Poverty affects nearly one in seven American children — about 10.5 million kids in total — and it has a long reach. Children living in poverty experience more stressors that can hinder healthy emotional and cognitive development and contribute to well-documented disparities in education, employment, and physical and mental health.

Researchers studying how poverty and adversity affect children’s development often track how negative experiences — be they poverty itself or factors such as having an incarcerated parent — affect decision-making, stress levels or aspects of brain function. But Seth Pollak, a psychologist at the University of Wisconsin–Madison, says that most of these efforts miss a crucial but long-overlooked component: children’s perceptions of their experiences.

Pollak studies how early experiences, especially negative ones, shape how kids think and feel about the world, and alter the brain circuitry in charge of those processes. In a 2020 article in Annual Review of Developmental Psychology, he describes a strategy for studying how young children learn to understand emotions. He argues that if you want to understand the effects of adversity, it’s time to move beyond looking at “objective” measures of kids’ environments, and instead take a more anthropological approach by asking children how they felt about their experiences. After all, he says, “something isn’t ‘adverse’ until a person experiencing it decides it’s adverse.”

Pollak spoke with Knowable Magazine about the importance of studying individual differences in experience. This conversation has been edited for length and clarity.

What led you to study how adversity and poverty affect children?

In the 1980s, when I was in graduate school, psychologists were just starting to explore the idea that experience affects brain development, and research on brain plasticity was coming into vogue. As an undergraduate, I had studied anthropology, where the whole discipline is about how the experiences we’re immersed in affect an individual. So I was conceptually ready for what ended up becoming a big thing in the field.

At the time, I wasn’t particularly interested in adversity, but my mentor had a set of data from kids who had experienced severe abuse and neglect. For me, it was an opportunity to look at how atypical experiences influence people.

How have researchers traditionally studied how poverty and adversity affect kids?

Early on, research on child adversity was very behaviorally oriented: things like observing kids on the playground or focusing on the personality characteristics of the parents. There was really no connection to biology. Also, research methods were not very sophisticated — administering parent or teacher questionnaires, or watching kids play — and, in fact, the field did not have tremendous respect within the biobehavioral sciences.

But then, as neuroscience started infusing it, the field got really hot. It became a way to look at nature versus nurture, a potential model for how a biological system interacts with an experience like child abuse. People were excited because it was sort of a double whammy. You had a great way to ask basic science questions about human development and how people become the way they are. And there also were practical, clinical and applied educational implications, because this was a public health problem. The idea was, if we can figure out the mechanism, we may find a target for intervention in kids that have had these terrible experiences.

How well has that worked out?

The problem is that researchers started embracing neuroscience-based methods for studying the human brain and its functioning, but they stuck with the old style of questionnaires from the ’80s and ’90s. It was like we got all this specificity with neuroscience, but we were using these really blunt, unsophisticated instruments that put children into adversity categories based on checklists measuring how many bad things have happened in their life, whether their mom went to college, or what the family income was.

There are thousands of papers reporting that “adversity” affects the brain, but the effects they find are all over the place because of these sloppy, imprecise, fuzzy ways of measuring children’s experiences. It’s unlikely that a broad variable like whether your mother went to college is going to be specifically tied to a neural system.

A photograph of children playing on the grass.

A happy, carefree childhood or a stressful, disruptive one? Objective measures of hardship don’t capture the individual child’s truth, making it hard for researchers to determine the long-term effects of adversity.


How might the field move beyond that?

Much of the research is modeled on classic studies in neuroscience showing that, for example, blocking vision in one eye early in life affects the visual cortex of kittens. But there’s a big difference between that kind of very precise study and asking, in a broad way, what a child’s parents were like and expecting to see a measurable effect in the brain.

But also, there is a lot of individual variability in how different kids respond to adverse circumstances, and these methods do not capture it. From animal studies, we know that when two mice encounter a predator, one can have a fight response and the other can have a flight response. Brain imaging in those two animals will show very different patterns of activation, yet they’re both encountering the same event.

The same goes for people. To understand how a life event affects someone’s brain, we have to dig into how that individual perceives it, not tabulate experiences that researchers have decided are stressful. 

I think the challenge here is that we’re asking scientists to move away from the instruments and approaches that they’re familiar with and that are accessible and easy to use. Most people in the field agree with the point that I’m making, but many get hung up on not knowing what they should do instead. And I understand that. A commentary I coauthored last year describing the need for this shift was the hardest paper I’ve ever had to write. It was hard to come out in print and say, after 25 years of work, that we have all been doing it wrong. 

Are you starting to do some of this work?

Yes. What we have started doing is getting children’s reports of their experiences and then waiting until the kids get older to see how they’re functioning. For a recent study, we interviewed elementary school children who had a range of life experiences relating to stress and adversity. Then, 10 years later, when they were young adults, we re-interviewed them and measured their brain-activity patterns during a gambling game in which they had to assess risk and make decisions in the lab.

It turned out that the stress in people’s current adult lives didn’t affect the brain mechanisms involved in decision-making that we measured. But the stress that they experienced when they were children did. The people who reported that they had very stressful childhoods were not attending to cues that were telling them they were about to lose a lot of money — warning signs that suggested they should change their behavior. And these observed brain responses to risk also correlated with what they reported about daily activities — like whether they wore a seat belt or a helmet.

Two brain scans with a small region colored blue on the left and another small region colored yellow on the right.

Early life stress may have a lasting influence on brain circuits involved in decision-making. In this research, investigators scanned the brains of young adults as they played a computer gambling game. Those who’d experienced high levels of stress as children had less activation in brain areas associated with risk assessment (blue) when a cue indicated they were about to lose money, compared to young adults who hadn’t experienced as much early life stress. After losing the money, the high-early-stress group had stronger activation (yellow) in brain areas related to feelings about the money loss.

These kids were from really high-stress backgrounds. There were kids whose parents had been shot through the window or were in jail, and there were kids who experienced horrific physical abuse and had been placed in foster care. I remember one little girl, who was about 8, scored very high on the stress measure because her family had recently moved. What was fascinating was, her family had stayed on the same street but just moved to a bigger place two blocks down. Same school, same friends, same neighborhood, but for her, that was a huge stressor.

There was another child in the study, a boy around 9, whose mom had been jailed for selling drugs and prostitution, and his grandmother had recently moved in. It certainly disrupted his life, but it wasn’t affecting his thinking. He did not report himself as being stressed.

If we objectively assess moving two blocks down on the same street, we might say that’s a low-stress event. Mother incarcerated — that’s a high-stress event. But when you look at these people’s view of the world, that’s not how they experienced it.

How can you tell whether what a child says truly reflects how events or circumstances affect them?

You ask them. It sounds simple, but people don’t do it. You say, “How much were you really upset about this? Is this really bothering you? Do you think about it a lot? Do you consider this a really big thing in your life?” Kids can engage with that. Some might say, “Yeah, I’m crying all the time and it really bothers me,” and others are, like, “No, I’m actually more upset about what I got on my math test this week.”

I think of it almost like ethnography — a return to my anthropology roots. For this kind of analysis, you want to hear people’s stories, and your data are not so-called objective measures, but instead, it’s what people are choosing to tell you about what mattered to them.


Are there other recent studies, from other labs, that show the value of this approach?

Yes. In one, researchers asked a group of low-income tweens and teens how supportive and responsive their parents were. Then the researchers used brain imaging to measure their brains’ activity in young adulthood. Kids who had parents they perceived as being very supportive had really normal-looking brain circuitry when they grew up. But in the kids who did not perceive their parents as engaged and supportive, brain circuits that regulate cognition, emotion and behavior were less well-connected. Traditionally, we think of comparing kids living in poverty to a comparison group of children not living in poverty. Here, the two groups experienced the same economic circumstances but differed based on the child’s own perspective.

Another study I love looked at identical twins in families with low socioeconomic status, determined by their parents’ income, education and occupation. They asked the young adults, all sets of twins, about their families’ standing in the community. The individuals who viewed their own families as having low socioeconomic status (even though all the families were in the same boat) had a much higher risk of drug abuse and mental health problems. But their identical twins — the siblings of these individuals — who didn’t view their families as being that poorly off were doing fine and had much better outcomes in early adulthood.

It’s fascinating that two kids in the same family who see their situations differently can have such different outcomes.

How might this new approach yield interventions for children who experience adversity?

That’s a great question. I don’t think, in child adversity, that we’ve cleared enough of the brush away yet to know what underlying brain systems are affected. Most interventions for stress-exposed kids are like the equivalent of Tylenol — they are very general and they focus on decreasing symptoms, because we do not yet understand what specific system has broken down.

To come up with more targeted interventions — for example, a drug or behavioral therapy directed specifically at say, motivation or decision-making — we need to first find the right way of categorizing adverse experiences. Currently, we do this based on events in children’s lives — physical abuse, neglect, witnessing domestic violence, experiencing parental loss. But I don’t think that’s going to map onto any particular brain circuitry.

Instead, I think looking at dimensions that capture how people understand their own experiences will be more fruitful in elucidating how their biology is responding to what has happened in their life. We might focus on factors that influence how people interpret events in their lives, such as the degree of social support a child receives (if something bad happened, were others around to support and comfort you?) or the degree of predictability in the child’s environment (were bad things anticipated, or did events feel random or surprising?).

These factors are more likely to tie into the brain circuitry relating to things like attention problems, stress regulation or substance abuse. So we will be able to match interventions to the underlying biology.