Parental Alienation Science Stumbles Along

 

Over the last few years, a major goal of proponents of the parental alienation belief system seems to have been to amass a series of weakly-designed and implemented “scientific investigations” to allow the proponents to argue that there is scientific support for their views. This preference for quantity over quality is understandable, as serious empirical investigations are time-consuming and costly in planning and resources, and family courts are more impressed by hearing about dozens of studies than interested in understanding how meaningful those studies are.

Amy Baker and William Bernet have for some years been the PA principal investigators of note. Their work has featured absent or questionable control groups, retrospective self-report information in many cases, and an insistence on forcing ordinal data into Procrustean parametric analyses.

Now the Colorado psychologist Jennifer Harman is taking the lead in weak studies, whose publication allows her to assert repeatedly that children involved in ambiguously-defined parental alienation cases are victims of family violence. She has recently added to her assertions by declaring that rejected parents are the objects of coercive control by the ex-spouse.

I will briefly describe and comment on two recent Harman publications.

Harman, J., Saunders, L., & Afifi, T. (2021). Evaluation of the Turning Points for Families (TPFF) program for severely alienated children. Journal of Family Therapy, DOI: 10.1111/1467-6427.123666. (Please note, this publication is not the same as American Journal of Family Therapy, which also publishes pro-PA material.)

Harman et al. collected data about TPFF, an intensive program run by the LCSW Linda Gottlieb. Like other PA interventions, TPFF requires that children be court-ordered into the program or that an agreement between the parents be approved by the court. Like other PA treatments, TPFF lasts about 4 days, involves prohibition of contact between child and preferred parent, for at least 90 days, and requires “aftercare” for both parents and child or children, performed by a a PA-approved therapist. TPFF also requires the preferred parent to write a letter to the child acknowledging his or her attempts at alienation. Gottlieb videotapes therapy sessions, and the videotaped material was made available to Harman et al. Gottlieb, incidentally, claims almost 100% success in creating positive relationships between children and previously rejected parents; as this rate of success has been questioned, it was desirable from Gottlieb’s viewpoint to have a positive report from Harman..

Harman was interested in evidence that TPFF was safe for children, as I and other authors have questioned whether this is so. On the grounds that none of the observed 32 children from 15 families ran away or carried out any self-harm during the program, and that Gottlieb states that none of the previously-treated children with whom she keeps in touch have done so, Harman concluded that TPFF was safe for children. She did not report on other possible aspects of harm, for example the PTSD diagnosed in a child following another PA intervention. Harman stated her belief that children who have reported distress and harm from PA interventions were simply still alienated, and would not have made such reports if their alienation had been repaired.

Harman also sought to know whether TPFF was an effective treatment for PA cases. To do so appropriately, of course, she would have needed to compare a group of TPFF-treated children with a matched control group who received no treatment or some other form of treatment. To compare behavior before and after treatment (as of course has also been done by Richard Warshak and other PA proponents) is to beg for confounding of variables so that it is impossible to know whether any changes were actually caused by the intervention.  Reports following Family Bridges experience point to threats from therapists as reasons for behavior change, but it is unknown whether such threats occur during TPFF.

Like some other PA proponents, Harman asked her staff to evaluate child behaviors on a Likert scale, rating behaviors from 1 to 5. Again like other PA proponents, Harman took these evaluative data and performed a statistical analysis which would only have been acceptable if the measurement method met certain criteria which were not met. The conclusion that TPFF is an effective treatment thus remains open to question.

There are many concerns about the basic data as well as about the data analysis. The videotapes evaluated by Harman’s students were made by Gottlieb and certainly in her presence (it is not clear whether other people than parent and child were in the room at the time). This is far from an independent set of data, as Gottlieb’s presence signaled to both parents and children what behaviors were desired, and indeed necessary to bring the program to an end. The correct method would have been to have the recording done by a person who was not aware of the purpose of TPFF and who observed  parent and child without Gottlieb being present. Without wishing to stress that videorecordings can be cherry-picked, I would also note that the recordings should have stayed in the possession of the neutral recorder until handed over for analysis.

The research report by Harman, Saunders, and Afifi thus adds to the collection of weak studies of PA interventions, but in fact does not allow any clear conclusions about the safety or effectiveness of TPFF.

A second recent publication by Harman is this:

Harman, J., Maniotes, C, & Grubb, C. (2021). Power dynamics in families affected by parental alienation. Personal Relationship. (this journal is unfamiliar to me and the pdf I downloaded did not contain any DOI information.)

In this study, Harman and her colleagues approached special interest social media groups like Facebook divorce or parental alienation groups. They provided a survey and asked for emails from people who would be willing to be interviewed, and interviewed 50 fathers and 29 mothers about their experiences with their ex-spouses and children following divorce. The point of the study was to examine how these experiences fit the concept of coercive control and how power dynamics were related to parental alienation—interestingly, the very features that protective parents have been attributing to parents now rejected by their children. Harman et al. appear to have collected information closely related to parental alienation concepts, for example, that “in some cases the adultification [of children] took the form of allowing the children to decide whether they wanted to have their parenting time with the targeted [sic] parent or by sharing inappropriate information with the children”, statements frequently found in discussions by PA proponents.

It would seem that reviewers or others must have queried Harman’s personal commitment to PA ideas and her ability to be objective on the topic. Harman stated that “the first author contends that her experience as an alienated step-parent provides a unique perspective to the study and has helped to gain the trust and confidence of the parents that were interviewed, as many were afraid or concerned about their experiences not being believed.” She noted also that the second and third authors were included because they did not  have PA experience and could provide “more objective interpretation of the data”.

There are a number of concerning issues here. The first is that interviewees were sought from organized groups who were likely already to share certain views of post-divorce events, particularly views of parental alienation as they felt it had negatively affected their lives”.  This is a matter not so much of preaching to the choir but of being preached to by a choir that has memorized the hymns for the season. Members of social media groups are likely to share beliefs both before they join (this is why they join), and afterwards, when they have thoroughly informed each other of their opinions and experiences. One would imagine that anyone who planned to use regression methods would want to include a number of “nonbelievers” to show a comparison to a different power dynamic among them, but this did not happen.

That Harman thought her personal experiences would set interviewees at ease is worrisome, as it suggest that she told the interviewees about her own life, thus introducing various types of bias in the forms of the wish to please her,  social conformity to the standards she supplied about views of post-divorce relationships, and increased memory and reporting of events that could be interpreted as PA. Harman’s statement that the second and third authors would be more objective in interpretation of data seems to be an acknowledgment that she herself would not be objective—although it is difficult to know how one can be subjective in reporting statistical results.

Once again, Harman and her colleagues have added yet another questionable study to the trove already provided in courts of law as “evidence” to support PA concepts regarding identification and treatment of a posited disorder.