Psychological Testing in Child Custody Evaluations: An Ethically Informed Approach

Robert A. Simon, Ph.D

Editor's Note: Robert A. Simon, Ph.D. is a forensic psychology consultant with over 30 years of experience. His office is located in San Diego, California.

Child custody evaluations are complex, multi-modal and multi-factorial investigative forensic assessments. When done well, they can be helpful and useful aids to judicial fact-finding and decision-making. When done poorly, child custody evaluations can mislead the Court and have the potential to play a tragic role in ineffective judicial decision-making, thereby potentially harming families. A well-done child custody evaluation depends on the skill of the evaluator with regard to the multiple components of the evaluation process. Given that the conclusions reached by the evaluator should result from the convergence of data gathered in the evaluation process*, it is important that each component be carried out with skill and care. 

Psychological testing is a frequent (but not required) component of custody evaluations. This article presents an ethically informed approach to the use of psychological testing in child custody evaluations and an ethically informed framework for the role of psychological testing in child custody evaluations. While the point of view expressed in this article is not embraced by many child custody evaluators and while this is not a settled issue, this article sets forth the view of the author and is an attempt to encourage open and vigorous discussion of the use of psychological testing in child custody evaluation. 

Quinnell and Bow (2001)* report on the frequency with testing was used and with which various psychological tests are used in child custody evaluations along with the amount of time spent by evaluators conducting, scoring and interpreting tests. They found that approximately 90% of adults and 60% of children are tested. They learned that an average of three hours was spent in testing with adults and two hours with children. Thus, a substantial amount of time (and therefore money) is spent on psychological testing in custody evaluations according to this research. The authors found that the MMPI-2 was the most frequently used test, with 92% of evaluators having used MMPI-2. Next most frequently used was the MCMI II/III, with 52% of evaluators having used this test. These results are not surprising given the ease with which these tests are administered and given the ready availability of machine scoring and computerized interpretation of these tests.*

While the findings of the survey research of Bow and Quinnell are interesting and of importance, these authors do not address how testing is best used in custody evaluations. This topic is the subject of some controversy and has been addressed by Brodzinsky* (1993) as well as Otto, Edens and Barcus (2000)* who discuss individual tests and their applicability to the custody evaluation process. None of these authors make the case that testing is a necessary or essential component. What follows in this article is an approach to the use of psychological tests that relies on ethical principles and practice standards and guidelines with regard to the use of psychological tests in Child custody evaluations.

The Ethical Principles of Psychologists and Code of Conduct of the American Psychological Association (2010)* offers crucial guidance in the use of psychological testing. Standard 9.02 reads in pertinent part:

9.02 Use of Assessments

(b) Psychologists use assessment instruments whose validity and reliability have been established for use with members of the population tested. When such validity or reliability has not been established, psychologists describe the strengths and limitations of test results and interpretation.

The Model Standards of the AFCC, Standard 6.4, previously referenced herein, reads in pertinent part:

6.4 Proper Use of Assessment Instruments
(b) Evaluators shall not use instruments for purposes other than those for which they have been previously validated. Evaluators shall be mindful of cultural and language diversity and the impact that these may have on test performance and the resultant data

The Specialty Guidelines for Forensic Psychologists of the American Psychological Association (2011)* states in pertinent part:

10.02 Selection and Use of Assessment Procedures

When the validity of an assessment technique has not been established in the forensic context or setting in which it is being used, the forensic practitioner seeks to describe the strengths and limitations of any test results and explain the extrapolation of these data to the forensic context Because of the many differences between forensic and therapeutic  contexts, forensic practitioners consider and seek to make known that some examination  results may warrant substantially different interpretation when administered in forensic  contexts.

It is fair to say that evaluators and judicial offers take seriously their responsibility in making recommendations and findings and orders that impact the lives of children.  It is in this context that psychological testing has the potential to be utilized improperly and in a manner that can ultimately be misleading. This is because psychological testing is not determinative or dispositive. It merely raises hypotheses. When evaluators and judicial officers use it in a determinative/dispositive manner, the likelihood of error is significant.

Psychological tests are not the precise and exact instruments many think they are–the many including some attorneys, some judicial officers and, yes, some mental health professionals. Psychological tests measure psychological constructs or concepts. Before a test can truly be thought of as being fully accurate, experts have to agree as to what the construct is. Such consensus is often difficult to come by. One can measure physical quantities such as weight or distance precisely. Measuring less objective and concrete psychological concepts and constructs has a degree of measurement error contained within it. Also, psychological tests are constructed differently and different tests utilize different response modalities. Some are paper and pencil and are self-administered. Some are administered by an examiner with responses are given verbally. Some are true/false. Some are multiple choice. Different types of tests and different modalities for presenting the stimulus or questions exist. Different ways of responding exist. In other words, not all tests are alike and the differences introduce something psychologists call “noise” or “error”. 

Psychological tests are probabilistic. They are not absolutely predictive but, instead, suggest the likelihood of something existing or not existing. Psychological tests must be “normed” or “standardized”.  To do this, the test developer chooses a reference population or populations to standardize the test on. This means that to the extent that the test is valid, it is valid for the reference population only. Tests are also developed for particular purposes. For example, a test might be developed to assess general personality functioning of adults between 18 and 30 who are seeking outpatient psychotherapy. If that is the normative population, then the test results are limited in their validity to that specific population. 

In the domain of child custody, there are no tests on the market that are psychometrically valid and reliable, that were designed and intended to be used to assist in making the kinds of distinctions and decisions that are required in a child custody evaluation and that were normed on a population of child custody litigants. Looking back at the Standards and Guidelines cited above, what emerges is that the use of psychological testing in a dispositive manner in custody evaluations is ill advised and not supported by the Standards and Guidelines that inform child custody evaluations.

To put it all together, tests are probabilistic in nature. There is always uncertainty, “noise” and measurement error with psychological tests. Also, tests are best used on populations for which they are normed and for the purposes for which they are developed. We don’t have any tests for use in custody evaluation that are both reliable and valid that are normed on child custody litigants. So, why are tests so frequently used in custody evaluations? It is my sense that this is done in part because evaluator’s frequently believe that testing is expected in evaluations even though there are no standards or guidelines that require testing. Also, because evaluators as well as attorneys and judicial officers want as much certainty as possible and because evaluators often lean on the (false) precision of testing to guide their conclusions, tests are included. Also, psychological testing and the domain of test development is something “owned” by the profession of psychology. Thus, psychologists use testing, in part, because it is a part of their identity as a profession. 

That said, psychological testing can have a useful place in child custody evaluations. Because of the ways in which testing can be misleading and the ways in which it may lead to a false sense of precision and certainty, it appears wise to use a substantial measure of caution when using psychological testing results in custody evaluation. Specifically, it is my position that testing is best used to generate hypotheses (Stahl and Simon, 2013)*.  Remembering that one of the key ways in which forensic psychological assessment proceeds is by hypothesis generation (see, for example, Gould & Martindale (2007)* and that hypothesis testing is also a fundamental way of controlling for bias in custody evaluation (see for example, Stahl and Simon (2013)), the results of psychological testing can be extremely helpful in generating hypotheses to be evaluated using other data gathered during the evaluation process. This line of reasoning means that testing should not be used to reach conclusions or findings and that test results should also not be used as part of converging data that leads to conclusions or findings.  Instead, psychological testing results can be wisely, safely and properly used if the use is strictly limited to hypothesis generation.

*Martindale, D. A., Martin, L., Austin, W. G., Drozd, L., Gould‐Saltman, D., Kirkpatrick, H. D. & Stahl, P. M. (2007). Model standards of practice for child custody evaluation. Family Court Review, 45(1), 70-91.
* Quinnell, F. A., & Bow, J. N. (2001). Psychological tests used in child custody evaluations. Behavioral sciences & the law, 19(4), 491-501.
* Despite the ready availability of computerized interpretation of tests, the AFCC Model Standards of Practice for Child Custody Evaluation advises great caution in the use of computer generated test interpretations.  This author echoes this caution and takes the position that computer generated interpretations are NOT best practice—the reason being that an individual should only use tests that they are competent to interpret themselves.  If an evaluator is competent to interpret tests, a computerized interpretive report becomes unnecessary. 
* Brodzinsky, D. M. (1993). On the use and misuse of psychological testing in child custody evaluations. Professional Psychology: Research and Practice,24(2), 213
* Otto, R. K., Edens, J. F., & Barcus, E. H. (2000). The use of psychological testing in child custody evaluations. Family Court Review, 38(3), 312-340.
* Stahl, Philip M. and Simon, Robert A. *(2013)  Forensic Psychology Consulting in Child Custody Litigation.  Chicago:  American Bar Association.
* Gould, J. W., & Martindale, D. A. (2007). The art and science of child custody evaluations. Guilford Press.

Family Law on facebook LACBA twitter LinkedIn family law section homepage link