RESEARCH DOSSIER
Religious Socialization and Development Across the Lifespan
Abstract. Religious socialization comprises the acquisition, interpretation, enactment, and revision of religious concepts, practices, commitments, and identities across changing social relationships and historical contexts. This fundamental research monograph synthesizes developmental, social-learning, testimony-based, family-relational, life-course, and intergenerational theories with evidence from child studies, multigenerational panels, longitudinal youth cohorts, aging research, comparative surveys, and qualitative investigations. It examines childhood religious learning, parental and grandparental transmission, schools and peer networks, adolescent exploration, emerging-adult transitions, midlife changes, later-life trajectories, and cohort-based transformations. The evidence demonstrates heterogeneous developmental pathways and divergent trends across attendance, personal prayer, perceived importance, and affiliation. The analysis critically evaluates construct validity, retrospective recall, age–period–cohort inference, attrition, causal identification, cross-cultural generalizability, and the growing significance of digital and artificial environments. It concludes with an integrated research agenda grounded in multi-indicator longitudinal and mixed-method study.
Keywords: religious socialization; religious development; lifespan psychology; intergenerational transmission; childhood religious learning; adolescent religiosity; religious identity; emerging adulthood; family religion; religious trajectories; religious switching; religious aging; cohort replacement; religious testimony; cross-cultural psychology
Scientific Overview: Religious Socialization as a Life-Course Process
Religious socialization is the developmental process through which people encounter, interpret, enact, negotiate, retain, or revise religious symbols, narratives, identities, practices, and relationships in situated social environments. It includes explicit instruction and indirect learning through family routines, community participation, ritual performance, conversation, observation, testimony, and the opportunities and constraints of historical circumstances. Development across the lifespan refers to continuing changes in the organization of those elements, from early childhood learning through adolescence, adult role transitions, and later-life reflection. The research object is therefore a changing relation between persons and religious social worlds, rather than a single score of religiosity. The life-course perspective emphasizes linked lives, socially timed transitions, and historical context Elder (1998).
This monograph asks how religious concepts are acquired; which family, peer, educational, and communal relationships predict transmission; how people differentiate inherited practices from personally endorsed identities; what trajectories are found in longitudinal evidence; and how historical cohorts, migration, minority status, and institutional change modify the developmental picture. A further question concerns method: whether a measured decrease in attendance is evidence of reduced belief, an altered institution, changing opportunity, or some combination. Recent longitudinal findings demonstrate that communal attendance, personal prayer, subjective spirituality, and religious coping can follow different age-related trajectories Dew et al. (2020) Nelson et al. (2024).
The central empirical premise is heterogeneity. Religious socialization can support continuity, selective reconstruction, intensified engagement, mixed identification, or disengagement. Individuals may retain inherited ritual skills while revising theological commitments, or maintain religious identification while reducing congregational participation. Growth-mixture analyses of intergenerational families have found stable similarity alongside trajectories in which children become less religious, become more religious, or return toward parental levels later in life Min et al. (2018). These descriptions are statistical classifications of observed data and do not establish fixed psychological types.
Religious socialization operates at several nested levels. Individual development involves memory, inference, imitation, trust in testimony, emotion, and identity work. Family interaction includes parental modeling, conversations, relationship quality, and relatives across generations. Peer and institutional environments provide alternative reference groups, rituals, sanctions, belonging, and public narratives. Historical and national environments condition the cultural availability of religious and nonreligious pathways. The recently published Cambridge synthesis by Lewis-Jong and Bullivant (2026) explicitly organizes transmission evidence across social, parental, minority, educational, and national contexts Lewis-Jong & Bullivant (2026).
A developmental account must specify what is being transmitted. Religious affiliation is a category; religious commitment is a judgment of importance; observance is patterned behavior; spiritual experience is a subjective report; religious knowledge comprises acquired concepts and narratives; and family religious culture includes shared expectations, language, objects, and practices. These indicators overlap but are neither interchangeable nor equally sensitive to age, culture, and institutional opportunity. The American longitudinal analysis of Dew and colleagues observed falling attendance alongside increasing reported religious importance and comparatively stable prayer during ages 12 to 34 Dew et al. (2020). A one-dimensional account would misdescribe this combination.
The investigation uses an integrative critical review of independently identified historical works, longitudinal studies, comparative surveys, developmental experiments, and research syntheses. It is a scholarly narrative synthesis: no preregistered exhaustive search, standardized risk-of-bias protocol, pooled effect estimate, or systematic-review claim is made. Where source access was limited to scholarly abstracts, the evidence reported remains within the methods and findings visible there. The explanatory claims developed below distinguish observed associations, interpretation, mechanisms proposed by theories, and future hypotheses.
Conceptual and Historical Foundations
Modern psychological accounts of religious development emerged through several overlapping traditions: developmental theories of belief and meaning, social-learning approaches to the acquisition of practice, sociological accounts of generational continuity, and lifespan theories of changing social roles. Their historical contributions address different questions. Theoretical stage approaches describe possible transformations in how people organize faith-related meaning; social-learning theory identifies mechanisms of behavioral acquisition and reinforcement; and longitudinal family research asks whether specific forms of similarity persist and under what conditions. Placing these traditions in sequence exposes an important shift from normative descriptions of mature faith toward empirically observed, heterogeneous trajectories Fowler (1981) Bandura (1977) Elder (1998).
James W. Fowler's Stages of Faith, published in 1981, provided an influential account of the structural development of faith as meaning-making Fowler (1981). Its intellectual importance lies in the proposal that persons reorganize commitments and interpretive structures over time. The model's stage language invites specific tests: whether the posited ordering holds across traditions; whether interview judgments measure the same construct among religious and nonreligious participants; and whether apparent stages follow personal experience rather than educational access or assumptions embedded in scoring. In this monograph, Fowler's model is treated as a historical theoretical framework, not an empirically established universal ladder.
Albert Bandura's Social Learning Theory (1977) offers a different explanatory entry point. Children and adults can acquire behavioral repertoires by observing admired or authoritative models, remembering their actions, encountering consequences, and adapting performance to context Bandura (1977). Applied to religion, this suggests why ritual repetition, household conduct, parental consistency, and the visibility of everyday moral practices may matter even when doctrine is rarely taught explicitly. The theory supplies general mechanisms, while religion-specific claims require independent evidence. Observed similarity between a child and a parent alone cannot identify imitation, instruction, shared genetic influences, selection into neighborhoods, or institutional reinforcement.
Glen H. Elder Jr. formulated the life course as a developmental framework attentive to social timing, historical circumstance, and interconnected biographies Elder (1998). Religious commitments are particularly suitable for this analysis because affiliation, attendance, identity, and practice can change when individuals leave a parental household, migrate, form partnerships, parent children, experience loss, or enter old age. These transitions need not exert uniform effects. An event can intensify participation in a community for one person and occasion questioning or disengagement for another. The explanatory problem is to observe transitions prospectively and identify the accompanying institutional and relational mechanisms.
Long-term cohort work transformed the field by following family members rather than asking a single generation to recall childhood retrospectively. Bengtson, Putney, and Harris synthesized multigenerational findings in Families and Faith (2013), paying particular attention to warmth and the quality of relationships connecting religious parents with their children Bengtson (2013). The later study by Min, Silverstein, and Gruenewald tracked 1,084 dyads across eight waves between 1971 and 2005 and found multiple patterns of relative parent–child religiosity Min et al. (2018). These studies helped establish that transmission is neither automatic repetition nor a uniformly one-directional childhood effect.
Contemporary work combines this tradition with interview-based family research, comparative cohorts, and evidence on children learning invisible and metaphysical propositions. Smith and Adamczyk's 2021 study used 215 in-depth interviews with religious parents and analyses of two nationally representative US parent surveys to examine how parents intentionally organize religious upbringing Smith & Adamczyk (2021). Harris and Koenig's 2006 developmental analysis of trust in testimony addressed how children learn claims that cannot be established through direct perception alone, including scientific and religious claims Harris & Koenig (2006). Together these traditions motivate an account of transmission as learning, relational interpretation, and continuous revision within cultural systems.
Constructs, Operational Definitions, and Explanatory Levels
Affiliation denotes reported identification with a religion, denomination, tradition, or with no religious group; participation denotes attendance or activity; devotion includes personal prayer or other independent observance; importance refers to evaluative centrality; knowledge denotes familiarity with concepts and practices; and spirituality concerns reported experiences or orientations that respondents may or may not associate with institutions. Operational definitions should record question wording, time frame, response scale, and the respondent's own interpretation. Comparing religious-development findings requires separating nominal affiliation from actual behaviors and inner orientation. National surveys show that changing one's stated religious affiliation and changing the intensity of religiousness are distinct processes Pew Research Center (2025a) Pew Research Center (2025b).
Transmission can be specified as similarity of parent and child affiliation, participation, value priority, beliefs, or behavior at a particular age. A longitudinal transmission estimate additionally requires dates, family relationships, baseline levels, and an explicit model of change. Congruence is a relational measure, not a direct observation of psychological influence. Parent–child dyads may be concordant because they share institutions and exposure, because adolescents adopt parental models, or because parents adjust their practices in response to their children. Designs with repeated reports from both generations improve temporal interpretation, although residual confounding and selective attrition remain Min et al. (2018) Bengtson (2013).
Continuity, intensification, weakening, and return describe possible change patterns. They are operationalized through longitudinal trajectories of independently measured indicators. Mixture models are useful when participants follow different paths, but the number and shape of classes depend on indicators, time points, model assumptions, missing-data treatment, and selection criteria. A trajectory class is a statistical approximation rather than an immutable category. Petts (2009) identified stable and declining attendance trajectories from adolescence into young adulthood; those classes should be read in relation to attendance and that study's population, not automatically generalized to belief or spirituality Petts (2009).
A developmental mechanism is a hypothesized process connecting exposure to later outcomes. Observational learning, selective trust in testimony, affectively meaningful relationships, social identity formation, institutional opportunity, and reinforcement from peer networks are plausible mechanisms, but their evidentiary status varies. A model can predict that consistent parental practice increases a child's opportunities for learning while also predicting that an authoritarian or conflictual relationship changes how that practice is interpreted. Bengtson and colleagues reported an association between warm, affirming parental relationships and religious continuity; causal direction was not settled by that observation alone Bengtson (2013).
Culture shapes the meaning of each variable. Attendance at a formal congregation may be central in some settings, while household rituals, pilgrimage, family observances, or informal community practice may be more representative elsewhere. The same frequency item can therefore have different construct validity in different religious traditions. A comparison of Muslim and Christian youth in Germany revealed heterogeneous trajectories and different religious-development profiles across immigrant-origin groups Spiegler et al. (2025). This result makes group-specific measurement and context-sensitive interpretation essential components of psychological explanation.
At the highest explanatory level, religious socialization concerns the distribution of practices and identities across cohorts. At the individual level, it concerns learning and life histories. These levels are related through families, schools, religious communities, and wider institutions, but population-level change cannot be inferred directly from individual aging. The British analysis by Voas and Crockett (2005) argued for generational replacement as an important explanation of religious decline, whereas later-life longitudinal studies documented within-person changes as well Voas & Crockett (2005) Hayward & Krause (2013). A rigorous synthesis retains both processes and asks which observations identify each.
Early Childhood: Participation Before Explicit Doctrine
Religious socialization often begins before children can articulate propositions about divinity, sacred authority, or supernatural causation. Young children encounter adults' gestures, seasonal routines, stories, objects, music, special places, and expectations of reverence as recurring elements of domestic and communal life. These practices furnish recognizable sequences and emotional contexts that later support explicit interpretation. The early developmental task is therefore best described as acquiring access to a culturally organized repertoire. It would be premature to infer theological assent from a toddler's participation, or to infer the absence of learning because the toddler cannot explain symbolic meaning. Social-learning theory makes attention, modeling, and repeated enactment central mechanisms to investigate Bandura (1977).
Adult responses shape which behaviors are repeated. A parent who invites questions and makes explanations available provides a different learning environment from one who makes ritual primarily an occasion for discipline or anxiety. The long-term research summarized by Bengtson and colleagues associates warm, supportive relations with more frequent intergenerational religious continuity Bengtson (2013). That association is consistent with a mechanism in which trust and emotional security make parental models salient, although it may also reflect unmeasured features of families that are simultaneously warm and religiously cohesive. Studies must measure parental and child reports separately to reduce common-method distortion.
Children develop religious categories through conversation as well as participation. Names for deities, ancestors, sacred events, rules, and community identities can be learned through ordinary language practices. These categories carry distinct kinds of claims: some can be verified publicly, some are historical, and some concern entities or events beyond ordinary observation. Harris and Koenig (2006) showed why children's reliance on testimony matters to the acquisition of knowledge about scientific and religious domains Harris & Koenig (2006). Testimonial learning is a general developmental capacity. Its content depends on who speaks, the community's credibility practices, the consistency of accounts, and how alternative explanations are encountered.
Evidence from cross-cultural developmental work rejects a simplistic account in which an unchanging child mind supplies identical supernatural beliefs everywhere. Astuti and Harris (2008) investigated beliefs concerning death and ancestors in rural Madagascar. Their interview studies showed that participants' judgments about continuing bodily and mental processes were sensitive to the narrative context and type of process being considered; adults expressed some afterlife continuities more frequently than children Astuti & Harris (2008). The study supports investigating how local explanatory frames and developing conceptual capacities interact. It also cautions against treating a judgment elicited by a particular story as a comprehensive measure of the respondent's religious worldview.
Age comparisons in childhood research can confound growing cognitive ability with growing cultural exposure. An older child has had more years of religious instruction, greater linguistic resources, and more opportunities to hear different accounts. A suitable study measures developmental language ability, individual socialization histories, exposure to institutions, and the interview setting. It should also compare religious and nonreligious household cultures on comparable dimensions, including narrative transmission and parental responses to uncertainty. Such designs address whether emerging differences arise from conceptual development, the organization of testimony, or the interaction between them Harris & Koenig (2006) Boyatzis (2012).
The crucial theoretical consequence is that early religious acquisition is cumulative and interpretive. The child learns whom to ask, which explanations are socially authoritative, how celebrations are organized, and when concepts apply. Explicit endorsement can follow, differ from, or be absent despite this learning. An adult's religion may therefore contain implicit practical memories that persist after doctrinal change. Establishing this possibility empirically requires evidence from later-life interviews or repeated observation, rather than presuming it from childhood ritual history.
Family Transmission, Parenting, and Generational Relationships
The family is a central site of religious socialization because it combines repeated access to models, attachment-relevant relationships, ordinary conversation, and decisions about institutional participation. Parents can select religious schools, arrange attendance, communicate moral expectations, engage in prayer, and discuss belief; they also communicate attitudes through inconsistencies between stated values and daily conduct. Smith and Adamczyk (2021) examined the intentional efforts and experiences of religious parents using interviews and national survey data, emphasizing how religious conversation and parental priorities shape the immediate environment Smith & Adamczyk (2021). Their evidence concerns a US social setting and does not warrant a universal ranking of parents above all other influences in every culture.
The character of the parent–child relationship matters. In Families and Faith, Bengtson, Putney, and Harris distinguished warm and affirming parenting from distant or authoritarian, ambivalent, and strained patterns when examining religious continuity Bengtson (2013). Warmth may sustain opportunities for open conversation and reduce the perceived social cost of asking difficult questions. It can also make practices personally meaningful rather than merely obligatory. These explanations are hypotheses consistent with the observed pattern. Shared causes such as parental commitment, marriage quality, neighborhood institutions, and selection into religious communities need explicit consideration before assigning causal effects to warmth itself.
Transmission takes place across more than two generations. Grandparents and extended relatives may carry stories, ritual knowledge, and expectations that endure while parents alter participation. Long-running linked-family studies can reveal these processes because the same family is observed at several points across social transitions. Min, Silverstein, and Gruenewald (2018), using eight waves of the Longitudinal Study of Generations and 1,084 parent–child dyads, found stable similarity, weakening, strengthening, and return patterns Min et al. (2018). Their findings establish that intergenerational congruence changes over time and that a single cross-sectional similarity coefficient obscures meaningful diversity.
Children also shape family religious practice. Adolescents may question doctrines, request participation, stop attending, or introduce ideas from peers and other communities. Parents sometimes respond by increasing discussion, modifying household practice, or reassessing affiliation themselves. This potential bidirectionality follows from reciprocal socialization theory but demands direct measurement of who initiated each change. Repeated reports, family interaction observations, and cross-lagged models can help, provided researchers distinguish within-family change from stable differences between families. The shared history of parent and child also means that causal interpretations based on a single reporter should remain restrained Bandura (1977) Min et al. (2018).
Family structure and domestic transitions can change opportunities for religious transmission. Residential movement, custody schedules, work patterns, and differing parental traditions affect which practices children can observe. Studies of youth attendance trajectories have identified family characteristics alongside adolescent transitions as predictors of the probability of different participation patterns Petts (2009). The relevant mechanism may involve opportunity, institutional access, identity, or selection; a household category alone does not reveal which. Analysis should compare actual opportunities, child choice, conflict levels, and the reported religious orientations of all participating caregivers.
Religious socialization also occurs in families that are nonreligious or deliberately pluralistic. Their members can transmit norms of skepticism, secular rituals, freedom of conscience, cultural religious literacy, and respectful engagement with other traditions. A conceptual framework confined to successful reproduction of parental faith would fail to characterize these forms of upbringing. The Cambridge review of intergenerational transmission explicitly addresses nonreligious families and contexts of declining transmission Lewis-Jong & Bullivant (2026). The appropriate outcome is a documented developmental trajectory, whether it involves religious continuity, nonreligious continuity, or revision in either direction.
Schools, Peers, Ritual Settings, and the Wider Social Ecology
Religious development takes place in ecological contexts extending beyond families. Schools may supply formal religious instruction, comparative religious literacy, moral discussion, or secular forms of public belonging. Peers provide settings where inherited identities are affirmed, contested, or treated as peripheral. Community institutions create recurring participation, recognized roles, intergenerational mentorship, and social reinforcement. The developmental question is how these contexts interact with individual capacities and family histories. Boyatzis (2012) articulated a social-ecological perspective that places family and broader contexts in relation rather than reducing development to a single influence Boyatzis (2012).
Peers can affect religious practice through friendship selection as well as influence. Young people with similar commitments may become friends because they share institutions, or a group's established practices may alter newcomers' behavior. Longitudinal peer-network designs are needed to distinguish these processes. In survey research, a correlation between having religious friends and being religious is therefore descriptive until prior commitments and friendship formation are measured. Transitions into new schools, colleges, employment, or residential settings create useful natural opportunities to observe change, although each transition can alter several social exposures simultaneously Petts (2009) Dew et al. (2020).
Ritual participation is an especially important meeting point between embodied learning and social identity. A person can learn the words, gestures, schedules, material arrangements, and emotional expectations of a ceremony through rehearsal long before endorsing all the beliefs associated with it. Participation also generates opportunities for connection across ages. These functions help explain why attendance is developmentally informative but incomplete as an index of religiosity. In the longitudinal study by Dew and colleagues, attendance, private prayer, and perceived importance diverged during early adulthood Dew et al. (2020). A method focused only on attendance risks confusing changes in institutional contact with changes in inner conviction.
Formal education introduces conceptual alternatives. Learning about other traditions, historical scholarship, science, and philosophy can encourage refinement, synthesis, doubt, or stronger commitment. The direction depends on prior identity, explanatory models, the social position of the learner, and whether questioning is treated as legitimate. A sound educational psychology of religion should examine quality of instruction, religious literacy, dialogic practice, and students' understanding instead of equating exposure to competing explanations with predetermined outcomes. Testimony research supplies one cognitive starting point for studying evaluation of information from authoritative and conflicting speakers Harris & Koenig (2006).
Minority and immigrant communities may attach religious practice to heritage language, family solidarity, and social recognition. This makes religion a possible vehicle for intergenerational continuity alongside a distinct object of belief. Spiegler, Jonsson, and Bracegirdle (2025) used seven waves of German adolescent data and identified different trajectory patterns for Muslim, immigrant-origin Christian, and non-immigrant Christian youth Spiegler et al. (2025). The contextual contrast is a strong warning that the religious trajectory commonly observed among young people in one national majority tradition cannot simply be treated as a universal developmental law.
Digital environments have become additional contexts for religious learning and identification. Young people can encounter teachings, argument, rituals, influencers, skeptical communities, and cross-tradition conversations without relying on family or local institutions as exclusive gatekeepers. Such environments may widen choice while also shaping information credibility and social feedback. The current evidence justifies treating internet exposure as a research variable in contemporary transmission studies; specific causal claims about platform effects require platform-level measurements, credible counterfactuals, and longitudinal follow-up Lewis-Jong & Bullivant (2026). The developmental unit is the person's changing pattern of interaction with multiple environments.
Adolescence: Exploration, Identity Negotiation, and Changing Authority
Adolescence expands the range of available religious and nonreligious reference groups. Young people encounter intellectual alternatives, develop more elaborate autobiographical accounts, and gradually acquire greater discretion over participation. Their questions may concern doctrines, social expectations, fairness, group boundaries, or the relationship between personal experience and inherited narratives. These processes overlap with general identity development and changes in relationships with parents and peers. A developmental analysis must distinguish cognitive evaluation from the public behavior of attendance: adolescents may remain at services for family reasons while privately reconsidering beliefs, or report strong convictions while attending irregularly because transportation or institutional access is limited Petts (2009) Dew et al. (2020).
Religious identity develops through an interaction between personal commitments and social recognition. Community membership can supply continuity, language for interpreting difficulty, and an intergenerational narrative. The same identity can become contested when a young person encounters peers who evaluate the inherited tradition differently. The developmental challenge includes deciding which inherited practices to retain, how to interpret authority, and whether belonging requires agreement on particular propositions. These are empirical questions about observed behavior and reported meanings. Researchers should avoid coding questioning itself as an automatic sign of either healthy maturation or declining well-being, because the significance of doubt depends on available relationships, institutional responses, and the individual's experience.
Petts (2009) analyzed trajectories of religious participation from adolescence into young adulthood and reported heterogeneity in patterns of attendance Petts (2009). The findings support studying different developmental paths rather than presuming that all adolescents gradually leave organized religion. However, frequency of attendance is affected by access, mobility, parental control, and calendar conflicts. Where a trajectory class is defined using attendance, it describes attendance. The distinction matters particularly in communities where institutional participation is not the primary expression of religious life, or where family obligations produce attendance without private endorsement.
Young people can simultaneously be members of families, friendship networks, religious institutions, and wider national communities. Expectations can reinforce each other or conflict. Religious minority adolescents may interpret identification as a connection to ancestry or solidarity even when their level of observance changes. Spiegler, Jonsson, and Bracegirdle (2025) used a person-oriented approach and seven longitudinal waves to examine trajectories among Muslim and Christian youth in Germany Spiegler et al. (2025). They identified different developmental patterns across the studied groups. These results highlight the interaction between developmental timing and context, while the sampling frame and measurement choices limit simple extension to other nations or religious populations.
Questions about religion are also questions about knowledge. Adolescents gain tools for evaluating sources, considering contradictions, and understanding that people may sincerely maintain divergent explanations. Earlier testimonial learning remains relevant, but the authority structure becomes more complex as teachers, friends, family members, clergy, social media, and scientific institutions offer competing accounts. Harris and Koenig (2006) analyzed children's acquisition of science and religion through testimony, supplying a general conceptual bridge to later authority evaluation Harris & Koenig (2006). Tests of adolescent mechanisms should measure which sources are consulted and how credibility judgments change rather than assuming the content of a belief reveals its source.
Adolescent socialization consequently involves continuity and negotiation. A decisive measurement issue is the interval between observed change and later retrospective interpretation. A person may remember adolescence as the moment of conversion or doubt even when the recorded trajectory shows years of gradual change. Repeated measures, event histories, qualitative interviews, and contemporary diary entries can illuminate the discrepancy. Interviews explain what an event means to the participant; longitudinal measures show when recorded indicators changed. The strongest developmental accounts combine these forms of evidence rather than letting one substitute for the other.
Emerging Adulthood: Mobility, Education, Partnerships, and Independent Practice
Early adulthood often brings changes in household structure, educational setting, occupational demands, financial autonomy, and social networks. These transitions reorganize opportunities for religious participation and sometimes redistribute the authority previously exercised by parents. Young adults may retain inherited beliefs while leaving a congregation, adopt new practices after relocating, or join communities through friendship and partnership. Life-course theory describes these developments through the timing and institutional conditions of transitions Elder (1998). Their religious significance must be established empirically for the person and context under study; the transition itself does not prescribe one direction of change.
A major longitudinal contribution comes from Dew, Fuemmeler, and Koenig (2020). Their analysis followed participants from adolescence into their thirties and examined religious attendance, subjective importance, and prayer separately Dew et al. (2020). Across the observed developmental period, attendance declined, importance increased, and prayer showed comparatively little aggregate change. These distinct patterns are incompatible with using a single attendance measure as a comprehensive proxy for religious development. They also illustrate why survey outcomes need parallel measurement and why aggregate trends should be separated from heterogeneous individual trajectories.
Chan, Tsai, and Fuligni (2015) studied changes in religiosity over the transition to young adulthood and examined individual and contextual patterns Chan et al. (2015). Their longitudinal approach contributes to a literature treating religious involvement as an unfolding process during emerging adult roles. Studies in this tradition need to account for educational selection, the kinds of institutions attended, household mobility, and local denominational opportunities. Students who leave home for college differ from those who do not in many measured and unmeasured respects, so a simple difference between groups is an inadequate causal estimate of the effect of education on religious belief.
Friendships and romantic partnerships become increasingly consequential. A new household can coordinate attendance, holiday observance, the education of children, and patterns of interaction with extended family. In partnerships with differing religious commitments, individuals may construct mixed arrangements or negotiate conflicts about ritual and identity. The empirical study of this process must observe both partners and the sequence of decisions, rather than treating their eventual similarity as proof that one persuaded the other. Religious organizations also vary in whether they offer accessible roles to unmarried adults, parents, students, migrants, or people working irregular hours; access can influence observed behavior independently of conviction.
Some changes are interpreted by individuals as conversion, deconversion, rediscovery, or intensification. Those narratives are psychologically important, but the RSP-09 research object focuses on developmental antecedents, trajectories, and the social environments of change. The detailed mechanisms and institutional consequences of conversion and disaffiliation belong to the separate research dossier on religious conversion, deconversion, and disaffiliation. The present analysis treats those events as possible outcomes of processes that begin earlier, including changes in networks, autonomy, explanatory commitments, and family expectations. National switching data provide useful background without directly measuring those mechanisms Pew Research Center (2025a) Pew Research Center (2025b).
Studies of young adults must include those who remain religious, those who remain nonreligious, those who change affiliation, and those whose private practices diverge from public labels. This inclusive sampling is important for studying development rather than a single institution's retention rate. Retention is an institutional outcome; developmental continuity includes values, habits, memories, narratives, and relationships that may persist outside organizational membership. The question of continuity therefore requires construct-specific interpretation and repeated assessment of both self-understanding and social practice.
Religious Diversity, Secular Development, and Differential Life Paths
Religious development should be studied as a set of possible cultural trajectories, including those of people raised without religion. Nonreligious socialization can transmit explanatory norms, attitudes toward authority, social commitments, family traditions, and practices of meaning-making. Children in religious and nonreligious homes both learn from testimony, example, and shared rituals; their social environments differ in the specific content and institutions through which learning is organized. Research restricted to religiously affiliated households cannot establish the frequency or mechanisms of nonreligious continuity, nor can it explain the full range of changes in religious identification Lewis-Jong & Bullivant (2026).
The meaning of religious importance is context-dependent. In a minority community, the category 'religious' may simultaneously signal belief, heritage, family obligation, community protection, and experience of discrimination. A majority member may report the same affiliation while experiencing little pressure to perform it publicly. This distinction means that comparative models need measurements of social position and perceived group boundaries. A single questionnaire translated into several languages is not sufficient to establish conceptual equivalence. Researchers should investigate whether participants use terms such as religious, spiritual, believer, practitioner, and secular in comparably organized ways across settings.
Spiegler and colleagues' German cohort investigation offers a concrete demonstration of heterogeneous developmental pathways Spiegler et al. (2025). Differences among the youth groups examined in that study should be interpreted alongside immigration histories, minority positions, and the availability of institutions, rather than attributed automatically to properties of a religion. Its person-oriented analysis addresses trajectories within identifiable sampling frames. Extending those distributions to other places, periods, or traditions would require representative new samples and appropriately tested measures.
Religious and nonreligious identities may be combined with mixed forms of practice. People can identify as nonreligious yet participate in family festivals, or identify with a religious heritage while treating doctrines metaphorically. Personal spirituality can remain significant in the absence of congregation membership. Such configurations make categorical switching statistics useful but incomplete. Pew Research Center's 2025 United States Religious Landscape Study distinguishes current religion from childhood religion and documents substantial switching Pew Research Center (2025a). Because recollection of childhood affiliation is retrospective, the comparison cannot by itself determine precisely when beliefs changed or how childhood influences operated.
Life paths are also shaped by the intersection of development with social identity and inequality. Lamb, Stawski, and Dermody (2021) investigated religious and spiritual development from adolescence to early adulthood, including differences associated with sexual orientation Lamb et al. (2021). Their study reinforces the need to specify which population and which dimension of religious involvement are being assessed. Social acceptance, family conflict, and perceived institutional fit are plausible variables linking identity to participation. Explanations of their causal influence require explicit design and should not be reduced to group labels.
Cross-cultural scientific explanation therefore begins with common analytical questions rather than a presumption of uniform outcomes. How are concepts learned? Which models are credible? What practices can be performed? What relationships support belonging? When do people gain agency to revise traditions? What events alter opportunity? What counts as identification in each institutional and linguistic setting? Comparative answers require locally valid instruments, genuine replication, and a clear separation between contextual description and proposed universal mechanisms.
Adulthood: Role Transitions, Reorganization, and Personal Continuity
Religious development continues after young adulthood as individuals organize commitments within changing personal and institutional roles. Marriage, parenting, separation, work, migration, caregiving, illness, bereavement, and retirement may change the rhythms of communal participation or the importance assigned to religious and spiritual practices. These transitions are analytically distinct from maturation itself. Becoming a parent can create new opportunities for religious education and ritual, while employment and geographic mobility can restrict attendance regardless of preference. Life-course analysis situates each event within previous commitments and the historical context in which it takes place Elder (1998).
Longitudinal evidence makes clear that patterns cannot be predicted from chronological age alone. McCullough, Enders, Brion, and Jain (2005) investigated religious development in adulthood using longitudinal data to examine change and the assumptions of a rational-choice approach to religious behavior McCullough et al. (2005). Their contribution is particularly relevant to questions of how perceived benefits, opportunities, and life circumstances interact with commitments over time. The theoretical distinction between religious demand and institutional supply is useful, but observed changes in attendance cannot be attributed uniquely to either without information about beliefs, alternatives, competing demands, and community accessibility.
Parenting can reactivate practices that were less visible in earlier adulthood. Adults may renew attendance to offer children familiarity with rituals, cultural heritage, or a structured community; others deliberately adopt a different approach from their own parents. Smith and Adamczyk (2021) studied the ways US parents attempt religious transmission, identifying practices through which family environments are intentionally structured Smith & Adamczyk (2021). The parent's intention is a relevant psychological variable but is not itself a measure of successful transmission. Outcomes depend on children, relatives, peers, local institutions, and later experiences that exceed parental control.
Repeated observations of linked families offer a means to investigate long-run continuity amid such transitions. Min and colleagues' (2018) finding of multiple parent–child religiosity patterns over decades establishes that similarity can decrease, increase, or recur after earlier divergence Min et al. (2018). This result complicates the idea that adult religious development merely consolidates the attitudes formed in adolescence. It also encourages time-sensitive modeling. Investigators should ask whether an apparent return reflects more regular attendance, stronger subjective commitment, or changing responses to a survey item. Without that distinction, the label 'return' lacks a sufficiently precise psychological referent.
Religious participation may be embedded in changing family relationships even when private belief is stable. Care responsibilities and the religious wishes of partners or older relatives can affect where, when, and with whom rituals occur. Participation may also provide opportunities for assistance and social recognition. A researcher examining the effects of role transitions must therefore measure family structure, access to communities, caregiving obligations, and satisfaction with participation alongside age. Longitudinal models with individual fixed effects or person-centered random effects can reduce confounding by some stable characteristics, but substantial unmeasured time-varying factors remain.
Adulthood also allows reflective reconstruction of childhood meaning. People may reinterpret inherited religion in light of later experiences while retaining emotionally salient memories and selected practices. Biographical interviews can reveal the vocabulary and events used in that reconstruction, while prospective panel measurements can assess whether the reported timing matches earlier responses. The interpretive account is valuable even when it differs from a questionnaire trajectory, because identity narratives are themselves developmental outcomes. The two forms of evidence answer related but distinct research questions.
Later Life: Aging, Reflection, Social Relationships, and Religious Change
Later-life religious development occurs within aging bodies, changing social networks, and new configurations of institutional opportunity. Bereavement, retirement, declining mobility, caregiving, and grandparenthood can alter the functions and accessibility of practice. Some individuals may intensify private prayer as attendance becomes difficult; others may disengage after experiences of disappointment or social loss. The age of a respondent is thus an imperfect indicator of developmental process. It matters whether studies have observed the same people over time, compared successive cohorts, or sampled older persons once Hayward & Krause (2013) Nelson et al. (2024).
Ingersoll-Dayton, Krause, and Morgan (2002) conducted open-ended interviews with 129 adults aged 65 and older, identifying stable, increasing, decreasing, and curvilinear trajectories of religiosity in participants' retrospective life-course accounts Ingersoll-Dayton et al. (2002). Participants connected changes to experiences including childrearing, adversity, and disappointment with religious communities. This qualitative evidence establishes the variety of meanings that older persons attach to religious transitions; it does not estimate the prevalence of these trajectories in the general population. The study is particularly useful for theory-building because it identifies events and social forces to measure prospectively.
Hayward and Krause (2013) analyzed long-term patterns of religious service attendance across the life course using longitudinal evidence Hayward & Krause (2013). Such research can distinguish within-person change from differences between individuals observed at separate ages more effectively than a single cross-section. Yet attendance can be constrained by mobility, accessible transportation, congregation availability, caregiving demands, and health status. Aging-related decreases may reflect changes in capacity or access rather than weakening personal conviction. Interpretations that attach a single psychological meaning to an attendance trend therefore require auxiliary information about the participant's circumstances.
Nelson, Joiner, Martinez, and Bergeman (2024) applied an accelerated longitudinal design to the study of religiosity and spirituality in midlife and older adulthood Nelson et al. (2024). Their work illustrates why multiple indicators and changes across repeated observations are needed to characterize development. An accelerated design combines trajectories of participants entering at different ages to estimate a wider developmental window, but inferences depend on how age, entry cohort, and calendar period are modeled. When those factors are difficult to separate, explanations of a change as 'aging' require additional qualification.
Grandparenthood can create opportunities for intergenerational transmission through family celebrations, storytelling, mentoring, and regular contact. Extended relatives may maintain continuity when parental commitments change, or they may support pluralistic arrangements among family members. Bengtson and colleagues' multigenerational perspective and the linked-family data studied by Min and colleagues help establish why a two-generation parent–child model can miss these relationships Bengtson (2013) Min et al. (2018). Future dyadic research should record the frequency, quality, and direction of grandparent–grandchild contact, rather than inferring influence from family position alone.
Late-life development also includes nonreligious and individualized forms of meaning-making. Some older adults retain inherited categories with changing interpretation; others form identities or practices that depart from childhood religious settings. Research should assess changing identities, personal values, and subjective meaning with instruments appropriate to older respondents, including those with sensory or cognitive limitations. Accessible interviewing and attention to residential institutions are crucial for representative sampling. The general conclusion from the studies considered is pluralism of trajectories, with the explanatory mechanisms remaining sensitive to relationships and institutional context.
Historical Cohorts, Secularization, and Intergenerational Population Change
A population's religious distribution changes through several processes. Individuals can change as they age; all age groups can respond to historical events; cohorts can differ because they were socialized in different environments; and population composition can change through migration, differential mortality, and fertility. These processes are analytically distinct. A single cross-sectional finding that older adults are more religious than younger adults cannot determine which processes produced the difference. An age–period–cohort analysis faces structural identification problems because age, period, and birth cohort are linearly related, requiring explicit modeling assumptions and external evidence.
Voas and Crockett's 2005 analysis of religion in Britain advanced an account in which generational replacement was a major factor in declining religious involvement Voas & Crockett (2005). In this explanation, the most consequential differences are those established between successive cohorts as they enter adulthood and remain relatively distinctive over time. Such an account predicts that within-cohort observation and comparisons between cohorts are indispensable. It also suggests that changes in childhood institutional exposure can have long-term aggregate consequences. The argument should be interpreted as pertaining to measured British religious indicators and periods, not as a universal rule governing every tradition.
Lewis-Jong and Bullivant's 2026 Cambridge synthesis reviews the psychology and sociology of intergenerational religious transmission, emphasizing parental socialization while examining national environments, peers, educational settings, religious communities, digital environments, and both religious and nonreligious families Lewis-Jong & Bullivant (2026). Its central research challenge is to explain how parent–child continuity varies across settings amid broader religious decline. The review connects micro-level family processes to macro-level transformations without implying that one mechanism has the same magnitude in every national context.
Large national surveys can document the scale of changing affiliation. Pew Research Center's 2025 Religious Landscape Study included comparisons between respondents' childhood religious identities and their current affiliations, revealing extensive switching among US adults Pew Research Center (2025a). This evidence characterizes reported change in the surveyed population, but the childhood measure is retrospective and does not provide prospective observations of the same person's religious orientation at successive ages. Its results should therefore complement, rather than replace, longitudinal family and youth cohort studies. Moreover, switching between labels can coexist with stability in private practice, moral commitments, or family attachment.
Explanations of why Americans leave or retain religious affiliations can be informed by respondents' reported reasons, including their descriptions of belief, experience, and relationships. Pew Research Center's later 2025 analysis of religious switching surveyed adults about reasons for changes and continuity Pew Research Center (2025b). Such accounts are valuable for identifying personally salient mechanisms and for constructing hypotheses, but a retrospectively selected reason is not equivalent to a controlled estimate of the causal contribution of that factor. The same person may be influenced by changes in belief, social belonging, and opportunity over overlapping periods.
Intergenerational change also depends on the institutional ecology in which parents try to transmit practices. The availability of congregations, schools, clubs, peer networks, minority institutions, and religiously diverse social environments can facilitate or constrain continuity. Two parents with similar levels of commitment may have different opportunities to organize a child's environment because of geographic location or economic resources. A multilevel research design should therefore consider household exposure and neighborhood or school context jointly. Such analysis links the life-course framework to the social structures that make some developmental pathways more accessible than others Elder (1998) Lewis-Jong & Bullivant (2026).
Empirical Research Designs and the Logic of Developmental Evidence
The scientific study of religious development requires designs that identify change at the appropriate level. Cross-sectional surveys compare people at one point in time and establish distributions and associations. Prospective longitudinal panels observe the same respondents repeatedly and can estimate within-person change. Intergenerational dyad or multigenerational designs relate relatives' reports across time. Experiments can isolate specific learning processes such as the effects of source testimony, while qualitative interviews reconstruct interpretations and personal histories. Every method carries particular inferential strengths and limitations; an integrated account needs to distinguish them rather than treating all findings as interchangeable evidence of causality Harris & Koenig (2006) Min et al. (2018) Ingersoll-Dayton et al. (2002).
Prospective repeated-measure research is especially useful because religious identity may change gradually, fluctuate around events, or diverge across domains. Dew and colleagues (2020) analyzed four waves of responses from more than 9,000 participants across ages 12 to 34 Dew et al. (2020). Their different aggregate trajectories for attendance, prayer, and importance demonstrate the value of multi-indicator panels. Still, the timing of observations determines which changes can be detected. A change that reverses between two widely spaced waves may be invisible. More frequent measurements help locate transitions, while longer intervals can better address durable continuity and generational effects.
Person-centered models examine whether observations can be summarized by multiple developmental patterns. Petts (2009) estimated trajectories of attendance across adolescence and young adulthood; Min and colleagues (2018) examined parent–child trajectories across eight waves; and Spiegler and colleagues (2025) identified religious-development profiles among German youth Petts (2009) Min et al. (2018) Spiegler et al. (2025). These studies have different units of analysis: individual attendance, family similarity, and individual multi-wave profiles. Their classes should not be compared as though they measure the same latent psychological entity. Model fit, class size, uncertainty of classification, robustness to alternative specifications, and available time points all matter.
Age is only one temporal dimension. Calendar events may change schooling, community access, social expectations, or religious institutions across all ages. An accelerated longitudinal design combines overlapping age segments of several cohorts to estimate broader developmental paths, as Nelson and colleagues (2024) did in their study of religiosity and spirituality Nelson et al. (2024). These designs can improve coverage when following a single cohort from childhood to old age would take decades. Their claims depend on assumptions about cohort comparability and missing data. A well-designed analysis therefore tests sensitivity to entry cohort, contextual variables, and alternative models for time.
Qualitative investigations reveal meaning and mechanisms that closed response scales may miss. Ingersoll-Dayton and colleagues' interviews with older adults identified varied life-course narratives connecting religious increase and decrease to experiences and relationships Ingersoll-Dayton et al. (2002). Smith and Adamczyk's interview and survey program investigated how parents consciously shape the religious environments of children Smith & Adamczyk (2021). Interview data can clarify perceived authority, interpretations of conflict, and the purposes of ritual. They do not establish statistical representativeness unless sampling and analysis justify it, and retrospective accounts can be shaped by the person's current identity.
Experiments on trust in testimony address a basic component of acquisition rather than the complete developmental trajectory of religion. Harris and Koenig (2006) connected children's knowledge of science and religion to testimony, supplying a framework for direct tests of who children believe and why Harris & Koenig (2006). Astuti and Harris (2008) showed that interviews about mortality and ancestors can elicit responses that depend on the specific narrative and conceptual context Astuti & Harris (2008). Experimental tasks improve control over immediate explanatory factors, but effects measured in a short interview may differ from socialization in a household or community over years. The methodological challenge is connecting micro-level learning processes to real-life development.
Psychometric Validity, Statistical Inference, and Comparative Measurement
Construct validity is fundamental in studies of religion because the same word can denote different experiences across traditions, generations, and individuals. A person may understand 'attendance' as a weekly service, an occasional temple visit, collective Friday prayer, seasonal pilgrimage, household ceremony, or an online gathering. Measures framed around one institutional model can systematically understate another form of practice. Researchers should document the operational target before choosing the measure and should test whether response categories describe relevant behaviors in the sampled communities Lewis-Jong & Bullivant (2026) Spiegler et al. (2025).
Developmental measurement additionally requires longitudinal comparability. If respondents interpret a questionnaire item differently at age 12 and age 30, an apparent change in score may reflect changed meaning rather than changed religiosity. Measurement invariance tests investigate whether a factor structure, loadings, and intercepts are sufficiently comparable across groups or waves. Qualitative cognitive interviews can identify wording problems before quantitative analysis. The same caution applies when comparing religious and nonreligious respondents: an item about prayer frequency may have structural nonapplicability, while an item about existential meaning might index a broad domain only weakly associated with religious socialization.
Research designs must specify whether the outcome is a single behavior, a latent construct, a configuration of practices, or a relationship between two people. Min and colleagues' intergenerational congruence analysis pertains to similarity in religiosity across linked family members Min et al. (2018). That relation cannot be reduced to the religiosity of a child alone. A model of dyadic change should consider the level and trajectory of each person as well as how similarity varies. Analyses that average over families with different baseline patterns may conceal both stable and reversals of relative commitment.
Missing observations are a serious concern in long panels. Respondents who move, disengage from institutions, experience conflict, or become ill may be more difficult to follow, and these experiences may be related to religious outcomes. Multiple imputation, full-information maximum likelihood, or inverse-probability weighting are potential strategies under different assumptions, but none automatically corrects bias from nonrandom loss. Reports should compare retained and lost respondents, state assumptions, and show sensitivity analyses. Such requirements are especially important when tracking transitions across several decades or comparing minority populations with different mobility rates.
Associations between religion and mental health require careful interpretation. Dew and colleagues examined the correlates of religious-development trajectories, including psychiatric outcomes, but their observational study does not establish that religious attendance causes psychological improvement or deterioration Dew et al. (2020). Changes in health can alter opportunities to participate, while existing social networks, socioeconomic conditions, and family context may affect both religion and outcomes. Longitudinal measurement improves temporal description; credible causal inference also requires an explicit intervention or counterfactual framework, plausible assumptions, and assessment of confounding and selection.
The strongest reporting practice states the population, instrument, age range, observation dates, statistical target, uncertainty, and generalizability limits alongside the claim. It also separates what was observed from what is proposed as an explanation. Where an article is available only through an abstract or a book summary, a monograph should not supply unverified effect sizes, coefficients, subgroup sample counts, or detailed statistical procedures. The present synthesis follows that restriction, while identifying the full publication for each work so that readers can examine the primary source.
Critical Scientific Discussion: Competing Models and Explanatory Limits
Several explanatory models organize the evidence. Social-learning accounts emphasize modeled practice, perceived authority, and consequences Bandura (1977). Testimony-based developmental accounts emphasize how children evaluate knowledge from other people Harris & Koenig (2006). Family-relational accounts examine the conditions under which commitments are communicated and accepted Bengtson (2013). Life-course frameworks address the timing of transitions and linked biographies Elder (1998). Cohort-based analyses explain population change through differences in socialization between generations Voas & Crockett (2005). These models operate at different levels and may explain complementary portions of the same trajectory.
The evidence does not currently identify one mechanism that explains all religious continuity. Parents are important agents of socialization, as examined in Smith and Adamczyk's research and the 2026 Cambridge synthesis, but young people also participate in institutions, peer networks, educational contexts, and digital communities Smith & Adamczyk (2021) Lewis-Jong & Bullivant (2026). The resulting pathways can be additive, mutually reinforcing, or in tension. A family's strong religious identity may encourage institutional participation, while institutional experiences can subsequently transform how the family's tradition is interpreted. The direction and strength of influence are empirical matters requiring appropriately timed measurements.
Stage theories and trajectory analyses answer different questions. Fowler's structural model proposes changes in how faith is organized and interpreted Fowler (1981). Repeated-measure studies such as Dew and colleagues' report changes in selected behavioral and evaluative indicators Dew et al. (2020). One cannot infer the stage of faith from an attendance trajectory, nor assume that increased religious importance represents progression to a structurally more complex form of meaning-making. The historical importance of stage accounts is in their questions about transformation; their contemporary explanatory value depends on cross-cultural tests and clear operationalization.
Social-context differences challenge uniform claims about secularization as an age effect. The German longitudinal study by Spiegler and colleagues reported distinct youth trajectories among Muslim, immigrant-origin Christian, and non-immigrant Christian participants Spiegler et al. (2025). In Britain, Voas and Crockett explained much of religious decline through generational replacement Voas & Crockett (2005). In American data, Dew and colleagues found decreasing attendance alongside increasing importance Dew et al. (2020). These results refer to different populations, calendar periods, instruments, and research designs. Their combined significance is that theories must specify which outcome and environment they explain.
Retrospective stories about religious change are indispensable for understanding meaning and incomplete as causal histories. The older participants interviewed by Ingersoll-Dayton and colleagues described adversity as relevant to both increasing and decreasing religiosity Ingersoll-Dayton et al. (2002). That bidirectional pattern makes a simple 'adversity causes religion' explanation inadequate. The developmental effect of an event depends on preceding commitments, the interpretation of the experience, available relationships, and institutional response. Prospective event-centered designs can test these interacting conditions more rigorously.
A comprehensive model must also represent the possibility of secular continuity and religious change in either direction. The empirical field includes parents transmitting nonreligious identities, adults revising inherited commitments, young people developing religious engagement in largely secular environments, and older persons reinterpreting longstanding practices. Treating one of these trajectories as the normative endpoint would bias recruitment, outcome definitions, and interpretation. Scientific developmental description instead asks which processes are observed, how they are measured, which populations they describe, and which counterfactual explanations survive scrutiny.
Integrative Developmental Synthesis and Boundary Conditions
The literature supports a multilevel model of religious socialization in which learning, relationships, institutional opportunities, and historical context jointly shape developmental trajectories. Its elementary unit is a person interacting over time with a changing social ecology. Information about sacred concepts is learned through testimony and cultural representation; behaviors are acquired through observation and participation; affiliation emerges through identification and recognition; and commitments are revised in response to later experiences. General social-learning and testimony frameworks articulate plausible psychological mechanisms, while long-term cohort and linked-family studies establish the observed heterogeneity of outcomes Bandura (1977) Harris & Koenig (2006) Min et al. (2018).
One strong empirical conclusion is that religiosity is multidimensional across development. Dew and colleagues' four-wave analysis found that attendance, prayer, and perceived importance did not share the same mean trajectory Dew et al. (2020). Nelson and colleagues' accelerated longitudinal study further illustrates the need to investigate age-related religious and spiritual factors separately Nelson et al. (2024). Therefore, developmental change cannot be defined by a single measure without narrowing the object being studied. This principle applies equally to research on religious continuity, affiliation loss, secular identity, and spiritual practice.
A second conclusion concerns heterogeneity across relationships and settings. Family transmission is important, yet the longitudinal dyadic evidence includes stable similarity, increasing difference, convergence, and return Min et al. (2018). Cross-national and minority comparisons reveal differences in the forms and timing of development Spiegler et al. (2025). These findings support theories with multiple pathways moderated by relationship quality, available institutions, identity, and cohort position. They do not support assigning a universal developmental destiny to an individual solely from the religion of their parents or their age.
A third conclusion is that historical change modifies opportunities for personal development. Population-level religious decline can reflect intergenerational replacement even when some individuals grow more engaged as they age Voas & Crockett (2005) Hayward & Krause (2013). The 2026 synthesis by Lewis-Jong and Bullivant directly connects parent–child transmission to national context, peers, schools, communities, and the internet Lewis-Jong & Bullivant (2026). Research that compares age groups without historical context risks confusing cohort-specific socialization with within-person maturation. The same concern applies to studies extrapolating a local youth trajectory to the world at large.
A fourth conclusion concerns interpretation. The same life event can accompany different changes in reported religiosity, as older adults' life-course accounts illustrate Ingersoll-Dayton et al. (2002). Religious learning and later revision cannot therefore be explained solely by the occurrence of the event. Studies should measure the person's preceding commitments, relationships, available practices, and understanding of what changed. The evidence makes room for continuity of family knowledge even when institutional affiliation changes, and for continuity of identification amid altered patterns of participation; those specific continuities should be measured rather than assumed.
These conclusions set limits on the scope of this dossier. It addresses religious socialization and age-linked development as a process; systematic analysis of conversion and institutional disaffiliation requires its own research object. The psychometrics of religiosity across traditions, religious coercion, coping under adversity, and religion-related mental health outcomes also require focused bodies of evidence. Their overlap with development is genuine, and a shared vocabulary permits productive cross-reference, but a single dossier should not imply that findings in one specialized domain settle questions in another.
Future Research: Causal Identification, Cultural Coverage, and Artificial-Era Environments
Future longitudinal research should follow children and caregivers prospectively across significant developmental transitions while measuring belief, participation, private practice, affiliation, religious literacy, identity centrality, and nonreligious commitments separately. This design would reveal whether changes begin in opportunities for participation, reported meanings, or family relationships, and how the sequences differ across populations. Where feasible, multiple household reporters and observation of actual practices could reduce single-informant bias. Regular follow-up around changes in schooling, residence, partnership, and parenthood would help identify the temporal structure of developmental adjustment Dew et al. (2020) Min et al. (2018).
Causal research needs mechanisms and comparison strategies. If an intervention changes the opportunity to participate in a religious community, its effects on behavior may occur without corresponding belief change. If a family communication program changes the quality of discussion, the relevant outcomes include mutual understanding, voluntary engagement, and the expression of disagreement as well as later affiliation. Natural experiments involving school policy or institutional access can help, provided researchers document plausible control groups and potential simultaneous changes. Ethical research must avoid treating children's private beliefs as outcomes that adults are entitled to impose.
Cross-cultural work should include different religious traditions, mixed and secular households, migrant and minority communities, and contexts with distinct legal and institutional relations between religion and public life. Measures must be adapted through conceptual consultation, qualitative testing, and statistical invariance checks. The German youth study exemplifies why a design based on one majority tradition can miss other trajectories Spiegler et al. (2025). Future multinational panels should report sampling frames, local category meanings, translation procedures, and the degree to which trajectories replicate within and across groups.
Digital and artificially mediated environments create new settings for learning about religion and nonreligion. Search engines, social networks, online congregations, and conversational artificial systems can offer explanations, simulate discussion, recommend communities, and change which sources appear credible. Their relevance to religious development is a research question requiring direct data on exposure, user interpretation, recommendation systems, and changes over time. The Cambridge intergenerational-transmission review already includes the internet among its contexts of investigation Lewis-Jong & Bullivant (2026). The Artificial Era adds the specific challenge of understanding how artificial interlocutors mediate testimony, authority, and self-exploration across developmental stages.
Artificial interlocutors should be analyzed through observable interaction features and testable hypotheses. A system might reproduce a tradition's language, provide comparative descriptions of beliefs, or elicit disclosure from a young user. None of these behaviors by itself establishes that the system has religious experience or that its responses cause a durable change in the user's convictions. Appropriate studies would record advice content, disclosure patterns, source-checking behavior, parental and institutional mediation, and later changes under controlled conditions. Particular care is needed with minors, confidentiality, manipulative personalization, and the uneven accessibility of reliable information.
Methodological advances should combine within-person longitudinal analysis, family network data, qualitative process tracing, and studies of actual media use. Open materials and preregistered hypotheses would improve comparability where the research design permits them; long-term data stewardship would reduce avoidable loss of evidence. The larger scientific task is to explain how social environments make certain forms of understanding and practice available while individuals reorganize commitments over the lifespan. That task connects the psychology of religion with developmental science, social psychology, educational research, and the study of human–artificial relations.
Conclusion
Religious socialization and development across the lifespan constitute a continuous, socially embedded process of learning, participation, identification, and revision. Its beginnings are visible in childhood encounters with family practices and testimony; its adolescent forms include questioning, changing peer relations, and emerging autonomy; its adult forms are reorganized through households, partnerships, institutions, and life transitions; and its later-life forms encompass renewed practice, reflection, continuity, and change. The empirical record is characterized by multiple pathways rather than one developmental sequence Harris & Koenig (2006) Petts (2009) Ingersoll-Dayton et al. (2002).
Several lines of evidence jointly support this account. Multigenerational and dyadic studies show that family religiosity may remain similar, diverge, or converge over decades Bengtson (2013) Min et al. (2018). Longitudinal research demonstrates that attendance, importance, and prayer can move in different directions Dew et al. (2020). Comparative youth research reveals differences among social and cultural groups Spiegler et al. (2025). Research on adult and later-life development confirms that chronological age alone is insufficient to characterize religious or spiritual change McCullough et al. (2005) Nelson et al. (2024). Historical cohort research further connects population transformations to patterns of transmission between generations Voas & Crockett (2005) Lewis-Jong & Bullivant (2026).
The primary methodological implication is to define the precise psychological object before choosing a measure or a theory. Family affiliation, individual conviction, practical observance, community belonging, and religious knowledge require distinct indicators and may have different developmental histories. The strongest evidence will emerge from designs that observe individuals and relationships repeatedly, establish comparability across contexts, and test mechanisms against alternative explanations. Historical stage theories, observational learning, testimony research, life-course frameworks, and population studies contribute different pieces to that explanation when their inferential boundaries remain explicit.
Religious development is therefore a case of general psychological development within historically organized cultural worlds. The continuing expansion of digital and artificial environments changes the channels through which testimony, instruction, and identity negotiation occur; it creates new empirical questions about source authority and relational mediation. The established research supports careful investigation of those questions without predetermining their answers. A comprehensive psychology of religious socialization follows persons, practices, and relationships through time, recognizing variation as part of the phenomenon itself.
References
- Astuti, R., & Harris, P. L. (2008). Understanding mortality and the life of the ancestors in rural Madagascar. Cognitive Science, 32(4), 713–740. https://doi.org/10.1080/03640210802066907
- Bandura, A. (1977). Social learning theory. Prentice Hall. https://www.ncbi.nlm.nih.gov/nlmcatalog/101562611
- Bengtson, V. L. (2013). Families and faith: How religion is passed down across generations (with N. M. Putney & S. C. Harris). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199948659.001.0001
- Boyatzis, C. J. (2012). Spiritual development during childhood and adolescence. In L. J. Miller (Ed.), The Oxford handbook of psychology and spirituality (pp. 151–164). Oxford University Press. https://doi.org/10.1093/oxfordhb/9780199729920.013.0010
- Chan, M., Tsai, K. M., & Fuligni, A. J. (2015). Changes in religiosity across the transition to young adulthood. Journal of Youth and Adolescence, 44(8), 1555–1566. https://doi.org/10.1007/s10964-014-0157-0
- Dew, R. E., Fuemmeler, B., & Koenig, H. G. (2020). Trajectories of religious change from adolescence to adulthood, and demographic, environmental, and psychiatric correlates. Journal of Nervous and Mental Disease, 208(6), 466–475. https://doi.org/10.1097/NMD.0000000000001154
- Elder, G. H., Jr. (1998). The life course as developmental theory. Child Development, 69(1), 1–12. https://doi.org/10.1111/j.1467-8624.1998.tb06128.x
- Fowler, J. W. (1981). Stages of faith: The psychology of human development and the quest for meaning. Harper & Row. https://library.strathmore.edu/Record/23806
- Harris, P. L., & Koenig, M. A. (2006). Trust in testimony: How children learn about science and religion. Child Development, 77(3), 505–524. https://doi.org/10.1111/j.1467-8624.2006.00886.x
- Hayward, R. D., & Krause, N. (2013). Patterns of change in religious service attendance across the life course: Evidence from a 34-year longitudinal study. Social Science Research, 42(6), 1480–1489. https://doi.org/10.1016/j.ssresearch.2013.06.010
- Ingersoll-Dayton, B., Krause, N., & Morgan, D. (2002). Religious trajectories and transitions over the life course. International Journal of Aging and Human Development, 55(1), 51–70. https://doi.org/10.2190/297Q-MRMV-27TE-VLFK
- Lamb, K. M., Stawski, R. S., & Dermody, S. S. (2021). Religious and spiritual development from adolescence to early adulthood in the U.S.: Changes over time and sexual orientation differences. Archives of Sexual Behavior, 50(3), 973–982. https://doi.org/10.1007/s10508-021-01915-y
- Lewis-Jong, J., & Bullivant, S. (2026). The intergenerational transmission of religion. Cambridge University Press. https://doi.org/10.1017/9781009707152
- McCullough, M. E., Enders, C. K., Brion, S. L., & Jain, A. R. (2005). The varieties of religious development in adulthood: A longitudinal investigation of religion and rational choice. Journal of Personality and Social Psychology, 89(1), 78–89. https://doi.org/10.1037/0022-3514.89.1.78
- Min, J., Silverstein, M., & Gruenewald, T. L. (2018). Intergenerational similarity of religiosity over the family life course. Research on Aging, 40(6), 580–596. https://doi.org/10.1177/0164027517723076
- Nelson, N. A., Joiner, R. J., Martinez, B. S., & Bergeman, C. S. (2024). Religiosity and spirituality development: An accelerated longitudinal design. The Gerontologist, 64(10), gnae097. https://doi.org/10.1093/geront/gnae097
- Petts, R. J. (2009). Trajectories of religious participation from adolescence to young adulthood. Journal for the Scientific Study of Religion, 48(3), 552–571. https://doi.org/10.1111/j.1468-5906.2009.01465.x
- Pew Research Center. (2025a). Religious switching. In Religious Landscape Study. https://www.pewresearch.org/religion/2025/02/26/religious-switching/
- Pew Research Center. (2025b). Why do some Americans leave their religion while others stay? https://www.pewresearch.org/religion/2025/12/15/why-do-some-americans-leave-their-religion-while-others-stay/
- Smith, C., & Adamczyk, A. (2021). Handing down the faith: How parents pass their religion on to the next generation. Oxford University Press. https://doi.org/10.1093/oso/9780190093327.001.0001
- Spiegler, O., Jonsson, J. O., & Bracegirdle, C. (2025). Religious development from adolescence to early adulthood among Muslim and Christian youth in Germany: A person-oriented approach. Child Development, 96(1), 141–160. https://doi.org/10.1111/cdev.14151
- Voas, D., & Crockett, A. (2005). Religion in Britain: Neither believing nor belonging. Sociology, 39(1), 11–28. https://doi.org/10.1177/0038038505048998