IRA DANIEL TURKAT, PH.D., Venice, Florida

There are few activities in which a mental health professional can engage that are more emotionally arousing, controversial, and potentially damaging than performing a custody evaluation. The process is hampered by two key facts. First, the scientific literature on custody determination is woefully inadequate. Second, for every competent professional evaluator, there may be many more incompetent ones.

For these reasons, it is important that attorneys be attuned to some of the strengths and weaknesses inherent in mental health experts' custody recommendations. This article will outline a variety of issues that one might consider when questioning such an expert's report on custody determination.


Perhaps the greatest impediment to the mental health expert's ability to provide fair, competent, and appropriate custody recommendations is the absence of a sound scientific body of research in this area. Currently, there is no clear-cut body of scientific data about some of the basic questions that underlie a custody recommendation. For example, there is an absence of strong scientific evidence regarding precise parenting characteristics that guarantee "good parenting." Similarly, we lack sound research data regarding the effects on the future of a child who's been placed with the "wrong" parent. Given the absence of well-established scientific data on these issues, this leaves the mental health professional with tremendous leeway in regard to how he or she decides to go about doing a custody evaluation and in the interpretation of the data collected for that investigation. As such, what one mental health expert might see as critical, another similarly trained professional might see as trivial. This leaves the court in a terrible quandary--one of which the court, at times, may not even be aware.

The problem is well illustrated when a mental health expert who has an excellent courtroom presentation style is actually incompetent. The court, relying on someone believed to be an expert, actually receives poor advice that the court does not have the expertise to evaluate. Unfortunately, as noted above, many individuals engaged in the business of providing custody recommendations probably have no business doing so. Hopefully, the reader will find the information contained in this article useful in evaluating such a mental health professional's report.


The psychiatrist (M.D.), psychologist (Ph.D.), and social worker (M.S.W.) are the experts in our society regarding issues of mental health. Unfortunately, this author is not aware of any scientific evidence proving that an M.D. or Ph.D. or M.S.W. is any more competent in making a good custody decision than a well-seasoned family court judge. However, since these experts are often called upon to make custody recommendations, I will present my biases regarding some of the "ideal" characteristics to be found in the professional custody evaluator.

The examiner should be able to document having accomplished a significant number of custody decisions that reflect an unbiased record (e.g., the evaluator does not always recommend that the mother get the children).

Of the three professions noted in the above paragraph, the PhD. in clinical psychology is typically the one distinguished by the most training in basic science. Having a critical attitude about data of any kind, and the tools by which to understand the scientific literature pertinent to the case, the Ph.D. is probably the preferred professional. However, this is not always the case. It is especially not the case when the person doing the evaluation has no specific training in the area of child development. Key to making a custody recommendation is the ability to predict the future development of the child if placed with a particular parent. As such, the professional with only one graduate level course in child development would appear to have an insufficient background. Instead, the ideal mental health professional for doing a custody evaluation has significant training in the area of child development, with particular understanding of the stages of development and how they interact with the positive and negative attributes of each parent.

In addition to the qualifications above, the ideal evaluator has significant training in psychopathology, with particular understanding of the factors that promote the development of mental disorders. In particular, the evaluator should have special expertise in the areas of personality and personality disorders. The clinician who lacks top-notch diagnostic skill is potentially short-changing the family and the child.

The ideal evaluator has many years of experience as a practicing clinician. However, in addition, the examiner should be able to document having accomplished a significant number of custody decisions that reflect an unbiased record (e.g., the evaluator does not always recommend that the mother get the children).

Finally, the mental health expert in this area must make it clear to the court that his or her tests, interpretations, and recommendations are potentially subject to error. The expert should not only make the court aware of this, but should give some indication to the court regarding the range and degree of possible errors involved.


We are all subject to biases, some of which we are not even aware. Even the most impartial examiner can find himself or herself at times favoring a particular parent in a custody dispute for reasons that have little to do with the quality of the parenting offered. Given that this is potentially the case with even the most "objective" of examiners, the implications for the "hired gun" are obvious: the hired gun should not be allowed to give a custody recommendation.

A second key issue in regard to the relationship between the expert and the parties involved is what is known as the problem of "dual relationships." This exists when the expert has a relationship with one or more of the parents independent of the custody evaluation. For example, assume the wife has been seeing a particular therapist for individual psychotherapy and then offers his services to the court to provide a custody evaluation. In this instance, the examiner has a clear history of being supportive and helpful to, and most important, being on the side of the wife. This unquestionably biases the therapist, whether or not he or she wants to admit it. Such an individual should not be permitted to also then evaluate the entire family for a custody determination.

Another example of the problem of dual relationships occurs when one of the parties enters into court-ordered therapy with a particular therapist and then afterward that therapist recommends that a full family evaluation be done. While the recommendation itself might be appropriate, it is inappropriate for the court-ordered therapist of one party to then perform the family evaluation.

Trying to be therapist and evaluator at the same time increases the risk of biased "professional" opinions.


Psychologists are divided themselves about the importance and appropriateness of psychological testing in making a custody determination. The fact that there are thousands of tests available in the psychological literature only compounds the problem.

The use of IQ tests is one clear-cut area of controversy. Some consider IQ tests to be required in every case, whereas others consider IQ tests as highly wasteful and inappropriate for a custody evaluation. To this author, it would seem that if the intelligence of the children is an issue in regard to which parent might be better able to foster that child's attributes, then perhaps IQ tests are relevant. Some evaluators may be viewed as adding them merely as a way to "pad the bill." The IQ test is one of the most well-known and well-established assessment instruments associated with the work of the clinical psychologist. Accordingly, the reader should not be surprised that tests of less stature may be even more controversial.

In recent years, several psychologists have attempted to develop "objective" measures that are specific to custody evaluations. For example, Bricklin has developed the Bricklin Perceptual Scales, and he claims a 90 percent agreement rate between his scales and judges' determinations. Similarly, the Ackerman-Schoendorf Scales for Parent Evaluation of Custody (ASPECT) have recently been introduced and the authors claim a 90 percent agreement with judges' custodial decisions as well. The mental health expert who does not use these tests should be queried as to why not. Those examiners who use these scales should be questioned regarding the amount and quality of research on these scales, what specific validity evidence exists in the research literature about them, and what are the known sources of error in these scales, as they relate to the case at hand.


While the choice of psychological test is controversial enough, interpreting test data can be even more controversial. In my own practice, both clinically and in training doctoral-level psychologists, I have often seen two professionals look at the same data and come up with quite different interpretations. Some psychologists may present computer interpretations to imply a more "objective" presentation, but there are problems here as well. First of all, the computer interpretation is limited by what the human being puts into the computer interpretation program. While the computer will be consistent in applying the rules it has been given, it does not mean that the rules it has been given are necessarily correct. Further, there are so many interpretive computer programs available to psychologists these days, the amount of scientific validity evidence for them varies tremendously. As such, the psychological examiner should be examined on this particular issue.

If the psychologist has a particular choice in mind as to who should receive custody, he or she may examine the psychological test data searching for evidence to confirm if as opposed to looking at evidence that might disconfirm it.

Additionally, the two most well-known and widely used major personality inventories, the Minnesota Multiphasic Personality Inventory (MMPI) and the Millon Multiaxial Clinical Inventory (MCMI), when given to the same person, may yield two very different computer interpretations. The bias of the examiner may come further into play when he or she examines the two discrepant personality inventories on the same individual. Clearly, the tendency to selectively pick out results becomes a problem.

Laypersons and psychologists alike are prone to search data looking for instances that confirm one's ideas. In other words, if the psychologist has a particular choice in mind as to who should receive custody, he or she may examine the psychological test data searching for evidence to confirm it as opposed to looking at evidence that might disconfirm it. He or she may then selectively pull out those pieces of test data to present to the court but inadvertently (or perhaps not inadvertently) leave out those data points that might contradict the psychologist's recommendations.


Many custody evaluators do not specifically and systematically observe the interaction between each parent and each child involved in the custody evaluation. The psychologist who fails to make systematic observations of how each child and each patent interact on various tasks may be doing the family a disservice. In this case, the examiner restricts information to that which is secondary to the basis of the custody question: namely, the child-parent interactions that are most likely to produce the best benefits for the child.

The bottom line is that the psychologist who does not provide a thoroughly convincing and compelling argument for the interpretation of his or her interview, test data, and observations to provide a custody determination is not doing the family or the court a useful service.

An excellent way to assess parent-child interactions is to videotape each parent with each child independently engaged in a variety of tasks. This permits the psychologist to review those interactions when alone, and also to invite other colleagues to view them and give opinions. This expands the psychologist's ability to assess the case beyond interviews, psychological tests, and records. The psychologist who does not do this should be prepared to defend why it hasn't been done.

A key point in picking tasks for the parent-child interactions to be observed is that they should not be biased toward either parent. For example, if one parent knows how to sew and the other does not, it would be unfair to instruct each parent to teach the child how to sew on videotape. Tasks such as playing a game, solving a problem, or teaching the child something new are useful indicators.


It is not uncommon to read a psychological report of 10 to 15 pages covering the interview, test data, and so on and then read the final section in which recommendations are made that have no clear, logical connection to all of the preceding information. The court should expect the recommendation section to provide a logical, clear, and convincing justification for why one parent should be named the primary residential parent and the other should not

Psychological examiners, like other human beings, often make errors in logic. For example, assume all of the data collected on the father and the mother are relatively equal except that the mother has an elevation on her depression subscale on the MMPI. To award primary residential custody to the father purely because the mother has one test score suggesting an elevation in depression represents an inadequate justification by itself. In fact, even if one parent has some type of psychopathology, that does not therefore mean that that parent is a poor parent. The bottom line is that the psychologist who does not provide a thoroughly convincing and compelling argument for the interpretation of his or her interview, test data, and observations to provide a custody determination is not doing the family or the court a useful service.

Another problem regarding the rationale for a custody recommendation is that some evaluators misrepresent the value of their recommendations. For example, one psychologist noted in his report to the judge that the method he adopted for conducting the custody evaluation was based on a method used by a former president of the American Psychological Association (APA). This implies that the psychologist's report has merit. Unfortunately, this is a misrepresentation of the report's value for at least three reasons. First, the man who became APA president made a major political achievement. However, there is no evidence that this APA president's approach to custody determination won him the election. Second, even if there is merit to this APA president's custody determination procedure, there is no evidence that the particular psychologist who used the method has the same competency as the APA president in administering and interpreting this method. Finally, and most importantly, there is no strong scientific evidence that this APA president's procedure is the method of choice for performing a custody evaluation. It is the scientific literature that should guide custody evaluation procedures, not rhetoric nor the charisma of any one statesman. The rationale underlying a custody recommendation has the most value when scientific data is provided to support it.


Custody battles are heart-rending. Children's futures are at stake. Scientific evidence to guide custody decisions is nowhere near the necessary level for experts who need support in making such decisions. Until the scientific literature on custody determination matures, the report of the mental health "expert" should be viewed with a critical eye.

Ira Daniel Turkat, PhD., is the chief psychologist at Venice Hospital and is on the clinical faculty of the University of Florida College of Medicine. He maintains a private practice in Venice, Florida.