Rethinking Child Custody Evaluations A Paradigm Shift: Introducing ‘The Chicago Model’

by Mark D. Mosk, Ph.D.

Working in the arena of child custody can be highly rewarding, and sometimes equally arduous and extremely challenging as well. The history of collaboration between legal and mental health professionals in Family Court is a rich and well-established one. With a unique set of knowledge, skills and expertise, mental health professionals have been invited into the courtroom to provide hopefully useful and meaningful information about medical-legal matters, and to assist the court in its judicial decision-making process, including in Family Court where psychologists and psychiatrists may opine on what ultimately is in the best psychological interests of the child.

As mental health specialists operating in the role of a forensic scientists, we are expected and I would offer, indeed are obligated to conduct a scientifically-based investigation of the presenting facts, using rigorous and empirically sound methods and data analysis procedures, to provide a clear and cogent understanding of the critical medical-legal issues, and in doing so, to educate the court about the presenting topics. For the forensic psychologist and psychiatrist, this evaluative approach regularly is applied in numerous specialty areas, from child custody and disability, to workers’ compensation and criminal matters. Regrettably, many of my fellow mental health colleagues would agree that these high professional standards are not consistently met in child custody evaluations.

As a caveat, the reader should note that most of my 25 years as a practicing forensic examiner have been spent in disability and workers’ compensation matters in California, having conducted nearly 1,000 comprehensive evaluations, and having reviewed and critiqued hundreds of forensic mental health reports. While my professional experience in the child custody arena is relatively new, dating back about five years, I nevertheless have had the opportunity to review a substantial number of child custody evaluation reports. Over that time, many family attorneys have shared openly that, at times, the conclusions and recommendations offered by custody evaluators merely confuse, obfuscate and distort, rather than inform, educate and enlighten. The core issue here is that reports inadvertently may convey inaccurate and misleading information to the Court, which very well could have an adverse long-term impact on families’ lives.

Despite the well-meaning intentions of and the very real efforts by several national professional organizations (e.g., American Psychological Association, Association of Family and Conciliation Courts, American Academy of Forensic Psychology, and American Academy of Child and Adolescent Psychiatry) to enhance the general quality and value of child custody evaluation reports, the quality and caliber of our custody evaluation reports can fall short of expectations, and sometimes, dramatically. I am not alone in recognizing this state of affairs, as many psychologists and family lawyers alike have begun raising red flags about this very issue. In fact, this topic was highlighted by our legal and mental health contemporaries as recently as this past summer, in workshops presented at the AFCC conference here in Chicago.

The purpose of this article is not to disparage the performance of the mental health community working in this important field, but to raise awareness of these issues and to acknowledge the continuing challenges that face us in today’s custody evaluation process. It is hoped that a meaningful discussion will ensure that leads to improvements in a system that affects the lives of thousands of Illinois residents yearly.

What follows below is a discussion of some of the challenges inherent in the child custody evaluation process and an elucidation of several possible underlying causes of those difficulties. In Part 2 of this article, I then offer for the reader’s consideration a potential substantive solution to many of these recognized problems by way of a novel and innovative conceptual model and framework for conducting child custody evaluations, one that could benefit not only the legal community and mental health evaluators, but the families and children that we serve as well.

I begin this discussion in a conceivably unexpected place, in the area of psychological testing. As the reader likely is aware, psychological testing often is included in child custody evaluations and is a community standard amongst psychologists (Quinnell & Bow, 2001). Not all psychologists administer psychological tests during their evaluations, for various reasons, and conducting such tests is not mandatory. Likewise, psychiatrists that wish to obtain psychometric testing data in custody related matters may retain a psychologist to do so.

There literally are hundreds of published psychological tests that purportedly derive accurate and useful information about adults, children and family dynamics. However, the vast majority of these tests are inappropriate for use in custody settings. This complex issue has been elucidated by a number of respected authors (e.g., Flens, 2005; Ackerman, 2006; Erickson, Lilienfeld & Vitacco, 2007) and is best saved for a separate discussion. Certainly, if there is suspicion on the part of the evaluator about possible psychopathology in one or both parents, the administration of an objectively-based symptom assessment instrument, such as the MMPI-2 or the MCMI-III, could be crucial in determining the presence of a serious emotional condition.

Generally speaking, the results from psychological tests can generate information about a person’s emotional state, perceptions, attitudes, cognitive abilities, and personality traits. However, even in circumstances where the proper tests are used in a custody evaluation, those tests with rare exception measure simply intrapsychic events — symptoms, attitudes, feelings, beliefs, etc. None of these tests measure specific, observable, operationally defined parenting behaviors. Consequently, the evaluator is required to initiate a process of hypothesis generation and testing to address the issue of acceptable parenting. The evaluator must make assumptions, theorize, postulate, extrapolate, and develop corollaries from the testing data about the individual’s strengths and weaknesses as a parent, and then hopefully test those hypotheses by whatever means she or he deems reasonable.

It is important that the reader understand that there is no direct relationship between the results from any currently available published psychological test and parenting capacity. This is unfortunate because, typically, the only source from which strictly objective data are derived during a custody evaluation is from psychological testing. Thus, absent the direct testing of parenting behaviors, one deductively can surmise that the evaluator’s assessment data, including clinical observations but with perhaps with the exception of confirmed historical information garnered from records, are strictly subjective in nature. In other words, the conclusions drawn and recommendations made by the custody evaluator can be heavily influence by personal, subjective opinions.

Since it goes to reason that different evaluators can develop vastly different impressions from large bodies of information, they therefore can and typically do have widely varying opinions about an individual’s psychological state. Yet, as forensic examiners charged with the task of conducting a systematic and scientifically-based investigation, it behooves us, and undeniably, we are mandated to employ as much scientific rigor as possible in the evaluation process. Subjectivity must be minimized, and if possible, entirely avoided, in scientific study.

Since it is the case that no direct relationship exists between psychological testing results and parenting capacity, one must ask the question, Why not?‚ The answer to this question is two-fold: (a) the lack of a true definition of parenting capacity‚ and (b) the absence of a valid and reliable psychological test capable of assessing directly for this construct.

The reason for the absence of a direct relationship between testing results and parenting capacity pertains broadly to a problem in measurement. First, while we desire to measure‚ parenting capacity, the term itself never has been adequately defined, and certainly not from a strictly behavioral and quantifiable standpoint. While the legal profession rightly imposes its definition of this term, even those definitions may be perceived as fairly vague and open to some interpretation. Certainly, in the mental health realm, there is not a universally accepted operational definition of the term.

One evaluator’s definition of parenting capacity can be, and typically is, different from that of another. If we evaluators are charged with the task of addressing the question of parenting capacity in our examinations, and yet do not have an explicit agreement amongst ourselves, let alone with our legal colleagues, on what the term, parenting capacity, means, then how can we as evaluators possibly choose an appropriate psychometric instrument to test for that construct? How does one select an appropriate measurement tool without knowing exactly what it is that one needs to measure?

I present this issue of psychological test selection as a means of illustrating one of the key challenges facing us in this field‚ the absence of an unmistakable definition of competent and effective parenting (Gould, 2006.) Without full consensus on precisely what the term, parenting competence‚ or, parenting capacity‚ actually entails, it follows then that it is almost irrational to expect consistent interpretations of data in relationship to parenting proficiency. Yet, it is exactly this task that evaluators are asked to perform. When we opine on this construct without clearly defining its constituent parts and parameters, we introduce error into the equation. Thus, it is not surprising that wildly differing opinions about the psychological attributes of the same family members, and even when using the same data set, can emerge across evaluators when they operate under these conditions.

In this context, let us examine the type of empirical data that evaluators often do collect in an attempt to address the question of parents’ functionality. As stated above, the primary and usually exclusive source of objective information is from psychological tests. Yet, do either of the two most commonly administered psychological tests in custody assessments, the MMPI and MCMI, for instance, both of which measure individual’s symptoms, provide explicit information that directly translates to functional behaviors? The simple answer is‚ No. As discussed above, the evaluator must postulate theories and generate inferences about the relative behavioral assets and deficits of each parent from those testing data.

The greater the degree of disparity between the type and genre of psychological testing data gathered (e.g., symptoms, attitudes, emotions, etc.) and effective parenting skills, the more the evaluator is required to hypothesize and rely upon conjecture. The fact of the matter is that symptoms are not behaviors, apples are not oranges, and our most commonly administered psychodiagnostic instruments do not measure parenting capacity, nor do these tests purport to do so. The predicament that we face as professionals in this arena is that the process of drawing inferences about specific behaviors from a person’s symptoms, feelings, and attitudes requires taking an extraordinary and possibly quite dangerous leap of faith. There are no guiding principles or an instruction manual on how a well-trained psychiatrist or psychologist might go about generating such interpretations.

Of course, the evaluator supposedly could generate a multitude of different hypotheses suggested by the data, and then systematically test each hypothesis. However, based upon my professional experience, this is not regularly practiced in the field. Still, behavioral descriptions generated from such hypotheses have the potential to help verify information about parenting behaviors that were obtained through other collateral sources.

The process of drawing meaningful conclusions about the parent’s behavior from non-behavioral data can lend itself to tremendous variability. This process accordingly introduces another type of error into the assessment process. Thus, due to the lack of a clear definition of and an objective means for measuring parenting capacity, we oftentimes find dramatic differences in opinion across evaluators.

Ostensibly, what we are facing today in this field is a systemic problem in the custody evaluation process. First, the field is filled with ambiguous terms that are used freely and frequently when discussing and in reaching judgments about parent and child behaviors. Terms such as best psychological interest of the child, nurturance, parenting capacity, attachment, enmeshment‚ and even, parental alienation‚ often are found in evaluators’ reports. Yet, we, as professionals, do not have 100% agreement on exactly what these terms mean. If evaluators insist on using such poorly defined terms, it at least would be helpful if they provided a detailed description of the specific observed behaviors that they believe constitutes that construct or phenomenon.

Second, reports have a tendency to incorporate significant subjectivity and inconsistency. As discussed earlier, with perhaps the exception of gathering objective information from psychological testing and identifying stable behavioral patterns from comprehensive historical data, the assessment process otherwise tends to involve primarily the amalgamation of a myriad of subjective impressions. Reports often offer little more than regurgitated histories by litigants with questionable credibility, often precipitating a proverbial, He Said-She Said‚ diatribe. While some evaluators sometimes actually describe their clinical observations, and they should be encouraged to do so, one cannot necessarily remove the influences, unintentional or purposeful, of the evaluator’s personal biases, values, and governing morals on those reported observations.

The role of bias entering into the evaluation process has been discussed at length by many esteemed experts in the field. For instance, Martindale (2005) suggests that this threat to objectivity can embody two different forms. Confirmatory bias reflects the desire on the part of the evaluator to obtain and emphasize data that supports a pre-existing, previously developed hypothesis about an individual(s). On the other hand, confirmatory distortion generally speaks to the evaluator’s excessive confidence in the accuracy of his or her pre-existing or quickly formed hypotheses that can lead to intentional selection of data that is considered for presentation to the Court as well as a misrepresentation of those data.

Third, the great variability in the quality and consistency of custody reports may be a function of the many different professionals backgrounds from which evaluators emerge, including psychiatry, psychology, and social work. Even amongst psychologists alone, there are numerous schools of training that produce developmental psychologists, educational psychologists, clinical psychologists, industrial psychologists, social psychologists, and others.

Further complicating the matter, there understandably exist highly diverse theoretical orientations within and across mental health professionals. For instance, psychiatrists can be trained as psychoanalysts or with an orientation towards biological psychiatry. Psychologists may receive training orientated towards humanism, family systems, Rogerian methods, cognitive therapy, behaviorism, play therapy, or others. This wide diversity of training backgrounds and theoretical orientations tints the color of the lenses through which evaluators perceive the world. This in turn shapes their behavior and helps guide their investigation process, naturally influencing the methods, procedures, and processes used to conduct their assessments. With this tremendous heterogeneity across so many dimensions, it is no wonder that each report may appear to be different from the next.

Evaluators can vary greatly in their reliance upon and importance they assign to the different types of data available to them. This may be influenced by the theoretical model that they follow. But other factors also can affect the type of information that is collected and analyzed, including the feasibility of accessing that data and the evaluator’s skill set in evaluating that data.

The methods used by evaluators to gather data often differ and are inconsistent. For instance, some evaluators prefer to conduct observations of parent-child interactions in their office, while others do so in the parents’ home. This practice raises an entirely different issue, specifically, the lack of unobtrusive observation of parent-child interactions. The potential influence on the behaviors of parties by the mere presence of an observer has been well established (Kazdin, 1979). The possible impact of this variable on the observational data often is not even mentioned by evaluators in their reports.

Another complicating factor is that different terminology and language often is found from report to report. Related to the discussion above regarding the definition of, parenting capacity, the use of ambiguous terms, technical language, psychobabble, and even simple English prose whose meaning is lost in the context of the written sentence is prolific. For the family law attorney and sitting Judge, reviewing 604(b) and 604.5 reports very well might feel as if one has been teleported to the Tower of Babel of biblical times, where all of the inhabitants spoke different languages. This lack of clear communication forms the bedrock of misunderstanding.

Furthermore, more often than not, evaluators fail to establish a clear nexus between their recommendations and the data upon which they supposedly relied to formulate those conclusions. Failing to present such a logical flow in writing suggests that the evaluator indeed may not have had an identifiable, reasonable, empirically sound, and reliable rationale for reaching the stated conclusions. Instead, one could surmise that the evaluator’s recommendations perhaps were based upon conjecture and supposition alone.

There is no doubt that evaluators recognize conceptually that they must opine about the aptitude of the examinee to fulfill a parenting role. However, reports uncommonly support their recommendations with detailed discussions that comprehensively outline the parents’ specific abilities to parent, with supporting behavioral data explaining each individual’s strengths and deficits, across a wide array of parenting functions.

Finally, on those occasions when a litigating parent is encouraged or even ordered to enter into psychotherapy, such recommendations rarely incorporate a clear, specific and defined therapeutic goal as it pertains to the individual’s expected ability to exhibit desired parenting skills post-treatment. It is unlikely that psychiatric intervention is recommended because the Court has interest in the individual’s ability to gain personal insight or attain self-actualization. Even if a recommendation for therapy is due to the presence of a psychiatric disorder, such as anxiety or depression, it is not the symptom relief that is paramount. Rather, at issue is the ability of that individual to perform more effectively in one or more aspects of the parenting role.

Thus, I submit that a combination of prolific use of ambiguous terminology and misleading language, extraordinary variety of methods and procedures employed, glaring lack of objectivity, exceptional contradictions across evaluators, corruption of reports by evaluators’ personal biases, and a frequent absence of a clear nexus between the evaluator’ data and recommendations characterize much of the custody evaluation system today. From my personal communications with many family lawyers, confusion, befuddlement and frustration frequently describe their efforts in trying to decipher the true meaning and gauge the real value of a given report.

This almost complete lack of uniformity in and standardization of evaluation reports has not gone entirely unnoticed. For example, the American Psychological Association, the AFCC, and the AACAP have established guidelines in an attempt to introduce some degree of structure and consistency into the field. However, these guidelines merely are suggestions, strategies, and advisories. They are not enforceable directives or statutes. Ostensibly, the evaluator is free to ignore these guidelines, and oftentimes, they do.

In my professional opinion, both the mental health profession and the legal profession share responsibility and equally are culpable for the current untenable and problematic situation as it pertains to today’s child custody evaluation reports. Our respective fields must remain accountable to remedy the untenable state of affairs that exists presently. Sadly, neither of our professions has yet to offer the Court a substantive and comprehensive solution to this dilemma. It is time for a change.

I suggest that a new conceptual and operational model is needed that employs a uniform and standardized functional assessment approach to the child custody evaluation process, which (1) uses evidence-based methods validated in the forensic mental health arena, (2) embodies standard methods for systematically gathering and analyzing specific objective information about critical parenting abilities, (3) is founded upon an underlying, valid conceptual framework that guides the evaluator’s investigation and seamlessly integrates relevant clinical data with recommendations, and (4) presents conclusions in a cohesive, logical, and rational fashion. Hence, I would like to introduce and propose for consideration such a model ‚Äì “The Chicago Model.” In Part 2 of this article, I will describe the components of “The Chicago Model,” its theoretical and evidence-based roots, its application, and the numerous benefits that it can offer attorneys, Judges, and mental health practitioners alike working in the arena of family law.


Ackerman, M.J. (2006). Clinician’s Guide to Child Custody Evaluations (3rd ed.). Hoboken, NJ: John Wiley & Sons.

Erickson, S.K., Lilienfeld, S.O. & Vitacco, M.J. (2007). A critical examination of the suitability and limitations of psychological tests in family court. Family Court Review, 45(2), 157-174.

Flens, J.R. (2005). The responsible use of psychological testing in child custody evaluations. In J.R. Flens & Drozd, L. (Eds.), Psychological testing in child custody evaluations (pp. 31-48). New York: Psychology Press.

Gould, J.W. (2006). Conducting Scientifically Crafted Child Custody Evaluations (2nd ed.). Sarasota, FL: Professional Resource Press.

Kazdin, A. E. (1979). Unobtrusive measures in behavioral assessment. Journal of Applied Behavior Analysis, 12(4), 713-724.

Martindale, D.A. (2005). Confirmatory bias and confirmatory distortion. In J.R. Flens & Drozd, L. (Eds.), Psychological testing in child custody evaluations (pp. 31-48). New York: Psychology Press.

Quinnell, F.A. & Bow, J. N. (2001). Psychological tests used in child custody evaluations. Behavioral Sciences and the Law, 19(4), 491-501.


Beginning his career as a clinical researcher in the Department of Psychiatry at the UCLA School of Medicine, Dr. Mosk has been a practicing forensic psychologist for over 25 years. He specializes in forensic work product review, child custody and disability evaluations, case consulting & litigation support, and expert witness work. His areas of concentration include psychological assessment and testing.

He can be reached at (847) 686-1350 or at