February 23, 2014

Child custody lore: The case of the runaway woozle

The bond between infant and mother is the bedrock of healthy child development When parents divorce, shared custody arrangements destabilize this primary attachment bond, leading to serious developmental problems in children. In general, mothers should maintain sole physical custody of children up until the age of four.

This is the consensus of a growing body of research. 

Or is it?

Do you remember when Winnie the Pooh and his friends became obsessed with the fear that a dread woozle was stalking them in the woods, only to realize that they were seeing their own footsteps? In science, a woozle is much the same. It’s a belief or claim that gains traction due to repeated citation, despite its lack of empirical support. Often, it’s an idea that appeals to members of the news media, politicians or the general public, because it fits with conventional wisdom or is politically expedient.

In the realm of child custody policy, the idea that shared parenting is a bad thing is a behemoth woozle that’s been trotting around the globe virtually unchallenged of late, according to Linda Nielsen, a professor of education at Wake Forest University in North Carolina and an outspoken proponent of shared parenting.

Nielsen’s case, methodically argued in the current issue of Psychology, Public Policy and Law, is pretty convincing. When you peel back the layers of the onion, the much-touted “body of research” about the dangers of shared parenting plans for infants and toddlers consists primarily of one severely flawed study. It's been so oft-repeated in academic journals, news media reports, and legislative hearing rooms that it has gained an aura of ultimate truth even as contradictory evidence from other studies -- finding either no ill effects or even developmental benefits to shared custody -- has been side-railed.

The much-ballyhooed study, by clinical psychologist Jennifer McIntosh and colleagues, was part of a report commissioned by the Attorney General’s office in Australia in 2010. According to Nielsen, the so-called “preschooler study” was driven by outmoded theoretical assumptions about mother-infant attachment that are not supported by recent empirical studies. McIntosh leans heavily on the work of neuroscientist Allan Schore. She quotes Schore as claiming that small children do best when they have only one primary caregiver tending to their bedtime routines, and that women’s brains are more neurologically equipped than men's for communicating with and forming attachments to infants.

Nielsen starts by pointing out some obvious problems (which others in the field have noted) with generalizing from the Australian preschooler study. Sample sizes were small, the majority of the parents had never been married to each other (one-third hadn’t even lived together), etcetera.

But it’s when she drills down into the intimate details of the study’s procedures that things get interesting. The woozle’s claim that “overnighting” (spending nights with their father) is bad for children rests on four negative findings – increased watchfulness; irritability, persistent gazing, and frequent wheezing. Yet, the methods used to measure these constructs were novel, and lacked any established reliability or validity. To take just two examples:
  • Wariness/ watchfulness in the mother’s presence: The rationale for measuring watchfulness was that some attachment theorists believe it to be a sign of insecurity and anxiety. The researchers created a “visual monitoring scale” by cribbing three items from a longer instrument. Mothers were asked how often the infant: (1) looked at her to see if she was watching, (2) tried to get her attention when she was being inattentive, and (3) tried to get her to notice or look at interesting objects. The researchers concluded that the infants in the frequent overnighting group – who scored higher on this novel scale -- were exhibiting signs of stress. This is problematic on its face, since this scale has not been established as a reliable or valid measure of insecurity, anxiety, stress, or attachment. But, more fundamentally – and quite ironically – the Communication and Symbolic Behavior Scales from which the items were drawn is intended to assess infants’ communication skills and language readiness. Thus, high scores on these items are interpreted as positive rather than negative -- indicating an infant is more developmentally advanced and poised to begin talking. Hardly evidence of impaired attachment and the perils of shared parenting. 
  • Wheezing: The researchers proposed wheezing as a sign of stress due to a “negative emotional environment.” They measured wheezing by asking the mother one yes-or-no question: “Does your child wheeze at night more than four times a week?” Setting aside the fact that single-question instruments are known to be unreliable, the researchers went with the a priori assumption that wheezing is a psychosomatic symptom, ignoring significant evidence of alternate causes. Wheezing, indeed, has well established genetic, physiological and environmental components “having nothing to do with stress or family dynamics,” Nielsen points out. These include low parental income and – logically enough -- exposure to pollutants, cigarette smoke, pets, cockroaches, mildew and the like. According to Nielsen, none of the three studies that the authors cite to support their hypothesized link between attachment stress and wheezing do in fact support such a conclusion.
Despite these and other flaws as meticulously deconstructed by Nielsen, the preschooler study has been enormously influential in professional organizations, legal settings, and public policy discourse around the globe. From Australia to the United Kingdom to Israel and the United States, “it has been cited as grounds to set limits on shared parenting, to the exclusion of almost all of the other studies that have examined outcomes for children in shared parenting families."

Beyond the issue of child custody law and policy, this article is a great teaching tool applicable to other areas of psychology-law in its illustration of how social science data can “woozle” academics and the general public alike into swallowing things that are not true.

Next up, I’m hoping someone will conduct a similar scholarly analysis of the perplexing problem of woozles’ kinfolk, the Heffalumps of Winnie the Pooh's psychedelic nightmare.

* * * * *

The article, “Woozles: Their Role in Custody Law Reform, Parenting Plans, and Family Court,” can be requested from the author (HERE).

 Hat tip: Mark Worthen, PsyD