The Reliability and Validity of Gardner's Indicators of Pedophilia

Terence W. Campbell^*

ABSTRACT: Despite the professional respect Richard Gardner deservedly enjoys, his "Indicators of Pedophilia" are grossly unreliable as a result of their ill-defined ambiguity.  Because of their inherent unreliability, Gardner's indicators rely excessively on clinical judgment.  In view of the extent to which clinical judgment invites arbitrary conjecture, self-styled "validators" — who rarely encounter allegations of sexual abuse they do not uncritically endorse — can easily exploit the ambiguity of Gardner's indicators to serve their own biased agenda.  The readiness with which Gardner's indicators provoke unreliable speculation necessitates persistent challenges directed at disqualifying them as a basis for expert testimony.

In the midst of the growing hysteria related to allegations of child sexual abuse, Richard Gardner stands out as a courageous figure wisely advising careful deliberation in situations that too often provoke reckless abandon.  His 1987 book, The Parental Alienation Syndrome and the Differentiation Between Fabricated and Genuine Child Sex Abuse (), perceptively identified allegations of child sexual abuse as minefields which mental health professionals were obligated to navigate with skeptical caution.

In particular, Alienation Syndrome prompted considerable attention for Gardner's "Sex-Abuse Legitimacy Scale" (SAL Scale).  His 1992 book, True and False Accusations of Child Sex Abuse (), further outlined developments and refinements related to the SAL scale leading to his definition of the "Indicators of Pedophilia."  Despite Gardner's well-deserved reputation as a paragon of reason confronting powerful but frequently irrational forces, this article contends that his "Indicators of Pedophilia" create more problems than they solve.

To his credit, Gardner acknowledged the potential shortcomings related to his SAL Scale and the "Indicators" which evolved from it. In his introduction to True and False Allegations, he explained:

I recognized that the SAL Scale was a first offering and that, ideally, before publishing such an instrument, one should work 15 or 20 years on its development and pull together the findings of many groups of investigators.  On the other hand, there was an urgency in that people were being jailed then and we could not wait until "all the data was [sic] in" before formulating an objective scale (1992, p. xxxiv).

Unfortunately, however, Gardner's commendable sense of social responsibility appears to have motivated the development of a seriously flawed evaluative procedure.

For cases of sexual abuse allegations directed at adult males, Gardner has developed 24 criteria that presumably discriminate between pedophiles and non-pedophiles.  He indicates that mental health professionals evaluating allegations of sexual abuse should determine whether each of the criteria are — or are not — satisfied for a given suspect.  In turn, Gardner advocates that clinicians summarize their findings in terms of how many separate criteria supposedly indicating pedophilia a particular suspect satisfies.  Basically, then, Gardner proposes his indicators as discriminative indices for differentiating between pedophiles and non-pedophiles in a population of adult males suspected of child sexual abuse.
  

Considerations of Reliability and Validity

The utility of any discriminative index ultimately rests upon acceptable levels of its reliability and validity (Anastasi, 1982).  The reliability of any index involves the extent to which evaluators obtain consistent results when using it.  The validity of an index refers to the accuracy with which it can discriminate between a particular class of events compared to another class (e.g.) is this an instance of a legitimate allegation of sexual abuse, or is this allegation false?).

The accuracy of any discriminative index inevitably reflects, among other factors, the reliability with which it can be applied to particular cases.  For example, if we find it difficult to determine whether some two-legged creature walks like a duck, and if we feel uncertain whether it really looks like a duck, and if we also wonder whether it sounds like a duck, the accuracy (or validity) with which we classify this creature as a duck — or not a duck — will be seriously compromised.  Similarly, then, if Gardner's indicators cannot be applied in a reliable manner to cases of alleged sexual abuse, their corresponding validity levels will be unacceptable.

The most important reliability issue related to Gardner's indicators involves considerations of inter-rater reliability.  Inter-rater reliability specifies the extent to which two or more evaluators apply an index in the same manner to the same case.  High levels of inter-rater reliability indicate that evaluators use an index in a consistent manner.  Conversely, low levels of inter-rater reliability indicate that evaluators use an index haphazardly; and this outcome reflects random errors committed by the evaluators.  Indices that resort to vague and ill-defined terms (e.g., "levels of felt experience") increase random error because of their ambiguity, and they are unreliable as a result.

The clarity and precision of an index reduces random error and increases its inter-rater reliability.  Well-defined decision rules applied to an index increase its reliability by virtue of their clarity and precision.  For example, assume that we wish to predict marital failure (as defined by divorce) and choose neuroticism as a predictive index.  The reliability of this index depends on the clarity and precision with which we define neuroticism.  Consequently, we could develop a decision-making rule and specify neuroticism as defined by any MMPI-2 profile with T-scores on scale 7 — and any one of scales 1, 2, or 3 — exceeding 65.  Whether this definition is valid is entirely another matter, but it would nonetheless result in perfect inter-rater reliability when defining neuroticism.

Inter-rater reliability is necessary but not, in and of itself, sufficient for establishing the validity of an index.  Therefore, unreliable indices can never validly predict another variable; but reliable indices may also prove invalid.  Unfortunately, the extent to which most of Gardner's indicators are grossly unreliable because of their ill-defined ambiguity will become clearly evident throughout the remainder of this article.  Though Gardner insists that his indicators are valuable when used "properly," in fact, their inherent unreliability precludes any such use.  Moreover, the few indicators that qualify as reliable still appear less than valid except for some isolated examples.
  

Indicators of Pedophilia in Adult Males

1) History of Family Influences Conducive to the Development of Significant Psychopathology

Gardner contends that family of origin pathology increases the probability of pedophilia in males.  As examples of familial pathology, he cites conditions such as a history of violence, alcoholism, drug abuse, psychopathy, serious psychiatric disturbance, and suicide.  With the exception of suicide, which is obviously an event usually identified with considerable accuracy, all of the remaining conditions resist reliable definition to the degree that they invite distortion and bias.

For example, this index raises issues of how does one define family violence, and is all familial violence essentially identical in its dynamics and effects?  Extreme circumstances in which someone is hospitalized as a result of another family member's assaultiveness obviously qualify as significant family violence.  On the other hand, how do we define a brief shove that terminates as rapidly as it started, or an angrily thrown but poorly aimed newspaper, or a pointed finger that violates one's sense of personal boundaries while stopping short of physical contact especially when episodes such as these occur only very infrequently?  These kinds of hypothetical situations practically defy reliable classification according to whether they are conducive to the "development of significant psychopathology."

Additionally, the time has come for mental health professionals to carefully reexamine their prevailing assumptions related to the impact of parental alcoholism on children.  Because the majority of alcoholics cease their excessive drinking by the age of 40 (Peele, 1989), most children exposed to the alcoholism of a parent also observe that the parent overcame his or her drinking problem.  Related research indicates that adolescents of recovered alcoholic fathers rate the atmosphere of their family as quite similar to that of other adolescents whose parents never exhibited a history of alcoholism (Jacob, Krahn, & Leonard, 1991).  In fact, Jacob et al. emphasized, "... the environmental impact of alcoholism is of a current and dynamic nature, rather than reflecting fixed and irreversible effects on the child's level of functioning" (p. 176).  Obviously, then, the relationship between family of origin alcohol abuse and an adult's subsequent psychopathology remains quite ill-defined (Logue, Sher, & Frensch, 1992), and these same considerations likely apply to parental drug abuse.

As with circumstances of family violence and parental alcoholism, considerations of psychopathy create the same problem — how do we reliably define this condition?  A cynic might be tempted to suggest that anyone whose parent sold used cars grew up in a psychopathic atmosphere.  One could also argue, however, that such cynics are merely indulging their own biases.  Similarly, how do we define a serious psychiatric disturbance"?  Does this classification necessitate a reaction of psychotic proportions, or would neurotic and/or personality disorders suffice?  Given the notoriously poor reliability of too many formal diagnoses (Ziskin & Faust, 1988), this factor also invites massive distortion.

Gardner's use of this index assumes that family of origin psychopathology validly discriminates between the population of males who are pedophiles and those who are not.  As he acknowledges himself, however, his assumption remains mere conjecture without the support of validation data.  Even more importantly, the inherent unreliability of this index practically precludes ever establishing its validity via empirical research.  Without well-specified decision rules for defining family violence, familial alcoholism, psychopathy, and serious psychiatric disorder, the ill-defined ambiguity of these terms guarantees their inconsistent application in practice.
  

2) Longstanding History of Emotional Deprivation

Gardner indicates that a history of emotional deprivation involving rejection, humiliation, and other betrayals characterize the formative experiences of male pedophiles.  Nevertheless, this index suffers from the same problems as the index of family of origin psychopathology — how do we reliably define a "longstanding history of emotional deprivation"?  Assessing such a history inevitably relies on retrospective data which are notoriously unreliable (Gerlsma, Emmelkamp, & Arrindell, 1990; Green & Hall, 1984).  Mood-congruent memory effects (Matt, Vazquez, & Campbell, 1992; Singer & Salovey, 1988) in the here-and-now can also profoundly distort any individual's recall of past events.  For example, depression increases the probability of remembering one's parents as rejecting and relying on negative controls; but this outcome reflects the contemporary influence of mood more than accurate recall of the past (Lewinsohn & Rosenbaum, 1987).

This index also invites distortion because of the eagerness with which mental health professionals leap to pejorative conclusions about the families of their clients.  When therapists described various members of their clients' families in the articles of a professional journal, for example, more than 90% of the descriptions were negative and less than 10% were positive (Campbell, 1992a).  The family members were portrayed as intellectually dull, critical and intrusive, and cold and withholding to cite but a few examples.  Especially alarming is the fact that the therapists arrived at these conclusions without ever having seen the people whose character they irresponsibly assassinated.

Too often, then, mental health professionals assume the worst about the families of their clients.  As a result, they question clients in such a manner that they eventually find "evidence" which they interpret as verifying their original assumptions (Arkes & Harkness, 1981).  Given these considerations, Gardner's "Longstanding History of Emotional Deprivation" is so subject to bias that it likely cannot be applied reliably either across different cases by the same evaluator, or to the same case by two or more evaluators.  Therefore, the prospects for this index validly discriminating between pedophiles and non-pedophiles are less than promising.
  

3) Intellectual Impairment

Gardner cites examples and rationales for both impaired intelligence, and high intelligence, predicting pedophilia.  He then indicates, "My own guess is that — in 20th-century Western society — pedophiles, on the average, are less likely to be of high intelligence; rather they are more likely to be of average or below-average intelligence" (1992, p.53).  To belabor the obvious, the significance of this issue is so important that it deserves considerably more than guessing.  Unfortunately, Gardner's thinking related to this question amounts to speculative nonsense.

Formally stated as an hypothesis, Gardner suggests that the incidence of pedophilia increases in the population of males who exhibit below-average intelligence.  Despite the reliability with which intelligence can be defined, empirically testing the validity of this hypothesis would be more difficult than we might otherwise think.

This hypothesis assumes that known pedophiles are a randomly selected sub-set of the population corresponding in most respects — including intellectual efficiency — to all pedophiles.  Nevertheless, pedophiles who manage to avoid detection may do so because of their superior guile and cleverness.  Consequently, attempting to assess the intelligence levels of known pedophiles to extrapolate that data to all pedophiles is ill-advised.  Indeed, we can only guess whether using an index of intellectual impairment as Gardner recommends would result in greater numbers of false-positive or false-negative classifications when attempting to identify pedophiles.
  

4) Childhood History of Sex Abuse

Gardner unequivocally states, "Pedophiles are more likely to have been sexually abused in childhood than those who do not exhibit such behavior" (p.53).  Considerations of both social learning theory and analytic formulations predicting the identification of victims with their aggressors make this an appealing hypothesis.  Despite its intuitive appeal, however, the limited data that appear to support this hypothesis are less than convincing.

As with the indices considered previously, this index also suffers from the limited reliability with which it can be defined.  Specifying exactly what constitutes sexual abuse is not an easy task (Campbell, 1992b).  Moreover, this index implicitly assumes that sexual abuse is a homogeneous experience leading to essentially similar outcomes for all male children who endure it.  Obviously, however, sexual abuse involves very heterogeneous experiences — ranging from one episode of brief fondling to repeated episodes of sadistic penetration — and equally heterogeneous outcomes.  At the very least, the reliability of this index necessitates that it specify the severity of prior sexual abuse.

We must also remember that when self-admitted pedophiles undergo treatment, therapy can persuade them that they were sexually abused as children when — in fact — they were not (Campbell, 1992b).  In circumstances such as these, the pedophile responds to his therapist's influence and invents a history of his own sexual abuse via imagination; while simultaneously, he remains convinced he has uncovered repressed memories.  Additionally, the willingness of pedophiles to consciously fabricate a history of sexual abuse warrants consideration (Peters, 1989).  Any pedophile who reports such a history portrays himself as a victim deserving sympathy instead of the scorn usually associated with perpetrators.  Therefore, given the ambiguities frequently surrounding self-reported histories of sexual abuse, this index also offers little promise for validly discriminating between pedophiles and non-pedophiles.
  

5) Longstanding History of Very Strong Sexual Urges

Gardner contends that pedophiles often exhibit a history of intense sexual urges dating back to early childhood.  This particular index, however, is another example of a definitional nightmare.  It raises questions such as: (1) How intense must one's sexual urges be to qualify as "very strong"? and (2) At what point in time should those urges have manifested themselves in a suspect's developmental history to qualify as "long-standing"?  The obviously ill-defined characteristics of this index severely limit its reliability; and in turn, the chances of ever empirically establishing it as a valid index are remote at best.
  

6) Impulsivity

Gardner indicates that pedophiles frequently exhibit impulsive life-styles, and moreover, such impulsiveness allows them to disregard the inner controls that might otherwise inhibit their sexually abusing children.  This is an interesting theoretical formulation, but it ultimately rests upon anecdotal evidence culled from case studies.  Gardner's thinking related to this index also borders on the circular.  He reasons that pedophilia is an impulsive act; therefore, pedophiles manifest impulsiveness as a general personality trait.  As a result, Gardner advocates that evaluators look for evidence of a suspect's impulsiveness unrelated to any pedophiliac urges.

Gardner's position presumes a stable consistency of behavior across a variety of life situations that is unsupported by the relevant evidence.  Instead, the accumulating data persuasively demonstrate that individuals often exhibit very different traits in different situations (Mischel, 1973; Ryff, 1987; Wilkenson & O'Connor, 1982).  Consequently, pedophiles may demonstrate impulsiveness as they sexually abuse children but exhibit forethought and planning in other situations not associated with their pedophilia.

As has also been previously pointed out, pedophiles who escape detection may not necessarily correspond in traits and characteristics to the population of known pedophiles.  Any discriminative value of this index additionally suffers from the problem of how do we reliably define impulsiveness?  Without well-defined examples of impulsiveness serving as operational definitions, various evaluators could come to very different conclusions regarding the impulsiveness of any particular individual suspected of pedophilia.
  

7) Feelings of Inadequacy and Compensatory Narcissism

Gardner indicates that pedophiles develop narcissistic traits to compensate for their underlying feelings of inadequacy.  In turn, this narcissism supposedly predisposes pedophiles to seek out the unconditional, idealized love that only a child can express.  Theoretically elegant as this hypothesis may seem, it also involves serious reliability problems.  Specifically, this index poses difficulties related to how do we reliably define "feelings of inadequacy" and "compensatory narcissism"?  Without adequate definitions for these phenomena, different evaluators could come to very different conclusions about the felt inadequacy and corresponding narcissism of the same subject.  These considerations profoundly undermine the accuracy with which this index could ever discriminate between pedophiles and non-pedophiles.
  

8) Coercive-Dominating Behavior

Gardner contends that pedophiles are often anti-social individuals whose corresponding aggressiveness leads them into assaultive behaviors.  He attributes this antisocial aggressiveness to the underlying needs of pedophiles to dominate and control others via either overt physical intimidation or covert psychological manipulativeness.  This index involves multiple categories of behavior (antisocial, aggressiveness, overt and covert domination) which are so poorly defined that they defy reliable classification.  In discussing this index, Gardner provides no decision-making rules (e.g., assaultiveness is defined as two or more arrests, or one or more convictions, or an assault related charge) to determine whether a subject satisfies it.  As a result, applying this index to actual cases of suspected pedophilia can only result in a wide range of unreliable opinion between evaluators.
  

9) Passivity and Impaired Self-Assertion

Gardner also indicates that pedophiles are sometimes characterized by passivity and a related inability to assert themselves.  This index raises questions of what kinds of overt behaviors reflect passivity and impaired self-assertion and even more importantly, how does an evaluator accurately identify those behaviors?  Gardner neglects to answer these questions and consequently, this index invites conjecture and speculation which can only result in grossly unreliable assessments.
  

10) History of Substance Abuse

Gardner emphasizes that alcohol and/or substance abuse often undermine the already impaired behavior controls of pedophiles.  In discussing this index, Gardner cites data supporting his assumptions related to pedophilia and substance abuse (Haugaard & Repucci, 1988; Peters, 1976; Wakefield & Underwager, 1988).  Nevertheless, this index also presents problems involving how to reliably define substance abuse.  Again, Gardner proposes no decision-making rules to increase the reliability of this index (e.g., any individual arrested for two or more DWI offenses, or arrested for one or more drunk and disorderly charges, is defined as exhibiting a history of substance abuse).

Moreover, there are no standardized psychometric instruments that reliably identify a history of substance abuse and/or addiction proneness.  Despite its undeserved reputation to the contrary, for example, the MacAndrew alcoholism scale of the MMPI misclassifies 85% of the population it designates as "alcoholic."  Only 15% of the subjects identified by the MacAndrew as alcoholic are actually affected (Gottesman & Prescott, 1989).  The unavailability of well-defined procedures for accurately identifying a history of substance abuse reduces evaluators to arbitrary and unreliable judgments when using this index.
  

11) Poor Judgment

Gardner merely speculates that pedophiles exhibit poor judgment.  Beyond his intuitive impressions, he offers no examples or decision-making guidelines for assessing this index.  Consequently, two or more evaluators examining the same subject could reach markedly different conclusions regarding his judgmental appropriateness.  Obviously, then, any potential value of this index is lost by virtue of its gross unreliability.
  

12) Impaired Sexual Interest in Age-Appropriate Women

Gardner outlines how pedophiles feel too inadequate to pursue heterosexual relationships with females their own age.  Again, however, this index is not defined well enough to allow its reliable application in practice.  It raises questions of exactly how do we define "impaired sexual interest"?  Extreme cases related to this index (e.g., a 30-year-old male who has never married and rarely dated) pose few classification problems.  On the other hand, how do we apply this index to the case of a 45-year-old male, divorced for 10 years, who occasionally enjoys the companionship of women, but seems to prefer the company of his male hunting and fishing friends?  Given the greater frequency with which evaluators encounter these gray area cases, this index is less reliable than we might think when considering more extreme cases.
  

13) Presence of Other Sexual Deviations

Gardner indicates that pedophilia more often than not co-occurs with other sexual deviations.  He lists examples such as voyeurism, exhibitionism, frotteurism, sadomasochism, rape, and other paraphilias as occurring more frequently in the population of pedophiles compared to non-pedophiles.  Unlike the other indices previously discussed, this particular index is more conducive to reliable definition.  As a result, this index deserves research to empirically test its discriminative power.  Given the availability of relevant data, it might be possible to develop an actuarial formula specifying which paraphilias, and/or in what combination, significantly increase the likelihood of pedophilia.  Without validation data to support it, however, this index only amounts to interesting speculation.
  

14) Psychosis

Gardner suggests that the impaired reality contact of psychotics increases the likelihood of their engaging in pedophiliac acts.  Recently published data related to predicting the dangerousness of a population suffering from a major mental disorder (Monahan, 1992) lends some credibility to Gardner's hypothesis.  In view of the moderately improved reliability for defining psychotic conditions afforded by DSM-III-R criteria, it is at least possible to empirically test Gardner's hypothesis in the future.  Nevertheless, his assumptions regarding the relationship between psychosis and pedophilia remain merely speculative without the availability of validation data.
  

15) Immaturity and/or Regression

Gardner points out how pedophiles prefer relating to children because of the pedophiles' fixation at, or regression to, juvenile stages of development.  He cites examples of fixated or regressed immaturity as including bed wetting, avoidance of day-to-day responsibilities, disregard for the feelings of others, and low frustration tolerance.  Some of these examples, such as bed wetting, lend themselves well to reliable definition.  Other examples, however, could solicit biased responses reflecting the personal values of an evaluator more than anything else.

For instance, I consider anyone who throws a golf club out of frustration as exhibiting juvenile immaturity.  Nevertheless, my reaction is likely biased by the resentments I quietly endured as an adolescent caddy obligated to retrieve the clubs of numerous temperamental golfers.  Without well-defined, behaviorally-based definitions of exactly what qualifies as immaturity and regression, this index invites arbitrary classifications reflecting the random error of idiosyncratic values.
  

16) Large Collection of Child Pornographic Materials

This index lends itself fairly well to reliable definition.  Nevertheless, it raises questions of how extensive must a pornographic collection be to qualify as "large."  Additionally, there are also problems related to exactly what qualifies as pornography.  In one case known to this writer, a child protection professional — and the police officer with whom she worked — concluded that a divorced father accused of sexually abusing his daughter satisfied this very indicator.  They arrived at this conclusion because of the many pictures the father had taken of his daughter in her bathing suit at a beach during her visitation with him.  Consequently, the potential for applying this index in a grossly distorted manner cannot be ignored.
  

17) Career Choice That Brings Him In Contact with Children

Gardner indicates that some pedophiles — though certainly not all — make career choices that maximize their opportunities for interacting with children.  This particular index is obviously more conducive to reliable definition than the majority of Gardner's other indices.  Nevertheless, its potential for false-positive classifications related to pedophilia is distressing to say the least.

Simply consider the fact that any male who works with children endures a greater risk of encountering false allegations of sexual abuse.  Without appropriate data to support the discriminant validity of this index, the probability of it resulting in an unacceptable frequency of false-positive classifications is alarmingly high.  For example, this index casts suspicion on Gardner himself by virtue of his status as a child psychiatrist.
  

18) Recent Rejection by a Female Peer or Dysfunctional Heterosexual Relationship

Gardner contends that adult female rejection provokes pedophilia.  He also indicates that, "Men with no pedophilic tendencies will not resort to such behavior, no matter how many rejections they suffer" (1992, p.63).  Attributing any discriminative power to this index defies fundamental considerations of logic.  If many males experience rejection from adult females and never sexually abuse children, how can this index accurately identify pedophiles?  Even if we overlook the reliability problems associated with defining "recent rejection," and "dysfunctional heterosexual relationship," the illogic of this index warrants disregarding it.
  

19) Unconvincing Denial

Gardner insists that known pedophiles often resort to feeble and/or obviously fabricated denials.  In contrast, he portrays suspects falsely accused of pedophilia as expressing more convincing denials by virtue of their feelings of impotent rage.  To belabor the obvious, this index involves massive reliability problems related to how to reliably discriminate between "weak and/or obviously feigned denials" and denials that are "convincing."  This is another instance of Gardner neglecting to provide any decision-making rules for clarifying the ambiguity of vague concepts such as "unconvincing denial."

More alarmingly, evaluators who attempt to use this index are acting as if they possess the discriminative power of polygraphs.  Such an assumption, however, identifies anyone who indulges in it as presumptuously pompous.  The related research literature clearly demonstrates that mental health professionals perform quite poorly when they attempt to discriminate between honesty and deceit (Ekman & O'Sullivan, 1991).  Consequently, if polygraph data are inadmissible evidence in courts of law, this index warrants emphatic repudiation by all legal forums.
  

20) Use of Rationalizations and Cognitive Distortions That Justify Pedophilia

Gardner describes how known pedophiles often rationalize their behavior seeking to minimize any sense of responsibility for their actions.  Accordingly, pedophiles may claim, "She enjoyed it, so what's wrong with it" (p.63), or "a child who does not physically resist really wants to have sex" (p.64).  This index does not qualify as an indicator of pedophilia; instead, it conclusively confirms pedophilia when a suspect satisfies it.  Suspects who seek to rationalize their pedophilia have, by definition, acknowledged that they are pedophiles.  Basically, then, this index merely identifies the cognitive strategies used by known pedophiles to rationalize and minimize the consequences of their abusive behavior.
  

21) Resistance to Taking a lie Detector Test

Gardner contends that actual pedophiles generally refuse to undergo polygraph evaluation.  He also indicates that falsely accused individuals "are often (but not always) eager" (p.64) to take a lie detector examination.  Gardner's observations underscore the enormous flaws inherent in this index.  When falsely accused suspects decline polygraph tests, this index misclassifies them as false-positives.  In view of the voluminous literature demonstrating that polygraph errors lead to more false-positive outcomes than false-negatives (Lykken, 1979; Saxe, Dougherty, & Cross, 1985), any suspect who refuses a lie detector evaluation should be able to do so without enduring the prejudice of an evaluator.
  

22) Lack of Cooperation in the Evaluative Examination

Gardner insists that bona fide pedophiles frequently respond to evaluations in an uncooperative and obstructionistic manner intending to circumvent efforts at examining them.  In comparison, he describes falsely accused suspects as welcoming evaluations by qualified professionals.  Once again, this index suffers from the reliability problems attendant to defining uncooperative, obstructionistic, and circumventing behaviors.  Without well-defined decision-making rules specifying how to apply this index, it can readily provoke unsubstantiated conjecture.

In another case known to this writer, for example, an evaluator concluded that a suspect satisfied this index because he had failed to keep his scheduled appointment for an evaluation.  In fact, however, the suspect thought that this evaluator's role only involved reviewing the voluminous documents related to his case.  Once the suspect's misunderstanding was clarified, he promptly appeared for his next appointment.  Nevertheless, the evaluator persisted in his opinion that the suspect had been both uncooperative and obstructionistic, citing this very index to support his conclusion.  Any suspect who encounters an evaluator this biased is likely well-advised to protect himself via uncooperative and obstructionistic actions.

Indeed, attorneys representing suspects they genuinely regard as falsely accused of child sexual abuse could legitimately object to their clients undergoing an evaluation by any professional who relies on Gardner's indicators.  Given the serious shortcomings of these indicators outlined herein, one could argue that the only sane response to an evaluator who uses them is to flee from that evaluator as rapidly as possible.  Otherwise, suspects who voluntarily subject themselves to such evaluations risk satisfying Gardner's indicator for poor judgment.
  

23) Duplicity Unrelated to the Sex-Abuse Denial and psychopathic Tendencies

Gardner indicates that pedophiles generally exhibit a long-term history of deception that extends beyond their pedophilia.  Nevertheless, he does not specify what kinds of deceptions pedophiles supposedly demonstrate and as a result, his assertion creates more questions than answers.  For example, are pedophiles especially inclined to distort their educational backgrounds, or are they more predisposed to deceit regarding their employment histories?  Unfortunately, Gardner offers no answers for questions such as these; and this index remains vague and ambiguous as a consequence.

Gardner also cites Haugaard and Repucci (1988) as concluding that pedophiles frequently exhibit elevated "psychopathic deviant" [sic] scores on the MMPI.  Gardner seems to suggest that the Psychopathic Deviate scale of the MMPI discriminates between pedophiles and non-pedophiles.  In fact, however, the relevant data do not support his impressions.  In their lengthy review of expert testimony in child sexual abuse litigation, Myers et al. (1989) emphasized:

Numerous methods of psychological assessment are used to evaluate sex offenders.  It is important to emphasize, however, that there is no psychological test or combination of tests than can determine whether a person has engaged or will engage in deviant sexual activity.  Psychological tests and instruments are useful aids to diagnosis and treatment, but they cannot be used to determine whether an act occurred (p.133).

Additionally, Myers et al. specifically pointed out that the MMPI research related to pedophiles is predominantly characterized by negative or inconsistent results.

Without well-defined decision rules specifying what qualifies as "duplicity" and "psychopathic tendencies," the reliability of this index can only flounder upon the conjecture and speculation it invites.  As a result, two or more evaluators could reach very different conclusions regarding the duplicity and/or psychopathic tendencies of the same suspect.
  

24) Excessively Moralistic Attitudes

Gardner contends that some pedophiles embrace moralistic attitudes with such dogmatic intensity that they readily censure any and all who do not endorse their fundamentalistic conceptions of right and wrong.  When pedophiles exhibit these characteristics, Gardner attributes their behavior to reaction formation condemning others for the socially unacceptable impulses which they deny but chronically struggle with themselves.  This is another interesting theoretical speculation, however, it also succumbs to enormous reliability problems.

Specifically, this index raises questions of how to accurately define "excessively moralistic attitudes"?  Without any decision-making rules for applying this index, we can only wonder if all members of the moral majority, or perhaps tree hugging environmental zealots, qualify as excessively moralistic?  This index encourages evaluators, who are so inclined, to indulge their values at the expense of others whose thinking they disdain.

In addition to the massive reliability problems associated with this index, evaluators who use it should also prepare themselves to defend the psychodynamic assumptions upon which reaction formation is premised.  Given the legions of critics who have effectively dismantled psychodynamic thinking, defending it is indeed a daunting task.  As long ago as 1966, for example, Marmor characterized analytic theory as provoking unbridled conjecture and wild speculation from its disciples.  In the 1970 edition of their authoritative text on personality theory, Hall and Lindzey outlined the glaring deficiencies of analytic theory emphasizing: "... the question then arises as to why psychoanalytic theory is taken seriously by anybody and why it was not relegated to oblivion long ago?" (p.71).  In his well-respected 1982 book, Narrative Truth and Historical Truth: Meaning and/or Interpretation in Psychoanalysis (), Spence demonstrated how analytic therapists regularly distort data to conform with their theoretical convictions.

In view of how vulnerable analytic theory is to a variety of devastating indictments, well-prepared attorneys could make cross-examination a memorable experience for any mental health professional who uses this index.  Consequently, evaluators who employ this index in a forensic setting risk losing their credibility as experts, and simultaneously, they also resort to a concept whose ill-defined ambiguity assures its unreliability.
  

Discussion

Anyone inclined to defend Gardner's indicators might argue that this review has focused excessively on the specific features of each index and overlooked the incremental validity of all the indices when combined together as a total composite.  Any set of discriminative indices achieves incremental validity when each index satisfied increases the chances of accurate classification.  Accordingly, one might argue that the number of Gardner's indicators satisfied — more than the particulars of which indicator — increase the likelihood of actual pedophilia.  This contention, however, rapidly falls apart when carefully examined.

The incremental validity of any set of indicators requires that each indicator index stand independent of the other indices.  Unless the condition of index independence is met, satisfying one index can spuriously lead to satisfying other related indices (Wolfner, Faust, & Dawes, 1993).  As a result, simply summing all the indices satisfied for an indicator set can result in very misleading outcomes.

For example, consider the following ten indices from Gardner's indicators: family of origin pathology, intellectual impairment, strong sexual urges, impulsivity, coercive-dominating behavior, history of substance abuse, poor judgment, immaturity and/or regression, lack of cooperation, and psychopathic tendencies.  Any psychologist who has ever worked in a correctional setting immediately recognizes all of these indices as corresponding to an elevated 4-9 profile of the MMPI.  In other words, all of these indices are inter-correlated with each other; consequently, satisfying any one of the ten indices listed above significantly increases the probability of satisfying the other nine.  In turn, an evaluator who overlooks the relationships between these ten indices can report each of them as satisfied and attribute significance to that finding — but, in fact, such findings merely amount to an artifact attributable instead to the non-independence of Gardner's indicators.

Without clearly specified decision-making rules, or well-defined examples, to enhance the reliability of Gardner's indicators, their application relies excessively on clinical judgment.  Unfortunately, the substantial research literature evaluating the accuracy of clinical judgment reports consistently discouraging results (Ziskin & Faust, 1988).  For example, mental health professionals often question clients in a manner that biases the information they obtain (Robins, 1985; Rosenthal, 1966).  Assumptions about a client's family of origin, drinking history, or level of immaturity, for example, increase the frequency of questions directed at those topics.  In turn, asking enough questions about a particular topic allows clinicians to think they have found evidence confirming their original assumptions — but their thinking in this regard is often seriously mistaken (Arkes, 1981).

Clinicians frequently insist that symptoms consistent with their diagnostic impressions were present in an interview; when in fact, they were not (Arkes & Harkness, 1980).  Conversely, they are less likely to recall symptoms actually exhibited in an interview but inconsistent with their diagnostic impressions.  Additionally, more experienced clinicians cannot legitimately claim greater judgmental accuracy by virtue of their years of experience (Dawes, 1989; Garb, 1989).

Despite the extent to which the relevant research underscores the dismal status of clinical judgment, clinicians still overconfidently assume that their judgments are exceedingly accurate (Dawes, 1989).  Nevertheless, it is also important to note that the degree of confidence expressed by clinicians in their judgments is unrelated to their actual judgmental accuracy (Faust, 1989).  Clinicians often reach their diagnostic conclusions very early in an interview, and then they cling to their original impressions even when confronted with contrary evidence (Robins & Helzer, 1986).  Simultaneously, they commonly overestimate the amount of information they process during their evaluations.  Clinicians assume that they weigh multiple factors in making their judgments, but the relevant evidence demonstrates that they rely on minimal data (Fisch, Hammond, & Joyce, 1982; Gillis & Moran, 1981).

Given the frequency with which clinical judgment too often results in arbitrary and unsubstantiated conclusions, the potential for abusing Gardner's indicators is quite alarming.  This writer has recently encountered cases where self-styled "validators" (Ewing, 1992; Gardner, 1991) — who rarely encounter allegations of child sexual abuse that they do not uncritically endorse — interpret Gardner's indicators to conform with their own biased assumptions.  In particular, these validators exploit the ambiguity of the indicators and apply them in a manner that appears to validate their prejudiced thinking regarding the suspects they examine.  The validators then invoke Gardner's reputation — and his indicators — portraying themselves as paragons of objectivity while defending their unwarranted conclusions.

When validators bend Gardner's indicators to suit their own purposes, clinicians who approach allegations of child sexual abuse with greater care and caution can find themselves in a difficult situation.  They often encounter unrelenting cross-examination regarding the extent to which Gardner is a respected figure in the area of child sexual abuse allegations.  Therefore, the ease with which validators can capitalize on Gardner's reputation and use his indicators to serve their own preexisting agenda should not be underestimated.
  

Conclusions

Gardner's previously acknowledged reputation as a courageous figure deserves continued respect, but his "Indicators of Pedophilia" do not.  By virtue of his psychiatric education and training, Gardner's unfamiliarity with the technical details of reliability and incremental validity is understandable.  Unfortunately, however, his oversights related to these issues threaten to provoke a major crisis in child sexual abuse litigation that demands corrective action.

Assiduously examining Gardner's indicators clearly demonstrates that they cannot support expert testimony in forensic settings.  Legal tests for determining the admissibility of expert testimony traditionally demand that it rely on evidence or principles that enjoy general acceptance in the relevant scientific or professional community (Ziskin & Faust, 1988).  The massive reliability and validity problems endemic to Gardner's indicators preclude their general acceptance by the scientific and professional communities on a prima facie basis.  Therefore, contending with the potentially disastrous consequences of these indicators necessitates persistent challenges directed at disqualifying them as a basis for expert testimony.

The issues outlined and discussed in this article also raise questions related to the appropriateness of clinically examining suspects in cases of child sexual abuse litigation.  Clinical examinations typically involve psychological testing which is often irrelevant in these cases, clinical interviews which can be horribly unreliable, and clinical judgment which can be exceedingly inaccurate.  Rather than invest enormous resources in examining suspects, evaluating the legitimacy of the allegations themselves may return more reliable information.

False allegations of sexual abuse can be considered unsubstantiated rumors which acquire an unwarranted status of facts via socially constructed realities (Campbell, 1992c).  Mental health professionals can assist courts in these cases by examining the history of the allegations and assessing the extent to which they correspond to rumor formation and rumor dissemination (Campbell) 1992d).  Instead of resorting to the speculation and conjecture that frequently characterize examinations preoccupied with what transpires within suspects, evaluating the interpersonal transactions surrounding allegations of sexual abuse addresses what transpires between people.  Because interpersonal events between people are more observable than what occurs within them, this latter endeavor enjoys greater objectivity.

Child sexual abuse litigation has too long served as a stage showcasing the incompetence of too many mental health professionals.  These clinicians naively embrace unbridled conjecture, premised upon intuitive impressions, while blithely disregarding fundamental considerations of scientific evidence and logic.  And as they indulge their ill-informed biases, their testimony qualifies as theater of the absurd.  Assessment procedures resorting to anatomically detailed dolls and ill-defined concepts such as repression, accommodation syndromes, and symbolic play have betrayed untold numbers of juries and jurists alike.

Ethically responsible mental health professionals are obligated to discredit practices which threaten to misinform the legal system, and unfortunately enough, this obligation extends to Gardner's "Indicators of Pedophilia."  Those who would defend these indicators do so on behalf of an evaluative procedure as seriously flawed as the practices favored by self-styled validators.  Consequently, then, such defenses seem ill-advised.
  

References

Anastasi, A. (1982). Psychological Testing (5th Ed.) ()(). New York: The Macmillan Co.

Arkes, H. R. (1981). Impediments to accurate clinical judgment and possible ways to minimize their impact. Journal of Consulting and Clinical Psychology, 49, 323-330.

Arkes, H. R., & Harkness, A. R. (1980). Effect Of making a diagnosis on subsequent recognition of symptoms. Journal of Experimental Psychology: Human Learning and Memory, 6, 568-575.

Campbell, T. W. (1992a). Therapeutic relationships and iatrogenic outcomes: The blame-and-change maneuver in psychotherapy. Psychotherapy, 29, 474-480.

Campbell, T. W. (1992b). Diagnosing incest: The problems of false positives and their consequences. Issues In Child Abuse Accusations, 4(4), 161-168.

Campbell, T. W. (1992a). False allegations of sexual abuse and their apparent credibility. American Journal of Forensic Psychology, 10(4), 21-35.

Campbell, T. W. (1992b). Allegations of sexual abuse II: Case example of a criminal defense. American Journal of Forensic Psychology 10(4), 37-48.

Dawes, R. M. (1989). Experience and validity of clinical judgment: the illusory correlation. Behavioral Sciences & the Law, 7, 457-467.

Ekman, P., & O'Sullivan, M. (1991). Who can catch a liar? American Psychologist, 46, 913-920.

Ewing, C. P. (July 1992). Judicial Notebook: Child sexual abuse "validation" on trial — and retrial. APA Monitor, p. 14.

Faust, D. (1989). Data integration in legal evaluations: Can clinicians deliver on their premises? Behavioral Sciences & the Law, 7, 469-483.

Fisch, H. U., Hammond, K. R., & Joyce, C. R. (1982). On evaluating the severity of depression: An experimental study of psychiatrists. British Journal of Psychiatry, 140, 378-383.

Garb, H. N. (1989). Clinical judgment, clinical training, and professional experience. Psychological Bulletin, 105, 387-396.

Gardner, R. A. (1987). The Parental Alienation Syndrome and the Differentiation Between Fabricated and Genuine Child Sex Abuse (). Cresskill, NJ: Creative Therapeutics.

Gardner, R. A. (1991). The "validators" and other examiners. Issues in Child Abuse Accusations, 3(1), 38-53.

Gardner, R. A. (1992). True and False Accusations of Child Sex Abuse (). Cresskill, NJ: Creative Therapeutics.

Gerlsma, C., Emmelkamp, P. M., & Arrindell, W. A. (1990). Anxiety, depression, and perception of early parenting: A meta-analysis. Clinical Psychology Review, 10, 251-277.

Gillis, J. S., & Moran, T. J. (1981J. An analysis of drug decisions in a state psychiatric hospital. Journal of Clinical Psychology, 37, 32-42.

Gottesman, I. I., & Prescott, C. A. (1989). Abuses of the MacAndrew MMPI Alcoholism scale: A critical review. Clinical Psychology Review, 9, 223-242.

Green, B. F., & Hall, J. A. (1984). Quantitative methods for literature reviews. Annual Review of Psychology, 35, 37-53.

Hall, C. S., & Lindzey, G. (1970). Theories of Personality (2nd Ed.) (). New York: John Wiley & Sons.

Haugaard, J. J., & Repucci. N. D. (1988). The Sexual Abuse of Children (). San Francisco: Jossey-Bass.

Jacob, T., Krahn, G. L, & Leonard, K (1991). Parent-child interactions in families with alcoholic fathers. Journal of Consulting and Clinical Psychology, 59, 176-181.

Lewinsohn, P. M., & Rosenbaum, M. (1987). Recall of parental behavior by acute depressives, remitted depressives, and nondepressives. Journal of Personality and Social Psychology, 52, 611-620.

Logue, M. B., Sher, K. J., & Frensch, P. A. (1992). Purported characteristics of adult children of alcoholics: A possible "Barnum effect." Professional Psychology, 23, 226-232.

Lykken, D. T. (1979). The detection of deception. Psychological Bulletin, 86, 47-53.

Marmor, J. (1966). Psychoanalysis at the crossroads. In J. Masserman (Ed.). Science and Psychoanalysis, Vol.10. New York: Grune & Stratton.

Matt, G. E., Vazquez, C, & Campbell, W. K. (1992). Mood-congruent recall of affectively toned stimuli: A meta-analytic review. Clinical Psychology Review, 12, 227-255.

Monahan, J. (1992). Mental disorder and violent behavior: Perceptions and evidence. American Psychologist, 47, 511-521.

Mischel, W. (1973). Toward a cognitive social-learning reconceptualization of personality. Psychological Review, 80, 252-283.

Myers, J. E., Bays, J., Becker, J., Berliner, L., Corwin, D. L., & Saywitz, K. J. (1989). Expert testimony in child sexual abuse litigation. Nebraska Law Review, 68, 1-145.

Peele, S. (1989). Diseasing of America: Addiction Treatment out of Control ()(). Lexington, MA: Lexington Books.

Peters, J. J. (1976). Children who are victims of sexual assault and the psychology of offenders. American Journal of Psychotherapy, 30, 398-421.

Peters, J. M. (June 1989). 'Why prosecute child abuse?" Paper presented at the 17th Annual Child Abuse and Neglect Symposium. Keystone, Colo. Cited by Wakefield, H. & Underwager, R. (1991). Female child sexual abusers: A critical review. American Journal of Forensic Psychology, 9(4), 43-69.

Robins, L. N. (1985). Epidemiology: Reflections on testing the validity of psychiatric interviews. Archives of General Psychiatry, 42, 918-924.

Robins, L. N., & Helzer, J. E. (1986). Diagnosis and clinical assessment: The current state of psychiatric diagnosis. Annual Review of Psychology, 37, 409-432.

Rosenthal, R. (1966). Experimenter Effects on Behavioral Research (). New York: Appleton, Century, Croft.

Ryff, C. D. (1987). The place of personality and social structure research in social psychology. Journal of Personality and Social Psychology, 53, 1192-1202.

Saxe, L., Dougherty, D., & Cross, T. (1985). The validity of polygraph testing: Scientific analysis and public controversy. American Psychologist, 40, 355-386.

Singer, J. A., & Salovey, P. (1988). Mood and memory: Evaluating the network theory of affect. Clinical Psychology Review, 8, 211-251.

Wakefield, H., & Underwager, R. (1988). Accusations of Child Sex Abuse ()(). Springfield, IL: C.C. Thomas.

Wilkenson, C. B., & O'Connor, W. A. (1982). Human ecology and mental illness. American Journal of Psychiatry, 139, 985-990.

Wolfner, G., Faust, D., & Dawes, R. (199j). The use of anatomically detailed dolls in sexual abuse evaluations: The state of the science. Applied and Preventive Psychology, 2, 1-11.

Ziskin, J. S., & Faust, D. (1988). Coping with Psychiatric and Psychological Testimony (4th Ed.) (). Los Angeles: Law and Psychology Press.