A body of research in developmental psychology has built up gradually around an uncomfortable observation: children begin forming a coherent sense of themselves, as capable or incapable, likable or difficult, confident or tentative, in the first years of life, and the people raising them play a significant role in what that picture looks like. By age four, the evidence suggests, the picture has already started to take shape, with parental behavior and parental beliefs leaving measurable marks on how children understand themselves.
The implication that follows is one the research does not state directly, but toward which the evidence points: most people’s intuitions about what they are good at, what they are suited for, and what they should want from life have roots in a period they cannot remember, shaped partly by other people’s expectations at an age when parental influence and a child’s own developing sense of self are both already active.
This is worth approaching carefully. It is easy to overclaim, and also easy to dismiss.
What the research on early self-concept shows
A 2009 study by Geoffrey Brown, Sarah Mangelsdorf, Cynthia Neff, Sarah Schoppe-Sullivan, and Cynthia Frosch, published in the Merrill-Palmer Quarterly, followed fifty families across the ages of three to four. At age three, the researchers observed child temperament, mothers’ and fathers’ parenting behavior, and the quality of the broader family interaction in the home, including how all three family members interacted together. At age four, children’s self-concepts were assessed using a picture-based tool established for use with young children.
The study found that parenting behavior shaped self-concept in two distinct ways. When mothers and fathers engaged in certain parenting behaviors, the child’s temperament appeared to amplify or dampen aspects of their self-reported self-concept, such that a temperamentally distress-prone child looked quite different depending on how the parents engaged with that distress. More strikingly, the quality of how the family interacted as a unit made independent contributions to children’s self-concepts, separate from temperament and separate from individual parenting behavior. This is a small study, fifty families, and its findings should be treated as suggestive rather than definitive. But it was longitudinal, it tracked development across the precise window in question, and it used direct observation of family behavior rather than self-report. The finding that parenting shapes self-concept this early is not, on its own, a settled conclusion. It is one careful look at a process that appears to begin before most people expect it to.
The parental expectations piece
Parenting behavior and parental expectations are related but not the same thing. A parent can be warm and consistent in their interactions while still holding limited expectations of what a child will become, and vice versa. The research on parental expectations specifically, as opposed to parenting behavior more broadly, has primarily focused on academic domains and tended to follow slightly older children. But it reaches back earlier than is commonly appreciated.
A 2014 study by Daniel Briley, K. Paige Harden, and Elliot Tucker-Drob, published in Developmental Psychology, drew on data from the Early Childhood Longitudinal Study, which followed large, nationally representative samples of American children born in 2001 and in the kindergarten cohort of 1998. Using both twin data and longitudinal cross-lagged models, the researchers found evidence of bidirectional transactions between parental educational expectations and child characteristics, including academic behaviors and learning approaches, that were detectable even before children entered school. The age four wave of the Early Childhood Longitudinal Study fell within this window. Parental expectations partly reflected genetically influenced characteristics of the child, and child characteristics in turn influenced what parents came to expect. The relationship was transactional from the start, with the influence running in both directions rather than simply from parent to child.
What Jacquelynne Eccles and her collaborators have traced across decades of longitudinal research is what happens next: how parental beliefs about what a child is capable of get taken up by the child as beliefs about what they themselves can do. That research, rooted in the expectancy-value framework Eccles developed across multiple studies, consistently shows that parental ability beliefs predict children’s own ability self-concepts, and that this relationship is measurable from the earliest years of formal schooling. A 1998 analysis by Pamela Frome and Eccles, published in the Journal of Personality and Social Psychology, found that parents’ perceptions of their children’s abilities exerted a stronger influence on children’s own self-perceptions in sixth grade than the children’s actual grades did.
Parents, in other words, are not simply responding to what they observe their children doing. They are partly constructing the lens through which children come to interpret what they are doing.
The timing is the uncomfortable part
What makes this literature difficult to sit with is not the general finding, which seems intuitive enough once stated. Of course parents shape how children see themselves. The difficulty is the timing.
By four, children have no independent basis for evaluating their own capabilities from the outside. They have not yet failed at enough things to develop calibrated self-assessments, or succeeded at enough to know what genuine competence feels like in a domain. What they have, instead, are the expectations that the adults around them carry into ordinary interactions: how long a parent stays engaged when a task gets hard; whether difficulty is treated as evidence of incapacity or as something to push through; which activities a parent pursues with a child and which ones quietly never come up again.
None of this is necessarily deliberate. The communication of expectations is mostly implicit, happening in the texture of how parents interact with children rather than in anything they explicitly say. That is part of what makes the research unsettling: the influence is not reducible to parenting choices that could simply be made better. It is embedded in the ordinary patterns of daily life that accumulate before either parent or child is paying attention to them.
By the time children can articulate what they want, they are already working within a self-concept that has been forming for years. The preferences they express are not constructed from scratch. They are filtered through a view of themselves that someone else helped to build.
What the research does not say
None of this means that a self-concept formed in early childhood is fixed. Self-concept shifts throughout development, in response to new experiences, new relationships, new evidence about what one is actually capable of. Teachers, peers, mentors, and unexpected successes in new domains can all revise what a child, or an adult, believes about themselves. The research is not a counsel of determinism.
The Eccles framework explicitly includes teachers’ beliefs alongside parents’, and recognizes that the relationship between parental beliefs and children’s self-concepts runs in both directions: children who succeed are more likely to develop positive self-concepts, which shape future parental and teacher beliefs in return. The Brown et al. study found that temperament and parenting behavior interact, that neither alone produces a fixed outcome. The Briley et al. study’s central contribution was precisely demonstrating that children influence parental expectations just as parental expectations influence children.
These results do not support a picture of parents simply writing a self-concept into a passive child. They support a more complicated and genuinely bidirectional picture in which parental expectations are one of several early inputs into a developing sense of self, and among the earliest that leave measurable traces.
An article can name a pattern. It cannot explain any one family’s life, and a body of research on group-level averages cannot tell a parent what their specific child’s self-concept is or what shaped it.
What remains worth noticing
What this research pushes back against is a comfortable assumption: that preferences are discovered rather than partly received, that what people want from their lives reflects something freely arrived at rather than something shaped by the circumstances of their earliest years. The evidence suggests the picture is more entangled than that. The view a child holds of themselves at five or six carries traces of what the people raising them believed about them at three or four, and that view shapes the questions a child learns to ask about themselves for a long time afterward.
That is not a settled conclusion about mechanism or magnitude. It is a finding from a field still working out how these early processes unfold, across what timescales, and how much they can be revised by later experience. What the research supports is more limited than a sweeping account of parental determination: the influences are real, they begin early, and they operate mostly below the level of conscious choice, on both sides. The rest is still being worked out.
We are writers and parents, not developmental psychologists or child development specialists. This article is a reading of the available research, not clinical guidance. Parents with specific concerns about their child’s development are best served by a pediatrician or developmental specialist.
