and Van IJzendoorn studied the tapes, documenting the behaviour of each mother-child pair. In the children’s case, they were interested in signs of externalizing behaviour — talking back, hitting, throwing food and toys, fits of anger, hyperactivity, any outward display of excessive aggression or unchecked energy. For mothers, the focus was a bit different. Less attention was paid to actions, and more to interactions. Bakermans-Kranenburg and Van IJzendoorn wanted to gauge something called maternal sensitivity, a variable that measures how adept mothers are at anticipating, interpreting, and responding to their children’s needs.
Mothers’ actions were rated on the Maternal Sensitivity Scale, a sociological measure comprised of four categories: a mother’s awareness of her child’s signals of needs or wants, her accurate interpretation of those signals, the appropriateness of her response, and how promptly she responded. Mothers who scored on the upper half of the scale — meaning they responded quickly and appropriately to their children’s needs — were placed in the high-sensitivity group, and mothers on the lower half — those who responded slowly or infrequently to their children, or who misinterpreted their needs — were classified low sensitivity.
To acquire the study’s final variable, researchers took DNA samples from each child participating in the study and determined which polymorphism (or variation) of the DRD4 gene they possessed. More specifically, they wanted to know whether or not the children had the infamous 7-repeat allele. Bakermans-Kranenburg, Van IJzendoorn, and their observation team were kept clear of the results of these tests until after they had reviewed the tapes and submitted their appraisals of mother and child behaviour. Sequestering data is critical in these types of studies, where objectivity is a goal that is exceedingly difficult to reach, and the slightest potential for bias in the researcher can poison the entire experiment. If the researchers charged with scoring mother-infant behaviour knew which children had the volatile 7-repeat allele, they could — either consciously or unconsciously — rank them higher on the externalizing behaviour scale in order to prove their hypothesis correct. This problem — called confirmation bias — is a constant hazard to experiments in the social sciences, and researchers must guard against it vigilantly.
When all the data was in place, Bakermans-Kranenburg and Van IJzendoorn put it through rigorous analyses. The results were significant. For children with the 7-repeat allele, mothers’ behaviour greatly influenced their odds of acting out. The 7-repeat children with low-sensitivity mothers (mothers who misinterpreted their children’s needs, or who responded indifferently to them) exhibited a great deal of externalizing behaviour, while 7-repeat children with high-sensitivity mothers (mothers who understood their needs and met them promptly) barely showed any externalizing behaviour at all. They were less than half as likely as 7-repeat children with low-sensitivity mothers to talk back, act out, or show signs of ADHD. Children without the 7-repeat allele fell somewhere in the middle, though they were closer to the well-behaved 7-repeat high-sensitivity children than the defiant 7-repeat low-sensitivity children. Their mothers’ level of sensitivity didn’t make much difference to their odds of displaying externalizing behaviour.
Bakermans-Kranenburg, M.J., and Van IJzendoorn, M.H. (2006). “Gene–Environment Interaction of the Dopamine D4 Receptor (DRD4) and Observed Maternal Insensitivity Predicting Externalizing Behavior in Preschoolers.” Developmental Psychobiology, 48(5), 406–409.
Keep in mind that the women in the low-sensitivity group were not necessarily bad mothers. The study did not recruit parents accused of child abuse or neglect, nor did it focus on families from groups generally considered to be high-risk (impoverished, uneducated, suffering from alcoholism or drug addiction, prone to violence, etc.). Low-sensitivity mothers didn’t leave their children home alone for hours on end to go party, or drink themselves sick, or bully their children with taunts or smacks or insults. Some of them were absent-minded. Many of them were overworked and exhausted and didn’t have the energy or patience required to meet their children’s every immediate need, but loved their children all the same and wanted nothing but the best for them. The vast majority were probably doing their best, and would be horrified at the thought that their actions might be hurting their children. And yet, their children were over twice as likely to display aggressive or violent behaviour as those of high-sensitivity mothers — assuming, of course, the children in question had the 7-repeat allele of the DRD4 gene. That is not a small discrepancy, and it shows that in parenting, the little things matter just as much as the big ones.
Fixing the Problem
Having established a link between genes, maternal sensitivity, and children’s behaviour, Bakermans-Kranenburg and Van IJzendoorn strove to use this knowledge to help those children who were most at risk. Their study showed that children with the 7-repeat allele and low-sensitivity mothers were by far the most likely to display signs of externalizing behaviour. Having just one of the two traits — 7-repeat but a high-sensitivity mother, or a low-sensitivity mother but no 7-repeat — was pretty much as good as having neither. Therefore, it stood to reason that changing one of those two variables in vulnerable children would lower their risk considerably. Genes can’t change. Maternal behaviour can.
The two researchers recruited 157 families with children who exhibited signs of externalizing behaviour, such as aggression, hyperactivity, and hostilely defiant behaviour toward authority figures. The families were randomly assigned to either a test or control group. Families in the test group underwent an intervention program Bakermans-Kranenburg and Van IJzendoorn had developed.
Each family in the program received six home visits from female consultants called interveners. During these visits, the interveners observed how the family interacted, appraised their parenting strategies, and offered tips and suggestions for improved communication between parents and children. Each visit was themed, focusing on different skill sets, such as responding promptly to children’s signals, sharing emotions, and using positive reinforcement. The last two sessions were review, where interveners helped bolster previously learned skills. At the end of the program, parents received brochures filled with tips and exercises on the key issues of the intervention.
Meanwhile, the control group participated in a largely insubstantial program called a dummy intervention. Researchers contacted parents by phone and discussed parenting issues with them. Parents were encouraged to talk about their children’s development, but the “interveners” did no intervening whatsoever; they did not observe the participants’ parenting strategies, nor did they offer any tips on how to more effectively parent children with externalizing behaviour disorders. The entire point of the control was to make parents think they were participating in an intervention program without actually providing one. Bakermans-Kranenburg and Van IJzendoorn met with both groups a year after the program’s completion and noted the change, if any, in children’s externalizing behaviours.
The results were, at first glance, lukewarm. As expected, 7-repeat children benefited from the real intervention, but the effect was marginal; 7-repeats who underwent the intervention program scored an average of 2.5 points lower on the externalizing behaviour scale than 7-repeats in the control group, who received the deliberately ineffective “dummy” intervention. Non-7-repeat children varied even less. Although 2.5 points is enough to be statistically significant, given the respectable sample size of the study, it is hardly earth-shattering. The intervention program seemed like a dud.
Bakermans-Kranenburg and Van IJzendoorn were not to be dissuaded. Pushing the control group aside, they considered their data from a new angle. Perhaps, they thought, merely participating in the program wasn’t enough. A far more influential factor would be whether or not the parents who received this information actually did anything with it, or whether they promptly fell back into old habits after the intervention program concluded.
The two researchers set out to ascertain exactly that. They took parents who had participated in the real intervention, re-observed them interacting
