement when using our criteria? The latter is obviously the answer.
If two months later we give that twenty-five percent the five cases once again, as Dr. Rueda did as well, and asked them a second time how the cases fit our criteria, would that be a good measure of reliability? Or is it a good measure of how much our tiny sample remembered participating in our studies two months ago? Dr. Rueda nonetheless claims that a PAS diagnosis is now reliable because his truncated sample of therapists (undefined) used his expert criteria (undefined) in a similar fashion two months later when they rated the five cases again.
Reliability relates to the overall consistency of a measure. If a test indicates that a sample of children is autistic, that test is reliable if the sample is diagnosed as autistic once again in the future. It does not however prove that the test is a good measure of autism, only that the test scores are similar when repeated. Many factors can also influence reliability, including remembering a test and then responding consistently.
Complicating reliability even further is the fact that some traits vary over time, anxiety for example. Is a test unreliable if on a second assessment a subject, during summer break for example, is less anxious? Or was the subject more relaxed when tested a second time?
If one purports to have truly discovered a trait or syndrome, one obviously needs to demonstrate that it can be reliably assessed. Without more information on the reviewers, the criteria, and the 5 cases, plus a more robust sample of therapists, it’s hard to know what Dr. Rueda achieved, but it certainly wasn’t a study of the reliability of PAS.
Dr. Rueda also claims that he had conducted an initial test of content validity, though he uses this term interchangeably with reliability. Where content validity is concerned, nothing could be further from the truth.
Content validity refers to expert opinion that a measure has been designed to assess every facet of a construct. As a proxy for expert opinion, Dr. Rueda used instead the opinions of his self-selected and undefined number of experts. That Dr. Paul Fink, a past President of the American Psychiatric Association, had previously described PAS as junk science would alone make content validity of PAS impossible.
How does one objectively define and quantify absurd rationalizations for the deprecation. Even if this criterion weren’t easily reducible to a quantitative measure, one would still have to demonstrate that this criterion, in conjunction with other criteria, has predictive validity if PAS is in fact a legitimate syndrome.
To do this, one would have to first construct a valid measure of PAS (an impossible task in and of itself) and then demonstrate, for example, that the test predicts which parents, at some future date, will alienate their children from a co-parent in a bitter custody battle. Large, representative samples of newlyweds would suffice, and the predictive validity of the test could then be evaluated according to how it predicts divorcing parents behavior during custody battles. Armchair theorizing, and imprecise data may be useful in a social setting for therapists, but as soon as it is introduced into the courtroom, it must rise to the standard of scientific knowledge. The costs of misdiagnoses are extreme.
That is the real issue herein. It relates not only to PAS, but to tests of pedophilia too, both of which are often used in conjunction. PAS purports to eliminate false claims of child sexual abuse, and tests of pedophilia purport to exonerate men wrongly accused of sexually molesting a child.
If this were simply a turf battle among warring academics, an ever-present phenomenon in the academy, there would be no reason for alarm; theories and research regularly fall from grace, with no societal impact whatsoever. When theories and research are introduced in a courtroom however, it’s another matter entirely, particularly if the error rate is high.
Diagnostic value is largely determined by two factors, improving upon the base rate and the costs of errors. This is no less true of a biomedical marker than it is of a psychological one. A diagnostic test must therefore improve upon guesses that rest upon base rates alone, but more importantly diagnostic tests must also avoid catastrophic errors.
Admitting expert evidence in a court of law for a syndrome (PAS), or a psychological test of pedophilia, both of which are void of scientific credence, is an open invitation for appalling consequences for children. Pedophiles are going to gain sole custody of their kids, and they are going to be exonerated for child sexual abuse too.
This is not to say that a mother has never falsely accused a father of child sexual abuse to gain sole custody of a child. Such cases most certainly exist. Men on the other hand are more likely to hide assets in divorce proceedings, and engage in domestic violence, both of which can have devastating impacts on child welfare too. Therapists have every reason to be concerned about vicious custody battles and the psychological health of children.
It is nevertheless extremely shortsighted to use dubious diagnostic measures that never reach the threshold of scientific knowledge, and are accompanied by tragic errors, in matters involving the welfare of children. Eliminating testimony about psychological tests of pedophilia and PAS from courtroom deliberations would be a huge step in rectifying a significant problem underlying the resolution of custodial issues in divorce.
Tania L. Abramson, MFA
Dr. Paul R. Abramson
Professor of Psychology, UCLA