The Cost of Relying on Behavioural Indicators of Sexual Abuse
by Terence W. Campbell, Ph.D.

Mental health professionals too often assume they can identify sexually abused children (or adults sexually abused as children) by relying on behavioural indicators. In the December 2001 issue of the BFMS Newsletter, for example, an accused father described the risk assessment undertaken in his case.

It relied on behavioural patterns described as "classic symptoms". In respect of our daughter these included self-harm, alcohol abuse, mental instability and childhood delinquency. It concluded that if these are coupled with allegations, however made, they prove abuse.

In the same issue of the BFMS Newsletter, Katherine Mair's report of attending a Ritual Abuse Information Network and Support (RAINS) conference indicated:

Valerie Sinason even suggested that some unexplained physical problems occurring at the present time could be seen as evidence that earlier abuse had occurred. For example, gynaecological pain could indicate earlier sexual assault, and food allergies could be a reaction to earlier enforced cannibalism.

Excessive Reliance on Clinical Judgment
Determining whether behavioural patterns or symptoms are clinically significant relies on clinical judgment. Clinical judgment, however, too often amounts to little more than conjecture and speculation. It is necessary to ask, for example, at what point does some behaviour reach the threshold of self-harm? Will cigarette smoking suffice? Or should the degree of self-harm be more exaggerated? Similarly, at what point does excessive drinking amount to alcohol abuse? How does a mental health professional reliably assess mental instability? At what point do incidents of childhood misconduct become delinquency? How frequent and severe must gynaecological pain be to indicate a history of childhood sexual abuse? At what point does an unpleasant reaction to some food amount to a food allergy?

There are no well-defined criteria available allowing clinicians to answer these questions in a consistent and reliable manner. Consequently, clinicians can only resort to their clinical judgment when addressing these questions. Clinical judgment, however, frequently responds to the biasing effects of "sharpening" and "levelling". In response to their preconceived expectations, clinicians can "sharpen" - or emphasise - any evidence suggesting that someone exhibits significant self-harm, alcohol abuse, mental instability, childhood delinquency, significant gynaecological pain, or food allergies. Simultaneously, these same clinicians can "level" - or de-emphasise - any evidence suggesting that these behavioural characteristics are absent or insignificant.

Even if someone exhibits a constellation of behavioural indicators suggesting childhood sexual abuse, relying on such indicators leads to an inordinate frequency of mistaken classifications. This problem of misclassification becomes evident when considering issues of (1) base rate, (2) the logical error of affirming the consequent, and (3) levels of sensitivity and specificity.

Base Rate Issues
Relying on behavioural indicators to identify sexual abuse neglects to consider what is known as base rates. Children who have been sexually abused may exhibit a broad rate of non-specific symptoms. In fact, however, the vast majority of children who exhibit non-specific behavioural symptoms have not been sexually abused. In the terms of mathematical psychology, relying on behavioural indicators to identify sexually abused children involves using high base-rate behaviours to identify a comparatively low base rate event.

When loosely defined, non-specific behavioural symptoms such as neglecting one's health (inviting interpretation as self-harm), incidents of excessive alcohol consumption (inviting interpretation as alcohol abuse), poor judgment (inviting interpretation as mental instability) and childhood misconduct (inviting interpretation as delinquency), are high base-rate behaviours. As children or as adults, most people exhibit some non-specific symptoms such as these at some point in time.

Fortunately, sexual abuse is a comparatively low base-rate event. Depending on the definition of sexual abuse, only 10-20% of children are ever sexually abused. Therefore, relying on high base-rate behaviours (behavioural symptoms) to identify a low base-rate event (childhood sexual abuse) inevitably leads to a substantial frequency of false-positive classifications, mistakenly concluding that a child has been sexually abused when she has not.

Consider, for instance, the following example of mistakenly relying on high base-rate behaviours to identify a low base-rate event:

a. Sexually abused children regularly walk, talk and drink water.
b. This child regularly walks, talks and drinks water.
c. Therefore this child has been sexually abused.

To belabour the obvious, these assumptions result in an inordinate number of false-positive classifications.

Logical Error of Affirming the Consequent
The following is an example of the logical error of affirming the consequent.

a. All humans who are pregnant are female.
b. This human is a female.
c. This female is pregnant.

Correspondingly, then, it is also mistaken to assume the following:

a. All sexually abused children exhibit behavioural symptoms.
b. This child exhibits behavioural symptoms.
c. This child has been sexually abused.

In fact, if this person is pregnant, she must be female; but if she is female, she is not necessarily pregnant. Relatedly, if this child has been sexually abused, she may exhibit behavioural symptoms; but if a child exhibits behavioural symptoms, she has not necessarily been sexually abused.

Levels of Sensitivity and Specificity
The sensitivity of any indicator (behavioural or otherwise) refers to how accurately the indicator rules-in some condition. The specificity of any indicator (behavioural or otherwise) refers to how accurately the indicator rules-out some condition. For example, consider the following indictor for identifying prostate cancer in males: Diagnose all males 50 years and older as having prostate cancer.

Because 99% of the population with prostate cancer are males 50 years and older, the sensitivity of this indicator would be approximately 99%. The obvious flaws of this diagnostic procedure only become evident when considering the male population without prostate cancer. Because only about 10% of males even develop prostate cancer, this classification procedure will misclassify the 90% of males without prostate cancer. The sensitivity of this procedure is therefore quite good (99%), but its specificity is unacceptably low (0%). In other words, this procedure can only rule-in prostate cancer. It cannot accurately rule-out prostate cancer.

Relatedly, behavioural indicators of sexual abuse can only rule-in sexual abuse. These indicators cannot rule-out sexual abuse. As a result, behavioural indicators of sexual abuse are systematically biased because of their rule-in emphasis.

Conclusions
Relying on behavioural indicators to identify a history of childhood sexual abuse defies fundamental considerations of logic and common sense. Such reliance inevitably leads to an inordinate frequency of mistaken classifications. In particular, these indicators result in a high frequency of false-positive classifications, i.e., mistakenly concluding childhood sexual abuse has occurred when, in fact, it has not.

About the author: Dr. Terence W. Campbell has served on the Professional and Scientific Advisory Board of the US False Memory Syndrome Foundation since 1993. Much of Dr. Campbell's work has been directed at the legal aspects of child abuse accusations. As a forensic expert in sexual abuse litigation, he has raised the issue of malpractice in psychotherapy. In 1992 the American Psychological Society made Dr. Campbell a fellow, in recognition of his "distinguished contribution to scientifically-oriented psychology.

His first book - Beware the Talking Cure: Psychotherapy May Be Hazardous to Your Mental Health - was published by SIRS Publishing in 1994. His second book - Smoke and Mirrors: The Devastating Effect of False Sexual Abuse Claims - was published in Insight Books in 1998, and is now distributed by Perseus Books. Dr. Campbell's third book - Cross-Examining Experts in the Behavioral Sciences, co-authored with D.J. Lorandos, Ph.D., J.D. was published by the West Group in December 2001.