Children Need to Learn Context to Know When to Stop

CNS 2014 Blog: Q&A with Christopher Chatham

It’s happens to all of us: times when we simply cannot stop ourselves mid-action, whether running a yellow (or red) light, making an inappropriate comment, or reaching for our buzzing phone during dinner. Most of the time, adults can overcome these impulses. Children, however, are notoriously different – most of the time seemingly incapable of curbing impulsive responses. New research suggests the problem is not an inability to stop but rather that they lack awareness of their environment.

Our ability to proactively monitor our surroundings is what allows us to stop mid-action, says Yuko Munakata of the University of Colorado Boulder, who will presenting new work at the CNS conference on Tuesday. She, along with Nicolas Chevalier of the University of Edinburgh and Christopher Chatham of Brown University, tested this idea in 7- to 9-year-olds, seeing if monitoring their environments improved their ability to subsequently stop behaviors compared to simply telling them to stop.

Chatham, who will be co-chairing the session at the CNS meeting about response inhibition, talked with CNS about this new research and how it fits into the big picture of differences between children’s and adults’ capabilities.

CNS: “Perseveration” and “abstract goal representation” are key terms in your body of work. Can you please describe these ideas?

Chatham: Perseveration is the tendency to stick with old behaviors even when circumstances have changed and those old behaviors are no longer appropriate. A key mechanism by which perseveration can be overcome is to simply keep in mind the intended goal – and to do so independently of the specific objects currently in the environment, or the specific actions that might be required to achieve that goal. Such a goal would be abstract in the sense that it could drive attention towards any number of more concrete things in the world that are relevant to achieving that goal. We see precisely these kinds of “abstract goal representations” develop by themselves in our computational models, as a seemingly natural by-product of realistic neural architectures undergoing learning.

CNS: In one of your papers, you give the example of a child searching for her toy. What is another example of the kind of behavioral change you are studying and why they are important?

Chatham: Perseverative tendencies like repeatedly checking some empty location for a toy may not, in principle, be wholly unlike the kinds of perseverative tendencies that are defining characteristics of brain damage. This is particularly true of frontal damage, among the worst kinds in terms of quality of life prognoses. So, understanding the origins of perseveration has dire clinical importance. Children are a natural group to study for this purpose because they can be so quickly and reliably induced to perseverate. Children are also highly plastic, meaning they may be sensitive even to weakly efficacious
interventions. This increased sensitivity may give us a better chance to find potential interventions that we might subsequently refine. And there are also much more basic, if long-term motivations: Children are simply incredible learning machines. The process of cognitive development is the process of training the deepest neural network ever known. As much as children learn from us, there may be as much – or even more – we can learn from them.

CNS: In your 2012 review paper in Current Directions in Psychological Science, you lay out three key transitions for children in becoming more flexible in their behavior. At what ages do these transitions occur? Are there clear markers in the brain for each transition?

Chatham: We along with many others have tried to emphasize how many of the seemingly abrupt, stage-like transitions of development could actually result from more continuous and graded neural changes. Take for example the transition from habit-based to goal-directed behavior: While most children are capable of flexibly switching between two simple card-sorting rules by age 4, they will once again fail to switch if the rules involve some further conditionality. Children who succeed in that case may likewise fail when asked to switch between two interpretations of an utterance. Yet even much younger children can correctly answer questions about how they should be behaving, so long as the questions are carefully phrased!

These phenomena complicate stage-like frameworks, and suggest that we consider how incremental neural changes might yield developmental dissociations rather than looking for time points where transitions could be said to be “complete” in a more absolute sense.

As for neural measures, our computational models offer some clear predictions about the neural mechanisms that are at play, but direct neural tests in the critical 0.8-to-8-year-old window are exceedingly challenging to perform. However, there is a growing number of intriguing results that dovetail with core features of our models.

For example, Amy Finn and colleagues have shown that potential neural measures of reactive control might persist even until adolescence. This raises the possibility that the graded changes we’ve emphasized might be even more protracted and continuous across age than anticipated, even if our behavioral measures may lack sufficient sensitivity in adolescence. The labs of Silvia Bunge and Bruce Morton have also reported key markers of immaturity in the corticostriatal circuitry thought to be responsible for “proactive working memory” updating. This, too, could suggest neural markers that accord with the computational mechanisms we and others have proposed underlie difficulty in young children’s transitions to more adult-like abilities in regulating their behavior.

CNS: I as an adult still find myself sometimes driving mindlessly to my child’s school on days even when school is not in session, out of habit. What affects our ability to break free from such habitual behavior as adults?

Chatham: While age-specific factors undoubtedly contribute to behavioral control, there is also surprising continuity across ages. For example, Adele Diamond has shown that even adults can incur subtle reaction time costs in the very same conditions that half of 4-year-olds pass by typical standards.

Of course, adults have at their disposal a much richer variety of strategies and control mechanisms for dealing with these challenges. Adults can learn to distract themselves from inappropriate thoughts or feelings merely by thinking of something else, a strategy many children fail to use. Adults can also engage in proactive behaviors to help them achieve their goals in the future, such as putting one’s keys in an unusual place to trigger subsequent recall of a delayed intention. And, relevant to the symposium topic, adults can also monitor their environment for important contextual cues – a capacity that we think is only nascently present in children and which contributes to their difficulty in regulating impulsive behaviors.

CNS: Can you explain your and others’ new intervention work to teach children to stop mid-action?

Chatham: When driving towards an intersection, even one with a green light, good drivers will stare pretty intently at the traffic light. You do so in case it changes to yellow and you need to adjust your behavior. We think a similar kind of sustained monitoring could be a critical factor in tasks that are said to assess response inhibition and impulsivity. To continue with the metaphor, it’s typically assumed that populations with impulsivity problems have a problem with their “brakes,” but they might just as easily have problems with knowing when the traffic lights have changed color – or watching them intently enough to begin with. Of course, one can easily tell these apart if you just ask people to mash the gas pedal, instead of slam on the brakes, when the lights change.

As we showed in 2012, the analogous laboratory procedure shows that key neurophysiological phenomena commonly interpreted in terms of stopping are in fact more prominent when people had to go, and stopping is not required at all. This surprising result actually converges with recent results from Adam Hampshire and David Sharp, but our newest round of experiments, headed up by Nic Chevalier, really takes it a step farther.

Nic’s study tested a counter-intuitive prediction for intervention that was motivated by these observations. The intervention was tested on a notoriously impulsive population: young children. We reasoned that if monitoring, as opposed to stopping, is the key act of control in these laboratory tasks, then we should be able to improve stopping by training children to monitor for key environmental signals.

Intriguingly, our data show this benefit holds even when the training works at cross purposes: Our training is effective even though it encourages children to commit, rather than stop, rather than stop, an action in response to this key environmental signal! In some ways, that kind of monitoring training may work even better than actually giving children more practice with stopping.

CNS: What are the implications of the work for helping to improve children’s response inhibition?

Chatham: A very foundational implication is that key results long interpreted in terms of stopping must now be turned on their head. By a new view, stopping may actually be a relatively automatic and effortless process, and, in some sense, a mere by-product of being appropriately mindful of environmental change. Instead of stopping, the central role is occupied by the ability to attend vigilantly to features of the world that might demand changes in behavior. In addition to being a richer and more positive view of how behavioral control occurs, this new perspective motivates a redesign of interventions for treating impulsivity not only in children but also patients and other populations with impulsivity problems. Nic’s data confirms such interventions can indeed be effective, particularly when monitoring is encouraged in the midst of ongoing behavior.

CNS: Does your research yet suggest what parents or schools can do to help children respond mid-action?

Chatham: Our research cannot yet motivate specific recommendations for schooling programs, though some work in Yuko’s laboratory is beginning to expand into such broader issues. However, one clear take home is that impulsivity interventions may be particularly effective when they encourage sustained monitoring during action, as opposed to focusing solely on the suppression of unwanted behaviors. We suspect that the often disappointing results from prior interventions for response inhibition might reflect the misplaced focus on stopping per se, as opposed to monitoring.

-Lisa M.P. Munoz

Yuko will present this work at the CNS symposium, “Mechanisms of Response Inhibition,” on Tuesday, April 8, at 3pm.

Follow the meeting on Twitter: @CogNeuroNews, #CNS2014 and read our blog for ongoing coverage.