Gender · three theories, not oneHow kids build a gendered self
Few developmental accomplishments are as conspicuous, or as early, as a child's arrival at a gendered sense of self.
Long before they can define the word, most children sort the social world into "boys" and "girls," announce which one they are, and begin policing the boundary with the fervor of a small border guard. The rough timetable is well established. Around age two to three children reach gender identity — the ability to label their own and others' gender correctly. By roughly age four comes gender stability, the recognition that gender endures over time (a boy grows into a man, not a woman). And around age six or seven children attain gender constancy, understanding that gender is not altered by superficial changes in hair, clothing, or activity (Kohlberg, 1966; Slaby & Frey, 1975). Rigid, almost caricatured gender-typing — "girls can't be firefighters," "pink is a girl color" — peaks in the preschool years and softens with age, a curve that any adequate theory of gender has to explain.
What is not controversial is that gender develops; what is fiercely contested is how. Three classic accounts compete over the mechanism, and each was built partly to repair a weakness in the one before it. The instinct to crown a single winner is worth resisting. The three theories answer different questions — about the source of gendered behavior, about its timing, and about how active the child is in producing it — and the modern consensus treats them as complementary layers of one self-socializing system rather than rival hypotheses to be knocked down one against another (Ruble, Martin, & Berenbaum, 2006).
1. Social learning theory (Bandura, Mischel) — children learn gender through reinforcement and modeling. Boys get praised for "boy" behavior, watch same-sex models, and imitate them. Explains why gendered behavior tracks the environment — but treats the child as fairly passive. 2. Cognitive-developmental theory (Kohlberg) — the child is an active thinker who must first achieve gender constancy: understanding that gender is stable over time and situations (a boy in a dress is still a boy) around age 6–7. Only then, Kohlberg argued, does the child seek out same-gender models. 3. Gender schema theory (Bem; Martin & Ruble) — the modern synthesis. Kids build a schema — a mental network of what goes with "boy" and "girl" — as soon as they can label their own gender (age 2–3), well before constancy. The schema then acts as a filter: kids attend to, remember, and prefer schema-consistent information. This is why gender schema theory largely won: it explains the very early rigid gender-typing that Kohlberg's timeline can't.
The three theories of gender development. Tap one to see its core claim.
Aligning with current APA guidance, keep three constructs cleanly separate. Gender identity = your internal sense of being a boy, girl, both, neither, or another gender. Gender expression = how you present it outwardly — clothes, hair, mannerisms — which is culturally coded and varies across time and place. Sexual orientation = who you're romantically or sexually attracted to. These vary independently: a transgender woman may be straight, gay, or bisexual; a boy who loves dresses may grow up to be any orientation and any identity. Conflating them is the most common error in this unit, and the APA's practice guidelines are explicit that they must be distinguished (American Psychological Association, 2015).
Mechanisms · the three accounts up closeLearned, reasoned, or filtered?
The callout above sketches the three theories; the argument between them rewards a closer look, because each names a genuinely different engine of gender development.
The oldest of the three, social learning theory, locates gender in the ordinary machinery of reinforcement and imitation (Bandura, 1977). Children are rewarded — with praise, approval, and the easy social traction of fitting in — for gender-typed behavior and are discouraged, sometimes pointedly, from crossing the line; boys in particular are policed hard for behavior coded as "for girls." Alongside this direct shaping runs observational learning: children watch what same-gender models do and reproduce it, frequently with no reward at all. The account's great strength is that it explains why gendered behavior tracks the surrounding culture so closely and shifts when the culture shifts. Its classic weakness is that it can make the child sound like a passive sponge, absorbing whatever the environment pours in. The revised version, social cognitive theory (Bussey & Bandura, 1999), answers that charge by adding self-regulation: once children internalize gender standards, they begin to guide their own conduct through self-approval and self-criticism, becoming agents of their own socialization rather than mere recipients of it.
The cognitive-developmental account (Kohlberg, 1966) turned the causal arrow around. Rather than gendered behavior being stamped in from outside, the child's own developing understanding of gender is what drives the search for models. On this view the pivotal event is gender constancy: only once children grasp that gender is permanent and unchangeable — typically around six or seven — do they become motivated to behave like, and to attend to, members of their own gender. There is real evidence for the mechanism, in that a child's level of gender constancy predicts how much a preschooler selectively watches same-gender models (Slaby & Frey, 1975). But the theory has an awkward problem with timing. Robust gender-typed preferences appear years before constancy is achieved, which is precisely the evidence the third theory was built to accommodate.
Gender schema theory (Bem, 1981; Martin & Ruble, 2004) kept Kohlberg's active, thinking child but discarded the requirement that constancy come first. A schema is an organized mental network of associations — everything a child has filed under "boy" and everything filed under "girl." As soon as children can reliably label their own gender, around age two or three, they begin building this schema and, crucially, using it. The schema then operates as a perceptual and motivational filter: children attend more closely to, remember better, and prefer objects and activities marked as fitting their own gender, and they will even distort memory to keep it schema-consistent, misremembering a picture of a girl using a hammer as a boy. In Bem's (1981) original studies, strongly sex-typed adults processed information — including information about the self — through the gender schema more readily than less sex-typed peers. This readiness to see the world in gendered terms is why the theory largely displaced Kohlberg's timeline as the dominant account: it explains the very early, rigid gender-typing that a constancy-first model cannot.
None of this makes the earlier theories wrong so much as incomplete. Contemporary developmentalists describe gender development as a self-socializing loop: biological inputs and social modeling supply the raw material, the child's own schema does the organizing, and self-regulation locks the pattern in (Bussey & Bandura, 1999; Ruble et al., 2006). The interesting question is no longer which single mechanism is correct but how the mechanisms feed one another.
Sex & gender · terms and biologyThe biological thread, handled honestly
Underneath the developmental story sits a distinction the theories mostly take for granted and that is worth making explicit. In careful usage, sex refers to the biological categories of male and female, indexed by chromosomes, gonads, hormones, and reproductive anatomy, whereas gender refers to the social roles, behaviors, and identities a culture builds on top of those categories (Ruble et al., 2006). The two usually align, but they are conceptually separate — and, as noted above, gender identity, gender expression, and sexual orientation are three further constructs that each vary independently of the others (American Psychological Association, 2015).
For a stretch of the twentieth century the pendulum swung hard toward pure socialization, on the assumption that gendered behavior was written entirely by reinforcement and modeling. The evidence accumulated since makes a socialization-only account untenable. Prenatal hormones, in particular, leave a measurable mark on later gendered behavior. The clearest natural experiment comes from congenital adrenal hyperplasia (CAH), a condition in which the fetus is exposed to unusually high levels of androgens. Girls with CAH, followed into childhood, show markedly more interest in toys coded as "for boys" and in rough-and-tumble play than their unaffected female relatives, despite being reared as girls (Berenbaum & Hines, 1992). Because the comparison is within families, the pattern is hard to explain by differential socialization and points to an early biological contribution to at least some sex-typed preferences.
The honest reading of this literature is interactionist, not deterministic. Biological inputs bias probabilities; they do not hard-wire outcomes, and the great majority of gendered behavior is shaped, amplified, or muted by culture and learning. Group averages, moreover, say nothing about any individual child: the same prenatal-hormone findings that reveal a genuine biological thread also reveal enormous overlap between groups and wide variation within them. The defensible position is that gender development is biopsychosocial — nature and nurture are not competing explanations but interacting causes, with hormonal predispositions channeled through exactly the learning and schematic processes the three theories describe (Ruble et al., 2006).
The old "born this way" versus "raised this way" framing is a false choice. The CAH data show that biology contributes; the cross-cultural and historical variation in what counts as masculine or feminine shows that culture contributes just as clearly. Modern developmental science treats these as interacting rather than rival forces — prenatal hormones may tilt a child toward certain activities, but the surrounding schema, models, and reinforcement decide what those activities mean and how far they are encouraged. Any claim that gender is "purely" biological or "purely" social is out ahead of the evidence.
Self-concept · the shape of the shiftFrom "I have a bike" to "I'm kind"
The self-concept is the set of beliefs a person holds about who they are — the answer to "describe yourself." It is distinct from self-esteem, the evaluative feeling of worth attached to that self-concept; one is the description, the other the verdict (Harter, 2012). Both are under active construction throughout childhood, and the direction of the change in self-concept is the part worth zooming in on, because it maps directly onto the child's expanding cognitive machinery.
The trajectory runs from the outside in. Early self-descriptions are anchored in what is visible and countable — possessions, physical features, observable skills — and only gradually reach inward toward stable psychological traits. This is not merely a matter of vocabulary; it reflects the child's growing capacity to abstract a consistent "me" out of a stream of particular behaviors, and to see that "me" through the eyes of other people.
Young children describe themselves in concrete, physical, categorical terms: "I have a red bike," "I have brown hair," "I can run fast." They rarely integrate traits and tend toward unrealistic optimism ("I'm the best at everything"). By middle childhood the self-concept becomes psychological and trait-based — "I'm kind of shy," "I'm good at math but not sports" — and increasingly social and comparative, defined relative to peers. The engine is cognitive development plus social comparison: as kids get better at perspective-taking, they start seeing themselves partly through others' eyes.
Self-esteem · not a single dialHow good it feels to be you
Self-esteem is often pictured as one master switch — high or low — but the developmental evidence describes something more textured.
Beginning in middle childhood, children evaluate themselves separately across distinct domains — scholastic competence, athletic ability, physical appearance, social acceptance, and behavioral conduct — and these domain-specific judgments do not simply average into a single global feeling of worth (Harter, 2012). Instead, overall self-esteem is weighted toward the domains a child personally cares about: a child who prizes athletics and struggles at sports takes a real hit to global self-worth, while an unathletic child who never valued sports in the first place is largely unbothered. Self-esteem, in other words, lives in the gap between how competent children feel and how competent they wish to be in the areas that matter to them.
Of all the domains, one predicts global self-worth more powerfully than the rest: perceived physical appearance. Across childhood and adolescence the correlation between how children rate their looks and how they feel about themselves overall is strikingly high and remarkably stable, a pattern that holds from grade school into adulthood (Harter, 2012). The direction of causation is genuinely two-way — feeling worthy can improve how one appraises one's appearance — but the tight linkage helps explain why the body-focused pressures of later childhood and adolescence carry such weight for global self-regard.
The developmental curve also matters. Young children's self-esteem tends to be globally, unrealistically high: a four-year-old will cheerfully claim to be the fastest, smartest, and strongest child in the room, partly because they cannot yet coordinate being good at one thing with being bad at another, and partly because they conflate wishes with abilities. As cognitive development brings social comparison and perspective-taking online, self-evaluations become more differentiated, more accurate, and, on average, more sober; the dip is usually largest around major school transitions, when the comparative reference group abruptly changes (Harter, 2012). A modest downward correction from those inflated early self-views is normal and even adaptive — it means the child is beginning to see themselves as others do.
Parenting · the missing fourth styleBaumrind's three became four
Baumrind's (1971) model, built from close observation of preschoolers and their families, gives three styles — authoritative, authoritarian, and permissive — defined by how parents balance demandingness (the limits and maturity expectations they impose) against responsiveness (the warmth and attunement they offer). It was a genuine advance over the crude "strict versus lenient" continuum it replaced, because it recognized that warmth and control are separate dials rather than opposite ends of one.
But the model worth knowing is the 2×2 that Maccoby and Martin (1983) built on top of it — and it contains a fourth style the three-style version literally cannot represent. Once warmth and control are treated as two crossed dimensions rather than three points along a line, a fourth combination falls out automatically, and it turns out to be the one where the worst outcomes cluster.
Cross two dimensions — warmth/responsiveness (high vs. low) and control/demandingness (high vs. low) — and you get four cells. High warmth + high control = authoritative. Low warmth + high control = authoritarian. High warmth + low control = permissive. And the cell Baumrind's three-style list has no name for: low warmth + low control = uninvolved / neglectful. This fourth style — disengaged, few demands, little emotional investment — predicts the poorest outcomes across nearly every domain. A three-style list is missing the corner of the grid where the worst outcomes live.
The warmth × control grid (Maccoby & Martin, 1983). Tap a cell.
The "authoritative is optimal" finding comes overwhelmingly from White, middle-class Western samples, and it does not travel cleanly. Ruth Chao's (1994) work is the classic corrective: the Western "authoritarian" label misreads a distinct Chinese parenting concept, guan ("to govern," but also "to care for/love"), often glossed as "training." High control in that cultural frame is experienced as involvement and devotion, not coldness — and Chinese and Chinese-American children of "training"-style parents often show strong academic outcomes, breaking the Western pattern. Similar caveats appear for some other ethnic-minority and non-Western families, where firm control can signal love and safety rather than harshness. The honest takeaway: the meaning of a parenting behavior depends on its cultural context, so "best" is not a universal constant.
Evidence · what the grid predictsThe outcomes, and the causal fine print
The 2×2 would be an idle diagram if the four cells did not sort children differently, and in Western samples they do. When roughly four thousand adolescents were classified into all four styles on the basis of their parents' warmth and control, those from authoritative homes scored highest on psychosocial competence and lowest on internalized distress and problem behavior; those from neglectful (uninvolved) homes sat at the bottom on nearly every measure (Lamborn, Mounts, Steinberg, & Dornbusch, 1991). The two middle styles produced a revealing split: adolescents with authoritarian parents were obedient and conforming but carried poorer self-conceptions and more distress, while adolescents with indulgent (permissive) parents felt confident and socially at ease yet reported more school misconduct and substance use. The grid, in other words, does not just rank children from good to bad; it predicts which strengths and vulnerabilities go together.
A companion finding sharpens the picture: parental involvement in schooling boosts adolescent achievement far more reliably when it occurs inside an authoritative home than when it is grafted onto a colder or more indulgent one (Steinberg, Lamborn, Dornbusch, & Darling, 1992). Warmth and structure appear to be the soil in which specific parenting practices either take root or fail to.
Two cautions keep this from becoming a recipe. First, the evidence is overwhelmingly correlational: authoritative parenting travels with better outcomes, but children are not randomly assigned to parents, so shared genes, income, neighborhood, and a dozen other confounds ride along with the parenting style. Second, causation runs in both directions. Temperamentally easy, self-regulated children make it easier for parents to be warm and reasonable, while impulsive or oppositional children can pull otherwise-warm parents toward harsher control — so some of the correlation between authoritative parenting and child competence reflects the child shaping the parent, not only the parent shaping the child. The style a family settles into is partly an outcome, not just a cause.
Culture · the limits of "best"When universal claims meet local meaning
The cultural critique in the myth box above is worth extending, because it is not a footnote to the parenting literature so much as a challenge to how the whole framework is built. The dimensions of warmth and control were defined, and their "optimal" combination identified, within a specific cultural setting that prizes individual autonomy, self-expression, and negotiation between parent and child. Export those yardsticks unchanged and they can misread families organized around different values.
The point generalizes well beyond parenting styles. The same tension appears in attachment research, where the Western assumption that sensitive caregiving should produce an independent, confidently exploring child does not map cleanly onto cultures built around interdependence. In Japan, the relationship concept of amae — a warmly accepted dependence on another — is central to close bonds, and caregiving that fosters closeness and reliance rather than early autonomy is experienced as competent and loving, not smothering (Rothbaum, Weisz, Pott, Miyake, & Morelli, 2000). The very behaviors a Western coding scheme might score as over-involvement can be, in context, exactly what a "secure base" is supposed to look like. The lesson for parenting is the same one Chao (1994) drew: a behavior's developmental meaning is set by the cultural frame around it, so declaring one style universally best mistakes a local optimum for a global law.
None of this dissolves into pure relativism. Some regularities do appear to travel — grinding neglect, the low-warmth, low-control corner of the grid, predicts poor outcomes across a wide range of settings, and warmth in some form seems to matter almost everywhere (Rothbaum et al., 2000). What varies is how warmth and control are expressed and what they signify. Firm parental control reads as coldness in one culture and as devoted involvement in another; the child's outcome depends less on the raw behavior than on the meaning the surrounding culture, and the child, assign to it.
Pulling the three threads together, the same principle runs through this whole unit. Gender, self, and the family are not stamped onto a passive child by a single force. In each case biology and environment supply inputs, the child's own developing mind organizes them, and culture sets the meaning — which is why the honest answers here are interactionist, and why the most confident-sounding single-cause claims tend to be the ones to distrust.
SourcesCited in APA 7
American Psychological Association. (2015). Guidelines for psychological practice with transgender and gender nonconforming people. American Psychologist, 70(9), 832–864. https://doi.org/10.1037/a0039906
Bandura, A. (1977). Social learning theory. Prentice-Hall.
Baumrind, D. (1971). Current patterns of parental authority. Developmental Psychology, 4(1, Pt. 2), 1–103.
Bem, S. L. (1981). Gender schema theory: A cognitive account of sex typing. Psychological Review, 88(4), 354–364.
Berenbaum, S. A., & Hines, M. (1992). Early androgens are related to childhood sex-typed toy preferences. Psychological Science, 3(3), 203–206.
Bussey, K., & Bandura, A. (1999). Social cognitive theory of gender development and differentiation. Psychological Review, 106(4), 676–713.
Chao, R. K. (1994). Beyond parental control and authoritarian parenting style: Understanding Chinese parenting through the cultural notion of training. Child Development, 65(4), 1111–1119.
Harter, S. (2012). The construction of the self: Developmental and sociocultural foundations (2nd ed.). Guilford Press.
Kohlberg, L. (1966). A cognitive-developmental analysis of children's sex-role concepts and attitudes. In E. E. Maccoby (Ed.), The development of sex differences (pp. 82–173). Stanford University Press.
Lamborn, S. D., Mounts, N. S., Steinberg, L., & Dornbusch, S. M. (1991). Patterns of competence and adjustment among adolescents from authoritative, authoritarian, indulgent, and neglectful families. Child Development, 62(5), 1049–1065.
Maccoby, E. E., & Martin, J. A. (1983). Socialization in the context of the family: Parent–child interaction. In E. M. Hetherington (Ed.), Handbook of child psychology: Vol. 4. Socialization, personality, and social development (pp. 1–101). Wiley.
Martin, C. L., & Ruble, D. (2004). Children's search for gender cues: Cognitive perspectives on gender development. Current Directions in Psychological Science, 13(2), 67–70.
Rothbaum, F., Weisz, J., Pott, M., Miyake, K., & Morelli, G. (2000). Attachment and culture: Security in the United States and Japan. American Psychologist, 55(10), 1093–1104.
Ruble, D. N., Martin, C. L., & Berenbaum, S. A. (2006). Gender development. In N. Eisenberg (Ed.), Handbook of child psychology: Vol. 3. Social, emotional, and personality development (6th ed., pp. 858–932). Wiley.
Slaby, R. G., & Frey, K. S. (1975). Development of gender constancy and selective attention to same-sex models. Child Development, 46(4), 849–856.
Steinberg, L., Lamborn, S. D., Dornbusch, S. M., & Darling, N. (1992). Impact of parenting practices on adolescent achievement: Authoritative parenting, school involvement, and encouragement to succeed. Child Development, 63(5), 1266–1281.