Research figure: Albert Bandura · Written by Artificial Sapiens
Kay Bussey and Albert Bandura's social cognitive theory of gender development explains how children and adults acquire, interpret, reproduce, regulate, and sometimes change gender-linked behavior through interacting personal, behavioral, and sociostructural influences. Their landmark 1999 article, “Social Cognitive Theory of Gender Development and Differentiation,” presents gender development as a continuing process involving observational learning, direct experience, socially conveyed evaluations, self-efficacy, self-regulatory standards, and individual agency. The theory treats family members, peers, educators, media, and institutions as interdependent sources of gender-related information and opportunity. It was jointly authored by Kay Bussey of Macquarie University and Albert Bandura of Stanford University, and its proper scholarly attribution belongs to both researchers (Bussey and Bandura, 1999; Macquarie University research record).
The central claim is that gender-related patterns emerge through reciprocal relations among what individuals encounter, what they come to believe and value, what they do, and how social environments respond. The same framework explains how social conventions can persist across generations and how people can depart from prevailing expectations. Its value lies in specifying potentially testable pathways rather than proposing a single, universal sequence through which every child must pass. Evidence is strongest for particular processes—modeling, differential feedback, the development of gender stereotypes, and the influence of social contexts—while the full integrative theory remains a framework to be assessed mechanism by mechanism (Bussey and Bandura, 1992; Martin, Ruble, and Szkrybalo, 2002).
The Original 1999 Theory and Its Scientific Context
Published in Psychological Review, volume 106, issue 4, pages 676–713, the 1999 article is a theoretical synthesis of gender development and differentiation. Its publication in a peer-reviewed theoretical journal does not mean it was itself a new randomized experiment or a single longitudinal test. Bussey and Bandura assembled claims from research on learning, cognitive development, social interaction, self-regulation, and social structures into a connected explanation of how gender-linked practices become established and change. The paper builds on Bandura's broader social cognitive account of reciprocal causation and human agency, while giving specific attention to developmental findings and gender-related research developed with Bussey (Bussey and Bandura, 1999).
The scientific problem they addressed was broader than why a child imitates an adult of the same gender. Gender differentiation includes expectations about activities, appearance, attributes, emotional expression, competence, interpersonal responsibilities, aspirations, and occupations. It also concerns who can access resources, which behavior attracts approval or penalties, and how children form personal standards that continue to guide choices when an external observer is absent. Their analysis linked these individual processes to broader social arrangements instead of confining explanations to a child's immediate household (Bussey and Bandura, 1999; Bussey and Bandura, 2004).
The theory developed in a field shaped by multiple intellectual traditions. Lawrence Kohlberg's cognitive-developmental approach focused on how gender understanding becomes organized. Sandra Bem's gender schema account emphasized the way gender categories organize attention, memory, and interpretation. Social learning traditions investigated modeled actions, consequences, and environmental contingencies. Bussey and Bandura sought an explanatory synthesis in which knowledge, observed action, motivation, social relationships, and structural constraints were all relevant. Different accounts make partly overlapping predictions, a point that became important in subsequent debates (Martin, Ruble, and Szkrybalo, 2002; Martin, Ruble, and Szkrybalo, 2004).
What Gender Development Means in This Framework
Within the 1999 historical formulation, gender development concerns the construction and regulation of beliefs, preferences, evaluations, behavior, and opportunities socially associated with female and male categories. Gender differentiation describes the emergence and maintenance of patterns distributed differently across these categories. Gender typing is the process by which objects, activities, traits, and behaviors become associated with gender categories, whether or not those associations accurately reflect individual capacities. These terms describe patterns of development and social classification rather than clinical diagnoses (Bussey and Bandura, 1999).
Contemporary research requires several further distinctions. Gender identity concerns an individual's experienced understanding of their gender; stereotypes are generalized beliefs about social categories; gender expression concerns outward presentation or behavior; sex-related biological characteristics are a different set of variables. Self-concept, interests, capabilities, expectations, social status, and identity are related but separately measurable. The 1999 paper largely used the binary categories typical of its period's research base, so contemporary applications must not assume that all children identify within those categories or that a population-average association determines an individual's possibilities. Current developmental scholarship explicitly examines gender-diverse trajectories alongside conventional samples (deMayo, Jordan, and Olson, 2022).
A central methodological point follows: knowing that an activity is labeled masculine or feminine, liking that activity, expecting criticism for it, feeling capable of it, and actually participating in it are different outcomes. A theory claiming to explain gender development should specify which outcome is being predicted and under what conditions. Otherwise, an association between two broad measures of gender typing may hide several distinct causal pathways.
The Architecture: Reciprocal Social Cognitive Causation
Bussey and Bandura placed gender development within a reciprocal causal system involving personal factors, behavior, and environment. Personal factors include knowledge, anticipated outcomes, goals, self-efficacy beliefs, evaluations, and self-regulatory processes. Behavior includes selecting activities, imitating models, practicing skills, challenging expectations, or enforcing rules toward others. Environmental factors range from immediate rewards and criticism to group norms, media representations, institutional policies, and access to education or occupations. Influence can move in multiple directions across these components (Bussey and Bandura, 1999).
Consider a child who sees a skilled female engineer at work. Exposure may broaden a perceived range of possible careers. Whether it affects later interest depends on the child's attention, perceived relevance of the model, expectations of success, access to learning opportunities, and the responses of parents, classmates, and teachers. If the child then joins a robotics group, participation creates further opportunities for mastery and new social relationships. Those subsequent experiences can change both beliefs and the local environment. This example is an application of the theory, not a reported Bussey–Bandura experiment.
Reciprocity should not be interpreted as an empirical claim that all influences are equally strong or operate simultaneously. A community may impose substantial restrictions that individual motivation alone cannot overcome. The theory specifically recognizes sociostructural determinants, including institutions and unequal opportunities, and therefore treats agency as action exercised within conditions that may enable or constrain it. Demonstrating a reciprocal process requires data that can distinguish temporal order and reciprocal paths; a cross-sectional correlation between a belief and a behavior is insufficient (Bussey and Bandura, 1999).
Observational Learning and Gender-Linked Models
How a model becomes informative
Observational learning allows individuals to acquire information about actions and their likely consequences by watching other people. Children encounter gender-linked models among family members, siblings, peers, teachers, athletes, entertainers, public figures, and fictional characters. A model can communicate a behavioral sequence, display a valued outcome, indicate what a community approves, or demonstrate what appears possible for someone perceived as similar to the observer. Modeling is therefore more than the visible copying of a movement. It can change expectations, aspirations, perceived competence, and evaluative standards (Bussey and Bandura, 1999).
Bandura's general observational-learning account distinguishes attention to a model, retention of what is observed, the ability to reproduce an action, and motivation to perform it. Applied to gender development, these processes help explain why children exposed to similar models may act differently. One child may notice an activity but lack the opportunity to practice it; another may learn it yet avoid public performance because peers are likely to disapprove. The acquisition of a modeled response must therefore be distinguished from its enactment in a given social setting. An observer's failure to perform a behavior does not establish that no learning occurred.
Model identity, competence, and context
Gender category can influence which models attract attention or appear socially relevant, but this is one factor among others. A model's demonstrated competence, status, familiarity, relationship to the child, and consequences for the behavior can matter. Situational features also matter: the same act may be encouraged at home, discouraged by peers, or valued in a classroom. The theory predicts context-sensitive learning rather than automatic replication of everything displayed by a same-gender model. Claims about exactly how strongly model gender influences behavior require task- and population-specific evidence (Bussey and Bandura, 1999).
Some of the theoretical vocabulary came from Bandura's earlier research on observational learning. Those general experiments are historical foundations, but the 1999 gender theory is a distinct, jointly authored extension. It should not be reduced to the Bobo doll experiments or treated as if those aggression studies established the long-term development of gender identity.
Direct Social Experiences, Reinforcement, and Sanctions
Gender-related behavior is also shaped by consequences experienced directly. Children may receive praise, access to preferred activities, teasing, exclusion, disapproval, or correction depending on whether others consider an activity appropriate. Social reinforcement may operate subtly, through attention, encouragement, materials supplied, expectations communicated, or the absence of opportunities. Adults and peers can also communicate norms without administering explicit rewards or punishments (Bussey and Bandura, 1999).
Three processes require separate analysis. First, direct outcomes influence whether a behavior is repeated in a particular setting. Second, observing another person's reward or penalty can alter the observer's expectation of consequences; this is commonly discussed as vicarious reinforcement. Third, repeated social experiences may gradually contribute to internal standards that influence behavior even when immediate approval or punishment is unlikely. These mechanisms can interact but should not be treated as interchangeable.
Reinforcement findings do not justify a deterministic claim that parents or media mechanically manufacture an individual's gender identity. Social environments influence behaviors and expectations across varying contexts, with responses moderated by age, identity, peer relationships, and opportunities. Even studies observing a statistically reliable relationship between parental attitudes and children's choices do not, without suitable causal design, establish a unique parental cause. A longitudinal study of 501 families found some small associations between trajectories of parental gendered play attitudes and later child outcomes, illustrating both the relevance and the modest size of some measured relationships (Rogers and colleagues, 2024).
Self-Regulation: From External Feedback to Personal Standards
A distinctive feature of Bussey and Bandura's account is the transition from behavior that depends heavily on external reactions toward conduct guided by internalized standards. Children learn which actions are socially approved or criticized, develop evaluations of those actions, and anticipate self-approval, embarrassment, pride, or self-criticism. Once an evaluative standard is adopted, a child may continue to follow it privately or even when no parent or peer is present. At the same time, individuals can revise standards through new evidence and social experience (Bussey and Bandura, 1992).
This account is important because a repeated gender-linked choice may be sustained by different mechanisms at different ages. A younger child might avoid a toy after observing peer rejection; an older child might make a similar choice because a personally endorsed standard has become salient. The outward behavior looks similar, while the underlying regulatory mechanism differs. Valid measurement therefore needs to distinguish knowledge of a rule, the expectation of others' reactions, the child's own evaluative response, and observed action.
Internalization is not identical to conformity. Personal standards can support rejection of restrictive conventions when an individual values fairness, autonomy, or competence over prevailing expectations. Social cognitive agency includes the possibility of modifying environments, joining supportive groups, and participating in wider social change. Whether this happens depends partly on resources and perceived efficacy, not merely on knowing that a stereotype exists (Bussey and Bandura, 1999).
The 1992 Bussey and Bandura Study: Design, Results, and Meaning
The study's actual research question
An important empirical precursor to the 1999 synthesis was the coauthored 1992 article “Self-Regulatory Mechanisms Governing Gender Development,” published in Child Development, volume 63, issue 5, pages 1236–1250. The study examined how gender knowledge, gender constancy, personal standards, and situational cues related to children's gender-linked behavior. It tested predictions from a social cognitive account rather than serving as a comprehensive trial of every proposition later incorporated into the 1999 framework (Bussey and Bandura, 1992).
Children at four levels of gender constancy were assessed on gender knowledge, personal standards, and behavior in different situational conditions. Gender constancy broadly concerns understanding the stability of gender classification across time or superficial changes; it was central to influential cognitive-developmental interpretations. The researchers were interested in whether gender-typed conduct would depend on constancy levels and whether anticipation of social or personal evaluative consequences would better explain what children actually did.
What the researchers found
According to the original study's published abstract, children at every assessed level of gender constancy engaged in more same-gender-typed than cross-gender-typed behavior. Younger children responded to peers' gender-linked behavior in a stereotypic fashion but did not show equivalent regulation by their own personal standards. Older children more clearly anticipated self-approval for same-gender-typed behavior and self-criticism for cross-gender-typed behavior; these anticipatory self-sanctions predicted actual gender-linked behavior. In this study, gender knowledge and gender constancy did not predict the measured gender-linked conduct (Bussey and Bandura, 1992).
These findings supported the authors' developmental account in the particular sampled tasks and conditions. They were especially relevant to the claim that external social sanctions and internal self-sanctions need to be separately modeled. They did not establish that gender knowledge or constancy is unimportant to every dimension of gender development, and the paper's negative predictive findings should not be extended automatically to identity formation, other populations, or other measurement instruments. Later work argued that gender identity and stereotype knowledge have important early developmental roles, leading to a sustained scientific debate (Martin, Ruble, and Szkrybalo, 2002).
How to evaluate the evidence
The 1992 work measured several constructs rather than treating gender typing as one undifferentiated score. That is a strength for mechanism-oriented interpretation. At the same time, findings from particular observations and measured self-reactions cannot alone establish population-wide developmental stages or the long-term causal origins of personally endorsed gender norms. When inferring causation, investigators need to examine whether situational influences were experimentally varied, whether potential confounders were controlled, and whether observed outcomes persisted outside the study. The published abstract establishes the reported comparison and associations; more granular claims about randomization, sample composition, or precise effect sizes require direct inspection of the complete methods and results.
Self-Efficacy, Goals, and Participation
Self-efficacy refers to a person's judgment of their capacity to organize and carry out actions needed for a particular goal. In the gender-development framework, efficacy beliefs can affect which activities individuals approach, how persistently they practice, whether they anticipate success, and how they respond to obstacles. When a child receives few opportunities to practice a field described as inappropriate for their gender, perceived capability may remain weak even when objective aptitude has never been tested. Conversely, mastering a task can strengthen a sense of capability and alter future choices (Bandura, 1997; Bussey and Bandura, 1999).
Self-efficacy is distinct from objective competence, biological ability, self-esteem, interest, and actual achievement. A student may have the skills to complete a computer programming task yet anticipate rejection and decline a course. Another may feel confident but need further training. The theory predicts that efficacy beliefs participate in a larger system of opportunities, goals, expected consequences, and social responses. A correlation between self-reported efficacy and occupational aspiration cannot identify whether confidence preceded aspiration, arose through prior successful experience, or reflects an unmeasured supportive environment.
This distinction is consequential for interventions. Showing a diverse range of competent role models may change expectations, but lasting participation often also requires accessible instruction, supported practice, meaningful feedback, safe group norms, and institutional opportunity. The evidence for changing aspirations should be assessed separately from evidence for changing long-term educational enrollment or professional outcomes. A review of counterstereotypical role-model research found promising effects in some contexts while emphasizing limitations in long-term behavioral evidence and the importance of model relevance and active engagement (Olsson and Martiny, 2018).
Families, Peers, Schools, Media, and Institutions
Family interaction
Parents and caregivers provide early examples of activities, work roles, emotional expression, and interpersonal responsibilities. Their influence can arise through their own conduct, conversational expectations, available toys and books, responses to experimentation, and support for learning. Yet parents' stated gender attitudes and children's behavior are not interchangeable measures, and children can influence parental responses as well as be influenced by them. Prospective data are useful for estimating temporal relationships, especially where differences in family resources and social context may affect both variables (Rogers and colleagues, 2024).
Peers and group norms
Peer groups can reward conformity, ridicule departures from expectation, and create opportunities for collaborative learning. A child's anticipation of a classmate's reaction can determine whether a learned skill is expressed publicly. Peer norms also vary between groups and over time; the social cognitive framework expects sensitivity to particular relationships and contexts. Describing a child as “gender stereotyped” without recording the social setting risks confusing a locally strategic response with an enduring personal preference.
Teachers and institutional practices
Schools communicate expectations through curricular examples, classroom participation, feedback, access to extracurricular activities, and assumptions about competence. Institutions also influence who becomes visible in different roles. Representation can matter, but its impact is plausibly moderated by opportunities for genuine participation and recognition. When a school changes a restrictive policy or makes an activity equally accessible, the resulting shift may reflect a structural change rather than a change in individual beliefs alone. Empirical studies should distinguish these possible mechanisms.
Media and symbolic modeling
Screen media offer large numbers of symbolic models who can be observed repeatedly without direct interaction. Media portrayals may shape beliefs about desirable appearance, acceptable interests, gendered occupations, status, and relationships. A scholarly review of research from 2000 to 2020 reported small but consistent associations in three meta-analyses between television exposure and stereotypic gender beliefs. The authors also emphasized differences among media genres, measures, age groups, and research designs. Associations from ordinary viewing cannot, on their own, identify the causal effect of a particular message, although experimental exposure studies can test narrower hypotheses (Ward and Grower, 2020).
Gender Stereotypes, Ability Judgments, and Contemporary Evidence
A critical test of gender-development explanations is whether they accommodate the large variation across activities, ages, cultures, and historical periods. Stereotypes about who likes a task are different from stereotypes about who can perform it well, and both differ from children's actual performance. Reviews indicate that gender stereotypes can appear early, change with development, and shape both individual choices and others' expectations. However, the timing and magnitude vary by domain. Treating every stereotype as a single global belief would erase differences the data reveal (Master and colleagues, 2025).
A preregistered meta-analysis published in Psychological Bulletin in 2024 integrated 98 studies encompassing 145,204 children aged 4–17 across 33 countries and studies conducted between 1977 and 2020. On average, stereotypes favoring male ability in mathematics were small, about 0.11 standard deviations from neutrality, whereas stereotypes favoring male ability in computer science, engineering, and physics were stronger, about 0.51 standard deviations. Pro-female verbal-ability stereotypes were also substantial, about 0.46 standard deviations. The age pattern and subgroup differences were more complex than a simple increasing or decreasing trend. These are aggregate stereotypes, not estimates of innate ability or proof that exposure to one social influence caused the pattern (Miller, Lauer, Tanenbaum, and Burr, 2024).
This evidence is compatible with the broad premise that gender-related beliefs are learned and maintained through complex social and cognitive processes. The meta-analysis was informed by several theoretical traditions and does not independently validate every Bussey–Bandura mechanism. It also shows why a scientific account must examine subject-specific expectations rather than presume a universal belief about gender. Another meta-analysis of 78 U.S. Draw-A-Scientist studies, encompassing 20,860 children, found that representations of scientists became more gender diverse over decades while older children's drawings remained more male-typed. The drawings measure representations, not direct career behavior (Miller and colleagues, 2018).
Research on psychological differences between girls and boys, women and men, and people across gender identities also requires careful separation from stereotype measures. Janet Shibley Hyde's review of 46 meta-analyses argued for similarity across many psychological characteristics, with differences depending on domain and context. Stereotype strength cannot be treated as a reliable proxy for the actual distribution of aptitude or personality (Hyde, 2005).
Gender Identity, Cognitive Development, and Competing Accounts
What cognitive-developmental approaches emphasize
Cognitive-developmental accounts examine how children understand categories, develop gender identity, build representations of social groups, and use those representations to organize behavior. Gender schema theories emphasize the processes through which gender-linked information becomes especially salient in attention and memory. Cognitive research therefore addresses mechanisms that observational or reinforcement measures may miss: a child may select information that seems relevant to a self-category before any immediate reward is offered. A major 2002 review by Carol Lynn Martin, Diane N. Ruble, and Joel Szkrybalo assembled evidence about early category understanding, labeling, stereotype knowledge, and the development of cognitive gender representations (Martin, Ruble, and Szkrybalo, 2002).
In Bussey and Bandura's theory, cognition is already part of the explanation: observation is interpreted, outcomes are anticipated, and individuals form standards and goals. Consequently, the debate is not adequately characterized as a contest between social influence and thinking. The substantive questions concern which cognitive constructs are central, when they emerge, which findings establish necessary precursors, and whether the mechanisms specified are precise enough to discriminate rival predictions.
The documented 2004 exchange
The disagreement became explicit in Psychological Bulletin in 2004. Albert Bandura and Kay Bussey argued for a broader cognitive, motivational, and sociostructural approach in a comment on Martin, Ruble, and Szkrybalo's 2002 review. Martin, Ruble, and Szkrybalo responded that gender identity and stereotype knowledge had substantial empirical support, that their account had been misconstrued, and that theoretical integration required a more accurate treatment of existing cognitive research. The two published texts document real scholarly differences concerning explanatory emphasis and interpretation of developmental evidence (Bandura and Bussey, 2004; Martin, Ruble, and Szkrybalo, 2004).
The exchange establishes that the 1999 formulation was scientifically debated, not that one published commentary settled the matter. Some claims can be resolved by longitudinal or experimental studies testing specific predictions. Others involve the architecture of explanation: how to weigh identity, cognitive categorization, observed contingencies, self-sanctions, and institutional constraints. For current research, the productive conclusion is to operationalize these constructs separately and evaluate the added explanatory value of each.
Gender Diversity and the Historical Boundaries of the Theory
Bussey and Bandura's 1999 article organized much of its discussion around the gender categories and gender-role measures that dominated late twentieth-century developmental psychology. A contemporary scientific assessment must recognize the limits of transferring conclusions from binary-coded samples to all children and adults. Gender-diverse and transgender populations have been studied through additional lines of research, and their developmental trajectories require direct evidence rather than assumptions drawn from gender-typed toy choices or the behavior of cisgender comparison groups (deMayo, Jordan, and Olson, 2022).
The broader reciprocal-influence framework can help researchers ask how social affirmation, exclusion, representation, identity, and institutional context relate to opportunities and well-being. Applying it in this way is a contemporary extension of the framework, not an original clinical claim made by Bussey and Bandura in 1999. It also supplies no clinical diagnostic criteria. Neither a nonstereotypical interest nor nonconforming gender expression, considered alone, is a disorder or a measure of psychological impairment.
Gender development research benefits from observing how participants describe their own identities, using culturally appropriate measures, and avoiding a single norm that treats conformity as the criterion of healthy development. Ethical interpretation distinguishes descriptive patterns from prescriptions about how people ought to behave.
Research Methods: How the Theory Can Be Tested
The theory contains claims at several levels and needs correspondingly different methods. Laboratory studies can manipulate the gender of models, visibility of rewards, peer reactions, or opportunities to practice and then observe immediate responses. They offer clearer tests of specified situational mechanisms but may not establish persistence across years. Naturalistic observation shows what families, classrooms, and peer groups actually do, while increasing the challenge of distinguishing causal influences from selection and reciprocal responses.
Longitudinal designs can track the ordering of stereotype knowledge, self-concept, self-efficacy, social feedback, and later participation. Yet repeated observations do not by themselves prove a causal effect: stable unmeasured characteristics, socioeconomic conditions, or reciprocal relationships may explain part of a detected association. Where appropriate, randomized interventions can test whether changes in specific feedback, role-model exposure, or access to instruction produce changes in clearly defined outcomes. Such studies should measure whether initial shifts in expressed beliefs generalize to enduring choices.
Researchers should use multiple measurement modes rather than collapsing every construct into a single gender-typing index. Explicit questions reveal articulated beliefs; indirect measures address different aspects of associations; behavioral tasks assess choices under specified conditions; and institutional data capture access and participation. Different measurements can yield different patterns. A 2024 research program of three studies found that children's endorsement of gender stereotypes differed depending on whether investigators asked about interests, ability, or category membership, reinforcing the need to define the measured construct precisely (Miller, Wheeler, and Woods, 2024).
Cross-cultural work should assess the local meaning of activities, the distribution of opportunities, and whether measures function equivalently across populations. Historical shifts also matter. An activity's social classification may change faster than deeply embedded institutional arrangements, producing distinct trajectories for stereotype judgments and participation. A complete test of the theory would examine whether proposed personal, behavioral, and environmental pathways explain those divergences.
Evidence Assessment: What Is Established and What Remains Open
Established findings
There is substantial evidence that children develop gender-related category knowledge and stereotypes, that their content and strength vary across domains and age groups, and that models, feedback, peers, schools, and media participate in the social environments through which those beliefs are encountered. The original 1992 Bussey–Bandura study documented age-related differences between externally cued responses and measured anticipatory self-sanctions in its study conditions. Meta-analytic and review research confirms that stereotypes are measurable and historically variable, with domain-specific patterns (Bussey and Bandura, 1992; Miller and colleagues, 2024; Ward and Grower, 2020).
Supported but context-dependent propositions
Observational role models, parental attitudes, and contextual opportunities can be relevant to expectations and behavior, but effect sizes, developmental timing, and durability differ among settings. A demonstrated change in immediate stereotype endorsement is not equivalent to a demonstrated long-term change in occupational choice. Parent–child and media-exposure correlations cannot be treated as universal causal mechanisms. Such findings support bounded, testable elements of social cognitive explanations rather than every possible application of the integrated theory (Rogers and colleagues, 2024; role-model research review).
Limited or contested propositions
The relative explanatory priority of gender identity, gender constancy, social sanctions, self-regulation, and information processing remains sensitive to the outcome, age, measurement, and theoretical definitions used. The 1992 failure of gender constancy to predict a set of behaviors does not rule out a role for cognitive categorization in other domains. The 2004 published exchange makes the interpretive disagreement explicit. The full reciprocal causal model is too expansive to treat as a single experimentally proven statement; investigators should isolate claims with discriminating predictions (Bandura and Bussey, 2004; Martin, Ruble, and Szkrybalo, 2004).
Education, Family Practice, and Institutional Implications
A carefully applied social cognitive perspective directs attention to the range of activities made visible, the reactions children anticipate, and the opportunities available to acquire competence. In classrooms, this can mean offering varied examples of scientific, artistic, caregiving, leadership, and technical work; providing equitable access to practice; examining patterns of feedback; and measuring actual engagement rather than satisfaction alone. These are evidence-informed applications of the framework, while the effect of any specific program must be evaluated in the population and setting where it is used.
Within families, the framework suggests that the activities adults visibly perform, the options they make available, and the language they use about ability may matter. Parents can observe a child's interests without assigning the child a fixed trajectory from one choice. Peer settings likewise deserve attention: reducing ridicule and widening opportunities to collaborate may change the consequences of behavior in ways a motivational message alone cannot achieve. The practical question is whether an environment expands usable possibilities and supports learning and agency.
Institutional changes can be especially important when choices are limited by structural barriers. A child cannot enroll in a program that is inaccessible, regardless of confidence. Likewise, broad representation without resources may change beliefs without changing opportunity. Researchers and practitioners can examine both the proximal process—what a child perceives and does—and the distribution of resources that governs what actions become possible. This is consistent with the sociostructural ambition of the 1999 theory (Bussey and Bandura, 1999).
Intellectual Contribution and Scientific Legacy
Bussey and Bandura's contribution was to link gender socialization with a dynamic theory of learning, motivation, self-regulation, and agency, while situating these processes within families, institutions, and social structures. Kay Bussey's contribution is integral to this intellectual history: she coauthored the 1992 empirical study, the 1999 Psychological Review theory, and the later 2004 book chapter, alongside Bandura. Describing the gender-development account simply as “Bandura's theory” obscures an essential coauthor and an empirical research program in which Bussey played a named scholarly role (Bussey and Bandura, 1992; Bussey and Bandura, 1999; Bussey and Bandura, 2004).
Their perspective remains useful in explaining why knowing a stereotype, expecting a sanction, evaluating oneself, and carrying out a behavior are separable phenomena. It also offers concepts for analyzing social change: changing models can change expectations; new opportunities can produce mastery; altered group responses can reduce the cost of departure from convention; and agents can participate in changing norms. Current evidence encourages integration with research on gender identity, cognition, intersectionality, variation across cultures, and developmental diversity. The theory's historical standing is established; the scope of each empirical claim continues to depend on the quality of studies testing it.
Frequently Asked Questions
Who developed the social cognitive theory of gender development?
The influential 1999 theory article was jointly written by Kay Bussey and Albert Bandura. It appeared in Psychological Review under the title “Social Cognitive Theory of Gender Development and Differentiation.” Bussey and Bandura had also jointly published a 1992 experimental-developmental investigation of self-regulation and gender-linked conduct (1999 original publication).
What are the main mechanisms in their theory?
The central mechanisms include observational learning from models, direct and observed consequences, development of expectations and gender-related knowledge, internalized self-evaluative standards, self-efficacy, goal-directed activity, and reciprocal interaction with interpersonal and institutional environments. Their relative importance is expected to depend on context and developmental conditions.
Does the theory say that children copy only same-gender models?
No. The social cognitive account assigns importance to models' relevance, competence, social status, consequences, and the observer's expectations. Gender matching can be one cue, but neither matching nor exposure guarantees that an observed behavior is learned or performed.
Did Bussey and Bandura show that gender constancy never matters?
No. Their 1992 study found that gender knowledge and gender constancy did not predict the gender-linked behaviors measured in that design. Other cognitive-developmental research supports meaningful roles for gender identity and stereotype knowledge. The broader interpretation was disputed in the literature (Bussey and Bandura, 1992; Martin, Ruble, and Szkrybalo, 2004).
Is the theory scientifically proven?
Its historical existence, scholarly authorship, and influence are established. Multiple proposed mechanisms have empirical support, with different levels of confidence across populations and outcomes. The complete synthesis is a broad theoretical framework whose causal pathways require separate tests; no single paper proves all its propositions.
Does the theory explain gender identity or prescribe gender roles?
It aims to explain the development and regulation of gender-related beliefs and behavior under social and personal influences. It does not establish a moral requirement to conform to a gender role, and its historical binary-centered evidence base should not be extended automatically to gender-diverse populations. Contemporary developmental studies investigate identity and diversity using specifically relevant evidence.
Related Articles
Albert Bandura — The central scientific examination of Bandura's intellectual development, social learning, social cognitive theory, self-efficacy, and scientific influence.
Social-Cognitive Personality Theory: Reciprocal Determinism and Behavior — The broader reciprocal-determinism framework in personality science, distinguished from Bussey and Bandura's gender-development formulation.
Learning Theories in Education: Behaviorism, Cognitivism, Constructivism, and Sociocultural Learning — Educational theories and the distinct questions they ask about modeling, instruction, development, and social context.
Albert Bandura's Social Cognitive Theory — The mature framework of social, cognitive, and behavioral influences.
Albert Bandura's Social Cognitive Theory of Mass Communication — Symbolic modeling and the psychology of mass communication.
References
Bandura, A. (1997). Self-Efficacy: The Exercise of Control. W. H. Freeman. Author bibliography.
Bandura, A., and Bussey, K. (2004). On broadening the cognitive, motivational, and sociostructural scope of theorizing about gender development and functioning: Comment on Martin, Ruble, and Szkrybalo (2002). Psychological Bulletin, 130(5), 691–701. https://doi.org/10.1037/0033-2909.130.5.691.
Bussey, K., and Bandura, A. (1992). Self-regulatory mechanisms governing gender development. Child Development, 63(5), 1236–1250. https://doi.org/10.1111/j.1467-8624.1992.tb01692.x.
Bussey, K., and Bandura, A. (1999). Social cognitive theory of gender development and differentiation. Psychological Review, 106(4), 676–713. https://doi.org/10.1037/0033-295X.106.4.676.
Bussey, K., and Bandura, A. (2004). Social cognitive theory of gender development and functioning. In A. H. Eagly, A. E. Beall, and R. J. Sternberg (Eds.), The Psychology of Gender (2nd ed., pp. 92–119). Guilford Press. Macquarie University bibliographic record.
deMayo, B. E., Jordan, A. E., and Olson, K. R. (2022). Gender development in gender diverse children. Annual Review of Developmental Psychology, 4, 207–229. https://doi.org/10.1146/annurev-devpsych-121020-034014.
Hyde, J. S. (2005). The gender similarities hypothesis. American Psychologist, 60(6), 581–592. https://doi.org/10.1037/0003-066X.60.6.581.
Martin, C. L., Ruble, D. N., and Szkrybalo, J. (2002). Cognitive theories of early gender development. Psychological Bulletin, 128(6), 903–933. https://doi.org/10.1037/0033-2909.128.6.903.
Martin, C. L., Ruble, D. N., and Szkrybalo, J. (2004). Recognizing the centrality of gender identity and stereotype knowledge in gender development and moving toward theoretical integration: Reply to Bandura and Bussey (2004). Psychological Bulletin, 130(5), 702–710. https://doi.org/10.1037/0033-2909.130.5.702.
Master, A., Baker, S., Patel, K. S., Roy, P., and Butler, L. P. (2025). The role of stereotypes in gender development and disparities. Annual Review of Developmental Psychology, 7, 339–362. https://doi.org/10.1146/annurev-devpsych-111323-115554.
Miller, C. F., Wheeler, L. A., and Woods, B. (2024). A multidimensional examination of children's endorsement of gender stereotypes. Social Development, 33(2), e12725. https://doi.org/10.1111/sode.12725.
Miller, D. I., Lauer, J. E., Tanenbaum, C., and Burr, L. (2024). The development of children's gender stereotypes about STEM and verbal abilities: A preregistered meta-analytic review of 98 studies. Psychological Bulletin, 150(12), 1363–1396. https://doi.org/10.1037/bul0000456.
Miller, D. I., Nolla, K. M., Eagly, A. H., and Uttal, D. H. (2018). The development of children's gender-science stereotypes: A meta-analysis of five decades of U.S. Draw-A-Scientist studies. Child Development, 89(6), 1943–1955. https://doi.org/10.1111/cdev.13039.
Olsson, M., and Martiny, S. E. (2018). Does exposure to counterstereotypical role models influence girls’ and women’s gender stereotypes and career choices? A review of social psychological research. Frontiers in Psychology, 9, 2264. https://doi.org/10.3389/fpsyg.2018.02264.
Rogers, A. A., Shawcroft, J., Stockdale, L. A., Coyne, S. M., and Fraser, A. M. (2024). Trajectories of parents' gendered play attitudes during early childhood and implications for children's gender development. Child Development, 95(2), 428–446. https://doi.org/10.1111/cdev.13989.
Ward, L. M., and Grower, P. (2020). Media and the development of gender role stereotypes. Annual Review of Developmental Psychology, 2, 177–199. https://doi.org/10.1146/annurev-devpsych-051120-010630.