Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Religious, Spiritual, and Nonreligious Identities

Abstract. Religious, spiritual, and nonreligious identities are socially situated and personally interpreted patterns of self-definition through which people organize belonging, ultimate commitments, continuity, and the public presentation of belief or nonbelief. This fundamental critical narrative research monograph examines their conceptual history, social-identity mechanisms, centrality and salience, spiritual-but-not-religious identification, atheist and other nonreligious positions, developmental trajectories, family and cultural ecologies, intersecting social identities, discrimination, meaning, and psychological well-being. Its evidence combines original surveys, longitudinal adolescent studies, qualitative interviews, social-psychological experiments, and systematic conceptual comparisons. The synthesis distinguishes affiliation, belief, practice, salience, self-description, and identity processes; evaluates what survey categories can support; and separates social association from causal inference. The resulting account treats identity as an empirically investigable coordination of personal commitments and social recognition, with explicit attention to cultural variation, heterogeneous nonreligious identities, health-related confounding, and unresolved questions about change and measurement.

Keywords: religious identity; spiritual identity; nonreligious identity; spiritual but not religious; atheist identity; agnosticism; religious affiliation; identity centrality; social identity; religious development; religious nones; secular belonging; identity concealment; cultural identity; religious discrimination; psychological well-being

Introduction: The Psychological Problem of Identity

Religious, spiritual, and nonreligious identities designate organized ways in which individuals describe themselves in relation to traditions, conceptions of ultimate significance, communities, and the possibility or rejection of supernatural belief. Their psychological importance emerges when a self-description becomes a basis for continuity, affiliation, emotional commitment, interpretation of others, or social treatment. People can inherit a religious label, participate in practices, privately question doctrines, and nevertheless experience religious belonging as a durable component of self. Others can be deeply engaged with spirituality while resisting membership in organized religion. Still others define themselves through atheism, agnosticism, humanism, a generic lack of religion, or indifference toward religion. These positions concern different forms of self-organization and cannot be recovered from a single belief item. Foundational conceptual work by Zinnbauer et al. (1997), sociological analysis by Lee (2014), and survey evidence from Pew Research Center (2025b) establish the range of the empirical problem.

The scientific object of this dossier is the organization, experience, expression, and social recognition of identity as religious, spiritual, nonreligious, or some combination of these. An identity is investigated through self-categorization, salience, centrality, group attachment, narratives of continuity, enacted practices, social constraints, and the consequences of disclosure. The scope includes development and variation across culture, family, social networks, and situations; it also examines psychological correlates while attending to alternative explanations. A distinct dossier examines the processes of conversion and deconversion as changes in affiliation or religious worldviews, whereas the present study asks how positions are experienced and maintained, including when they are unstable or plural. Social identity approaches Ysseldyk et al. (2010) and multidimensional group identification Leach et al. (2008) guide questions about belonging, investment, and collective meaning.

The governing questions are how religious and spiritual categories become elements of personal identity; why membership, belief, practice, and personal importance sometimes diverge; how nonreligious identification acquires positive social content; how discrimination and recognition affect disclosure; how identities develop within families and social ecologies; and which observations support judgments about consequences for well-being. This is a critical narrative evidence synthesis informed by original empirical investigations, conceptual scholarship, and documented national surveys. It is not a registered systematic review, a meta-analysis, or a report of original human-subject experimentation. The central inferential discipline is to identify which claims are descriptive, which are correlations within selected populations, which are longitudinal associations, and which follow from experimental manipulations. The empirical and interpretive limits of each evidence type remain part of the scientific explanation.

Definitions and the Multidimensional Identity Matrix

Religious identity refers to a person’s self-definition in relation to a religious tradition, collectivity, set of practices, or perceived sacred order. That definition permits the identity to be central even when beliefs are unsettled, practice is intermittent, or institutional participation is limited. Spiritual identity refers to self-understanding organized around perceived transcendence, depth, ultimate meaning, connection, or a sense of the sacred, with meanings that vary substantially by culture and individual. Nonreligious identity refers to a self-understanding in which distance from, rejection of, indifference to, or alternative relations with religion have recognizable personal or social significance. These are working operational categories for research, not claims that one essence underlies every instance. Comparative evidence from Zinnbauer et al. (1997), Wixwat and Saucier (2021), and Lee (2014) demonstrates why respondents must be allowed to specify what their labels mean.

At least six dimensions must be recorded separately in a credible study: nominal affiliation, propositional belief, behavioral practice, affective commitment, identity centrality, and public identification. Affiliation locates a person relative to an institution or community. Belief captures assent, doubt, or uncertainty toward particular propositions. Practice measures activities such as worship, prayer, meditation, or communal participation. Affective commitment concerns what an identity means emotionally. Centrality asks how important the identity is to the overall self. Public identification asks which labels a person uses in interaction, a dimension influenced by anticipated acceptance and stigma. A seventh dimension, longitudinal continuity, tests whether a self-description persists across settings and time. Treating these coordinates as interchangeable invites conceptual error, especially when survey questions ask only for a religion or no religion. The different objects of the measures are apparent in the Religious Landscape Study chapters on identity and self-description Pew Research Center (2025b) Pew Research Center (2025c).

An individual may describe themselves as culturally Christian and personally nonbelieving, as Muslim and questioning, as Jewish by heritage while practicing little religion, as spiritual but not religious, or as an atheist who participates in religious family rituals. These combinations are analytically possible because heritage, metaphysical commitment, spiritual orientation, and social membership occupy separable coordinates. The prevalence or psychological consequences of any particular combination must be estimated rather than assumed. Moreover, labels are culturally translated classifications: the English expression 'spiritual but not religious' cannot be assumed to denote an identical position in every linguistic community. The field therefore benefits from examining identity profiles and within-person patterning rather than imposing mutually exclusive psychological types on answers to a single questionnaire item Ammerman (2013) Pew Research Center (2025a).

Historical Foundations of the Research Tradition

Scholarly interest in the subjective meanings of religion predates the contemporary language of religious identity, but systematic inquiry into self-labeling gained particular prominence as disciplines distinguished belief, institutional participation, personal experience, and individual commitment. The research genealogy involves sociology of religion, social psychology, developmental psychology, personality science, and ethnographic study of secular lives. Early discussions frequently organized the field around religious affiliation and personal religiosity. Subsequent research treated spirituality and nonreligion as categories with their own social histories and interpretive disputes. The empirical question shifted from whether a person is religious to how that person understands and communicates a religious, spiritual, or nonreligious position. The program advanced by Zinnbauer et al. (1997) and the later reconsideration of survey categories by Lee (2014) make this shift visible.

Zinnbauer and colleagues' 1997 study, 'Religion and Spirituality: Unfuzzying the Fuzzy,' investigated definitions and relationships between religion and spirituality among participants recruited from diverse religious and community settings. Its importance lies in treating respondents' meanings as data rather than treating a researcher’s prior binary as settled. The sample was assembled from selected groups, which makes the evidence informative about conceptual variability rather than a national population estimate. The authors' conceptual distinction should therefore be used to improve questions and classification, with national prevalence left to probability-based survey designs. Later, Wixwat and Saucier (2021) reviewed evidence on the 'spiritual but not religious' designation and emphasized both its heterogeneity and its cultural and demographic contingencies. The two works reveal a durable methodological lesson: psychological categories require empirical validation at the level of language and social practice.

Developments in nonreligion research likewise expanded the object from a residual census category to forms of identity and community. In a United Kingdom qualitative study, Lee (2014) demonstrated that generic responses such as 'no religion' can express diverse substantive positions rather than the simple absence of identity. Among U.S. atheists, Smith (2011) analyzed narratives through which participants came to recognize and publicly articulate a nonbelieving self; Frost (2019) later showed that narratives of nonreligious certainty and uncertainty are multiple. These investigations produced richer accounts of meaning and interaction, while their selected qualitative samples cannot measure population-wide prevalence. The contemporary field therefore has two complementary genealogies: large-scale survey demography that maps social categories and intensive interpretive research that clarifies what those categories mean in lived experience.

Social Identity Theory and Group Identification

Religious and nonreligious identities often function as social identities: people understand themselves partly through actual, anticipated, or contested relationships with groups. The relevant mechanisms involve self-categorization, perceptions of shared norms, emotional investment, recognition by others, and comparisons between communities. Religion may organize daily time, mutual obligations, symbols, friendships, and moral expectations; nonreligious communities may organize around intellectual commitments, civic practice, or shared experiences of exclusion. A social-identity perspective Ysseldyk et al. (2010) explains why an ostensibly private conviction may acquire public psychological consequences. It also prevents belief from doing all the explanatory work: community members may disagree on doctrine yet maintain a strong collective identity, and persons with comparable beliefs may report different degrees of belonging.

Multicomponent research on group identification provides a useful structural vocabulary. Leach et al. (2008) distinguished aspects of group-level self-definition and self-investment, including perceived similarity or homogeneity, solidarity, satisfaction, and centrality. Roccas et al. (2008) integrated several modes of group identification that reflect different ways individuals attach significance to membership. These models were not devised exclusively for religion, so their transport into religious settings is an empirical proposition that requires scale validation. Nevertheless, they show why a single 'strength of religious identity' item conceals distinguishable states: identification through attachment to fellow members can coexist with ambivalence toward authority, and centrality can be high even when participation is low. The predictive value of each component for actual behavior may differ across traditions, ages, and institutional contexts.

Group identities are also relational, since perceived boundaries are maintained through social recognition and contested claims to membership. Someone may use a tradition's name while other members deny that the person conforms to expected practices; conversely, an individual may avoid a label that outsiders routinely impose. The difference between self-ascription and external classification can affect social support, stigma, and perceived authenticity. In comparative research, it is especially important to ask who counts as an insider, how membership is established, and which practices institutions treat as evidence. A measured association between collective identification and adjustment can reflect access to community networks, family status, perceived exclusion, or prior psychological characteristics. Conceptual models Ysseldyk et al. (2010) Leach et al. (2008) clarify plausible mechanisms, but any causal account must identify the relevant social process rather than infer it from a label.

Centrality, Salience, Commitment, and Identity Coherence

Centrality denotes the enduring importance of a social identification to the self, whereas salience denotes its situational accessibility or relevance. A person can consider religion essential to personal biography while rarely thinking about it during routine work; a low-centrality nonreligious identity can become highly salient when religious observance is expected at an event. Emotional commitment concerns the value and attachment associated with the identity, while identity coherence concerns the experienced compatibility of one's positions across domains. These constructs generate distinct research questions. High salience under threat could coincide with distress even when the same identity ordinarily offers support, and high centrality could amplify both protective community attachment and conflict with external institutions. The conceptual differentiation follows the multidimensional group-identification analyses of Leach et al. (2008) and Roccas et al. (2008).

Coherence is not the same as rigid consistency. An individual can hold multiple identities without experiencing conflict when the associated practices and commitments are contextually coordinated. Conversely, apparently consistent questionnaire responses can obscure tension between private metaphysical uncertainty and publicly expected declarations. Researchers should therefore examine subjective integration, perceived conflict, and narratives of change in addition to category agreement across survey items. Longitudinal adolescent data provide particularly useful leverage because changes in personal identity can be compared with evolving family and ethnic identification. In a three-wave study of students from different ethnic backgrounds, Lopez et al. (2011) found that religious identity and participation followed different developmental trajectories and that within-person religious identity changes related to shifts in ethnic and family identities. This evidence supports an interactive developmental model while leaving the direction of influence unresolved.

Identity centrality has to be distinguished from religious importance questions that are answered on behalf of tradition, family, or public culture. Asking whether religion is 'important in your life' may reflect belief, practice, belonging, heritage, or an obligation to provide a normatively respectable answer. Salience can be studied using repeated experience sampling, event-specific reports, and diary measures; centrality through repeated surveys and validated components of group identification; coherence through narrative interviews and measures of perceived compatibility. Establishing discriminant validity requires demonstrating that these instruments capture distinguishable patterns rather than relabeling similar responses. Across religious and nonreligious identities alike, psychological explanation improves when the measurement design directly corresponds to the proposed process Leach et al. (2008) Lee (2014).

Spirituality, Sacred Meaning, and the SBNR Designation

Self-identification as 'spiritual but not religious' (SBNR) brings into focus the distinction between experiential meaning and formal organization. It can refer to intimate transcendent experience, an ethical outlook, a quest for connection, nature-related meaning, traditions interpreted outside their institutions, or a deliberate distance from ecclesiastical authority. The phrase does not indicate one uniform belief system. In qualitative work, Ammerman (2013) identified multiple conversational patterns through which participants constructed spirituality, including theistic, extra-theistic, ethical, and belief-and-belonging orientations. These patterns describe meaningful varieties of language and social practice, rather than mutually exclusive psychological diagnoses. The task is to understand why a self-description is chosen, how it organizes a person's life, and which comparisons its use permits.

The review by Wixwat and Saucier (2021) emphasizes that SBNR positions are heterogeneous and that some correlates reflect the demographic and cultural contexts in which the label has been studied. Two individuals using the identical phrase may diverge in belief in gods, mystical orientation, experiences of religious harm, attendance, community attachment, and interpersonal tolerance. A questionnaire that asks respondents to select a single category therefore needs follow-up dimensions if it is intended to explain psychological functioning. 'Spiritual' itself may mean proximity to a deity in one account and a non-theistic experience of connectedness in another. Measurement schemes developed in primarily Western populations cannot establish a universal distinction without cross-cultural testing. Conceptual variability becomes a reason for finer-grained observation, not grounds for treating the label as empty.

Population surveys provide valuable context without resolving this conceptual diversity. In Pew Research Center's 2023–2024 U.S. Religious Landscape Study, 53% of U.S. adults described themselves as both spiritual and religious, 21% as spiritual but not religious, 5% as religious but not spiritual, and 21% as neither, using the survey's operational questions and thresholds Pew Research Center (2025c). These figures are not the same as the proportion unaffiliated with organized religion, because one classification uses self-described degrees of spirituality and religiosity whereas the other concerns religious identity or affiliation. The figures are specific to the country, survey period, wording, and design. They show the importance of independently measuring two dimensions; they do not show that one category is a stable personality type or that it causally determines well-being.

Adolescent Development and the Differentiation of Identity

Adolescence provides an informative setting for studying identity because commitments acquired through family participation are increasingly subject to reflection, comparison, and independent public presentation. Religious identification can continue through this period, become less central, diversify, or be reinterpreted in light of peers and new intellectual contexts. It is therefore necessary to distinguish continuity of a label from continuity of motivation and practice. Developmental research indicates that these trajectories may diverge. In the three-wave high-school study by Lopez et al. (2011), religious identity was relatively stable across grades while participation declined, and changes within adolescents corresponded with changes in ethnic and family identity. The study involved 477 participants from Latin American, Asian, and European backgrounds in the United States, making it informative about these adolescents while not constituting a universal timetable of religious development.

A developmental perspective regards identity as a construction embedded in social obligations and opportunities. Adolescents may encounter different interpretations of inherited religion at home, school, and in peer networks; their felt capacity to disclose doubt can depend on perceived family response. Religious practice may sometimes be maintained as a form of belonging independently of propositional certainty. Conversely, an adolescent who participates little in communal activities may retain a profound private religious identity. Because survey items often collapse beliefs and community practice into one frequency score, the empirical distinction matters for interpretation. Historical measurement work that includes familial, cultural, communal, and religious dimensions Koteskey et al. (1990) and the later longitudinal findings Lopez et al. (2011) support studying these social contexts together while keeping their constructs distinguishable.

Developmental explanations should model earlier exposure, identity commitments, available alternatives, neighborhood and school composition, perceived social cost, and family relations. A design comparing a single cohort at two ages cannot cleanly separate maturation from changes in social environment; repeated measurements across multiple cohorts and cultural settings are more informative. Moreover, changes in religious category may follow shifts in psychological well-being, occur alongside them, or precede them. Any longitudinal association can still reflect unmeasured processes. A study of adolescent religious deidentification Hardy et al. (2026) shows the value of comparing prior functioning to subsequent change, while also showing why selective institutional settings and histories of family religion complicate causal attribution. The major developmental question is how persons negotiate continuity and revision under changing relational conditions.

Emerging Adulthood, Autonomy, and Individualized Commitment

Emerging adulthood is often marked by relocation, educational transitions, labor-market entry, partnership, and expanded access to different worldviews. These transitions can alter opportunities for religious participation while giving individuals greater authority over the labels they use. Research on personalized belief has emphasized that the religious lives of young adults may combine selected inherited elements with individual judgments. Arnett and Jensen (2002) described individualized religious belief among emerging adults, a formulation useful for understanding autonomy and personal interpretation without treating individualized belief as a universal historical trajectory. A participant's reasons for disengaging from an institution may involve transportation, workload, peer networks, ideology, or family history; only careful investigation can distinguish identity change from changes in circumstance.

The category 'spiritual but not religious' is particularly relevant because it may arise during the reorganization of family-inherited affiliation, but it can also appear without a prior institutional departure. The longitudinal analysis by Upenieks and Ford-Robertson (2022) used two waves of the National Study of Youth and Religion from 2005 to 2008 to compare changing and persistent spiritual and religious descriptions in emerging adulthood. In the reported analysis, consistently identifying as SBNR was associated with worse physical and mental health outcomes relative to consistently religious respondents; mediation models considered attendance, felt closeness to God, doubt, and life meaning. Such pathways are model-dependent statistical explanations rather than experimental demonstrations that adopting an SBNR label causes poor health. They encourage careful consideration of background experiences, transitions, and social integration.

Autonomy is also culturally embedded. In some communities, independent expression of religious preference signals responsible maturation; in others, continuity with family practice expresses care and ethical obligation. The meaning of leaving a congregation may depend on institutional rules, family relationships, and local minority status. Comparative identity studies can investigate autonomy without presuming that departure from organized religion is inherently more mature or that continuity is necessarily conformity. The relevant developmental outcomes include self-authorship, perceived coherence, relational continuity, and capacity to negotiate commitments. A design that combines repeated identity questions with interviews and context-sensitive measures can ask how these outcomes arise without treating affiliation itself as a direct index of agency Arnett and Jensen (2002) Lopez et al. (2011) Upenieks and Ford-Robertson (2022).

Family Systems, Heritage, and Intergenerational Belonging

Religious identity is frequently transmitted through routine family interactions: shared meals and observances, celebrations, naming practices, moral language, contact with relatives, and expectations about marriage or partnership. These practices can provide a familiar identity long before a young person formulates explicit doctrinal judgments. Family influence operates through direct instruction, imitation, emotional affiliation, and the practical availability of communities. It may be experienced as continuity, care, tension, or constraint, depending on the family and the person. Within-person links between religious, family, and ethnic identities observed by Lopez et al. (2011) make the family more than a background demographic variable; it is a changing relational environment within which religious self-understanding develops.

Family heritage also complicates cross-sectional labels. A person may identify with a religious tradition because it anchors ancestry and memory despite a naturalistic metaphysical outlook; another may believe in supernatural realities yet avoid the historical religion of their family. The distinction is analytically important in diasporic and minority settings, where religious and ethnic labels may be jointly imposed by outsiders or jointly embraced by community members. Treating a named tradition as a pure belief variable conceals these histories. Researchers can ask about childhood affiliation, family ritual, ethnic background, inherited language, current belief, and perceived acceptance as distinct variables. The relational emphasis in Koteskey et al. (1990) and the group-identity models of Roccas et al. (2008) offer complementary approaches to conceptualizing these ties.

Intergenerational differences are best studied as negotiated processes. A family's public religious practice might remain intact while its members increasingly disagree privately; a new nonreligious designation can reshape expectations about holidays and marriage without ending attachment to relatives. The particular effects depend on disclosure, sanctions, dialogue, social alternatives, and institutional norms. Longitudinal designs can ask whether family conflict precedes identity concealment or whether disclosure changes relational dynamics, while qualitative interviews can identify what different participants understand as loyalty and autonomy. Such a model recognizes that changes in labels can coexist with remarkable continuity in everyday obligations. Separate research on conversion and disaffiliation investigates the dynamics of changing affiliations; here, the concern is the organization of belonging and self-description within relationships Smith (2011) Frost (2019).

Narrative Identity, Moral Commitments, and Biographical Continuity

Identity is partly maintained through the story a person can tell about earlier and current commitments. Religious stories may center on family heritage, a formative experience, a perceived sacred calling, or belonging to a community. Nonreligious stories may center on inquiry, emotional distance from religious institutions, valued relationships, or a stable lack of interest in religious claims. Narrative inquiry can reveal why apparently identical survey selections carry different emotional meanings. In a sociological account of American atheist identity formation, Smith (2011) examined the construction of a meaningful nonbelieving self through interaction and biography. The findings illuminate interpretive work among the participants, while representative population estimates require independent designs.

An identity narrative can preserve continuity even when practices or propositions change. Someone may come to reinterpret a childhood ritual as cultural heritage rather than supernatural obligation, retaining its biographical value. Another person may retain belief in a deity while changing the community in which that belief is lived. Coherence depends partly on whether the person can integrate these layers with other commitments, such as professional responsibilities, political values, or family roles. Such integration is not a simple score of rational consistency; it includes emotional, social, and temporal meanings. Qualitative accounts of nonreligious Americans Frost (2019) show that certainty, uncertainty, and indifference support different narratives, so the researcher must avoid presuming a single atheist or unaffiliated pathway.

Researchers can connect narrative evidence to repeated ratings of centrality, congruence, stress, and belonging. For example, a report of increased coherence after redefining one's religious label should be evaluated against prior levels of conflict and concurrent changes in relationships. Interviews conducted after a transition may reconstruct earlier motives in light of present commitments, creating retrospective narrative bias. Diary designs, prospective qualitative follow-up, and contemporaneous reports can reduce dependence on a single retrospective account. The most useful synthesis treats narrative identity as the interpreted structure of personal continuity and group identification as a related but separately measured set of social processes Smith (2011) Frost (2019) Leach et al. (2008).

Identity Intersectionality, Migration, and Minority Position

Religious self-definition is often intertwined with ethnicity, race, migration, gender, linguistic community, and national belonging. These dimensions affect how an identity is publicly recognized and whether a person can express uncertainty safely. In minority settings, a religious designation can become salient through external classification even when it carries little private doctrinal importance. Individuals may also deliberately invest in religious identification because it sustains relationships, ancestry, or resistance to exclusion. A theoretically useful study must distinguish these processes rather than assume that all members of a named community have equivalent experiences. Among Arab Americans, Hashem and Awad (2021) examined religious centrality, public regard, perceived discrimination, and psychological distress, revealing the importance of the social reception of religious belonging.

In the Detroit Arab American Study examined by Hashem and Awad (2021), Muslim and Christian Arab American participants differed in some experiences of discrimination and negative religious public regard even as their measured levels of religious centrality and psychological distress were similar. These are observations among a particular regional population and cannot establish a universal ordering of group risk. Importantly, mediation models of public regard and perceived discrimination represent statistical pathways under stated assumptions. They cannot by themselves determine whether discrimination preceded all reported distress or rule out shared causes. Their scientific significance is the recommendation to measure public treatment and private centrality separately when studying the psychological correlates of religious identity.

Recent research with Muslim American college students also links immigrant-generation position, perceived discrimination, Muslim American identity, and mental health indicators Bonumwezi et al. (2026). The study's selected college sample makes its psychological and social pathways informative while limiting transfer to all Muslim Americans, especially older people and those outside higher education. Together, the two studies indicate why a named religious label is insufficient for understanding adjustment: neighborhood composition, historical marginalization, family migration trajectories, and exposure to social rejection matter. Future research should distinguish experiences linked to religion, ethnic stereotyping, and their interaction. The empirical object is the person situated in overlapping classifications, not a fixed group essence.

Individual Differences in Identity Organization

Personal identity profiles can differ by the importance of abstract beliefs, emotional investment in communities, tolerance of doubt, motivation for spiritual experiences, and preference for organized participation. These features should be treated as variables to be investigated rather than as components presumed to occur together. A person may engage in reflective religious inquiry without close ties to an institution; another may belong strongly to a congregation while remaining uncertain about metaphysical doctrine. The review by Wixwat and Saucier (2021) reports evidence that 'spiritual but not religious' tendencies are not reducible to one disposition, and longitudinal research into religiousness and spiritual seeking Wink et al. (2007) suggests that religious engagement and spiritual exploration can follow different personality-related patterns.

Individual differences interact with social conditions. Strong religious identity may support predictability and acceptance within a hospitable community but create demands for concealment in an antagonistic environment. Atheist identification may be accompanied by explicit reflection on epistemic commitments, but many unaffiliated individuals do not organize their daily life around a secular group label. Consequently, broad stereotypes about religious dogmatism or nonreligious rationality cannot substitute for measurements of cognitive style, social motivation, and individual history. The appropriate design distinguishes self-description from explanatory traits and tests whether correlations persist after accounting for education, age, local norms, and other confounders Wixwat and Saucier (2021) Frost (2019).

Longitudinal stability is likewise an empirical question. Individuals can retain a category across decades while their understanding of it changes, or change category without a major shift in values and social ties. Personality findings should not be translated into irreversible identity types. A stronger investigation would combine measures of centrality and group-identification modes with context-specific practices and periodic narrative accounts. Doing so would enable researchers to distinguish stable commitment from stable vocabulary and to examine whether identity predicts future behavior over and above its constitutive self-descriptions. The multidimensional framework developed in Roccas et al. (2008) and the longitudinal approach of Wink et al. (2007) illustrate complementary research resources.

Nonreligious Identification and the Diversity of Unaffiliation

Religious nonaffiliation is a survey response about belonging to a named religion, while nonreligious identity is an interpreted position that may carry personal, intellectual, or social meaning. The distinction is especially important for people who select 'nothing in particular.' Such respondents may be indifferent to religion, committed to secular values, spiritually engaged, uncertain about supernatural claims, or disengaged from religious institutions for practical reasons. Lee (2014) argued that treating 'not religious' as a residual category creates an interpretation problem: a negative survey answer can conceal positive understandings of identity, relationships, and everyday practices. In representative survey work, these differences are relevant to estimating both the size and composition of unaffiliated populations.

Atheism concerns the absence or rejection of belief in gods, but the personal importance of calling oneself an atheist varies. Agnosticism can express a judgment about the limits of knowledge, uncertainty toward specific claims, or a socially recognized affiliation label. 'Nothing in particular' can reflect low religious salience or a positive preference against available institutional categories. These positions have overlapping but distinguishable conceptual structures, and researchers must respect the wording used in a given instrument. Pew's Religious Landscape Study estimated that 29% of U.S. adults were unaffiliated in 2023–2024, including approximately 5% atheist, 6% agnostic, and 19% 'nothing in particular'; category totals are affected by rounding Pew Research Center (2025b). Those estimates describe U.S. respondents in that survey, not global nonreligion.

A nonreligious person can engage in prayer, endorse spiritual realities, participate in cultural rituals, or maintain religious family belonging. Conversely, a self-described religious person may doubt supernatural claims or have limited participation. These patterns call for distinctions between metaphysical stance, institutional affiliation, spiritual orientation, and enacted affiliation. In the 2025 global analysis of survey data from 22 countries, Pew Research Center (2025a) reported substantial spiritual belief among many religiously unaffiliated respondents, with striking cross-national variation. Its evidence rules out equating affiliation with a uniform worldview and highlights the importance of measurement-specific comparisons. A robust account models nonreligion as a set of empirically diverse identity possibilities rather than a single score on a religiosity continuum.

Atheist Identity, Agnosticism, Indifference, and Secular Narratives

Nonreligious self-definition can arise through different narrative orientations. Some participants emphasize certainty that theistic claims are false, some describe unresolved epistemic questions, and some indicate that religion plays little role in their everyday life. Frost (2019) analyzed variation in American nonreligious identity narratives, demonstrating why certainty and indifference should be investigated separately. These patterns can influence readiness to disclose a label, interest in secular organizations, and reactions to requests for religious justification. They are interpretive distinctions derived from qualitative research and require survey measures before their population distribution can be estimated.

Public atheist identification may be a deliberate social act in environments where the category carries stigma. Smith (2011) studied how Americans construct identity and meaning around rejecting theism; the process involves recognizing a label, interpreting previous experiences, and negotiating relationships. Yet an individual who lacks belief need not consider that absence a defining component of the self. The difference between a belief state and a salient group identity helps explain why self-reported atheist prevalence may differ depending on whether a survey asks about belief, identity, or religious affiliation. A study that treats these responses as equivalent has changed its operational construct without evidence that the responses share the same meaning.

Secular identity may express commitments to human relationships, science, freedom of conscience, particular moral values, or community organizations. Such commitments require study through direct reports and observable practices. They cannot be inferred from atheist or agnostic classification alone. The relationship between secular commitments and meaningful life is especially sensitive to measurement wording, because constructs of purpose that explicitly presuppose a supernatural source will classify nonbelievers as lacking by definition. The empirical analysis by Speed et al. (2018) addressed that risk by comparing items on life meaning, fatalism, and nihilism in American survey data. Its results support studying the quality and source of meaning separately from the religious label respondents select.

Secular Belonging, Community, and Public Recognition

Collective identity may develop around secular associations, humanist organizations, local social networks, or shared practices of discussion and mutual support. The psychological processes can include solidarity, recognition, group norms, and collective projects much like those encountered in other forms of group life. However, nonreligious individuals vary in whether they seek a secular community at all. Many may prefer private indifference or maintain relationships in mixed-belief settings. Applying the identification dimensions of Leach et al. (2008) to secular groups provides measurable questions about group centrality, solidarity, and satisfaction without assuming that membership is psychologically equivalent across religious and nonreligious organizations.

Belonging has practical dimensions: access to social assistance, recognized holidays, compatible intimate partners, public language for important life events, and opportunities to speak without concealment. In societies where religious organizations dominate local civic institutions, nonreligious individuals may experience exclusion regardless of whether they consider their position central. In more secular settings, a nonreligious label may be unremarkable and carry little personal meaning. This situational variation helps explain why data from a strongly religious region cannot straightforwardly predict outcomes in a less religious one. Research on the social reception of atheist identity Frost et al. (2023) and experiments about perceived atheist prevalence Gervais (2011) identify contexts in which public recognition and perceived normality matter.

Community studies should distinguish actual participation from desired participation. Someone may value a secular organization in principle without joining, or participate while retaining a private religious orientation. The same person may value friendships across belief differences more than a clear ingroup boundary. Relevant outcome measures include social support, loneliness, perceived acceptance, organizational involvement, relationship quality, and freedom to express beliefs. Causal assessment requires attention to self-selection: individuals with different prior social networks may choose different forms of belonging. The aim is to explain how recognition and participation shape identity experiences while acknowledging that a category label is a limited proxy for the surrounding social ecology.

Prejudice, Public Stereotypes, and Identity Concealment

Public stereotypes can change the costs of identification. A person who expects an atheist label to invite distrust may choose a less explicit description, while a religious minority may avoid identifying practices in a setting associated with hostility. Concealment can protect short-term social access but may impose monitoring demands and limit authentic self-presentation. It is therefore useful to separate perceived societal hostility, direct discriminatory experiences, behavioral concealment, and internalized evaluation. Each has different causes and potential consequences. Frost et al. (2023) used survey data and latent-variable modeling to distinguish perceived hostility and identity concealment among Americans who self-identified as atheists, demonstrating the value of treating these as related but separable constructs.

Experimental work offers a narrower causal test of how an externally supplied perception can affect judgment. In four studies, Gervais (2011) examined relations between perceived atheist prevalence and anti-atheist prejudice; the experimental components reported reductions in explicit or implicit distrust when participants were reminded of the prevalence of atheists. This demonstrates an effect of information framing under studied conditions, not that simply increasing the actual share of atheists will cause identical changes in all societies. Other experiments have investigated distrust as a central mechanism in anti-atheist prejudice Gervais et al. (2011). These results should be interpreted in relation to their populations, measures, and social contexts rather than generalized to all religious and nonreligious relations.

Experiences of discrimination among religious minorities also vary by external racialization, ethnicity, migration status, dress, local policies, and historical context. Hashem and Awad (2021) found that public regard and perceived discrimination were associated with psychological distress among Muslim and Christian Arab Americans in its study context; Tineo et al. (2021) investigated discrimination, acculturative stress, religious support, and Muslim identity in a Muslim college sample. Such findings clarify possible protective and harmful pathways but remain conditional on design and selection. Concealment itself may be adaptive in certain immediate situations and distressing across time; researchers should avoid prescribing disclosure as a uniform remedy. Ethical investigation prioritizes respondent control over sensitive identity information.

Cross-Cultural Classification and the Limits of Western Terminology

Survey categories travel imperfectly across languages and traditions. Terms such as religion, spirituality, belief, atheist, and agnostic carry historical associations that differ across societies. In some contexts, inherited rituals and relational obligations are classified as cultural practices rather than religious commitments; in others, naming a tradition may signal ancestry and citizenship as well as doctrine. These differences make literal translation insufficient for measurement equivalence. Researchers must investigate the actual comprehension of questions, the available response options, and the social consequences of choosing one answer. Lee (2014) and Zinnbauer et al. (1997) supply conceptual reasons for the practice, while Pew Research Center (2025a) illustrates its necessity in cross-national data.

In the Pew Research Center's September 2025 study, the group labeled 'nones' in 22 countries was defined by respondents identifying as atheist, agnostic, or 'nothing in particular.' Significant percentages of these unaffiliated participants expressed spiritual beliefs, and the frequency of those beliefs varied by country Pew Research Center (2025a). The report also documented that the question used to refer to 'God' was adapted to linguistic contexts. This is a useful reminder that measurement interpretation remains part of the result; a national percentage is a property of a specific instrument applied to a defined population. Cross-national comparisons should present denominators, survey years, response categories, and uncertainty as relevant.

To test identity constructs across cultures, researchers can begin with cognitive interviewing and local qualitative inquiry, develop translated items through collaborative adaptation, and examine whether the factor structure and item thresholds remain comparable. Differential item functioning can indicate that two respondents with similar underlying identification have different probabilities of endorsing an item because the wording carries different cultural implications. Measurement invariance testing can improve interpretability, but its adequacy depends on model assumptions and sample structure. A theoretically sound synthesis combines comparable items with locally meaningful descriptions. It refrains from declaring one civilization more or less spiritual solely from differences in labels whose conceptual referents vary.

Religious Identity as a Majority or Minority Social Position

Being religious is not associated with a single social rank. Religious traditions vary between majority and minority status across places, and an individual may belong to a majority in one context and a minority in another. These differences affect whether religious identity operates as ordinary background culture, a public source of prestige, a basis of perceived exclusion, or a focal point for collective solidarity. Similarly, nonreligion can be socially unremarkable where secular identification is common and contested where religiosity is strongly normative. Contextual analysis therefore treats local distribution and institutional power as variables. A study of one denominational or national setting cannot supply a universal psychological description of religion or atheism.

Public regard is an especially useful variable because it concerns how individuals think outsiders see their group. In the Detroit Arab American Study analyzed by Hashem and Awad (2021), negative religious public regard and discrimination helped account statistically for associations with distress. Among atheists, Frost et al. (2023) identified social locations and experiences associated with concealed identity, indicating that the anticipation of reception also matters outside traditional religious minorities. Context can change both the probability that a person will disclose an identity and the significance of that disclosure. A locally common label might be chosen casually; a marginalized label might express a deliberate statement about belonging.

Multilevel designs offer a route toward testing contextual explanations. Individuals can be nested within schools, regions, workplaces, and countries with measured variation in local affiliation patterns and normative environments. The strongest studies separate individual belief, group membership, and local institutional conditions instead of assuming one variable proxies for the others. Repeated measures can show whether a person changes public expression as they move between environments while retaining private commitments. These observations are consistent with social-identity theories Ysseldyk et al. (2010) and experimental evidence about perceived atheist prevalence Gervais (2011), but their generalizability depends on the quality of geographic, cultural, and interpersonal measures.

Digital Contexts and Emerging Identity Expression

Digital communities create additional settings for exploring religious and nonreligious identity. Online discussion allows individuals to encounter competing interpretations, observe peers who share uncommon experiences, and experiment with labels before presenting them in family or workplace settings. It also exposes them to conflict, stereotyping, and algorithmically amplified boundary disputes. The existing psychological literature on religious identity supports testable questions about online identity salience, audience segregation, public regard, and belonging; it does not justify assuming that digital platforms uniformly strengthen or weaken religion. A rigorous study would measure actual platform use, interaction networks, and self-descriptions over time, rather than treating all screen exposure as equivalent.

Online contexts can produce different public and private identities. Users may maintain multiple accounts, disclose a nonreligious identity to a sympathetic community, or present a religious label to relatives while questioning privately. These behaviors can be investigated through consent-based diary methods and confidential interviews, with careful attention to safety and the risks of linking accounts across contexts. The distinction between self-description, disclosure, and perceived hostility already developed in Frost et al. (2023) provides a conceptual starting point. Observed posting frequency should not be equated with centrality; highly central identities may be deliberately concealed, while situational disputes can lead to brief periods of intensive posting.

In the Artificial Era, language systems and automated recommendation tools may become additional interlocutors through which people articulate identity questions. Their potential role can be studied empirically as a change in access to information, perceived anonymity, conversational rehearsal, and social comparison. Such work requires explicit ethical safeguards, as religious beliefs and nonreligious convictions can be sensitive personal data. An Artificial Sapiens perspective can frame the future research agenda around human–artificial mediation of identity narratives while preserving a clear boundary between philosophical propositions and outcomes demonstrated in controlled psychological studies. At present, claims about the causal effects of these emerging interactions remain hypotheses needing longitudinal and experimental investigation.

United States Religious Identity Survey Evidence, 2023–2024

Large population surveys are especially useful for separating the public distribution of labels from theories about why individuals adopt them. The Pew Research Center's 2023–2024 Religious Landscape Study estimated that 29% of U.S. adults selected atheist, agnostic, or 'nothing in particular' when asked about religious affiliation Pew Research Center (2025b). This figure is drawn from that survey's national adult sample and classification design. It should be distinguished from the 21% who described themselves as spiritual but not religious on separate self-description questions Pew Research Center (2025c). Because the measures concern different dimensions, their percentages cannot be added to obtain one comprehensive nonreligious population. Many unaffiliated respondents also described themselves as spiritual, while some affiliated respondents described themselves as having limited personal religiosity.

The earlier Pew report devoted specifically to American religious 'nones' Pew Research Center (2024) examines diverse beliefs, reasons for nonaffiliation, and attitudes within an apparently single survey category. A nonreligious affiliation response can accompany spiritual beliefs and different appraisals of religious institutions. These findings are compatible with both negative definitions—absence of a particular institutional affiliation—and positive self-understandings such as atheism or a secular worldview, but the data do not prove that every unaffiliated respondent possesses a salient nonreligious identity. The scientifically defensible statement is that affiliation and identity salience are different variables. When comparing annual public estimates, researchers must also distinguish changes in survey wording, response rates, sampling frames, and weighting from shifts in actual belief or identification.

Cross-sectional prevalence describes how many respondents occupy a category at the time of survey, not how many have recently converted, deconverted, or changed internal commitments. Longitudinal transition analysis requires repeated observations of the same participants or careful reconstruction from reported histories. Nor do national averages establish uniform patterns among age cohorts, ethnic groups, or communities. The appropriate use of Pew Research Center (2025b) and Pew Research Center (2024) is to locate contemporary identities within a documented demographic landscape and then investigate mechanisms with complementary empirical designs. As a methodological principle, estimates of category frequency and explanations of identity formation should remain distinguishable in the published synthesis.

Comparative Evidence on Spiritual Belief Among Unaffiliated Populations

Cross-national surveys reveal that nonaffiliation does not uniquely determine views about supernatural or spiritual realities. Pew Research Center's 2025 analysis focused on 22 countries where the 2023–2024 surveys included sufficiently large unaffiliated samples for separate study Pew Research Center (2025a). Among religiously unaffiliated adults, beliefs in life after death, spiritual forces, and God varied widely between countries. Some respondents identified with no religion while sustaining beliefs or practices ordinarily categorized as religious. Such findings complicate interpretations that treat religion as a single dimension running from high engagement to categorical absence. They also demonstrate the difference between a population indicator and a psychological identity: two respondents with identical affiliations may hold divergent metaphysical commitments.

National context affects the meaning of a generic 'none' response. In societies with strong religious institutions, 'nothing in particular' may signal opposition to those institutions or the absence of a comfortable affiliation. Where formal membership is less central to community life, the same response may carry a different implication. The global evidence Pew Research Center (2025a) therefore challenges models that treat a U.S.-derived classification as automatically equivalent in every country. Researchers should report the official translated question, available answers, and distinctions between belief in God, a universal spirit, afterlife, and other spiritual claims. These concepts may show different levels of overlap within and across cultural groups.

Comparative data also warn against equating the size of the nonreligious category with the total population of naturalistic or secular worldviews. A self-described atheist may embrace one naturalistic worldview among many; an unaffiliated person who believes in God may have no institutional home; an affiliated person may privately reject supernatural beliefs. Any generalization requires the relevant joint distribution of affiliation, belief, practice, and centrality. Future research can combine these dimensions into interpretable profiles, examine which replicate across cultures, and test associations with demographic, relational, and psychological factors. Establishing that an identity profile generalizes is a substantive empirical finding, not an assumption built into an initial classification.

Evidence from Longitudinal Youth and Family Studies

Recent longitudinal studies add temporal information to the literature but retain substantial limits on causal interpretation. Hardy et al. (2026) analyzed three biennial waves from 652 U.S. adolescents who were religious at the first wave, focusing on the association of subsequent religious deidentification with indicators of positive and negative functioning. The authors reported that deidentifying adolescents already differed in some outcomes before a change in affiliation and that subsequent trajectories varied across measures and religious backgrounds. The sample and observation design make those findings informative about specified transitions while leaving room for unmeasured family circumstances, institutional costs, and prior experiences. The results belong to developmental research, and an identity dossier uses them to clarify the temporal relation between prior adjustment, self-labeling, and relational environment.

In the 2026 parent–child study by Clevenger and Dyer (2026), spiritual and religious identities were operationalized jointly to examine anxiety and depression. Baseline differences appeared between identity categories among children, including elevated measured symptoms for the spiritual-but-not-religious group. The longitudinal component is particularly valuable because it can distinguish observed initial differences from changes across the study interval, yet its regional dyad sample cannot stand in for all American families or global populations. A study combining categories into a four-way scheme must also recognize internal heterogeneity within each group. Before attributing symptoms to spiritual identity itself, researchers should examine attendance, perceived family conflict, prior distress, social isolation, and personal experiences that may influence both identification and health.

Earlier longitudinal evidence from Lopez et al. (2011) and Upenieks and Ford-Robertson (2022) reinforces the importance of distinguishing identity stability, participation, and adjustment. Lopez and colleagues observed stability in average adolescent religious identity amid declining participation, while Upenieks and Ford-Robertson examined within-person changes in SBNR identity in emerging adulthood. The studies use distinct ages, measures, settings, and outcomes and therefore cannot be combined as if they estimate one common effect. Their collective contribution lies in demonstrating the need for prospective measurement and careful interpretation of changing self-definitions. Future projects should compare multiple populations using common measurement cores accompanied by locally meaningful identity questions.

Contemporary Studies of Minority Identity and Psychological Context

Recent evidence on minority religious identities helps show how distress can arise through social circumstances rather than through a doctrine or label alone. Bonumwezi et al. (2026) examined immigrant generation, perceived discrimination, Muslim American identity, and depression or anxiety symptoms among a sample of Muslim American college students. The sample included 157 participants, and its recruitment and demographic composition limit population-level generalization. Its contribution is to consider immigrant-generation differences together with perceived social treatment and identity in one analysis. A finding in an educationally selected sample is best used to refine hypotheses about social context, with replication needed among nonstudents and across regions.

That research can be interpreted alongside Hashem and Awad (2021), which used data from the Detroit Arab American Study involving 1,016 Muslim and Christian Arab American participants. The authors examined public regard, perceived discrimination, centrality, and psychological distress and reported relationships among these dimensions. The studies differ in both ethnicity and religious composition and in their measurement of social conditions. These differences are analytically meaningful: public regard is an assessment of perceived societal views, while reported discrimination concerns experiences or their appraisal. Neither variable can be replaced by an assumed characteristic of all Muslim or Christian individuals. Joint consideration of identity and context improves the explanatory precision of psychological inquiry.

The contemporary research program should also address the social position of atheists, agnostics, and other nonreligious minorities. Frost et al. (2023) focused on patterns of perceived hostility and identity concealment among self-identified atheists in U.S. survey data, demonstrating observable variation in the management of a nonreligious identity. Social identity theory Ysseldyk et al. (2010) encourages comparable models across groups, but different historical vulnerabilities and disclosure costs should remain visible. The most useful integration examines the general processes of belonging, centrality, and external recognition while treating the cultural and political meanings of specific groups as empirical features of the cases being studied.

Research Agenda for the Artificial Era

The Artificial Era creates opportunities to examine how human religious, spiritual, and nonreligious self-descriptions are articulated in interaction with nonbiological information systems. Such systems can mediate the discovery of concepts, comparison of interpretations, composition of autobiographical narratives, and communication with communities. The psychological phenomena potentially involved include source credibility, perceived conversational privacy, identity rehearsal, social comparison, and perceived judgment. Their significance must be tested against concrete exposures and outcomes. The present empirical evidence on religious identity supports the formulation of these variables through existing theories of group recognition Ysseldyk et al. (2010), narrative construction Smith (2011), and stigma management Frost et al. (2023); it does not demonstrate a specific psychological effect of interaction with Artificial systems.

Future research can compare different forms of human–artificial interaction with ordinary information search, human conversation, and private journaling. Experimental designs could vary disclosure expectations, conversational framing, or the visibility of alternative beliefs, with appropriate consent and safeguards. Longitudinal studies could examine whether repeated interaction changes identity centrality, conceptual clarity, or readiness to communicate with family and peers. Such designs would need to rule out selection into different technologies and consider whether the instrument itself encourages a particular religious or secular vocabulary. The distinction between philosophical authorship of an Artificial Era framework and empirical claims about human psychological outcomes is essential for scientific accountability.

Religious and nonreligious identity involves sensitive information. Systems that process this material should minimize unnecessary data collection, avoid revealing concealed positions, and refrain from steering individuals toward predetermined metaphysical commitments. Ethical safeguards are especially important where family conflict, discrimination, or institutional pressure make disclosure consequential. A scientifically productive program also recognizes the possibility of positive outcomes, including increased access to explanations of unfamiliar traditions and opportunities for reflective articulation. These are testable prospects. The canonical contribution of the present dossier is the clarified human psychological research object, with Artificial Era implications delineated as a future research domain.

Identity, Meaning in Life, and Personal Values

Meaning in life concerns the perceived significance, coherence, and direction of one's existence, and these qualities can be cultivated within religious, spiritual, and nonreligious frameworks. Religiously organized identities may integrate personal purpose with sacred narratives, shared ceremonies, and community obligations. Nonreligious identities can organize meaning through relationships, ethical commitments, work, civic participation, naturalistic interpretations, or personal projects. The empirical question is which processes support meaningfulness under which conditions, rather than whether a particular metaphysical position has an exclusive capacity for meaning. Speed et al. (2018) analyzed nationally representative U.S. survey data and found no systematic association of atheism or religious nonaffiliation with elevated fatalism or nihilism in its tested measures, while reporting stronger endorsement of internally generated meaning among some nonreligious respondents.

Measurement can accidentally construct the answer it claims to discover. An item defining meaning exclusively through divine purpose will classify an atheist as lacking that particular form by construction; it cannot independently measure whether the person experiences life as valuable. Conversely, assuming all self-authored meanings are psychologically liberating would import another normative premise. Research should instead use construct-valid measures that are intelligible to people across positions. The distinction between the source attributed to meaning and the degree of meaning experienced is central. Findings from Speed et al. (2018) help formulate that distinction, but their single-wave observational design does not establish whether identity causes particular interpretations.

Identity coherence may be one pathway linking beliefs or worldview to felt meaning, while relationships and life circumstances provide additional pathways. Coherence can arise from a stable inherited tradition or from a consciously articulated secular orientation; it may also be pursued through spiritual exploration. Methodologically, a researcher could model identification centrality, social support, personal value alignment, and meaning as distinct time-varying constructs, controlling for baseline values and demographic background. There is a risk of circularity if identity coherence and meaning measures use nearly identical wording. Well-designed longitudinal studies and qualitative accounts can clarify whether apparent associations reflect overlapping definitions, social resources, or genuine changes in psychological functioning Upenieks and Ford-Robertson (2022) Arnett and Jensen (2002).

Psychological Well-Being and Mental Health Correlates

Research concerning religious and spiritual identity and mental health yields associations whose interpretation depends on the identity category, population, outcomes, and context. Some individuals experience group belonging and spiritual practices as resources for coping; others encounter conflict, discrimination, or expectations they cannot reconcile with private values. Measures of depression and anxiety are distinct from identity centrality and require appropriate instruments and clinical interpretation. No single identity label should be treated as a diagnosis, a clinical risk state, or a therapeutic mechanism. A longitudinal study of emerging adults Upenieks and Ford-Robertson (2022) associated persistent spiritual-but-not-religious identification with less favorable health outcomes relative to persistent religious identification under its modeling choices, while leaving open explanations involving community attendance, doubt, earlier circumstances, and selection.

In a 2026 study involving parent–child dyads sampled in the Western United States, Clevenger and Dyer (2026) analyzed religious and spiritual categories in relation to depression and anxiety across baseline and later observation. At baseline, children in the spiritual-but-not-religious category reported higher symptom levels in the study than several other categories. The authors' design offers a particular developmental and familial perspective, but selection and the meanings of categories within that sample limit generalization. The study contributes contemporary evidence that spirituality and religion measured separately can be more informative than a single summary scale; it does not establish that taking up a spiritual or nonreligious label causes psychopathology.

Longitudinal evidence also complicates simple interpretations of religious departure. Hardy et al. (2026) examined 652 adolescents who were religious at the first observation, comparing trajectories of young people who later deidentified with those who did not. The authors reported that some preexisting differences preceded deidentification and that subsequent changes in certain outcomes varied, including by religious background. These patterns are consistent with selection and context-dependent pathways rather than an undifferentiated effect of nonbelief. The design, sample composition, measures, institutional context, and analytic adjustments must inform interpretation. Research seeking causal explanations needs richer prior-functioning data, measures of family or community cost, and comparisons across social ecologies.

Religious and Nonreligious Identity Under Discrimination

Psychological distress may be linked with experiences surrounding an identity rather than the identity's content. People can encounter prejudice because others classify them as religious, visibly observant, nonreligious, or members of a historically marginalized group. Such experiences may heighten identity salience, undermine security, or encourage selective disclosure. Research must separate direct experiences from perceived societal norms and distinguish responses to events from enduring psychological vulnerability. Among Muslim and Christian Arab Americans, Hashem and Awad (2021) examined associations between religious centrality, negative public regard, discrimination, and distress. In Muslim college students, Tineo et al. (2021) included perceived discrimination and acculturative stress among the proposed pathways relating social context to symptoms.

The effects of public categorization can operate even when a person privately holds a different identity. A nonbelieving member of a religious minority may be treated as belonging to the minority by outsiders, and an atheist in a religiously homogeneous workplace may conceal their private belief state without strong involvement in organized atheism. This creates an important difference between identity ascribed by others, publicly enacted identity, and internal identification. A study capturing only private self-labels will miss experiences arising from external classification. The finding that perception of atheist prevalence can shift distrust in experiments Gervais (2011) suggests that community-level expectations can influence attitudes, but it does not determine any particular individual's psychological outcome.

Ethical and methodological considerations intersect in this area. Gathering detailed religious or nonreligious identity data can expose respondents in settings where disclosure has consequences. Researchers should establish confidentiality, minimize indirect identifiers, permit refusal, and avoid assuming that public recruitment through a community organization reaches concealed or unaffiliated members. Analyses should distinguish participants who actively use a label from those to whom it is assigned. The long-term objective is to explain how social recognition, self-definition, and material conditions jointly affect experience. Studies of perceived public hostility Frost et al. (2023) and discrimination-related distress Hashem and Awad (2021) provide informative models with different populations and outcomes.

The Measurement Problem: Categories, Scales, and Construct Validity

A religious-identity measure is valid for a specified purpose only when its items, scoring, and interpretations correspond to the construct under investigation. A denomination item can classify nominal affiliation; it cannot simultaneously serve as an exact measure of spiritual experience, belief, religious commitment, and group centrality. Identity research requires explicit operational definitions and sometimes more than one instrument. Koteskey et al. (1990) tested a relational approach to identity that included family, community, culture, and religion, while Leach et al. (2008) and Roccas et al. (2008) developed multidimensional accounts of group identification. These instruments and models are not interchangeable, and transferring an instrument across religious and secular groups requires evidence of equivalent item functioning.

Sampling and questionnaire mode also affect who appears in the dataset. For example, a study recruiting respondents through congregations is unlikely to capture the full spectrum of privately religious but unaffiliated people or persons who have ceased public participation. Surveys using only 'religious' versus 'not religious' categories may misclassify spiritual but not religious respondents or impose labels on individuals who reject the binary. The 2023–2024 Religious Landscape Study asks multiple questions about religious identity, spirituality, and religiosity Pew Research Center (2025b) Pew Research Center (2025c), illustrating the value of differentiating survey objects. However, even independently measured self-ratings do not directly reveal narrative meanings or intimate behaviors without additional evidence.

Credible psychometric practice evaluates item comprehension, dimensionality, test–retest reliability where meaningful, measurement invariance, and relationships with theoretically related and distinct variables. The same standard applies to atheism and agnosticism measures, which should distinguish uncertainty from indifference and a belief judgment from a positive social identification. Psychometric success within one American population cannot establish equivalent interpretation among non-Western respondents. Researchers may need qualitative interviews before establishing a shared item pool. In studies of identity change, repeated measures should minimize artificial changes due to altered question wording or response options. Such a design allows observed change to be meaningfully distinguished from instrument change Zinnbauer et al. (1997) Lee (2014) Pew Research Center (2025a).

Research Designs, Causal Inference, and Replication

Cross-sectional surveys, longitudinal panels, qualitative interviews, social-psychological experiments, and cross-national comparisons answer different scientific questions. A cross-sectional association between strong religious identity and well-being may reflect social support, prior mental health, family stability, or selection into communities. A longitudinal association can establish temporal ordering for measured variables while leaving unmeasured time-varying confounding. An experiment manipulating the perceived prevalence of atheists can identify an effect of the experimental informational treatment on measured prejudice in its setting, but it does not show the causal effect of becoming an atheist. Distinguishing these inferential targets is essential to sound synthesis Gervais (2011) Upenieks and Ford-Robertson (2022).

Qualitative work is particularly strong at showing how identities acquire meaning in local interactions. The studies of nonreligious identity narratives by Smith (2011) and Frost (2019) provide detail unavailable in categorical surveys, but selected interview participants cannot yield unbiased prevalence estimates for entire populations. Conversely, probability-based surveys can estimate the distribution of an explicitly worded category with known design limitations, while potentially missing identity processes that their response options do not represent. Triangulation combines these strengths by developing constructs through qualitative inquiry, testing them psychometrically, estimating distributions in representative samples, and examining temporal pathways through longitudinal evidence.

Replication requires more than repeating a headline comparison. Investigators should replicate the exact operational definition, assess robustness to alternative classifications, and determine whether findings generalize across age, language, region, and levels of religious marginalization. A study of adolescents leaving a high-cost religious community Hardy et al. (2026) addresses a different question from a study of secular identification in a setting where nonreligion is common. Both may be rigorous while yielding different associations. Preregistered analytic decisions, appropriately powered samples, transparent missing-data practices, accessible code, and explicit checks of statistical uncertainty increase trust. Where causal claims are central, natural experiments or carefully designed interventions may be informative, provided ethical conditions and plausible identification assumptions are satisfied.

Comparative Evaluation of Explanatory Models

Three broad explanatory traditions organize the field. Social-identity models focus on identification, group boundaries, commitment, recognition, and the consequences of belonging Ysseldyk et al. (2010) Leach et al. (2008). Developmental and narrative models investigate how a person relates inherited and revised commitments to a continuous biography Lopez et al. (2011) Smith (2011). Social-ecological models highlight the institutional, demographic, and cultural conditions that make a religious or nonreligious position more or less salient or socially costly Frost et al. (2023) Hashem and Awad (2021). The traditions intersect because narrative continuity can depend on community recognition, and the consequences of group centrality may depend on local power relationships.

Each model makes a different set of predictions. A group-identification approach expects solidarity and centrality to correlate with forms of engagement when group belonging is relevant. A narrative approach predicts that identity-related distress or well-being may vary with the ability to integrate personal experiences into a valued biography. A social-ecological approach predicts moderation by the norms and institutions of a given setting. These claims are testable if measures are designed to avoid the use of outcome language as part of the predictor itself. For example, defining 'positive religious identity' as feeling supported will inflate its correlation with perceived support. The study must operationalize identification independently from downstream psychological outcomes.

The empirical literature supports a multidimensional synthesis rather than exclusive adherence to one explanatory framework. Evidence from adolescents Lopez et al. (2011), adults describing nonreligious identity Frost (2019), survey work Pew Research Center (2025b), and research on discrimination Hashem and Awad (2021) makes it possible to identify partially independent levels of analysis. The synthesis itself is an integrative reading of published studies, not a newly validated causal theory. Stronger adjudication would involve head-to-head tests in multiple populations with longitudinal or experimental designs, comparable instruments, and preregistered predictions. Until such tests are available, explanatory claims should be matched to the evidence of their constituent mechanisms.

Scientific Controversies and Unresolved Questions

One dispute concerns whether spirituality and religion form separable dimensions, overlap substantially, or should be defined according to local language practices. Conceptual and empirical work Zinnbauer et al. (1997) Ammerman (2013) Wixwat and Saucier (2021) shows that self-described spirituality carries heterogeneous meanings, making universal classification difficult. A second dispute concerns how to interpret the rise of unaffiliation: survey categories can indicate institutional departure, private disbelief, changing norms of disclosure, or some combination. The Religious Landscape Study Pew Research Center (2025b) documents demographic change but cannot determine all psychological mechanisms responsible for it from aggregate category shifts alone.

A third controversy concerns mental health. Comparisons across identity categories can reflect social resources, adverse experiences, prior distress, family situations, and measurement choices. Longitudinal studies Upenieks and Ford-Robertson (2022) Hardy et al. (2026) and newer parent–child evidence Clevenger and Dyer (2026) offer more informative trajectories than simple group comparisons, yet they remain observational. Their populations and categories cannot be treated as interchangeable. Another open question concerns whether the psychological significance of minority identity is driven more by privately experienced centrality, external social classification, or ongoing discrimination. Studies of Arab American groups Hashem and Awad (2021) and atheist concealment Frost et al. (2023) indicate the importance of measuring each channel.

Research programs should prioritize valid cross-cultural instruments, representative panels, developmentally appropriate questions, and closer integration of qualitative and quantitative methods. More work is needed on people with mixed heritage and mixed belief positions, those for whom religion is largely irrelevant, individuals who maintain cultural participation while privately disbelieving, and identities that are concealed for reasons of safety. The growing importance of digital environments raises questions about how mediated interactions change public expression. These remain empirical questions about persons and societies, requiring data sensitive to political, historical, and institutional differences. A research program that welcomes variation in identity structure is better equipped to explain apparently contradictory results.

Conclusion

Religious, spiritual, and nonreligious identities are multidimensional arrangements of personal self-definition and social participation. They can organize affiliation, belief, practice, emotional investment, narrative continuity, belonging, and public recognition in different combinations. Their diversity is visible in original conceptual investigations Zinnbauer et al. (1997), social-identity models Ysseldyk et al. (2010) Leach et al. (2008), qualitative accounts of secular identity Lee (2014) Frost (2019), longitudinal developmental findings Lopez et al. (2011), and contemporary population surveys Pew Research Center (2025b) Pew Research Center (2025a). The evidence shows that nominal affiliation cannot stand in for the whole psychological organization of religious or nonreligious life.

Across the research literature, observed associations with well-being are shaped by a person's history and social ecology. Developmental studies distinguish identity centrality from participation, while studies of discrimination and concealment show why public reception matters Lopez et al. (2011) Hashem and Awad (2021) Frost et al. (2023). Longitudinal health studies offer evidence about temporal trajectories without establishing a universal causal effect of any identity category Upenieks and Ford-Robertson (2022) Hardy et al. (2026) Clevenger and Dyer (2026). The most defensible scientific conclusion is that identity-related outcomes arise through multiple measurable processes involving commitments, relationships, resources, expectations, and life conditions.

The central agenda for future work is to develop conceptually precise, culturally interpretable, longitudinally robust measures capable of connecting identity descriptions with enacted lives. Research should specify its categories, populations, methods, and inferential claims and preserve the voices of people whose positions do not fit conventional classifications. In the Artificial Era, new tools for mediated communication may alter how identity narratives are explored, but the psychological consequences must be investigated as empirical hypotheses. This dossier establishes a scholarly synthesis for that inquiry while retaining the independence of neighboring research on conversion, religious cognition, spiritual practices, community institutions, and clinical outcomes.

References