Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Religious Conversion, Deconversion, and Disaffiliation

Abstract. Religious conversion, deconversion, and disaffiliation involve changes in beliefs, practices, institutional affiliation, personal identity, and social relationships that can unfold on different schedules. This critical narrative research synthesis examines the history of conversion psychology, operational distinctions among religious switching and personal deconversion, social and cognitive mechanisms, developmental contexts, family consequences, personality and well-being trajectories, comparative population evidence, and the methodological problems of causal inference. Original longitudinal studies from the Netherlands, New Zealand, China, and the United States are evaluated alongside Finnish administrative data, comparative surveys, qualitative accounts, and research on high-demand communities. The evidence indicates substantial temporal and contextual heterogeneity: religious practice often changes around formal affiliation in ways that precede the recorded event; broad personality effects are generally modest and vary by study; psychological outcomes differ across persons and institutional contexts; and pretransition trends complicate simple causal accounts. The monograph defines research priorities for repeated transitions, cross-cultural measurement, network effects, longitudinal identification, and digital and artificial-mediated religious inquiry. It reports no original experiment, preregistered systematic review, statistical meta-analysis, or clinical treatment recommendation.

Keywords: religious conversion; deconversion; religious disaffiliation; religious switching; religious identity; religious exit; religious doubt; social networks; family relationships; personality change; religious attendance; life-course development; psychological well-being; longitudinal research; religion and spirituality; Artificial Era

Religious Transition as a Psychological Research Object

Religious conversion, deconversion, and disaffiliation are changes in the organization of religious commitment over time. Their psychological study concerns beliefs, commitments, ritual participation, group membership, identity, social relationships, and the interpretation of one's biography. Conversion can designate entry into a religious tradition, movement between traditions, or intensified commitment within an existing tradition. Deconversion refers to a substantive process of withdrawing or revising religious commitment, sometimes involving a departure from the inherited religious worldview. Disaffiliation specifically refers to relinquishing an organizational or self-reported religious affiliation. These events often overlap, but their observable indicators need not coincide. Rambo (1993) Streib (2021)

Scientific explanation begins with the temporal relation between a psychological process and the event used to classify it. A survey may record a person as Christian in one wave and unaffiliated in the next; that observation establishes a category transition during the interval, not the date at which conviction changed. A participant may stop praying years before formally leaving a congregation, while another may retain private belief long after disaffiliation. Bleidorn et al. (2024) followed changes in practices and belief around formal changes of Christian affiliation; its findings demonstrate why the transition event must be distinguished from the processes that precede and follow it.

The central questions of this monograph are how commitment develops or weakens, which social and cognitive mechanisms are implicated, which consequences are observable, and how research design limits causal claims. The review integrates historical theories, prospective cohort studies, survey transitions, qualitative accounts, comparative populations, and contemporary observational findings. It is a critical narrative synthesis rather than a preregistered systematic review or new meta-analysis. Its primary unit is the individual's changing relation to a religious community across time, analyzed within family, institutional, developmental, and cultural contexts.

Conversion, Deconversion, Disaffiliation, and Religious Switching

Religious switching is an operational population-level term for a difference between reported childhood religion and reported current religion. Its scale depends on the categories offered. A Catholic-to-Protestant movement counts as switching when denominations are distinguished, but may disappear when Christianity is represented by one aggregate response. Pew Research Center (2025a) explicitly excludes within-world-religion moves from its international large-category switching measure, while Pew Research Center (2025b) reports United States switching under a different set of classification choices. Percentage estimates drawn from different coding schemes therefore answer different questions.

Deconversion typically directs attention to belief, meaning, identification, and personal disengagement, including their possible emotional or existential significance. Disaffiliation identifies the relationship between a person and a formal or self-described collective. Someone who adopts a privately spiritual outlook while leaving a church can be institutionally disaffiliated without describing the experience as loss of the sacred. Likewise, a person may doubt doctrine and remain a member through family ties, cultural tradition, or habitual participation. Streib (2021) treats deconversion as a heterogeneous biographical phenomenon requiring both longitudinal variables and individualized histories.

Conversion need not be confined to dramatic testimony. Joining a congregation can precede full assent, while belief changes can occur without community entry. Withdrawal from one tradition may be followed by entry into another, producing sequential deconversion and conversion rather than a single binary event. These distinctions have consequences for sampling, outcomes, and reporting: a study of attendance cessation investigates behavior, a religious identification panel investigates category mobility, and a narrative interview investigates experienced transformation. Each method becomes informative when the construct is kept explicit.

The Emergence of Conversion Psychology

Late nineteenth-century researchers made religious transformation a subject of empirical psychological inquiry. Starbuck (1899) assembled accounts of religious consciousness and development, including conversion in adolescence, as part of an early effort to investigate religious experience systematically. William James subsequently organized lectures on religious experience around varieties of disposition, struggle, and transformation. These historical works are indispensable for identifying the original problems of the field: the timing of religious awakening, its remembered causes, and the possibility of enduring reorientation. Their methods relied strongly on self-report and selected testimony, conditions that require contemporary scrutiny.

The distinction between an event and an interpretation was apparent from the field's beginnings. A convert's narrative may organize many preceding changes around one decisive moment because that moment gives the history social intelligibility. The researcher must investigate both the narrated turning point and the longer temporal development. Retrospective reports contribute access to experienced meaning but cannot independently establish the exact sequence of prior beliefs or eliminate reconstruction after commitment. Historical testimony and modern event-time panels accordingly offer different kinds of knowledge.

Twentieth-century research added institutional contexts, social networks, and explicit comparisons of individual and group change. Lofland & Stark (1965) examined entry into a marginal religious perspective, emphasizing the interaction of personal vulnerabilities, social ties, and recruitment relationships. Its small, situated field setting made it influential as theory-building evidence rather than a general prevalence estimate. The later integrative work of Rambo (1993) located conversion within cultural context and relational encounters, opening a framework in which individual cognition and community structures could be examined together.

Conversion as a Contextual and Relational Process

Rambo's account describes contextual conditions, crisis, quest, encounter, interaction, commitment, and consequences as analytical dimensions of conversion. Rambo (1993) presents these stages to organize the complexity of religious change rather than to demonstrate that every individual completes a fixed linear sequence. Cultural expectations, availability of organizations, perceived opportunities for belonging, persuasive encounters, and the interpretation of experience shape what form a transition can take. The model directs attention to opportunities for participation and to the social arrangements through which commitment becomes practicable.

An individual may seek religious community because it provides rituals, interpersonal recognition, moral guidance, a vocabulary for adversity, or a setting for collective action. Another may enter through marriage, migration, friendship, or gradual participation before constructing a new account of the change. Psychological motives should therefore be tested at more than one level. Statements about spiritual need are self-reported reasons; social network density is an observable relational characteristic; subsequent ritual participation is behavior. Their association does not by itself identify an initiating cause.

Social pathways also create selection effects. Communities have contact with people already receptive to participation, and participants choose or remain in settings that fit their values and relationships. Lofland & Stark (1965) makes interpersonal access an explicit component of the conversion problem, while Paloutzian et al. (1999) cautions against treating profound subjective change as evidence for alteration of all broad personality dispositions. The theoretically useful object is a transaction between personal tendencies, opportunities, institutions, and interpreted experience.

Historical Accounts of Deconversion and Role Exit

Systematic attention to leaving grew as religious mobility and unaffiliation became increasingly visible to researchers. Streib (2021) reviews deconversion in relation to development, values, attachment, well-being, and changing religious fields. One productive sociological analogy is role exit: the departure from a socially recognized identity can involve doubts, alternative affiliations, a public turning point, and the construction of a former-member identity. Haire (2022) applies the role-exit tradition to the mental-health narratives of religious exiters. As a framework it names recurring processes but is not a universal clinical progression.

Exiting a group can require renegotiating everyday practices, schedules, marriage expectations, language, friendships, and assumptions about the future. The significance of a departure depends on the proportion of ordinary life organized around the former community. A low-salience denominational change may affect few relationships, whereas leaving an encompassing community can modify employment, kinship, housing, and intimate partnership opportunities. The magnitude of institutional entanglement is therefore an explanatory variable, not an automatic property of the label 'deconversion.'

Biographical accounts often distinguish a period of unresolved doubts from public departure and subsequent reorganization. These categories are useful for interview design, but retrospective chronology may be shaped by the current worldview and the questions asked. The investigation should collect multiple dated indicators and preserve the individual's own terms for the transition. Streib (2021) calls for complementary nomothetic and idiographic methods, connecting population-level patterns to the distinctive trajectories that summary categories can conceal.

Temporal Organization: Preparation, Transition, and Reconstruction

Religious changes often unfold across several clocks: belief, attendance, social identification, organizational membership, and narrative self-description. The clocks can move together or in different orders. In the Dutch longitudinal study, Bleidorn et al. (2024) observed that measures of religiosity differed from formal conversion and deconversion timing, with some trends preceding and extending beyond classification changes. This undermines an interpretation in which an affiliation response is the sole initiating point of psychological change.

Event-centered panel designs align observations relative to the first observed change of group membership. This design offers a way to compare pretransition levels, immediate level shifts, and longer posttransition trends. Its interpretive limits include the distance between survey waves, the number of observed events, temporary switching, and the need to define a stable comparison population. An apparent immediate shift may reflect change that occurred anywhere between two assessments. Repeated religious status questions improve temporal resolution but may also expose instability that a simple first-switch indicator ignores.

Longitudinal explanation additionally requires separating selection and consequence. People who leave may have lower participation long before departure, and people who join may already endorse relevant beliefs. Bleidorn et al. (2024) reports large baseline religious differences between future switchers and comparison groups even when formal change occurred years later. These data support a model in which religious transition is often the culmination of prior processes. More intensive within-person studies are required to locate the mechanisms and direction of change.

Belief Revision and the Cognitive Structure of Religious Commitment

Religious belief is a multidimensional psychological construct encompassing judgments about supernatural agents, moral and metaphysical propositions, trust in religious authorities, and confidence in the interpretation of experience. Belief change may involve a gradual decline in certainty, a transformation of particular doctrines, or a reassessment of the standards by which claims are accepted. This multidimensional view matters because religious identity can persist after endorsement of some doctrines declines. Streib (2021) discusses deconversion as a change of values and developmental orientation alongside changes in conventional affiliation.

Studies that measure only belief in God cannot establish which doctrines or explanatory commitments have changed. Conversely, attendance cannot serve as an unambiguous proxy for private conviction: people may attend for kinship, cultural attachment, aesthetic satisfaction, or social obligation. Bleidorn et al. (2024) measured belief in God, prayer, and attendance separately around changes in Christian affiliation and found divergent temporal patterns. A precise theory of belief revision needs item-level measurement and attention to whether the same question has the same meaning before and after transition.

Religious doubt can prompt information seeking and conversations that alter confidence, but it may also coexist with robust commitment. Researchers should distinguish experienced uncertainty from institutional exit and from the later endorsement of a secular or different religious explanation. When evidence comes from retrospective testimony, the study may reveal the individual's present account of change without uniquely establishing which cognition caused the departure. Designs with repeated measurements of doubts and subsequently observed decisions are needed to test directional hypotheses.

Meaning, Values, and Personal Life Narratives

Religious frameworks can organize personal goals, moral commitments, belonging, interpretations of suffering, and expectations concerning death. Changing the framework can therefore alter midlevel goals or autobiographical narrative without producing a measurable transformation of broad temperament. In an influential theoretical synthesis, Paloutzian et al. (1999) proposed distinguishing relatively stable dispositional traits from more changeable motives, attitudes, behavior, meaning, and identity. This conceptual distinction helps reconcile participants' accounts of life transformation with modest average shifts in global personality scales.

Narrative reconstruction is an active process of making an altered life intelligible to oneself and others. A person may reinterpret childhood religious education, describe a period of crisis, or identify an experience that now symbolizes transition. The narrative is itself a psychological outcome: it can organize memory and social identity, communicate membership in a new group, and establish continuity amid changed convictions. Haire (2022) analyzes religious exit narratives through role and identity transformation, emphasizing the emotional and social context in which those stories acquire their meanings.

Researchers should avoid treating narrative coherence as proof that the events happened in the order given. Autobiographical explanations can change with the audience, new attachments, and subsequent exposure to religious or secular narratives. Longitudinal interviews, life calendars, and contemporaneous records can strengthen temporal inference. Both individual accounts and standardized measures are necessary because meaning and identity may change substantially while brief global well-being indicators remain comparatively stable.

Identity Alignment, Fit, and Social Norms

Commitment is partly a relation between personal orientations and the expectations of a social environment. Religious groups can furnish identities that are publicly recognized and reinforced by shared practices. People whose beliefs diverge from their peers may encounter strain, while those with matching values may find group participation easier. Bleidorn et al. (2024) discusses person–religion fit and sociocultural motive perspectives as explanations of conversion and deconversion. A fit hypothesis predicts transitions in part from differences between personal orientations and local norms; an effect-of-transition hypothesis instead predicts new changes following participation.

These hypotheses cannot be separated using a single measurement after a change of affiliation. The same observed difference between converts and nonconverts could reflect preexisting dispositions, anticipatory adaptation, socialization, or the influence of a third event such as marriage or relocation. Repeated assessments of pretransition beliefs, motives, personality, and community exposure provide a stronger basis for identifying which differences precede the formal switch. Population context is indispensable because an identical personal orientation may imply conformity in one community and dissent in another.

Fit also helps explain the asymmetry between theological agreement and social attendance. Someone may continue participation to sustain family bonds despite doctrinal disagreement; someone else may join primarily through friendship before changing beliefs. Treating affiliation as an outcome rather than as a complete description of inner commitment makes those patterns visible. Mixed-method designs can test predicted aggregate associations while using interviews to establish how participants interpret institutional membership.

Relationships, Recruitment, and Network Pathways

Conversion often involves access to a new community through friendships, family, partners, religious teachers, or organized outreach. Lofland & Stark (1965) placed interpersonal connections at the center of an early field account of recruitment, while Rambo (1993) integrated encounter and social participation into a broader process model. Networks can provide opportunities for discussion and rituals, models of religious behavior, normative pressure, and practical assistance. An interaction between network contact and personal receptivity is a more precise object than a general claim that community influence alone produces conversion.

Social influence has multiple mechanisms. Informational influence occurs when trusted contacts are treated as credible sources. Normative influence concerns the rewards and sanctions attached to accepted group conduct. Repeated participation produces behavioral familiarity and can change how individuals interpret their own commitments. These mechanisms can operate together but should be distinguished in measurement. An observational association between friendship networks and subsequent joining can also result from choosing friends with similar orientations, so network selection and socialization require separate tests.

Network effects after disaffiliation can run in the opposite direction. Leaving may reduce access to religious friends while increasing ties to former members or nonreligious communities. The practical importance of this shift depends on whether a new social network offers emotional, material, and informational support. Keller et al. (2026) studied social support and social-media engagement among former ultra-Orthodox Jews, showing that patterns of online coping and perceived support have different associations with mental health. The observed relationships are correlational and culturally situated.

Family Ties and the Cost of Changing Affiliation

Religious change can alter perceived family similarity, shared routines, and expectations about loyalty. Family members may treat a departure as a private choice, as a conflict over values, or as an event that changes communal identity. Hendricks et al. (2024) analyzed three waves of the National Study of Youth and Religion, involving 2,352 participants, and related changes in identification to subsequent parent–child relationship quality. Its longitudinal modeling found associations between deconversion and poorer relationships at specified waves, including pathways involving parental warmth and belief similarity.

These findings cannot establish that every religious exit damages families or that one direction of causality explains all observed associations. An already strained parent–child relationship may influence religious change, while changes in belief may provoke tension, and both can be shaped by developmental autonomy or other family circumstances. The use of repeated observations improves temporal ordering; however, unmeasured time-varying family factors remain possible. Distinguishing the source of affiliation change and the household's response is essential.

The same family system may support conversion or disaffiliation through encouragement, companionship, or conflict mediation. Intergenerational differences also influence whether an adult's new identity appears as continuity, innovation, or rupture. Studies should record whose belief or affiliation changed, who anticipated the change, the family's religious homogeneity, material dependence, and the degree of communication or exclusion. A blanket category of 'family rejection' obscures the substantial variation between negotiation, temporary disagreement, supportive disagreement, and sustained estrangement.

Developmental Windows and Life-Course Conditions

Adolescence and emerging adulthood are significant periods for religious identity exploration because relationships, education, autonomy, and institutional commitments are undergoing change. Starbuck (1899) made adolescent conversion historically prominent, although his population and method cannot determine modern age-specific incidence. Contemporary research must separate developmental change from historical cohort effects and the social setting in which a person comes of age. An apparent age gradient may arise from changing opportunities to leave home, employment transitions, or variation in religious institutions.

Lefevor & Skidmore (2025) followed LGBTQ+ Latter-day Saints over four years and found that younger participants deidentified more quickly, while greater engagement with LGBTQ+ communities predicted subsequent reductions in service attendance in cross-lagged analyses. The particular religious and social context is crucial: the finding identifies relationships within a specific population rather than an invariant developmental law. It also provides evidence for prospective association between access to alternative communities and change in religious participation.

Transitions later in adulthood may coincide with bereavement, illness, marriage, migration, retirement, or reassessment of identity. Because these events also affect social support and psychological well-being, they create difficult causal comparisons. A life-course approach asks what opportunity structures, family obligations, and historical conditions surround each change rather than imposing a single age-based sequence. Studies must follow both those who shift and comparable individuals who maintain affiliation to distinguish religious change from broader developmental trends.

Personality Change: Theoretical Levels and Causal Questions

One influential claim of conversion testimony is that the person becomes fundamentally different. Personality psychology requires specifying which level of organization the claim concerns. A change in religious self-understanding, lifestyle, or values can be profound for the individual even where broad trait measures detect only small mean changes. Paloutzian et al. (1999) reviewed this problem and distinguished relatively stable dispositional features from middle-level goals, feelings, and attitudes and more self-defining identity and meaning. These distinctions are conceptual, not evidence that all conversions have the same consequence.

The corresponsive principle of personality development suggests that experiences can reinforce tendencies that helped people select those experiences. If a person joins a community congenial to preexisting dispositions, subsequent roles might amplify those tendencies. However, a post-conversion trait shift can reflect repeated behavior, changing self-description, measurement effects, or concurrent life events. Stronge et al. (2021) and Bleidorn et al. (2024) offer unusually informative longitudinal investigations yet produce partly different trait patterns. Their juxtaposition discourages a universal personality-change narrative.

Claims about causes require comparisons against plausible counterfactuals. A person cannot be observed both joining a faith and not joining it under exactly the same circumstances, so panel studies use comparison groups and assumptions about trajectories to approximate this contrast. The timing of assessment, stable differences between groups, and social context alter interpretation. Personality scale invariance across transitions is also relevant: endorsing 'agreeable' or 'conscientious' statements may acquire different reference points when group expectations change.

The Netherlands: Eleven Waves Around Institutional Transition

Bleidorn et al. (2024) analyzed 11 waves from the Netherlands' Longitudinal Internet Studies for the Social Sciences, using a population panel of approximately 20,000 adults. In the event-time analytic groups, 181 participants newly affiliated with a Christian church and 450 left Christian affiliation. The investigators assessed belief in God, attendance, prayer, Big Five personality, and well-being before, at, and after the institutional transition. This design is an important advance because it observes individuals across time rather than asking them to recall an entire transformation in a single interview.

The converts increased service attendance and prayer as they approached conversion without a corresponding increase in stated belief in God. Deconverts showed declines in religious belief and practice spanning periods before and after disaffiliation. Some small shifts in personality traits appeared near the change of membership, but the directions were not uniformly consistent with intuitive expectations. Neither group showed a detectable change in measured well-being. The distinction between a change in institutional label and a long trajectory of belief and behavior is the study's most generalizable methodological contribution.

Interpretation remains bounded by the setting and event counts. The probability-sampled Dutch context supports analysis of those observed, but the conversion and deconversion subsamples are modest relative to the full panel. Events were defined by Christian and nonreligious affiliation, leaving other traditions and within-Christian switches outside the principal comparison. Observation intervals constrain the precision of event dates, and measures of meaning crises, coercion, and identity reconstruction were not available. The study was not preregistered, as its authors report. Bleidorn et al. (2024)

New Zealand: Character Traits Across Religious Mobility

Stronge et al. (2021) used a nationally representative New Zealand longitudinal sample spanning 2009–2017 and applied piecewise latent growth models to religious conversion and deconversion. The source reports 31,604 adults in the broader sample, with 540 converts and 886 deconverts in analyzed transition groups. The investigators examined Big Five dimensions and honesty–humility before and after changes of Christian affiliation. Its large population frame and multiyear observation provide evidence complementary to the Dutch panel while differing in religious context, trait measures, and model specification.

The study reported increases in honesty–humility, conscientiousness, and neuroticism after conversion; before deconversion it observed increasing honesty–humility and decreasing agreeableness. It did not observe personality changes before conversion or after deconversion under the specified analyses. These findings are useful precisely because their timing differs by transition direction and trait. Honesty–humility belongs to the HEXACO framework and should not be equated with a generic moral character score. Results depend upon the instrument and the definition of each latent trajectory.

Comparison with Bleidorn et al. (2024) shows that small trait effects are sensitive to timing, cultural setting, and modeling choices. The Dutch study reported limited level shifts at the transition and measured Big Five traits rather than the New Zealand study's special emphasis on honesty–humility. A difference in findings does not make either population invalid; it identifies which claims remain contingent. The synthesis supports careful measurement of personal goals and practices alongside broad traits and discourages interpreting the formal affiliation decision as proof of a uniform transformation in personality.

Chinese Prospective Evidence: Divergent Psychological Outcomes

Hui et al. (2018) conducted a three-year prospective study of 632 Chinese Protestant Christians, among whom 188 left the faith during follow-up. The authors examined personality, beliefs, values, and psychological symptoms, allowing comparisons between eventual exiters and those who remained. Prospective assessments provide a temporal advantage over entirely retrospective accounts, but the sample is religiously and culturally bounded. Religious exit in this design relates to the study's specific criteria for disengagement and cannot automatically be generalized to all forms of institutional or metaphysical deconversion.

The investigators found limited distinctive postexit personality change. Some measures of beliefs and values shifted relative to those who stayed, and eventual exiters displayed certain differences before departure. Growth-mixture analysis suggested that approximately half of exiters followed a trajectory of early symptom improvement and the other half early deterioration. Lower initial quality of life was associated with a less favorable postexit symptom trajectory. These findings directly challenge the idea that leaving produces a uniform improvement or decline in psychological condition.

The mixture-model classes are statistical summaries of heterogeneity, not permanent psychological types or diagnostic categories. Class proportions depend on model specification, measurements, and sample composition. The strongest substantive inference from Hui et al. (2018) is that trajectories vary and that relevant values may diverge before the observed exit. More detailed pretransition measurements and repeated observations across different institutional settings are needed before assigning a dominant causal mechanism.

Deidentification, Minority Experience, and Alternative Communities

Religious transitions intersect with sexuality, gender, family values, and the acceptability of personal identities within institutions. Lefevor & Skidmore (2025) investigated religious deidentification among LGBTQ+ Latter-day Saints using four years of longitudinal observations. The study associated younger age with faster deidentification and linked engagement with LGBTQ+ communities to subsequent changes in service attendance. Mental health and well-being were not significantly associated with religious deidentification in the reported analyses. Such findings illustrate why minority-specific relational environments require direct study.

A participant's religious exit can reflect doctrinal disagreement, conflict between personal identity and institutional norms, an expanded range of social alternatives, or combinations of these processes. Access to an affirming network may precede a reduction in religious attendance without necessarily representing an externally imposed withdrawal. Cross-lagged results are compatible with a temporal sequence, but observational cross-lagged associations still require assumptions to justify causal interpretations. A change in attendance may also coexist with continuing private belief.

Ethical interpretation depends on treating respondents' descriptions of identity and religion as empirical data rather than diagnosing one commitment or another. Minority status may alter the social risks of remaining as well as the costs of leaving, and these vary within a religious tradition. Comparative designs need adequate representation of those who remain religious while affirming their sexual or gender identity, those who change affiliation, and those who become unaffiliated. Lefevor & Skidmore (2025)

German–American Deconversion Cohorts and Long Intervals

Chen & Streib (2026) analyzed religious deconversion using an evidence base assembled over two decades in Germany and the United States. The study reports an analytic sample of 4,451 and a longitudinal subset of 502, employing cross-sectional and exposure-/outcome-wide longitudinal comparisons. Its operational contrast concerns people reporting that they deconverted from the religion of their upbringing versus people who stayed. The long intervals between assessments provide a perspective on lasting associations rather than on short-lived responses immediately around departure.

These data make an important methodological contribution because the field often relies on a single transition narrative or a small local sample. At the same time, the long-distance temporal comparison may conceal shorter cycles of doubt, temporary return, or successive affiliations. A self-reported status of having left the religion of upbringing also combines several possible trajectories, including adoption of another religion and secular disaffiliation. Claims about individual predictors and outcomes should therefore be linked to the actual measured variables and the study's analytic contrasts.

A robust synthesis should set results from this cohort alongside the denser event-time series of Bleidorn et al. (2024) and the three-year trajectories of Hui et al. (2018). Each design identifies a different temporal pattern. Wide-scope predictor analyses raise multiple-testing and selection concerns; event-centered models require enough transitions for stable estimates; and interviews offer mechanisms without identifying population-level prevalence. Convergence across these methods, rather than the apparent detail of any single method, is the appropriate standard for broad claims.

Disaffiliation and Population Religious Mobility

Population surveys describe the distribution of religious identities but must be read according to their definitions and observation periods. The 2023–2024 United States Religious Landscape Study, summarized by Pew Research Center (2025b), found that approximately 35% of adults reported a current religious identity different from the one in which they had been raised. The proportion includes moves between religions, changes to no affiliation, and movement from no affiliation to religious membership. It is a comparison between retrospective childhood affiliation and current identification, not a prospective count of conversion events observed as they occurred.

The study reported that 20.2% of all United States adults were raised in a religious tradition and currently identify as unaffiliated, while 3.5% moved from being raised unaffiliated to identifying with a religion or did not state their current affiliation. Those percentages use the 2023–2024 survey's adult population denominator; they are not proportions among people who had ever practiced or professed detailed beliefs. Retrospective classifications are susceptible to recall and category interpretation, but they reveal the large demographic scale at which religious life-course mobility matters.

Population-level net change cannot determine whether exit improved or worsened an individual's mental health. Aggregate movement combines generational replacement, immigration, differential retention, cross-category changes, and differences among social environments. A psychological explanation needs person-level measures collected across time, preferably including reasons for movement, relationship context, and beliefs as separate dimensions. Pew Research Center (2025b) makes the extent of switching visible; Bleidorn et al. (2024) investigates individual changes around institutional transitions. Their contributions are complementary rather than interchangeable.

Global Comparison, Cultural Variation, and Classification

Pew Research Center (2025a) analyzed religious switching across 36 countries and reported marked variation in the proportion identifying with a different broad religious category than in childhood. The report's international framework compares present identification with religious upbringing and excludes movement within a major world religion. It consequently differs in definition from some national denominational studies. It found that unaffiliation accounts for a large share of switching in many surveyed societies while other countries exhibit high retention. A global average would conceal substantial institutional and historical variation.

Religious traditions vary in the social meaning of membership. Some identities integrate ancestry, family, language, ritual, and nationality, while others emphasize voluntary assent, institutional participation, or conversion ceremony. A person who changes belief may continue identifying with a tradition as a cultural inheritance; another may change formal affiliation while retaining practices. The same checklist item can therefore represent different social facts. Comparative research requires translation work, local expert review, cognitive testing of questions, and a clear distinction between inherited identity and active commitment.

The global findings invite hypotheses about social regulation, educational opportunity, religious competition, and the social costs of switching. Their cross-sectional design cannot establish that national context caused the individual transitions it summarizes. Geographic differences may reflect cohort composition, religious history, legal regimes, or survey wording. Pew Research Center (2025a) is strongest as a descriptive comparative source; mechanisms require designs that compare contexts while measuring who changes, when, and under what interpersonal conditions.

Well-Being After Religious Exit: Competing Associations

Religious disaffiliation is frequently presented either as liberation from constraint or as loss of meaning and social belonging. Population studies reveal that both narratives can capture real experiences, while neither describes every participant. Fenelon & Danielsen (2016) used pooled General Social Survey observations from 1973–2012 to compare adults who left the religion of upbringing with persistent affiliates and those consistently unaffiliated. It found lower health and subjective well-being among disaffiliates in those comparisons, and the reported difference was statistically accounted for by attendance frequency.

Such mediation does not prove that declining attendance caused lower well-being. Attendance may index social ties, health capacity, normative compatibility, or unmeasured selection processes. The contrast includes people who leave for very different reasons and is not an experimental manipulation of attendance. Hui et al. (2018) additionally reported opposite postexit symptom trajectories within a prospective Christian sample. These studies should be interpreted together: mean differences and heterogeneous within-person change can coexist without contradiction.

Researchers should specify whether the outcome is life satisfaction, depressive symptoms, distress, psychological flourishing, self-rated health, or social isolation. These constructs share variance but have distinct time courses and response properties. The consequences of religious exit can be contingent upon remaining social ties, perceived safety, family reactions, alternative communities, and earlier well-being. A clinical interpretation of population-level differences without assessment of the individual case is unjustified.

Attendance Cessation and Health Trajectories

Jokela & Laakasuo (2024) analyzed people who went from at least monthly religious attendance to no active attendance in four prospective United States cohorts. The combined sample included 6,592 individuals and 37,743 person-observations gathered across studies with differing age ranges. Outcomes included psychological well-being, depressive symptoms, self-rated health, body mass index, and health behaviors. The investigators found no worsening of the ten-year trajectories after attendance cessation across their measured outcomes; less favorable trends were already observable during periods of attendance.

This design improves the interpretation of preexisting differences, but its exposure remains cessation of attendance rather than an exhaustive psychological definition of deconversion. Some participants may have stopped because of illness, relocation, changing family schedules, or altered religious motivation while retaining their faith. In addition, the relationship between attendance and health may depend on denomination, local social opportunities, and population age. It is important to retain the authors' specific outcome and event definitions.

The result counters simple claims that attendance cessation independently produces broad health deterioration, while leaving open effects on unmeasured outcomes, short-term grief, and particular groups. Fenelon & Danielsen (2016) and Jokela & Laakasuo (2024) use different observational contrasts, and their conclusions differ partly because one estimates between-group differences while the other studies trajectories around changes in behavior. This distinction is central to causal reasoning in the psychology of religion.

Psychiatric Outcomes and Finnish Population Registers

Reini et al. (2026) used Finnish population registers to examine religious switching and sickness allowance granted for mental disorders in young adulthood. Its prospective cohort followed more than one million people born from 1986 through 2003, using register-based observation between 2004 and 2023 and analyses that included sibling fixed-effect comparisons. The outcome was an administrative benefit associated with registered mental disorders, rather than a direct questionnaire of all symptoms or a diagnostic interview administered to every citizen.

The study found higher hazards of mental-health-related sickness allowance in people who switched religious affiliation and also observed elevated hazards in certain affiliation categories. Crucially, among switchers the rise in sickness-allowance incidence was steeper before the recorded change than after it. This temporal pattern is consistent with the possibility that health and social difficulties precede or contribute to switching; it also permits other common-cause interpretations. The observational design cannot determine one universal direction of causation.

Administrative records permit large samples and relatively reliable event dating, but they measure help seeking, certification, eligibility, and labor-market circumstances as well as illness. Public benefit systems differ across countries, and registered affiliation may miss privately held belief. Reini et al. (2026) should thus inform the discussion of temporal ordering and institutional context, not be read as establishing that religious switching itself causes psychiatric disorder. Findings are compatible with heterogeneous experiences in smaller prospective studies.

Resilience, Distress, and Exit from Demanding Communities

Leaving a high-demand religious community may involve the simultaneous loss of social support, identity, employment prospects, and familiar practices. Thoma et al. (2022) studied 622 former members of fundamentalist Christian communities using a cross-sectional online survey and latent profile analysis. It identified four distributions of well-being indicators labeled resilient, normative, vulnerable, and adverse. The proportions in that particular sample were about 25.7%, 36.4%, 27.2%, and 10.7%, respectively. The categories describe statistically derived profiles in a self-selected sample, not population prevalence among all former members.

The study associated more favorable profiles with resources such as social support, self-esteem, and sense of coherence; abuse or maltreatment as a reason for departure was particularly relevant. The associations cannot independently establish whether protective resources preceded exit, followed it, or arose from a third factor. An additional qualitative investigation by Luther (2023), based on interviews with ten former Jehovah's Witnesses, documented reported experiences of ostracism and disrupted social life. Such interviews establish concrete experiences and processes but do not provide a general population effect size.

These sources support studying the social conditions of departure as potential moderators of psychological outcome. They also warn against treating a single high-demand tradition as representative of religion generally or defining every exit as a disorder. Follow-up studies should compare voluntary and involuntary departures, retained friendships, social isolation, material dependency, institutional responses, and available support outside the group. The ethical task is to measure lived consequences carefully while respecting each participant's account.

Religious Exit, Ostracism, and Social Replacement

Some departures are accompanied by social exclusion, conflict, or the loss of shared resources. The relevant construct is the actual experience of social sanctions and relational loss, which cannot be inferred simply from a person's former religious label. Luther (2023) used interpretative phenomenological analysis of interviews with ten former Jehovah's Witnesses and identified themes of loneliness, distrust, disrupted opportunities, and continuing distress associated with reported shunning. Such qualitative work clarifies possible mechanisms and identifies experiences that standardized affiliation surveys may omit.

Exiters may also form new relationships that restore belonging or broaden access to support. Keller et al. (2026) surveyed 1,146 people who had left ultra-Orthodox Judaism and examined social support and social-media coping. Supportive problem-solving and perceptions of giving and receiving help correlated with more favorable outcomes, whereas avoidant and escapist online coping related to poorer mental-health measures. The associations do not prove that one online behavior directly improved mental health; distress may itself shape coping behavior and platform engagement.

A meaningful research design distinguishes emotional support, material assistance, community identity, and coercive social control. These dimensions can change separately. It must also account for who experiences an imposed exclusion as opposed to a voluntary and supported change of community. The social ecology surrounding the transition may determine whether a person experiences relief, grief, or both, and it is a strong candidate for repeated assessment in future prospective research.

Operationalization, Psychometrics, and Measurement Equivalence

Measurement begins by defining what counts as an observed transition. Studies may use a religious-affiliation item, attendance threshold, prayer frequency, self-identification, conversion ritual, degree of conviction, or explicit narrative of departure. These indicators have different reliability and meaning. A single retrospective comparison between childhood and present affiliation misses multiple transitions and their timing, while a panel response may register a change that occurred at any point between assessment waves. Pew Research Center (2025b) and Bleidorn et al. (2024) illustrate these contrasting operational strategies.

Construct validity requires distinguishing organizational exit from belief revision, spiritual change, identity transformation, and psychological distress. A participant can endorse nonreligious identification yet retain spiritual practices, or continue membership after revising essential beliefs. When scales compare groups across time, measurement invariance should be assessed before interpreting a score difference as a change in an underlying psychological property. Social desirability, local sanctions, language, and the perceived confidentiality of questions may influence both reporting and participation.

Research should record item wording, response options, missingness, religion-specific category definitions, and whether the person considers the reported move voluntary. Standard questionnaires make group comparisons efficient, whereas in-depth interviews can expose hidden assumptions of the categories. The optimal combination depends on whether the question is incidence, psychological mechanism, institutional process, or autobiographical meaning. Streib (2021) explicitly supports integrating broad quantitative and individual biographical approaches.

Research Designs, Confounding, and Evidence Evaluation

Longitudinal panels give change a temporal structure but remain observational unless participation or exposure is randomly assigned under a defensible intervention design. Event studies align measurements around transition; latent growth models estimate trajectories; cross-lagged models examine directional prediction; and sibling comparisons reduce some shared familial confounding. None automatically removes unmeasured time-varying causes. Bleidorn et al. (2024), Stronge et al. (2021), Lefevor & Skidmore (2025), and Reini et al. (2026) illustrate these analytical families with distinct questions, samples, and definitions.

Selection, attrition, and event rarity are persistent problems. Individuals who switch may differ from stayers years before the switch, and those experiencing the greatest social consequences may be underrepresented in long follow-up. Model estimates depend on the frequency of observation and the specification of comparison groups. A broad nationally representative panel may still contain few transition events, while a targeted exiters' sample can offer rich detail but cannot estimate population transition rates. All inferences must state these constraints rather than equating a large overall sample with a large event sample.

Evidence quality is also affected by selective reporting, cultural coverage, measurement overlap, and uncertainty about time. The current literature combines original empirical studies, methodological reviews, population reports, and qualitative narratives. This monograph does not combine effect sizes into an estimated pooled intervention effect because outcomes, time horizons, exposure definitions, and designs are substantively heterogeneous. Its conclusions emphasize directions supported by convergence and identify domains where the evidence remains conditional.

Cross-Cultural and Institutional Boundaries

Explanatory theories developed from Christian affiliation in secular Western countries need direct validation in other religious and social environments. Pew Research Center (2025a) documents marked international differences in the frequency and direction of religious switching, but these differences alone cannot establish which psychological mechanisms operate locally. The meaning of a change in religious identity depends on whether religious membership is inherited, legally registered, socially exclusive, individually chosen, or associated with cultural belonging beyond belief.

Within the same country, minority status, family organization, urban migration, religious education, and social control may alter the risks of making change public. The empirical literature reviewed here includes Dutch and New Zealand Christian panels, Chinese Protestant participants, American religious mobility, Finnish register data, former Jehovah's Witnesses, and exits from ultra-Orthodox Judaism. Their coexistence enables methodological comparison but does not confer global representativeness. Claims must be indexed to populations actually studied.

Research on groups experiencing legal or coercive restrictions raises specific ethical demands. Confidentiality and informed consent can be more consequential where disaffiliation endangers housing, family contact, or safety. Recruitment through public former-member networks can expose participants and overrepresent highly visible stories. Ethically robust inquiry should offer confidential pathways, avoid publishing identifiable transition narratives without explicit permission, and interpret differences as features of particular social ecologies rather than characteristics of entire traditions.

Religious Transition in Digital Social Environments

Digital communication changes the opportunities through which people encounter religious ideas and alternative networks. It can extend access to information, former-member communities, private discussion, and peer support, while also exposing people to harassment, surveillance, and intensified group conflict. Keller et al. (2026) provides direct observational evidence that different purposes for social-media use correlate differently with well-being among former members of ultra-Orthodox Judaism. This evidence concerns human behavior within networked platforms; it does not establish a universal therapeutic benefit of digital participation.

Online conversion accounts and deconversion testimony can offer models for interpreting ambiguous experiences. Such narratives may serve as sources of information, identity validation, recruitment, or social belonging. Researchers should distinguish self-selection into these spaces from effects of exposure and should examine the accuracy of online representations. Repeated surveys and platform-aware longitudinal interviews can track whether digital contact precedes changes in attendance, affiliation, or private belief while acknowledging the privacy risks of collecting individual activity traces.

Within the Artificial Era, interactive artificial systems are becoming part of the information environments in which people formulate questions about faith, identity, and institutions. Psychological hypotheses about AI-mediated exploration concern perceived epistemic authority, conversational reinforcement, privacy, and possible shifts in reliance on human social networks. These questions require prospective studies that directly measure interactions and outcomes. They remain an interdisciplinary research agenda, and the present evidence on human digital networks cannot be reclassified as established evidence of Artificial Sapiens-mediated religious change.

Critical Synthesis and Open Scientific Problems

The strongest longitudinal evidence establishes that changes in religious behavior, belief, affiliation, and broad personality are partly separable. Bleidorn et al. (2024) found religious trajectories extending beyond the formal affiliation change with small and varied trait effects and no corresponding detectable change in well-being. Stronge et al. (2021) found trait changes with a different timing and measurement profile. Hui et al. (2018) identified heterogeneous psychological symptom trajectories after exit. These studies jointly support a model of multiple linked processes rather than one standardized psychological event.

Social consequences vary with institutions and available alternatives. Hendricks et al. (2024) documents associations with family relationship quality, Thoma et al. (2022) identifies diverse well-being profiles among former members of demanding communities, and Luther (2023) describes lived effects of reported shunning. Jokela & Laakasuo (2024) and Reini et al. (2026) show why pretransition trends matter when interpreting health associations. No single result justifies claiming that conversion universally improves personality or that disaffiliation universally harms mental health. Interpretations must preserve the particular observed population and outcome.

Open questions concern the mechanisms linking doubt to action, the relation between change in stated belief and tacit practice, how friendship selection differs from network influence, whether trajectories vary across institutional costs, and which variables predict resilience without conflating prediction with cause. More work is needed on repeated transitions, return to religion, age-specific processes, religion-specific measurement invariance, cross-cultural representativeness, and the effects of digital mediation. These questions follow directly from the field's observed diversity and methodological constraints.

Conclusion

Religious conversion, deconversion, and disaffiliation are psychological transitions in which belief, practice, institutional membership, personal identity, and social relationships can reorganize on different temporal schedules. Formal religious switching is a useful demographic observation, while the full psychological phenomenon includes processes that may precede the observed membership change and continue beyond it. Historical research made personal transformation a problem for psychology; contemporary prospective designs make its timing, heterogeneity, and evidentiary limits measurable.

The evidence supports a differentiated picture. Many changes are gradual; broad personality shifts are generally modest and context-dependent; religious participation and belief do not move uniformly; and well-being trajectories vary with prior circumstances, family response, community structure, and resources after transition. Population surveys and individual longitudinal studies answer complementary questions and should be interpreted in their own methodological terms. Religious change cannot be reduced to a single mental-health outcome or a fixed sequence of narrative stages.

Future science can advance by integrating repeated belief and affiliation assessments, relational and institutional measures, culturally valid instruments, explicit causal models, and careful study of online and artificial-mediated environments. The resulting research program treats religious transition as an observable family of cognitive, developmental, social, and biographical processes. Its purpose is to understand how people construct, revise, enter, and leave systems of commitment, and how those transitions interact with the broader organization of psychological life.

References

  • Bleidorn, W., Lenhausen, M. R., Schwaba, T., & Hopwood, C. J. (2024). Psychological change before and after religious conversion and deconversion. Journal of Personality, 92(4), 1193–1210. https://doi.org/10.1111/jopy.12881
  • Chen, Z. J., & Streib, H. (2026). Characteristics, predictors and outcomes of religious deconversion: A cross-sectional and longitudinal analysis. The International Journal for the Psychology of Religion, 36(1), 97–114. (Published online in 2025.) https://doi.org/10.1080/10508619.2025.2520104
  • Fenelon, A., & Danielsen, S. (2016). Leaving my religion: Understanding the relationship between religious disaffiliation, health, and well-being. Social Science Research, 57, 49–62. https://doi.org/10.1016/j.ssresearch.2016.01.007
  • Haire, F. S. (2022). Born again: Narrating the mental health journeys of religious exiters. Sociology Compass, 16(9), e13030. https://doi.org/10.1111/soc4.13030
  • Hendricks, J. J., Hardy, S. A., Taylor, E. M., & Dollahite, D. C. (2024). Does leaving faith mean leaving family? Longitudinal associations between religious identification and parent–child relationships across adolescence and emerging adulthood. Journal for the Scientific Study of Religion, 63(1), 23–41. https://doi.org/10.1111/jssr.12876
  • Hui, C. H., Cheung, S.-H., Lam, J., Lau, E. Y. Y., Cheung, S.-F., & Yuliawati, L. (2018). Psychological changes during faith exit: A three-year prospective study. Psychology of Religion and Spirituality, 10(2), 103–118. https://doi.org/10.1037/rel0000157
  • Jokela, M., & Laakasuo, M. (2024). Health trajectories of individuals who quit active religious attendance: Analysis of four prospective cohort studies in the United States. Social Psychiatry and Psychiatric Epidemiology, 59(5), 871–878. https://doi.org/10.1007/s00127-023-02497-x
  • Keller, Y., David, Y., & Trachtenberg, E. (2026). On a new path: Social support, social media engagement, and well-being after religious disaffiliation. New Media & Society, 28(2), 739–764. https://doi.org/10.1177/14614448241302312
  • Lefevor, G. T., & Skidmore, S. J. (2025). Navigating faith transitions: A 4-year longitudinal examination of religious deidentification among LGBTQ+ Latter-day Saints. Journal of Counseling Psychology, 72(1), 15–29. https://doi.org/10.1037/cou0000765
  • Lofland, J., & Stark, R. (1965). Becoming a world-saver: A theory of conversion to a deviant perspective. American Sociological Review, 30(6), 862–875. https://doi.org/10.2307/2090965
  • Luther, R. (2023). What happens to those who exit Jehovah’s Witnesses: An investigation of the impact of shunning. Pastoral Psychology, 72, 105–120. https://doi.org/10.1007/s11089-022-01051-x
  • Paloutzian, R. F., Richardson, J. T., & Rambo, L. R. (1999). Religious conversion and personality change. Journal of Personality, 67(6), 1047–1079. https://doi.org/10.1111/1467-6494.00082
  • Pew Research Center. (2025a, March 26). Around the world, many people are leaving their childhood religions: Religious switching in 36 countries. https://www.pewresearch.org/religion/2025/03/26/around-the-world-many-people-are-leaving-their-childhood-religions/
  • Pew Research Center. (2025b, February 26). Religious switching. In Religious Landscape Study 2023–24. https://www.pewresearch.org/religion/2025/02/26/religious-switching/
  • Rambo, L. R. (1993). Understanding religious conversion. Yale University Press. https://yalebooks.yale.edu/book/9780300065152/understanding-religious-conversion/
  • Reini, K., Kolk, M., & Saarela, J. (2026). Religious switching and mental disorders in young adulthood: Evidence from Finnish population register data. American Journal of Epidemiology, 195(5), 1276–1283. https://doi.org/10.1093/aje/kwaf245
  • Starbuck, E. D. (1899). The psychology of religion: An empirical study of the growth of religious consciousness. Walter Scott. https://wellcomecollection.org/works/g2h63nys
  • Streib, H. (2021). Leaving religion: Deconversion. Current Opinion in Psychology, 40, 139–144. https://doi.org/10.1016/j.copsyc.2020.09.007
  • Stronge, S., Bulbulia, J., Davis, D. E., & Sibley, C. G. (2021). Religion and the development of character: Personality changes before and after religious conversion and deconversion. Social Psychological and Personality Science, 12(5), 801–811. https://doi.org/10.1177/1948550620942381
  • Thoma, M. V., Rohner, S. L., Heim, E., Hermann, R. M., Roos, M., Evangelista, K. W. M., Nater, U. M., & Höltge, J. (2022). Identifying well-being profiles and resilience characteristics in ex-members of fundamentalist Christian faith communities. Stress and Health, 38(5), 1058–1069. https://doi.org/10.1002/smi.3157