Research figure: Albert Bandura · Written by Artificial Sapiens
Albert Bandura and Richard H. Walters's Adolescent Aggression: A Study of the Influence of Child-Training Practices and Family Interrelationships (1959) is an early empirical monograph investigating how patterns of family interaction, parental responses to misconduct, identification, emotional relationships, and social control relate to aggressive and antisocial conduct among adolescent boys. Its central contribution was to make the family environment a systematic object of comparative psychological investigation and to connect aggressive behavior with social learning processes that Bandura and Walters would elaborate in later work. The book was published by Ronald Press in New York, and its 475-page main text and bibliographic identity are documented in the National Library of Medicine catalog (National Library of Medicine, 1959 catalog record).
The original investigation compared 26 boys identified as aggressive and antisocial with 26 comparison boys and examined information gathered from adolescents and parents. The findings were interpreted in the context of child-rearing, internalized controls, parent–child relationships, and pathways by which aggression may be learned and expressed. The design yields historically important evidence of group differences within a selected sample. Its evidence about causal origins of adolescent aggression is more limited: assigning existing adolescents to comparison groups and reconstructing family histories cannot establish that any particular parental practice produced subsequent behavior. A contemporary interpretation combines the 1959 contribution with longitudinal, experimental, and systematic-review evidence accumulated afterward (Gaier, 1961; Hoeve et al., 2009).
What Is Adolescent Aggression (1959)?
Full title, authorship, and publication
The full original title is Adolescent Aggression: A Study of the Influence of Child-Training Practices and Family Interrelationships. Albert Bandura and Richard H. Walters are the two authors. The Ronald Press Company published it in 1959 as part of its psychology series. The bibliographic description is xii preliminary pages plus 475 numbered pages, with illustrations. Robert R. Sears contributed a foreword, a role distinct from joint authorship. The National Library of Medicine, WorldCat, and an author-bibliography resource agree on the publication and principal authors (NLM catalog; WorldCat record; Albert Bandura books).
Its subtitle defines the primary scientific object: the associations of child-training practices and family interrelationships with forms of adolescent aggression. The book is a research monograph, meaning an extended treatment of a specific empirical program and its interpretive framework. It is neither a manual for diagnosing individual adolescents nor a universal theory of delinquency. Its importance lies in the questions it made empirically examinable and in the link it provided between Bandura's early developmental work and the subsequent social-learning approach.
The question behind the study
How can apparently similar families produce adolescents whose willingness to express aggression differs markedly? The investigators considered whether the patterns of discipline, emotional relationships, and behavioral models available in the family might help explain the acquisition, inhibition, and expression of aggression. Their question had two dimensions: the kinds of aggressive conduct that appeared and the forms of social regulation that adolescents experienced or internalized.
This distinction is foundational. Aggressive behavior refers to acts or dispositions involving hostility, threat, or harm, with the exact operational definition depending on the study. Antisocial behavior encompasses a wider set of norm-violating conduct; juvenile delinquency is a social and legal classification involving particular acts and contexts. A young person can be aggressive without meeting the criteria for a psychiatric disorder, and a reported history of delinquency is not itself a validated measurement of one psychological mechanism. The 1959 research dealt with historically defined groups and measures; modern interpretations must avoid collapsing these categories.
Historical Context: From Child Rearing to Social Learning
Child-training research in the 1950s
The monograph appeared when developmental psychology was examining how family authority, punishment, dependency, attachment, and socialization relate to children's behavior. The background included Robert R. Sears and colleagues' research on patterns of child rearing, which blended learning processes with questions traditionally addressed by personality and developmental theories. A contemporary review described Bandura and Walters as drawing on the investigative orientation of Sears's child-rearing research while directing it toward aggressive adolescents (Gaier, 1961).
Within that historical setting, punishment and approval were more than isolated stimuli. Their meaning depended on the person who administered them, the consistency of the response, the child's prior history, and the relationship between parent and adolescent. Questions about guilt, fear of external punishment, hostility, and identification were central to how researchers understood social control. These are historical explanatory constructs; contemporary personality science, developmental psychopathology, and behavioral assessment define and measure related phenomena using different instruments and research designs.
Albert Bandura and Richard H. Walters as collaborators
Bandura and Walters were research collaborators and coauthors. Attributing the 1959 argument solely to Bandura obscures the scholarly record. The pair also authored Social Learning and Personality Development in 1963, published by Holt, Rinehart and Winston. That later book developed a wider formulation of social learning and personality processes, whereas the 1959 monograph remained anchored to the narrower empirical question of aggressive adolescent conduct and family relationships (Bandura bibliography; Wellcome Collection catalog).
Bandura later became associated with observational learning, modeling, self-efficacy, and social cognitive theory, but those later contributions should not be read back into the 1959 book as though its authors had already formulated the mature theory. The historical sequence matters because concepts emerged from successive programs of research. The joint study of family relationships preceded the best-known laboratory studies of aggressive modeling and the later theory of personal agency.
The research was distinct from the Bobo doll experiments
Adolescent Aggression examined families and already-existing patterns of adolescent conduct. The 1961 Bobo doll investigation, conducted by Albert Bandura, Dorothea Ross, and Sheila A. Ross, experimentally exposed young children to models and measured subsequent imitative aggression in a controlled setting. It addressed an experimentally narrower causal question about the short-term effect of modeled behavior on behavior in the laboratory (Bandura, Ross, and Ross, 1961).
A retrospective account of Bandura's research can explain how family-based observations contributed to interest in modeling. The monograph and the Bobo doll work remain different studies with different participants, measurements, ecological conditions, and inferential strengths. Neither is the Little Albert experiment, which belongs to a separate research tradition associated with John B. Watson and Rosalie Rayner.
The Research Design
Two groups of adolescent boys
Contemporary scholarly discussion and a later historical review describe a comparison of 26 boys with a history of aggressive and antisocial behavior and 26 comparison boys, for a total of 52. The groups were selected and matched on relevant background characteristics. A later account specifies that the sampled boys were of average or above-average intelligence, came from intact homes, and had parents in steady employment. These selection restrictions define the population to which the results could most directly speak (Gaier, 1961; Kolvin, Ounsted, and Lee, 1967).
The selection was useful for comparing social histories while reducing certain obvious sources of heterogeneity. It simultaneously narrowed the scope of inference. The findings cannot automatically generalize to adolescent girls, diverse family structures, adolescents with intellectual disabilities, other socioeconomic conditions, or families living in different cultural and historical contexts. They also cannot establish the population prevalence of particular parenting practices or aggressive behavior, because the investigators intentionally selected groups defined by behavior rather than drawing a representative community sample.
Interviews, family histories, and psychological measures
A central method was intensive interviewing with the boys and their parents. Historical review evidence also describes the use of projective tests in the investigation. The assessment targeted child-training experiences, parent–child attitudes, feelings about authority, expressions of aggression, and related social and psychological characteristics (Gaier, 1961).
The various information sources gave the authors the opportunity to compare the adolescents' descriptions with parental reports and to connect present behavior with accounts of earlier family relationships. Such breadth was a methodological strength relative to a single informal observation. It did not eliminate errors in recall, socially desirable reporting, inconsistent interpretation of interview material, or the possibility that members of different groups described equivalent behavior differently.
Projective procedures were common in mid-century personality research. Their interpretation requires special caution today: an investigator's theoretical interpretation of a response is not equivalent to an objectively validated measurement of a stable latent trait. Without contemporary psychometric reanalysis, a 1959 projective result should be represented as historical assessment evidence, rather than treated as a modern diagnostic finding. This distinction applies even when the original investigators interpreted test results alongside detailed interviews.
What the study could compare
The design could identify differences between groups in reported family practices, emotional relationships, attitudes toward parents or peers, and patterns of aggressive expression. Where measures were scored and compared, it could estimate associations within this particular sample. Convergence between independent informants would strengthen an interpretation; disagreement would demand investigation of context and method effects.
The study could not independently manipulate the early family environment, randomly assign a boy to a particular style of upbringing, or prospectively observe the full sequence from childhood practices to adolescent behavior. Therefore, a group difference is compatible with several causal histories. A parental response might precede aggression, follow aggression, interact reciprocally with it, or correlate with other familial and social influences. These alternative structures are especially relevant when interviewing a family after a behavioral pattern has developed.
Main Findings and Historical Interpretations
Aggression as behavior expressed in social settings
A period review reported that the aggressive group expressed antagonism more directly, especially outside the home, and showed stronger conflict with authority figures and less favorable attitudes toward peers. The reviewers also described critical and distrustful attitudes toward fathers as prominent in the aggressive group. These are historical group-level descriptions reported through secondary contemporary coverage, not estimates that every member of the group showed each characteristic (Gaier, 1961).
The distinction between aggression at home and aggression elsewhere matters. Behavior changes across audiences, settings, anticipated sanctions, and relationships. A child who inhibits opposition in the presence of an authority figure may express the same disposition differently with peers. The pattern can be understood through contextual regulation without assuming a single aggression trait that operates identically in every setting.
Internalized standards, guilt, and punishment
The theoretical hypothesis emphasized differences in the internalization of parental standards. The authors and their reviewer considered whether adolescents with aggressive conduct relied more on externally administered consequences, such as fear of punishment, whereas comparison adolescents were more likely to experience internal restraints associated with guilt and accepted social standards (Gaier, 1961).
This proposal is historically consequential, but the interpretation has levels. A respondent's stated fear, a projective-test interpretation, the investigator's inference about internal restraint, and directly observed behavior are different types of evidence. Contemporary psychology would separate moral emotions, effortful control, executive functioning, social cognition, and compliance rather than assume that one measure of guilt explains them all. The 1959 work can be credited for asking how social standards become effective regulators of behavior while retaining caution about the precision of its original measurements.
Parental warmth, rejection, and hostile relationships
The book focused on family interrelationships as a possible environment for the development and regulation of aggression. Historical interpretations connected aggressive behavior with family hostility, rejection, and difficulties in identification with parental figures. These dimensions were considered together with disciplinary responses rather than as isolated parenting variables.
The evidence permits the claim that the research program investigated these patterns and found historically reported differences between selected groups. It does not license the rule that a particular aggressive child must have experienced parental rejection or that a particular parent's conduct inevitably produces aggression. Modern estimates demonstrate that familial processes are probabilistic, shaped by development and context, and accompanied by variability between individuals.
The problem of punitive responses and modeled conduct
A central conceptual tension runs through the book's research question: a parent may prohibit aggression while also modeling harsh, coercive ways of responding to disagreement. A youth may thus encounter conflicting behavioral information concerning aggression's acceptability and effectiveness. The 1959 investigation is part of the intellectual lineage that made this tension visible, and Bandura's subsequent experimental and theoretical work gave observational learning a more explicit formulation (Bandura, Ross, and Ross, 1961; Bandura and Walters, 1963).
Still, three mechanisms require separate evidence: direct reinforcement of aggressive acts, imitation of an aggressive model, and reactions to rejection or inconsistent discipline. They may coexist in families but are not interchangeable. A retrospective correlation between parental punishment and adolescent aggression cannot reveal which mechanism is operative, or establish whether all three are involved. Research designed to isolate one process must measure or manipulate it independently.
Fathers and identification
The work belongs to a period that gave particular attention to boys' relationships with fathers and to identification as a developmental concept. Historically, investigators sometimes interpreted hostility toward a father, perceived paternal rejection, or difficulty identifying with that parent as potential contributors to aggressive patterns. These explanations reflect specific theoretical and social assumptions of their time.
Modern interpretation requires both inclusion and restraint. A father may be important, and paternal warmth, involvement, behavioral regulation, and conflict can be measured; the psychological significance of caregiving and modeled behavior is not confined to a single parent or to male caregivers. Family roles vary, and social learning can occur through mothers, siblings, peers, teachers, media, and other influential figures. Findings from 52 selected boys in the 1950s cannot establish a gender-universal hierarchy of parental effects. The later parenting literature likewise documents differences in how much research has examined mothers and fathers (Hoeve et al., 2009).
From the 1959 Study to Social Learning Theory
A developing explanation, not a completed later theory
The early monograph explored how aggressive conduct related to the family as a social environment. The subsequent joint book Social Learning and Personality Development (1963) expanded learning-based explanations of personality, and Bandura's work on observational learning supplied more direct experimental tests. The canonical conceptual sequence is therefore family-focused comparative research, extended social-learning formulation, experiments on modeling, and the more elaborated social cognitive account (Bandura bibliography).
This chronology supports a coherent intellectual interpretation while preserving attribution. Richard H. Walters was coauthor of the first two books. Dorothea Ross and Sheila A. Ross were coauthors of the 1961 modeling paper. Bandura's later independent writings, including Aggression: A Social Learning Analysis (1973), involved a distinct authorship and explanatory scope. A citation to a later Bandura theory should clarify that it is a later formulation instead of presenting it as a verbatim 1959 conclusion.
Acquisition, inhibition, and expression
One enduring analytical distinction is between learning a behavioral repertoire and deciding whether to perform it. Someone may acquire aggressive scripts from observation while suppressing them because of expected sanctions, moral standards, anticipated social reactions, or contextual constraints. Bandura's 1965 work directly investigated how observed reinforcement contingencies influenced the acquisition and performance of imitative responses, advancing the separation between what observers learn and what they enact (Bandura, 1965).
Applied retrospectively to the 1959 family investigation, this distinction offers a useful interpretive lens: family discipline and models can influence both available behavior and its expression. It does not turn the earlier correlational data into an experiment on modeling. It also warns against defining aggression solely by a momentary observed action, since restraint and setting may alter whether a learned response appears.
Social learning and reciprocal influence
Later scholarship increasingly treated child development as transactional: parents influence children, and children's behavior influences parenting. This is compatible with the recognition that feedback and reciprocal adaptation occur in social relationships, but it must be evaluated with longitudinal measurements rather than deduced from the 1959 group comparison. The historical study opened questions later designs were better positioned to test, including the temporal ordering of parental responses and adolescent conduct.
What Later Research Confirms
Parenting and delinquency: a broad meta-analytic perspective
Hoeve and colleagues' 2009 meta-analysis synthesized 161 published and unpublished manuscripts on parenting and delinquency. The strongest links were reported for parental monitoring, psychological control, and negative aspects of support such as rejection and hostility, with these factors accounting for up to 11 percent of the variance in delinquency in the analysis. Moderators included the adolescent's age, the genders of parent and child, the identity of the reporter, and the type of delinquency. Fewer than one fifth of the studies focused on fathering, a relevant limitation when comparing modern evidence with father-centered historical explanations (Hoeve et al., 2009).
This synthesis corroborates the scientific relevance of family relational variables, while placing their explanatory magnitude and diversity in perspective. An association accounting for a fraction of outcome variation shows that additional developmental and contextual variables matter. The synthesis does not uniquely validate Bandura and Walters's 1959 sample, measures, or proposed internalization mechanism; it assesses a broader and methodologically heterogeneous literature.
Parenting dimensions and externalizing behavior
A major meta-analysis by Martin Pinquart brought together 1,435 studies of parenting and externalizing problems in children and adolescents. Parental warmth, behavioral control, autonomy granting, and authoritative parenting were associated with lower externalizing problems, while harsh and psychological control were associated with higher levels. Associations varied in magnitude and by methods and populations. Longitudinal analyses also found indications of bidirectionality between parenting and children's externalizing behavior (Pinquart, 2017).
Externalizing problems commonly encompass aggressive, rule-breaking, and disruptive behavior, depending on the instruments employed. They are related to but broader than the specific historically defined adolescent aggression group in the 1959 monograph. A finding about aggregate externalizing symptoms should not automatically be interpreted as a finding about violent criminal offending. The updated evidence supports the importance of parenting processes while making clear why contemporary work requires precise outcome definitions.
Physical discipline and developmental outcomes
Gershoff and Grogan-Kaylor examined spanking in a meta-analysis of 111 distinct effect sizes representing 160,927 children. Most examined mean effects showed associations with unfavorable outcomes, including aggression-related measures. The evidence addresses correlations and other designs across different contexts and supports concern about physical punishment as a developmental risk factor. It does not imply that the 1959 monograph itself experimentally demonstrated these later aggregate effects (Gershoff and Grogan-Kaylor, 2016).
This distinction is important because aggressive behavior might elicit increased harsh discipline and because common causes can influence both variables. Longitudinal controls, within-family analyses, and intervention evidence provide different leverage on causal claims. Their findings should be compared rather than reduced to a single unequivocal statement about one family's history.
Contemporary work on bidirectional development
A longitudinal investigation published in 2026 examined maltreatment and aggression across five annual waves in a South Korean youth panel of 2,267 participants aged roughly 10–14. It was explicitly designed to investigate reciprocal associations instead of treating adversity as a purely one-way predictor. Its setting and measurements differ substantially from Bandura and Walters's 1950s sample, but its research question illustrates the development of temporal designs in this field (Lee, 2026).
Such research cannot retroactively determine what caused the original group's aggressive behavior. It can improve the conceptual framework for interpreting that group's findings. Broadly, the evidence supports investigating interacting parent and adolescent behavior over time; the exact pathways, effect sizes, and generalizability depend on the study design and population.
Evidence Status: What Is Established, Limited, or Contested?
Established historical facts
The authorship, date, publisher, length, and research focus of the monograph are firmly documented. Historical coverage supports the comparison of 26 aggressive and antisocial boys with 26 comparison boys and an interview-based examination of participants and parents. It is also established that the book preceded the coauthors' 1963 social-learning monograph and the subsequent public prominence of the Bobo doll experiments (NLM catalog; Gaier, 1961).
Supported broad developmental associations
Modern meta-analyses support an association between parenting dimensions—including rejection, monitoring, harsh or psychological control—and several measures of delinquency or externalizing problems. The direction and strength vary with design, age, informant, and behavioral definition. This is credible evidence that social relationships belong in the explanation of adolescent aggression. It remains distinct from direct proof of a unique 1959 causal mechanism (Hoeve et al., 2009; Pinquart, 2017).
Limited original causal evidence
The 1959 study was a selected-group comparison relying substantially on retrospective and cross-sectional assessment. Its small all-male sample, its social and family inclusion restrictions, its era-specific measures, and likely overlap between reporter and group definitions restrict generalization and causal identification. An interpretation that a particular parenting pattern caused an individual's later aggression exceeds what this design alone can prove.
Contested or historically bounded constructs
Broad notions of identification, internalization, and guilt were plausible explanatory terms in mid-century personality research. Their measurement, discriminant validity, and proposed causal roles would require modern operationalization and testing. A psychological construct does not become empirically established merely because it organized an influential historical study. The same caution applies to retrospective explanations that describe one parent as the central source of aggression without direct mechanism-specific evidence.
Methodological Strengths and Limitations
Strength: A defined comparison question
The investigators placed different groups under a common empirical question: which family processes and patterns of social control differ across adolescents identified by aggressive behavior? An explicit comparison is more informative than isolated clinical anecdotes because it invites consistent measurement and systematic examination of alternative explanations. Intensive collection of adolescent and parental accounts also broadened the descriptive material available for analysis.
Limitation: Selection and range restriction
A deliberately chosen group is valuable for investigating contrast, but it distorts prevalence estimates and may exaggerate or obscure associations found in a general population. Boys from intact, employed households of at least average intelligence reflect a restricted historical sample. Adolescents outside those conditions can have different exposures, protective factors, institutional contexts, and behavioral pathways. A case–comparison design therefore supports conditional findings about the sample, rather than claims about all families.
Limitation: Recall and shared method variance
Adults remembering child-rearing practices after a son's difficulties have become visible may reconstruct the past in light of present events. Adolescents and parents may also disagree about whether punishment was fair, consistent, harsh, or frequent. If behavior-group membership influences reporting, the observed association between childhood practice and later conduct can be partly a product of recall or measurement. Converging interviews help but do not automatically remove shared method variance or retrospective bias.
Limitation: Developmental timing
Social learning and aggression unfold across childhood and adolescence, but a study concentrated on adolescents at one point cannot tell when each difference first arose. The meaning of parent–child conflict also changes with age, autonomy, peer exposure, pubertal development, and school context. Prospective cohorts are better suited to specifying sequences of change and cross-lagged relationships, though even they still require careful assumptions about unmeasured causes.
Limitation: Construct and cultural equivalence
Terms such as guilt, hostility, paternal identification, permissiveness, and antisocial conduct were embedded in a particular research and social environment. In an international encyclopedia, historical measurements should be situated in their setting. Modern research must assess whether the same instrument and criterion have equivalent meanings across gender, family structure, ethnicity, culture, and historical period. Generalization is an empirical proposition to test.
Alternative Explanations That a Modern Research Program Must Consider
Temperament, impulsivity, and self-regulation
Individual differences in emotional reactivity, impulsivity, attention, and regulatory capacities could influence both adolescent behavior and parental responses. These differences may amplify family conflict or modify how discipline works. A measured parenting–aggression association need not represent a single environmental cause. A convincing mechanism test would measure relevant child characteristics, temporal changes, and family processes rather than attribute all variation to parental practice.
Peers, school, and community environments
Adolescence expands opportunities for learning and reinforcement beyond the family. Peer approval, neighborhood conditions, school exclusion, exposure to violence, and access to supportive adults may reshape behavior. Some mechanisms could be consistent with observational learning, while others involve opportunity, stress, social position, or institutional response. An adequate developmental account must distinguish these possible pathways and how they combine within a person's environment.
Reciprocal parent–adolescent transactions
Aggressive conduct may lead caregivers to increase surveillance, punish more often, withdraw affection, or respond inconsistently under strain. Those responses may in turn alter behavior. Such cycles can produce robust associations even when no single event marks the origin. A cross-sectional group comparison sees their accumulated products. Longitudinal and experimental designs help determine which changes precede others, although reciprocal models still depend on the validity of measurement and controls.
Poverty, adversity, and structural constraints
Family processes unfold within employment, housing, security, discrimination, and access to services. The 1959 study's sample restrictions controlled some of these conditions by selection, but they also made social structural variation less visible. Contemporary analysis should examine those contexts directly. Otherwise, family interaction can be mistaken for an entirely isolated private cause while the pressures shaping it go unmeasured.
Implications for Research and Practice
Research design and interpretation
The most useful lesson from the 1959 monograph is the need to make pathways between social experience and behavior measurable. Researchers can compare child reports, caregiver reports, observational measures, school records, and longitudinal trajectories to avoid depending on a single perspective. Separately validated measures are needed for harsh discipline, warmth, monitoring, coercive exchanges, hostility, empathy, social norms, and aggressive acts. Analyses should explicitly identify confounders, selection mechanisms, and possible reverse causation.
A causal claim requires stronger design than a statistically significant group difference. Randomized family interventions, when ethically appropriate, can test changes induced by an intervention; longitudinal approaches can assess temporal precedence; natural experiments can sometimes provide additional leverage. No study type is sufficient by its label alone. The question is whether alternative explanations for the observed relationship have been addressed convincingly.
Clinical and educational relevance
An adolescent's aggression calls for contextual, developmentally informed assessment. Its form, frequency, severity, triggers, victim impact, and functions need clarification, along with safety concerns, family relationships, peer environments, potential trauma, and co-occurring difficulties. A historical explanatory theory does not supply an individual diagnosis. Psychological and educational responses should depend on contemporary evidence, appropriate assessment, and the specific person and setting rather than an inference that one parent caused the behavior.
Family relationships are legitimate targets for supportive interventions when coercive interactions, inconsistent responses, or conflict are present. Educational settings can also examine modeled behavior, reinforcement, prosocial norms, and opportunities for regulation. The 1959 work is a foundation for formulating research questions; the selection of particular treatments belongs to current disorder-specific and intervention-specific evidence, not to the authority of a classic book alone.
Avoiding individual blame
The presence of child aggression does not identify a guilty parent or establish a singular cause. Equally, the existence of family conflict does not justify harm toward children, and concerns about harsh punishment have an independent contemporary evidentiary basis. The practical value of learning-based explanations lies in identifying modifiable relational patterns while preserving a nuanced account of the adolescent's development, agency, and surroundings.
Intellectual Legacy
Why the book matters in Bandura's research trajectory
Adolescent Aggression documents an early effort to investigate aggression in a social developmental framework. It helped establish questions about learned conduct, contextual restraints, parental models, and the internalization of standards. Subsequent empirical work separated experimental imitation from broader developmental correlates and later theorizing addressed expectations, agency, and self-regulation. The book is therefore best understood as part of an evolving research program whose successive stages used distinct methods.
Why Richard H. Walters remains essential to its history
Richard H. Walters shared authorship of both the 1959 aggression monograph and the 1963 social-learning book. His bibliographic status is a substantive component of the scientific record, not a footnote to another psychologist's reputation. Accurate attribution also helps separate the jointly authored work from later Bandura writings and from the experiments performed with other collaborators. Such distinctions are necessary for reconstructing how psychological ideas actually developed.
What remains valuable today
The lasting scientific question is how family conduct, social control, internal processes, and wider environments interact as aggression develops. Modern meta-analyses continue to find meaningful associations between parenting and antisocial or externalizing behavior, while documenting heterogeneity, reciprocal influences, and limited effect sizes for individual factors. The 1959 study thus has a firm place in the history of psychological explanation and a more carefully bounded place in current causal evidence (Hoeve et al., 2009; Pinquart, 2017).
Frequently Asked Questions
Who wrote Adolescent Aggression in 1959?
Albert Bandura and Richard H. Walters were joint authors. Robert R. Sears wrote a foreword. The correct bibliographic record credits both researchers rather than describing the book as Bandura's single-author work (National Library of Medicine).
What did Bandura and Walters study?
They investigated aggressive and antisocial conduct in adolescent boys in relation to reported child-rearing practices, parental relationships, emotional attitudes, and forms of social control. Historical reviews report a comparison of 26 aggressive boys with 26 comparison boys, using interviews with boys and their parents along with psychological assessment (Gaier, 1961).
Did the study prove that harsh parenting causes aggression?
Its group comparison could establish reported associations in a selected sample, but it could not isolate a parenting behavior as a cause. More recent longitudinal and meta-analytic work links harsh control and other parenting dimensions to externalizing outcomes and recognizes reciprocal processes (Pinquart, 2017).
Is the book the same as the Bobo doll experiment?
No. The book is a 1959 family-focused adolescent investigation written by Bandura and Walters. The classic 1961 Bobo doll article is an experimental study of young children's observational learning by Bandura, Dorothea Ross, and Sheila A. Ross (Bandura, Ross, and Ross, 1961).
How did this work influence social learning theory?
It posed empirical questions about how social environments and learned controls relate to aggressive behavior. The coauthors' 1963 Social Learning and Personality Development developed a wider theoretical account, and later modeling experiments and Bandura's publications refined separate learning and performance mechanisms. This is an intellectual sequence, not an assertion that the entire later theory was experimentally proven in the 1959 study (Bandura bibliography).
What is the book's current scientific status?
Its bibliographic identity, historical significance, and selected-group comparative design are established. The family–behavior relationships it examined are broadly compatible with modern evidence. Its precise mechanism-specific causal explanations and generalizability are limited by sample selection, historical measurement, and retrospective assessment; the broader field continues testing reciprocal, developmental, and contextual mechanisms.
Related Articles
Albert Bandura — A comprehensive scientific account of Bandura's life, research program, original works, and intellectual influence.
Social Learning and Criminal Behavior — An examination of social learning approaches in criminology, including their theoretical distinctions from Bandura's early work.
Moral Disengagement and Criminal Offending — An applied treatment of moral disengagement and offending, connecting historical theory with a distinct criminological intent.
Albert Bandura's Theory of Moral Disengagement — Mechanisms of suspending moral self-sanctions during harmful conduct.
Albert Bandura's Aggression: A Social Learning Analysis — The original 1973 account of aggression and social learning.
References
Bandura, A. (1965). Influence of models' reinforcement contingencies on the acquisition of imitative responses. Journal of Personality and Social Psychology, 1(6), 589–595. https://doi.org/10.1037/h0022070
Bandura, A. (1973). Aggression: A Social Learning Analysis. Prentice-Hall. https://albertbandura.com/albert-bandura-books.html
Bandura, A., Ross, D., and Ross, S. A. (1961). Transmission of aggression through imitation of aggressive models. Journal of Abnormal and Social Psychology, 63(3), 575–582. https://doi.org/10.1037/h0045925
Bandura, A., and Walters, R. H. (1959). Adolescent Aggression: A Study of the Influence of Child-Training Practices and Family Interrelationships. Ronald Press. https://www.ncbi.nlm.nih.gov/nlmcatalog/53451
Bandura, A., and Walters, R. H. (1963). Social Learning and Personality Development. Holt, Rinehart and Winston. https://wellcomecollection.org/works/wcam8z4x
Gaier, E. L. (1961). Book reviews: Adolescent Aggression by Albert Bandura and R. H. Walters. International Journal of Social Psychiatry, 8(1), 71–72. https://doi.org/10.1177/002076406100800114
Gershoff, E. T., and Grogan-Kaylor, A. (2016). Spanking and child outcomes: Old controversies and new meta-analyses. Journal of Family Psychology, 30(4), 453–469. https://doi.org/10.1037/fam0000191
Hoeve, M., Dubas, J. S., Eichelsheim, V. I., van der Laan, P. H., Smeenk, W., and Gerris, J. R. M. (2009). The relationship between parenting and delinquency: A meta-analysis. Journal of Abnormal Child Psychology, 37(6), 749–775. https://doi.org/10.1007/s10802-009-9310-8
Kolvin, I., Ounsted, C., and Lee, D. (1967). Aggression in adolescent delinquents. British Journal of Criminology, 7(3), 296–314. https://www.kolvinpsych.net/node/318/
Lee, W. (2026). Bidirectional association between child maltreatment and adolescent aggression from early-to-mid adolescence. Aggression and Violent Behavior, 87, Article 102136. https://doi.org/10.1016/j.avb.2026.102136
National Library of Medicine. (1959). Adolescent Aggression [Bibliographic catalog record, NLM ID 0053451]. https://www.ncbi.nlm.nih.gov/nlmcatalog/53451
Pinquart, M. (2017). Associations of parenting dimensions and styles with externalizing problems of children and adolescents: An updated meta-analysis. Developmental Psychology, 53(5), 873–932. https://doi.org/10.1037/dev0000295