RESEARCH DOSSIER
Online Self-Presentation and Digital Identity Performance
Abstract. Online self-presentation and digital identity performance concern how people construct, revise, disclose, and sustain socially meaningful representations of themselves in environments characterized by persistent records, imagined audiences, variable privacy, and technologically mediated distribution. This fundamental research monograph integrates interactional sociology, impression-management psychology, hyperpersonal communication theory, original experiments, profile studies, qualitative work, longitudinal evidence, and developmental reviews. It examines identity claims, authenticity, anonymity, audience segmentation, disclosure, feedback, visual curation, social comparison, life transitions, algorithmic exposure, and artificial mediation. The evidence is critically evaluated for construct validity, research design, confounding, reproducibility, and generalizability. The synthesis separates communicators' intentions, produced evidence, technological distribution, audience interpretation, and subsequent self-evaluation. This is a critical narrative synthesis of published research, without original human-participant data or a new meta-analysis.
Keywords: online self-presentation; digital identity performance; impression management; computer-mediated communication; hyperpersonal interaction; context collapse; imagined audiences; authenticity; anonymity; privacy; self-disclosure; digital identity; profile curation; social media; identity development; adolescence; social comparison; algorithmic visibility; digital traces; human–artificial communication
Research Object and Scientific Scope
Online self-presentation is the situated production, selection, concealment, and circulation of information through which people make aspects of themselves available for interpretation in digital environments. Digital identity performance refers to the recurrent enactment of those identities through profiles, posts, images, conversations, visible affiliations, and platform-specific practices. The unit of analysis comprises a person, an actual or anticipated audience, a technological setting, a communicative act, and the feedback that can follow. A profile is simultaneously a record of past behavior, an interface for future encounters, and a source from which others infer personal characteristics. The distinction between an intended impression and an observed effect, developed within impression-management research, is fundamental to this dossier Leary & Kowalski (1990) Goffman (1959).
The governing questions are how communicators imagine their audiences, select identity-relevant evidence, decide what to disclose, interpret feedback, and manage coherence across different digital situations. These activities incorporate self-concept, motives for affiliation, reputation, authenticity, privacy, interpersonal trust, and the capacity of audiences to redistribute information. Their study requires treating self-report, behavioral traces, observer judgments, and platform affordances as separate evidentiary streams. A photograph chosen to communicate competence, for example, establishes a presentation choice; it does not prove that viewers perceived competence or that the author possessed the represented competence. The difference is both conceptual and methodological Hogan (2010) Zhao et al. (2008).
The scope is deliberately medium-conditioned identity performance. General face-to-face impression management, stable personality structure, clinical identity disturbance, dating-platform design, and adolescent development have neighboring scientific treatments; each enters here when digital mediation materially changes the phenomenon. This is a critical narrative synthesis of original studies, theoretical works, and selected reviews. It reports no newly collected participant data, estimates no pooled effect size, and claims no systematic-review protocol. Historical explanations are evaluated as conceptual foundations, experiments as tests of local mechanisms, and contemporary longitudinal research as evidence about changes over time, with inferential limits retained Avci et al. (2025) Chen & Mares (2025).
Historical Foundations of Public Identity
Erving Goffman's The Presentation of Self in Everyday Life described social encounters as settings in which actors manage expressive resources, social expectations, and the definition of a situation. His dramaturgical framework supplied enduring concepts for analyzing performance, audience, front region, backstage behavior, and the maintenance of a workable public identity. The analysis was primarily interpretive and interactional rather than a laboratory demonstration of a universal causal mechanism. Its relevance to networked environments arises from the continuing need to organize conduct for audiences whose expectations may conflict, although the mechanisms of visibility and persistence differ from those of co-present encounters Goffman (1959).
Social-psychological approaches refined the explanatory questions. Leary and Kowalski distinguished impression motivation, which concerns why an actor seeks to regulate impressions, from impression construction, which concerns the content and techniques used to produce an impression. This separates perceived stakes from actual messages: high concern about evaluation can lead to cautious silence, and elaborate posting can occur without an explicit goal of deception. Such distinctions remain necessary when studying edited biographies, profile photographs, highly public posts, and messages to small groups. Social identity is thus produced through choices under constraints rather than simply copied from an inner essence Leary & Kowalski (1990).
Research on networked identity emerged alongside computer-mediated communication. Studies of anonymous or relatively disembodied interaction investigated whether reduced nonverbal cues altered interpersonal engagement and expressive possibilities. Subsequent work examined social-network profiles attached to offline acquaintances, dating sites where strategic description mattered, and image-centered platforms that privilege discoverability and audience metrics. The historical sequence matters because findings from anonymous text forums cannot be transferred unchanged to persistent, searchable accounts connected to family, employers, or multiple communities. A coherent history follows the changing relation among personal expression, audience knowledge, and communicative infrastructure Bargh et al. (2002) Ellison et al. (2006) Marwick & boyd (2011).
From Interactional Performance to Digital Exhibition
Digital identity is partly performed in real-time exchanges and partly assembled as an exhibition available for later inspection. Hogan's distinction between performance and exhibition is analytically productive: an exchange depends on the particular audience and moment, whereas a curated archive can be viewed by people absent when its component objects were created. Users arrange photographs, achievements, status updates, and associations into records whose ordering and prominence may be controlled by platform design. The displayed identity consequently depends on editorial selection, technical ranking, and later acts of interpretation, even when the originating person has stopped communicating Hogan (2010).
An exhibition can combine evidence created under different intentions. An image posted for friends can become professional evidence when encountered by a recruiter; a personal disclosure intended for a support group can reappear in an unrelated community. The interpretive problem involves temporal displacement as well as audience displacement. Researchers studying digital performance must therefore record the intended audience at the time of production, the actual viewers who later gained access, and the transformations introduced by forwarding, screenshots, search, or ranking. The same material may support different inferences when detached from its initial conversation Marwick & boyd (2011) Davis (2012).
Identity performance is relational because the archive contains contributions by others: photographs in which a person is tagged, comments, endorsements, connections, and publicly visible interactions. Other people supply cues that may support or contradict an actor's preferred presentation. This makes complete authorial control implausible and introduces questions about consent, shared ownership, and reputation. The distinction between claims created by the subject and evidence contributed by third parties is especially important when evaluating profile authenticity. A careful study measures both types of signals and tests their influence on observers rather than treating an account as a direct psychological inventory Back et al. (2010) Zhao et al. (2008).
Impression Motivation, Self-Concept, and Anticipated Evaluation
People present identities for many purposes: affiliation, occupational opportunity, romantic contact, safety, recognition, play, political voice, support, and continuity with valued relationships. Impression motivation intensifies when an audience controls a valued outcome, when an identity is personally important, and when the person expects an impression can be changed. Yet motivation does not uniquely predict presentation content. The same desire for acceptance can produce disclosure among supportive peers and concealment before an audience perceived as threatening. Leary and Kowalski's model provides a basis for specifying audience value, perceived discrepancy, and impression efficacy as distinct explanatory variables Leary & Kowalski (1990).
Online identities also reflect self-knowledge and identity commitments. A person may select material that is factually accurate but incomplete, emphasize attributes that fit a current project, or experiment with a role not yet integrated into offline routines. None of these decisions necessarily indicates a fabricated personality. Research on Facebook profiles showed that observable profile information could correlate with judgments of actual personality rather than simply idealized self-descriptions, challenging a universal assumption of digital self-idealization. Because accuracy varies by feature, network, period, and observer, this finding supports a conditional measurement claim rather than a global guarantee of honesty Back et al. (2010).
Presentation feedback can influence the presenting person as well as the audience. Seeing one's own selected profile, receiving recognition, or maintaining a public commitment can make some self-descriptions more salient. Gonzales and Hancock experimentally examined exposure to one's Facebook profile and self-esteem, illustrating a route from digital display to self-evaluation. Such studies do not establish that routine platform use improves self-esteem for everyone. The relevant causal questions are whether composing, viewing, posting, and receiving feedback have distinguishable effects, how long effects persist, and which users or circumstances moderate them Gonzales & Hancock (2011) Bailey et al. (2020).
Computer-Mediated Communication and Hyperpersonal Processes
Walther's hyperpersonal model explains how computer-mediated interaction can support selective self-presentation, favorable receiver attributions, channel-enabled editing, and reinforcing feedback. Text-based or asynchronous exchanges may permit deliberate construction of messages with fewer immediately observable cues, while receivers can fill informational gaps with favorable expectations. The model is a conditional account of interactions under particular affordances and temporal constraints. Modern video calls, publicly networked profiles, and algorithmically distributed short-form content differ from the early contexts in which much foundational computer-mediated communication research occurred Walther (1996).
The selective sender component directs attention to the information communicators choose to release. An individual may rehearse a message, crop an image, delay a response, or avoid an identity-relevant question. The receiver component concerns how audiences interpret the available fragments; selective cues can produce idealization, skepticism, or indifference depending on trust and prior knowledge. Channel attributes influence what is editable and when, while feedback can strengthen or correct the developing impression. The explanatory value of the model lies in these separable pathways and their interactions, each of which can be tested with controlled message manipulations and recorded judgments Walther (1996) Ellison et al. (2006).
Online interaction often connects to embodied, offline relationships, which constrain hyperpersonal processes. Friends can compare a profile with known behavior, and dating partners may anticipate later verification of autobiographical claims. Identity presentation in these settings mixes selective positivity with incentives for credibility. Ellison, Heino, and Gibbs found that online daters negotiated accuracy and desirable self-presentation under expectations of future encounters. The generalizable mechanism is management of credibility under anticipated accountability, while its specific expression depends on site affordances, relationship purpose, and the possibility that inconsistencies will become evident Ellison et al. (2006) Toma et al. (2008).
Imagined Audiences and Context Collapse
Every presentation is oriented toward some conception of who can see it. Marwick and boyd's analysis of Twitter users highlighted the imagined audience: a practical mental representation of recipients assembled despite imperfect knowledge of actual reach. Context collapse occurs when communication environments bring audiences from previously separable social situations into proximity. A post visible to relatives, classmates, colleagues, and strangers may have to satisfy incompatible interpretive expectations. Consequently, a seemingly bland or generalized identity performance can be the product of sophisticated social risk management Marwick & boyd (2011).
Audience size, composition, and perceived heterogeneity have different psychological meanings. A person may tolerate a very large audience sharing similar norms yet hesitate before a small mixed audience that includes an employer and an intimate partner. Counts of followers cannot capture who matters, whose response is feared, or who has the power to forward content. Researchers should therefore elicit expected audiences, perceived disapproval, relationship importance, and actual exposure wherever feasible. Observed audience diversity can be compared with perceived context collapse, but the variables need not coincide Davis (2012) Chen & Mares (2025).
Audience strategies include self-censorship, selective posting, humor, coded expression, audience segmentation, use of multiple accounts, temporary content, and deliberate ambiguity. These strategies can support autonomy when users face legitimate threats to privacy. They can also produce identity strain when monitoring each audience requires continuing effort. Young people's interviews have documented perceived tensions between opportunities for expression and the risks of making a record accessible to multiple social circles. The appropriate empirical question is which strategy protects an intended boundary in a given environment, rather than whether a single level of openness should be considered universally authentic Davis (2012) boyd (2014).
Platform Affordances and the Conditions of Performance
Visibility, persistence, searchability, editability, synchronicity, and association among network ties shape the practical choices available for online self-presentation. These terms describe affordances in a particular user-platform relation rather than fixed psychological properties of every service. A self-description on a publicly indexed profile creates different possibilities from an encrypted message exchanged with a trusted friend. The actor's understanding of these possibilities also matters: people may overestimate privacy controls, underestimate persistence, or infer that ranking algorithms selectively show posts to favorable audiences Hogan (2010) Marwick & boyd (2011).
Editable profiles support curation, while time-sensitive exchanges make rapid interaction and social responsiveness more salient. The option to delete may reduce the perceived risk of a statement, although deletion cannot ensure that screenshots, caches, or memories disappear. Public reaction counters create additional evaluative cues, whereas private feedback can preserve a more intimate context. Studies should manipulate concrete interface properties and observe how message choices change, rather than infer affordance effects merely from the platform's brand name. Platform comparisons confound differences in design, user composition, cultural norms, and the activities that bring people to each service Walther (1996) Bazarova & Choi (2014).
User practices can counter or reinterpret intended affordances. People create private accounts on public platforms, operate several identities under a single profile, or use conventions that signal membership to a particular community. These adaptations show that identity performance emerges through reciprocal adjustment between people and technological conditions. A platform can make an action technically possible without making it socially safe or desirable. Analyses should combine feature audits with interviews and behavioral observation to learn which affordances users actually recognize, trust, and employ in their communication boyd (2014) Haimson (2018).
Profile Curation, Selective Evidence, and Personality Inference
Profiles combine self-authored declarations with photographs, activity histories, associations, and social endorsement. This mixture makes them attractive for person perception research, but the inference from a visible cue to a stable characteristic requires independent validation. Back and colleagues compared Facebook profiles with self-reported personality and observer judgments, providing evidence against the simple claim that all network profiles express an idealized rather than an actual personality. Their findings are bounded by the sample, platform, trait measures, and historical period. They do not transform an entire profile into a psychometric instrument Back et al. (2010).
Zhao, Grasmuck, and Martin examined Facebook identity construction in a setting tied to existing relationships. Their analysis emphasized implicit identity claims through photographs and social connections as well as explicit declarations. Such practices illustrate how digital presentation can communicate belonging and desirable affiliations without direct assertion. Interpretation depends on local conventions: the same photograph might express friendship, status, irony, aspiration, or a memory. Researchers need audience-based validation and contextual evidence before treating a particular visual sign as a reliable indicator of motive or identity Zhao et al. (2008).
An informative design differentiates three stages: selection of evidence by the profile owner, transmission or ranking of that evidence by a system, and interpretation by an observer. When those stages are collapsed, investigators may mistake successful self-selection for observer acceptance or mistake algorithmic exclusion for a decision to conceal. Longitudinal work can assess whether the same person changes presentation following employment transitions, relationship changes, or new audience composition. Supplementing profiles with consented interviews and independent behavioral measures permits more discriminating tests of representational accuracy Hogan (2010) Back et al. (2010).
Authenticity, Accuracy, and the Ethics of Selective Expression
Authenticity is frequently invoked as a standard for evaluating online identity, but it requires an operational definition. Researchers may mean congruence between online presentation and self-reported dispositions, consistency with personally endorsed values, willingness to express emotions, or others' perception that a person appears genuine. Those constructs can diverge. A thoughtfully edited description may accurately communicate valued commitments, while a spontaneous disclosure can exaggerate a momentary emotion. Measurement must specify whose judgment establishes authenticity, which traits or experiences are relevant, and the temporal interval over which consistency is assessed Bailey et al. (2020) Back et al. (2010).
Bailey and colleagues assessed authentic expression using the relationship between self-reported personality and personality estimates derived from Facebook Likes and status updates in a dataset of 10,560 users. More authentic self-expression was associated with higher reported life satisfaction. Their paper also described a preregistered longitudinal experiment involving authentic posting and short-term affect. The observational association does not establish that matching any externally inferred profile to a personality questionnaire causes lasting well-being. The experimental component supports a narrower within-person effect under its tested instructions and measurements. The two forms of evidence are complementary and should be kept distinct Bailey et al. (2020).
Authenticity also has a relational dimension. A person may communicate accurately with an intended audience while withholding information that another audience has no right to receive. Identity presentation becomes ethically problematic when it exploits others through material deception or unauthorized disclosure. A scientifically useful taxonomy separates accuracy, completeness, audience suitability, agency, and the possible harms of exposure. Such distinctions avoid treating privacy protection as dishonesty and avoid treating visible candor as sufficient proof of responsible behavior. Studies of mediated disclosure show that people manage these dimensions under specific social consequences, particularly when information is sensitive Andalibi & Forte (2018) Bazarova & Choi (2014).
Anonymity, Pseudonymity, and Multiplicity of Selves
Anonymity and pseudonymity alter how identities can be connected to consequences. Under anonymity, an audience may encounter an expressive act without a reliable link to an identifiable person; under pseudonymity, an enduring identifier can accumulate reputation without a public legal name. Neither condition guarantees freedom from social evaluation. Pseudonymous communities can maintain norms and sanctions, and personally identified environments can permit selective disclosure. Early studies of online expression suggested that altered visibility of offline identity cues can facilitate the communication of valued aspects of the self, a possibility that requires attention to participant selection and historical platform contexts Bargh et al. (2002) McKenna et al. (2002).
Multiple accounts can provide a practical method of audience segmentation. A public professional profile, a private friendship account, and a pseudonymous creative identity may express different roles of a single person. The appearance of inconsistency can result from audience-specific relevance rather than incompatible identities. Yet maintaining several representations creates demands for memory, boundary management, and anticipation of linkability. Research should ask what information travels between accounts, what users believe about discoverability, and which relationships are disrupted if previously separated audiences become connected Marwick & boyd (2011) Davis (2012).
Identity multiplicity is especially consequential for communities where disclosure can produce stigma or danger. Accounts with restricted audiences can provide opportunities to explore and articulate experiences before wider disclosure is possible. The methodological challenge is to distinguish voluntary experimentation from constrained concealment, and individual preference from unequal exposure to sanctions. Researchers should avoid treating a public account as a complete record of a person's social world. Studies of transitions and stigmatized disclosures indicate that platform choice, audience knowledge, and support conditions are integral to the meaning of identity performance Haimson (2018) Andalibi & Forte (2018).
Privacy Strategies, Boundaries, and Information Control
Online privacy management is a form of anticipatory social reasoning. Individuals evaluate what particular audiences might infer from a post, how likely material is to circulate, and what consequences unwanted access could create. This assessment is rarely a complete technical risk calculation. It draws on folk models of visibility, trust in platform defaults, and experiences of earlier disclosure. The scientific focus is the decision process connecting perceived audience, perceived risk, anticipated benefit, and disclosure behavior. A low frequency of posting can represent deliberate protection, disinterest, lack of access, or concern about judgment; these explanations require distinct measures Marwick & boyd (2011) boyd (2014).
Boundary strategies include restrictive settings, alternative identities, coded references, delayed sharing, conversational migration to private channels, and asking other people not to repost material. Their effectiveness depends on other participants' behavior and the platform's design. A private group is socially constituted by its members as well as technically constituted by its access controls. Screenshots and forwarded messages can convert an apparently bounded disclosure into enduring evidence in another setting. Researchers should observe both the initial disclosure choice and the subsequent circulation of information, while obtaining meaningful consent for sensitive digital-trace analysis Andalibi & Forte (2018) Hogan (2010).
Privacy and authenticity should therefore be analyzed jointly rather than placed on a single scale. A person can experience authentic expression within an intentionally restricted audience and perform a conventional identity in public without psychological incoherence. Conversely, pressure to maintain a publicly admired identity may lead to emotionally costly curation. The empirical question is when selective access supports autonomy and when it intensifies perceived surveillance or self-monitoring. Cross-sectional correlations between privacy concern and posting are insufficient to establish which pathway is operating; designs that vary audience control or follow actual account changes are more informative Davis (2012) Chen & Mares (2025).
Self-Disclosure, Intimacy, and Social Support
Self-disclosure involves communicating personally relevant information, including feelings, experiences, values, and vulnerabilities. On digital platforms, the social consequences depend on whether disclosure is addressed to a specific partner, a bounded community, or an indefinite audience. Bazarova and Choi extended a functional approach to disclosure motivation and characteristics on social-network sites, showing why disclosure should be investigated through its communicative purpose and context. A highly personal public statement can serve affiliation, explanation, advocacy, or support seeking; a less personal message can still be consequential for a relationship Bazarova & Choi (2014).
Online interaction may enable people to describe experiences that feel difficult to disclose in face-to-face settings. Early work on Internet relationships discussed the expression of personally important aspects of the self and the formation of meaningful connections through mediated communication. Such results do not imply that disembodied disclosure is always easier or that online relationships uniformly substitute for offline ties. Participants differ in social resources, expectations, community norms, and exposure to harmful responses. Studies should assess the perceived responsiveness of the receiving audience rather than using the volume of disclosure as a proxy for connection McKenna et al. (2002) Forest & Wood (2012).
Disclosure can also change an audience's response to the speaker. Forest and Wood examined how low self-esteem, negative disclosure, and the reception of Facebook posts relate to the benefits of online expression. Their research emphasizes the gap between perceived opportunity for disclosure and actually receiving supportive reactions. A communicator can disclose sincerely and nevertheless elicit responses that undermine hoped-for relational outcomes. The relevant outcome measures include perceived support, actual responses, subsequent interaction, and well-being, preferably separated in time. This prevents defining a disclosure as successful simply because the person pressed a publish button Forest & Wood (2012) Bazarova & Choi (2014).
Stigmatized Experience, Disclosure Risk, and Social Transition
Sensitive disclosures expose the dual character of networked identity performance: they can recruit recognition and care while also creating durable records that travel beyond an intended audience. Andalibi and Forte studied decisions about announcing pregnancy loss on an identified social network, revealing the importance of audience knowledge, anticipated reaction, and disclosure timing. Such qualitative evidence documents mechanisms and decision considerations within a specific experience and participant group. It cannot establish a universal ranking of disclosure strategies or show that every platform offers equivalent safety. The analytical emphasis should remain on who controls exposure and what social consequences are expected Andalibi & Forte (2018).
Social transitions can demand corresponding changes in online identity records. Haimson examined social media as social transition machinery through transition blogs and interviews, emphasizing how digital spaces can support self-understanding and social recognition while simultaneously exposing users to unwanted audiences. Transitions unfold over time, and a profile's historical material may conflict with a person's current self-description or reveal information they wish to keep bounded. Researchers must take temporal agency seriously: the capacity to revise a name, photograph, pronouns, network, or archive can be central to whether a digital environment supports a changing identity Haimson (2018).
An ethical account treats identity disclosure as an action with uneven consequences across social positions. The same act that provides belonging in one network may risk discrimination, violence, or family conflict in another. Safety is affected by local laws, cultural norms, economic dependence, and the platform's control mechanisms. Research should therefore record environmental exposure and participant-defined goals alongside conventional measures of authenticity or posting frequency. Longitudinal studies that combine confidential interviews with consensual behavioral records can illuminate how people revise their strategies after receiving support, encountering hostility, or moving between life stages Chen & Mares (2025) Haimson (2018).
Feedback, Social Validation, and Public Metrics
Digital self-presentations frequently produce visible responses: comments, reactions, follower counts, shares, and recommendation signals. These metrics can become social cues, but their meaning is contingent. A high reaction count may reflect approval, controversy, audience size, or automated distribution; a low count may indicate limited exposure rather than rejection. The sender's interpretation of feedback is an intervening psychological process that should be measured independently. Research on adolescents' friendship-network site use linked feedback and social self-esteem, inviting more precise examination of whether feedback quality, audience significance, or apparent popularity accounts for observed associations Valkenburg et al. (2006).
Feedback is also a possible mediator between a presentation choice and later self-evaluation. Posting a positive identity narrative can attract affirming responses, which may reinforce subsequent expression. The same narrative may receive hostile comments and trigger withdrawal. One six-wave study of LGBTQ+ young people found that positive identity-related presentation predicted later self-acceptance and self-esteem within the study's time frame, with supportive feedback relevant to some associations. These longitudinal relations support directional hypotheses while remaining vulnerable to unmeasured time-varying causes and selective platform choice Chen & Mares (2025).
Metric-driven presentation may encourage people to favor content expected to perform well in a particular recommendation environment. Yet the motivation to gain attention is empirically distinct from a wish to deceive. A user might simply publish more often about subjects with a responsive audience, or adjust visual style without changing the underlying claim. Testing platform-metric effects requires randomizing the visibility or availability of quantitative feedback, controlling actual audience size, and comparing message content before and after exposure. Observational popularity alone cannot distinguish adaptation to metrics from preexisting differences in audience and posting goals Hogan (2010) Valkenburg et al. (2006).
Visual Performance, Images, and Identity Signaling
Images can carry identity claims that are not expressed in accompanying text. Clothing, settings, companions, body posture, facial expression, and photographic conventions provide cues about activity, affiliation, aspiration, and social role. Their interpretation depends on cultural context and the receiver's expectations. Studies of Facebook profiles and MySpace practices showed that images and social connections occupied a prominent place in online identity construction. A visual display may communicate a plausible biography, a temporary role, or playful exaggeration; rigorous interpretation requires information about production context as well as recipient judgments Zhao et al. (2008) Manago et al. (2008).
Visual editing makes impression construction tangible. Selecting a flattering angle, cropping distracting details, or adding a filter can change what observers see without changing the factual identity of the person shown. More extreme alterations can create discrepancies relevant to body image, attractiveness judgments, and interpersonal trust. The research task is to differentiate editing behavior, the user's goal, the observer's inferred attributes, and measurable consequences for later interaction. Experimental profiles with controlled image manipulation can test perceived attractiveness or trustworthiness, while observational feeds establish which practices are common in situated communities Toma et al. (2008) Vogel et al. (2014).
Image-centered environments also intensify opportunities for social comparison. Observing curated images of others may alter a viewer's self-evaluation, while anticipating that comparison may shape what the viewer later publishes about themselves. These processes belong to separate positions in the communicative sequence: receiver comparison and sender performance. Vogel and colleagues investigated social-media comparison and self-esteem, illustrating why users' reactions to others' displays should be distinguished from their own motives to create favorable displays. The relation is reciprocal in ordinary life, but reciprocal processes require longitudinal or experimental designs for causal specification Vogel et al. (2014) Chou & Edge (2012).
Identity Claims, Credibility, and Deception
Presentation is not synonymous with deception. A short biography is necessarily selective; credibility concerns whether the information supplied is sufficiently accurate for its communicative purpose. Toma, Hancock, and Ellison compared online dating profiles with independently measured characteristics and found that some self-descriptions involved discrepancies. Such evidence demonstrates that accuracy can be assessed against external criteria and that motives for desirability coexist with pressures for future verification. It does not justify treating every participant as uniformly deceptive or assuming that misrepresentation follows a single motive Toma et al. (2008).
Credibility judgments may depend on signals not controlled by the profile owner. Mutual connections, archived interactions, external credentials, and third-party comments can confirm or challenge self-authored assertions. An audience's familiarity with offline behavior can further constrain plausible performance. Profile research indicates that identity is established by a combination of explicit statements and implicit evidence. Testing credibility therefore requires identifying the precise claim, a reliable criterion measure, and the observer's actual evaluation; generalized labels such as 'fake persona' obscure variation across claims and settings Zhao et al. (2008) Back et al. (2010).
Deception can also be misunderstood when a person's identity changes. A historical profile may accurately describe an earlier role while no longer representing current commitments; a pseudonym can preserve privacy without misleading a community about relevant conduct. The evaluative standard should reflect purpose, temporal reference, and foreseeable reliance by others. Methodologically, studies can compare time-stamped identity claims with independently verified behaviors and interviews concerning their intended audiences, while avoiding invasive verification of information that participants did not consent to disclose. Interpretation requires care with gender, cultural, and safety-related reasons for selective representation Ellison et al. (2006) Haimson (2018).
Adolescence, Exploration, and Self-Concept Development
Adolescence provides an especially informative setting because identity exploration, peer affiliation, sensitivity to evaluation, and widening social roles coincide with growing participation in networked environments. Digital profiles permit experimentation with interests, affiliations, and images while creating records that can be revisited by peers and family. The developmental question concerns whether particular activities support exploration, commitment, self-concept clarity, or distress over time. The amount of time spent using a platform is an incomplete proxy for these mechanisms. Posting, comparison, direct communication, and the management of audiences can have different implications for the same person Avci et al. (2025) Nesi et al. (2018a).
Avci, Baams, and Kretschmer reviewed 32 studies comprising 19,658 adolescents and young people. Active participation was associated with more identity exploration, while authentic presentation showed positive associations with self-concept clarity. Social comparison related to exploration and identity distress. The synthesis underscores the importance of what users do online, but the underlying studies vary in measures, platforms, ages, and inferential strength. These patterns should not be represented as causal developmental laws. Particularly needed are preregistered studies that distinguish exploratory posting from passive viewing and follow within-person changes across meaningful developmental transitions Avci et al. (2025).
The transformation framework of Nesi and colleagues describes how social-media environments alter the frequency, immediacy, intensity, and public character of peer interactions. These changes can make feedback more persistent, expand potential audiences, and produce new opportunities for comparison or conflict. The theoretical framework is useful for specifying processes, but a theoretical transformation does not automatically identify a beneficial or harmful individual outcome. Developmental studies should measure offline peer relationships, family context, age, pubertal and social transitions where relevant, and platform-specific exposure before attributing identity change to social media itself Nesi et al. (2018a) Nesi et al. (2018b).
Gender, Sexual Identity, and Differential Conditions of Expression
Identity performance operates within social expectations about gender, sexuality, and legitimate self-description. Manago and colleagues' qualitative research on MySpace described gender-related patterns and negotiations of self-presentation, offering a historically situated account of the conventions participants perceived. Such findings identify social meanings and locally reported strategies; they should not be converted into timeless categorical claims about women, men, or other groups. Platform conventions, age, cultural setting, and the opportunities available to respondents shape both reported behavior and its interpretation Manago et al. (2008).
Research on LGBTQ+ young people's online identities shows why audience support and disclosure safety deserve explicit study. Chen and Mares followed 337 U.S. participants aged 15 to 22 over six weekly surveys. Identity exploration was associated with frequency, positivity, and authenticity of LGBTQ+-related presentation, particularly when audiences were perceived as relatively homogeneous. Positive presentation predicted subsequent self-acceptance and self-esteem, whereas frequency and authenticity did not predict those outcomes in the same models. This distinction prevents treating all identity expression as one uniform psychological exposure or all positive outcomes as proof of a single mechanism Chen & Mares (2025).
Supportive feedback may modify the significance of a particular disclosure. Yet cultural and legal conditions can produce major differences in the risks of visibility, and an audience with mixed identities is not necessarily an unsafe audience. The meaningful variables are perceived attitudes, probable exposure, available privacy controls, access to support, and material consequences of unwanted identification. Ethical research must respect the reasons individuals maintain restricted or pseudonymous accounts. Future comparative studies should investigate these mechanisms without demanding public identification from people for whom visibility itself can create harm Chen & Mares (2025) Haimson (2018).
Friendships, Intimacy, and Relationship Formation
Online self-presentation is often designed for relationship formation and maintenance. People communicate hobbies, values, social availability, humor, and vulnerability in ways that invite particular responses. Early research on Internet relationships described how individuals may express personally important qualities through mediated interaction and form connections around shared identities or interests. A strong account distinguishes initiation, disclosure, perceived responsiveness, relationship maintenance, and later satisfaction. These are temporally ordered but not interchangeable stages, and each has its own alternative explanations McKenna et al. (2002) Walther (1996).
Social-network environments also make relational affiliations part of the identity record. Visible friends, mutual contacts, responses, and tagged images can create social proof or complicate self-authored narratives. A person's intended message may be interpreted differently depending on who endorses it or how others engage publicly. This form of co-produced identity is analytically distinct from dyadic intimacy: it involves a network of observers and potential contributors. Observational research can identify associations between network features and perceptions, while experimental variation of displayed endorsements can test whether those cues change judgments Zhao et al. (2008) Hogan (2010).
Relationships can become strained when presentation serves incompatible audiences. A private joke shared publicly, an old photograph displayed after a breakup, or a selective account of a conflict may be understood as disloyalty by one recipient and ordinary expression by another. Understanding such conflict requires the expectations of each relationship and the visibility architecture of the encounter. Studies that measure only the original poster's intention miss the interpersonal consequences. Dyadic designs, with consent from both partners and longitudinal follow-up, can connect presentation decisions to perceived respect, trust, and repair Marwick & boyd (2011) Bazarova & Choi (2014).
Dating Profiles and Anticipated Offline Verification
Dating environments represent a specific institutional setting for self-presentation. Participants seek to be considered attractive and compatible, frequently through concise biographies and selected images. Ellison, Heino, and Gibbs interviewed online daters and showed how people negotiated desirability and authenticity under expectations that they might later meet in person. The prospect of offline verification distinguishes dating presentation from contexts in which an audience may never encounter the person again. This prospective accountability influences what claims seem plausible, beneficial, or risky Ellison et al. (2006).
Direct comparison between dating profiles and independently measured facts provides a more precise approach to deception than asking users whether they believe others misrepresent themselves. Toma and colleagues investigated discrepancies in profile characteristics, demonstrating a design that links self-presentation to external criteria. The findings arise from a specific platform era and participant population; contemporary image editing, algorithmic matching, and safety practices require renewed study. Moreover, discrepancy does not always reveal intention: an outdated attribute, approximate description, and deliberate falsehood have different psychological mechanisms Toma et al. (2008).
Dating-platform research should include the changing social position of users. Those seeking safety may withhold details that could identify a workplace or home; others may disclose identity-related information early to establish compatibility. The ethical evaluation of these choices depends on what a prospective partner reasonably needs to know and what identifying information is unnecessary for an initial interaction. A strong methodology combines profile content, stated motivations, post-match communication, and follow-up outcomes under explicit consent. This keeps the study within digital identity performance while preserving the independent scientific scope of dating-specific research Ellison et al. (2006) Andalibi & Forte (2018).
Professional Reputation and Institutional Audiences
Employment, education, and professional networks create audiences with unequal power over opportunities. A public profile may be interpreted as evidence of reliability, competence, group membership, or judgment, even when its content was produced for friends. The resulting performance problem concerns boundary coordination across institutions. People may foreground credentials, restrict personal photographs, or separate professional and informal accounts because evaluation by employers or teachers has material consequences. These choices are forms of impression management under institutional stakes, not necessarily a stable trait of self-monitoring Leary & Kowalski (1990) Marwick & boyd (2011).
Digital traces are noisy evidence for professional judgment. Posts can be outdated, ironic, context-dependent, or selected by recommendation algorithms; people do not receive equal opportunities to control how others portray them. An employer's impression is an outcome distinct from the truth of a candidate's ability. Research should experimentally vary actual job relevance and the context accompanying online material while measuring evaluator judgments and discrimination risk. Field observations can supplement these tests but cannot by themselves establish that particular profile features validly predict performance. Privacy expectations and institutional policy also shape acceptable research and organizational use Hogan (2010) Back et al. (2010).
Professional identity may become a long-term project of self-narration. Researchers, creators, freelancers, and public-facing workers often build cumulative archives of achievements and expertise. The continuity can help audiences recognize skill, yet it also creates pressures to maintain coherent reputational claims through role transitions and changing circumstances. Longitudinal research could compare how people correct earlier accounts, acknowledge mistakes, and adapt presentation after receiving external evaluations. A scientifically useful model distinguishes truthful evidence of experience, strategic prominence of accomplishments, and fraudulent claims rather than aggregating them into a single professional-branding score Leary & Kowalski (1990) Hogan (2010).
Algorithmic Visibility and the Limits of Audience Control
On many platforms, communicators choose what to publish but do not completely choose who encounters it. Recommendation, ranking, search, moderation, and distribution systems determine the visibility of material. This introduces an additional process between self-presentation and impression formation: exposure allocation. A user can tailor a message for an imagined audience and then discover that it traveled to an unexpected group. Understanding performance therefore requires identifying the algorithmic and social paths by which content becomes visible, alongside the person's beliefs about those paths Marwick & boyd (2011) Hogan (2010).
Users may develop informal theories about what receives attention, when a post will be suppressed, or how platform metrics should be interpreted. Those beliefs can guide posting choices even when they are inaccurate descriptions of the underlying system. The psychological variable is the person's inferred distribution rule, not simply the platform's undisclosed code. A strong design elicits predictions about audience exposure, records actual reach where available, and then examines how changes in predicted visibility alter content selection. Experimental studies of displayed ranking cues or feedback availability can test local mechanisms without assuming access to complete proprietary algorithms Hogan (2010) Bazarova & Choi (2014).
Algorithmic selection complicates causal inference in observational datasets. Higher engagement with an identity category might result from users posting more about it, a recommendation system distributing it more broadly, audience preference, or all three. Comparing apparently successful posts with unsuccessful posts cannot isolate the author's impression motive. Researchers should seek random exposure variation, within-person comparisons, or controlled simulated interfaces, and should report when such designs sacrifice ecological realism. The key explanatory distinction is among production, distribution, reception, and feedback, each of which can alter subsequent identity performance Valkenburg et al. (2022) Avci et al. (2025).
Identity Across Platforms, Time, and Life Transitions
An individual's digital identity is rarely confined to one account. Different services organize different audiences, conversational norms, and opportunities for self-expression. A profile directed to professional colleagues may foreground expertise, while a community account highlights shared interests or personal experience. Cross-platform variation can reflect a coherent set of context-sensitive roles. To study that variation, researchers need consented linkage, a clear rationale for comparing accounts, and measures of role centrality and intended audience. Scraping unrelated accounts to assemble a supposedly complete personality risks imposing an external identity model on participants Davis (2012) Marwick & boyd (2011).
Temporal change adds another layer. A person may revise a public name, disclose a previously private experience, adopt a new occupation, or reconsider a belief. The record of earlier performances persists even while the person's current identity evolves. Haimson's research on social transition illustrates the importance of digital systems that allow histories, audiences, and presentations to change together. The scientific question is how available revision mechanisms influence psychological continuity, social recognition, and practical safety. A historical record is useful as evidence of an earlier act, but it should not be mistaken for an immutable description of the present person Haimson (2018).
Long-term identity studies can distinguish stability of values from stability of representations. They might ask whether the same underlying commitment appears in different styles across platforms, whether changing audiences prompt more selective disclosure, and how feedback affects subsequent self-descriptions. Designs using repeated consented snapshots, diaries, and qualitative interviews are better suited to these questions than a single captured profile. Researchers should also record platform redesigns and major changes in audience composition, since observed differences over time may arise from altered infrastructure rather than intrapersonal development Avci et al. (2025) Hogan (2010).
Operational Definitions and Construct Validity
Research on online self-presentation is vulnerable to construct inflation: studies may assign the same name to visible posting frequency, motives to impress, positive wording, claimed authenticity, or differences between online and offline behavior. A credible operationalization specifies the target construct and identifies its observable indicators. Presentation behavior can be coded from content and platform actions; motivations require participant reports or appropriately validated tasks; perceived impressions require receiver assessments; identity consequences require measures of self-concept or subsequent conduct. A single self-report survey cannot establish all four simultaneously Leary & Kowalski (1990) Bazarova & Choi (2014).
Authenticity illustrates the problem. A score derived from correspondence between linguistic features and self-reported personality assesses one kind of representational consistency, not the whole moral or phenomenological meaning of authenticity. A rating by friends assesses perceived genuineness and may reflect familiarity or approval. A diary asking whether a post expressed personally endorsed feelings provides yet another indicator. These measures can converge, but convergence must be demonstrated. Bailey and colleagues employed computational correspondence measures and reported associations with well-being, while Chen and Mares used repeated self-reports of identity-related expression; the findings therefore concern related but nonidentical constructs Bailey et al. (2020) Chen & Mares (2025).
A complete measurement model separates exposure opportunity, production decision, message features, audience response, and personal outcome. Indicators may include probability of posting under varying audiences, number of edited drafts, disclosure specificity, content alignment with self-report, third-party ratings, and change in self-esteem. Researchers should report reliability, validity evidence, missingness, platform restrictions, and the time frame captured. Because social environments change, a previously validated instrument may require renewed measurement-invariance analysis before use with different languages, generations, or technological settings. Construct validity is a continuing empirical task rather than a permanent property of a scale Avci et al. (2025) Valkenburg et al. (2022).
Experimental Paradigms and Causal Inference
Experiments can isolate mechanisms by randomizing the audience, the visibility of a post, editability, feedback, or the content of presented information. In an audience-salience experiment, participants might write a profile for close friends or for an unknown mixed audience. Investigators could then code disclosure specificity, evaluative tone, and revision behavior, while checking whether participants actually believed the audience manipulation. Receiver-side judgments should be collected separately, ideally from blinded observers. Without a manipulation check and distinct sender and receiver outcomes, a change in wording cannot be attributed confidently to an impression-management motive Leary & Kowalski (1990) Walther (1996).
Existing experimental studies illustrate specific components. Gonzales and Hancock manipulated Facebook self-exposure and assessed short-term self-esteem; Vogel and colleagues used observational and experimental approaches to social comparison. Bailey and colleagues reported a preregistered longitudinal experiment targeting authentic posting and affect. Each design identifies effects of a particular intervention under its sampling and measurement conditions. Results should be interpreted with attention to duration, preexisting attitudes, differential compliance, and whether the experimental activity resembles routine communication. Short-term affective change cannot be silently converted into long-term developmental benefit Gonzales & Hancock (2011) Vogel et al. (2014) Bailey et al. (2020).
Factorial experiments could separate three often-confounded influences: intended audience composition, displayed audience feedback, and apparent algorithmic reach. Participants might create and edit messages under independently randomized conditions, and a separate audience might judge the resulting messages. Outcomes would include self-disclosure, perceived authenticity, credibility, and willingness to publish. Ethical safeguards are essential if experiments involve sensitive identity cues; simulated disclosure should not cause involuntary outing or persistent public exposure. Preregistration, adequate power, transparent exclusion rules, and independent replication would increase the value of causal evidence Chen & Mares (2025) Andalibi & Forte (2018).
Longitudinal, Diary, and Experience-Sampling Designs
Identity performance is iterative, making repeated measurement indispensable. A person receives feedback, updates expectations about future audiences, revises presentation choices, and may later reinterpret earlier public claims. Cross-sectional correlations cannot establish this sequence. Longitudinal panels can test whether earlier identity-related posting predicts subsequent self-concept measures after accounting for prior levels. Diaries and experience sampling can locate decisions close to their occurrence, recording context, audience expectations, emotional state, and perceived consequences. These designs reduce some retrospective distortions while introducing participant burden and missing-data challenges Avci et al. (2025) Chen & Mares (2025).
Chen and Mares collected six weekly surveys from 337 U.S. LGBTQ+ participants aged 15 to 22, permitting distinctions between identity exploration, presentation characteristics, perceived audience heterogeneity, and later self-evaluation. Their results offer a concrete example of temporal ordering and heterogeneity of associations. Because respondents selected social contexts and measured exposures were not randomly assigned, the findings do not prove that positive presentation invariably causes improved self-esteem. Time-varying events, supportive relationships, and platform changes remain possible competing explanations. The study nevertheless advances beyond single-occasion correlations by testing directional predictions across repeated observations Chen & Mares (2025).
Longitudinal designs should specify a theoretically relevant interval. Changes in immediate affect may occur within minutes, identity experimentation over weeks, and major role transitions over months or years. A six-week design answers different questions from a multiyear cohort. Repeated digital-trace sampling can establish what was posted, but it may miss deleted drafts, private messages, and off-platform interactions. Combining behavioral logs with consented interviews and structured self-reports allows researchers to distinguish absence of publication from absence of experience, a distinction especially important when participants choose not to reveal sensitive identities Haimson (2018) Andalibi & Forte (2018).
Interviews, Ethnography, and Situated Interpretation
Qualitative research supplies evidence about motives and meanings that cannot reliably be read from posting frequency. Ellison and colleagues interviewed online daters about accuracy and desirable self-presentation; Davis interviewed young people about identity opportunities and constraints; Haimson studied social transitions through blogs and interviews. These studies document the strategies participants articulate and the situations in which those strategies become meaningful. Their explanatory strength depends on transparent sampling, reflexive interpretation, attention to discrepant cases, and clarity about what was observed rather than inferred Ellison et al. (2006) Davis (2012) Haimson (2018).
Ethnography can reveal platform-specific conventions, such as when an apparently exaggerated identity statement is understood by a community as humor, solidarity, or role play. The same phrase extracted from its interactional setting might be coded as deception or narcissism by an outside observer. Participant observation, contextual interviews, and comparison of messages with interaction histories help recover situated meaning. Researchers should nevertheless avoid treating participant retrospective explanation as a complete record of psychological causation. Verbal accounts and recorded behavior are distinct types of evidence that can be compared rather than forced into agreement Manago et al. (2008) Zhao et al. (2008).
Ethical procedures are central because a publicly retrievable post is not necessarily offered for unrestricted research redistribution. Linking quotations to accounts can make participants searchable, and presenting identifiable sensitive disclosures can expose them to unintended audiences. Responsible design considers consent, paraphrasing where appropriate, data minimization, secure handling, and the right of participants to clarify or withdraw sensitive material within the approved protocol. These protections are part of scientific quality: participants who fear unwanted exposure may systematically conceal the very practices a study seeks to understand Andalibi & Forte (2018) Chen & Mares (2025).
Computational Content Analysis and Digital Trace Validity
Large-scale platform data can characterize posting rhythms, linguistic choices, image conventions, and interaction patterns with greater temporal resolution than many surveys. Computational models may classify themes, estimate emotional tone, or infer personality-related attributes from language and Likes. These outputs are measurements produced by models, with training data, error profiles, and platform-specific assumptions. Bailey and colleagues used computational personality judgments as part of an authenticity operationalization, demonstrating the analytical possibilities and the need to interpret inferred traits carefully Bailey et al. (2020).
Trace data are incomplete representations of social life. Private exchanges, drafts, deletes, blocked accounts, unobserved algorithmic impressions, and strategic abstention are usually absent. Selection into a platform or into data-sharing consent can correlate with identity-related motives. Automated language measures can misread sarcasm, dialect, multilingual expression, and community-specific references. A large sample does not correct a biased construct definition. Researchers should document data provenance, sampling frame, classification validation, demographic performance differences where ethically measurable, and the boundaries of generalization across periods and platforms Avci et al. (2025) Hogan (2010).
Mixed-method validation provides an important safeguard. A model that labels a post as self-promotional should be evaluated against human judgments and participant-reported intentions on a relevant subsample. The study can then estimate when observers agree, when they disagree, and whether the disagreement has substantive significance. Analyses of exposure must distinguish impressions delivered by a platform from reactions actively made by viewers. When proprietary feeds prevent direct measurement of reach, investigators should report this as an uncertainty rather than treating likes or comments as equivalent exposure counts Bazarova & Choi (2014) Marwick & boyd (2011).
Cultural Variation, Inequality, and Generalizability
Digital self-presentation occurs within cultural expectations about modesty, direct self-promotion, privacy, family roles, and acceptable identity expression. A behavior coded as confident individuality in one setting may be interpreted as inappropriate self-assertion or ordinary group participation in another. Platform access, language, literacy, disability, age, and economic circumstances further influence who can create certain forms of identity evidence. Research should specify country, recruitment method, period, and social setting before treating a reported association as general to a population. The historical concentration of platform research in convenience samples limits the transportability of findings Manago et al. (2008) Avci et al. (2025).
Unequal exposure to surveillance and harassment can make selective disclosure a rational strategy. A participant who withholds an identity from a mixed audience may be responding to threats that another participant does not face. Data about authenticity cannot be interpreted independently of those conditions. Chen and Mares' U.S.-based study illustrates the importance of audience attitudes and support for identity-related self-presentation, while the authors' recruitment and context circumscribe generalization. Cross-national studies must adapt measures to local categories and protections rather than assuming identical social meanings for every identity label Chen & Mares (2025).
Measurement invariance across groups is essential when comparing scores that purport to capture authenticity, self-esteem, or reputational concern. Differential access to privacy tools and unequal familiarity with platform conventions may also change the behavior associated with the same psychological motive. Researchers should assess whether instruments function similarly, record structural constraints, and include marginalized participants in the interpretation of qualitative data. A theoretically defensible account examines both personal strategies and the institutions that distribute risks and resources across audiences Haimson (2018) boyd (2014).
Reproducibility, Heterogeneity, and the Evidence Hierarchy
The literature is methodologically heterogeneous. Classic interactional theory explains why public identity is organized for audiences; small qualitative studies document situated practices; laboratory experiments can isolate selected mechanisms; observational profile datasets reveal naturally occurring associations; systematic reviews evaluate patterns across studies. Their conclusions should be compared at the appropriate evidentiary level. It would be an error to treat a philosophical metaphor as a causal estimate or to promote a cross-sectional association into proof of an intervention effect. The evidence hierarchy here is question-specific rather than a single ranking independent of subject matter Goffman (1959) Avci et al. (2025).
Replication is especially difficult in fast-changing digital settings. Features are redesigned, cohorts adopt new conventions, and an intervention that once appeared credible may later become familiar or implausible. Conceptual replication should retain the relevant mechanism while adapting its operationalization, and direct replication should reproduce the essential protocol in a comparable setting. Reviews should examine sampling overlap, publication bias, measurement inconsistency, effect-size uncertainty, and the distinction between preregistered and exploratory tests. When these checks are unavailable, the manuscript describes the strength of evidence proportionately instead of manufacturing a pooled conclusion Valkenburg et al. (2022) Avci et al. (2025).
Competing explanations must be actively considered. A relationship between positive posting and later well-being might reflect prior mood, supportive friends, selection of safe platforms, or successful presentation itself. A relationship between public visibility and cautious expression might reflect reputational motive, privacy threat, differing norms, or poor understanding of interface controls. Discriminating studies should measure these candidate causes, manipulate mechanisms where ethically possible, and report null or contrary results. Longitudinal patterns and experimental evidence complement one another, but neither automatically resolves all remaining confounding or generalizability concerns Chen & Mares (2025) Bailey et al. (2020).
Artificial Mediation and the Changing Authorship of Digital Identities
Artificial systems increasingly participate in the composition, editing, translation, and distribution of online self-presentations. A user may request suggestions for a biography, revise a message with a conversational system, or publish an artificially generated portrait. Such acts introduce a co-production process into impression construction. The scientifically accessible questions concern which decisions remain with the human user, how machine suggestions alter self-disclosure or perceived authenticity, and whether audiences know the conditions under which messages were produced. Established models of selective presentation and computer-mediated communication supply mechanisms to investigate without presuming identical psychological organization in human and artificial actors Leary & Kowalski (1990) Walther (1996).
Artificially mediated content can improve expression for some users by providing language support, feedback, or alternative forms of communication. It can also introduce unsupported claims, homogenize personal narratives, or obscure who is responsible for a statement. The proper unit of analysis is an identifiable workflow: goal formation, system prompting, generated alternatives, human selection, publication, audience reception, and later correction. Experiments should compare unassisted authorship with disclosed and undisclosed assistance, measure actual factual accuracy, and distinguish reader perceptions of authenticity from users' own experience of agency. The relevant causal claims remain hypotheses until specific designs test them Bailey et al. (2020) Hogan (2010).
Within the Artificial Era, digital identity performance raises a further comparative research question: what would count as a persistent and accountable presentation by a nonbiological artificial agent? A defensible operational model would require explicit evidence of stable identifiers, continuity of goals or commitments, audience-sensitive communication, memory of prior public claims, and reliable procedures for correcting errors. These criteria provide a program of inquiry, rather than evidence that human motives can automatically be attributed to every artificial system. The distinction between human identity performance assisted by software and the identity organization of an artificial agent is necessary for rigorous comparison across different kinds of intelligent participants Walther (1996) Hogan (2010).
Scientific Synthesis and Research Priorities
The accumulated literature supports a multidimensional account. Online self-presentation involves a situated identity goal, an imagined audience, options for selecting and modifying evidence, a medium that conditions visibility, actual reception, and feedback that may modify subsequent conduct. Each element can vary independently. A person can post a true statement for a restricted audience, post a selective statement for a broad audience, or refrain from posting because the costs of circulation exceed expected gains. The explanatory strength of a research design depends on its ability to observe the relevant decision and separate it from later judgments Leary & Kowalski (1990) Marwick & boyd (2011).
Some propositions have comparatively direct empirical support. Profile information can contain valid personality-relevant cues under studied conditions; experimentally manipulated self-profile exposure or authentic-posting tasks can change specified short-term evaluations or affect; identity-related self-presentation is associated with developmental processes and perceived audience support. Other propositions remain more conditional, including the long-term causal effects of algorithmic visibility, the generalizability of authenticity effects across cultures and media formats, and the ways coauthored artificial content changes responsibility and perceived genuineness. The conclusions must preserve the designs and populations that generated them Back et al. (2010) Gonzales & Hancock (2011) Bailey et al. (2020) Avci et al. (2025) Chen & Mares (2025).
A cumulative research program should prioritize independent replications of audience manipulations, transparent longitudinal panels across transitions, measurement invariance across communities, experiments that separate posting from viewing and feedback, and ethically governed analyses of reach and persistence. Researchers should report effect sizes and uncertainty for the actual outcomes they estimate, archive protocols when consent permits, and distinguish intended identity from audience perception and lasting self-concept change. Comparisons across human and artificial communication should state which properties are observed and which remain theoretical, while maintaining clear provenance for generated material Valkenburg et al. (2022) Avci et al. (2025) Chen & Mares (2025).
Conclusion
Online self-presentation and digital identity performance are organized processes through which people communicate socially meaningful versions of themselves under conditions of variable visibility, persistence, editability, and audience uncertainty. The foundational insight of impression-management theory is that a presentation has motives, forms, recipients, and consequences. Digital environments extend this organization across time and networks, allowing identity evidence to be curated, redistributed, measured, and revisited by people beyond the original exchange. Goffman's interactional perspective, the two-component model of impression management, and theories of computer-mediated communication remain useful when their distinct assumptions are specified Goffman (1959) Leary & Kowalski (1990) Walther (1996).
Empirical research establishes that these processes are shaped by the properties of the medium and by the social circumstances of the person. Profile studies, interviews, experiments, computational analyses, systematic reviews, and repeated surveys illuminate different levels of the phenomenon. The strongest conclusions are conditional: authentic expression can be linked to well-being in defined settings; audience composition and perceived support can influence identity-related presentation; and adolescents' active participation can accompany identity exploration. These findings warrant careful attention to measurement, heterogeneity, alternative explanations, and the limits of causal inference Bailey et al. (2020) Avci et al. (2025) Chen & Mares (2025).
The central synthesis is that digital identity is enacted through relationships among personal commitments, communicative choices, audiences, and technological distribution. Its psychological significance emerges in the consequences of those relationships for recognition, agency, self-understanding, privacy, and social participation. A rigorous science of digital identity performance requires the study of intentions and observable behaviors together, the analysis of both benefits and costs, and clear distinctions between human self-presentation, artificially mediated production, and the possible identities of artificial participants. These distinctions provide a stable foundation for research across changing platforms and the broader conditions of the Artificial Era Hogan (2010) Haimson (2018).
References
- Andalibi, N., & Forte, A. (2018). Announcing pregnancy loss on Facebook: A decision-making framework for stigmatized disclosures on identified social network sites. Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems, Paper 158, 1–14. https://doi.org/10.1145/3173574.3173732
- Avci, H., Baams, L., & Kretschmer, T. (2025). A systematic review of social media use and adolescent identity development. Adolescent Research Review, 10, 219–236. https://doi.org/10.1007/s40894-024-00251-1
- Back, M. D., Stopfer, J. M., Vazire, S., Gaddis, S., Schmukle, S. C., Egloff, B., & Gosling, S. D. (2010). Facebook profiles reflect actual personality, not self-idealization. Psychological Science, 21(3), 372–374. https://doi.org/10.1177/0956797609360756
- Bailey, E. R., Matz, S. C., Youyou, W., & Iyengar, S. S. (2020). Authentic self-expression on social media is associated with greater subjective well-being. Nature Communications, 11, Article 4889. https://doi.org/10.1038/s41467-020-18539-w
- Bargh, J. A., McKenna, K. Y. A., & Fitzsimons, G. M. (2002). Can you see the real me? Activation and expression of the “true self” on the Internet. Journal of Social Issues, 58(1), 33–48. https://doi.org/10.1111/1540-4560.00247
- Bazarova, N. N., & Choi, Y. H. (2014). Self-disclosure in social media: Extending the functional approach to disclosure motivations and characteristics on social network sites. Journal of Communication, 64(4), 635–657. https://doi.org/10.1111/jcom.12106
- boyd, d. (2014). It's complicated: The social lives of networked teens. Yale University Press. https://yalebooks.yale.edu/book/9780300199000/its-complicated/
- Chen, Y. A., & Mares, M.-L. (2025). Social media self-presentation of LGBTQ+ youth in the United States: The role of identity exploration, context collapse, and supportive feedback. Journal of Computer-Mediated Communication, 30(6), Article zmaf020. https://doi.org/10.1093/jcmc/zmaf020
- Chou, H.-T. G., & Edge, N. (2012). “They are happier and having better lives than I am”: The impact of using Facebook on perceptions of others' lives. Cyberpsychology, Behavior, and Social Networking, 15(2), 117–121. https://doi.org/10.1089/cyber.2011.0324
- Davis, K. (2012). Tensions of identity in a networked era: Young people's perspectives on the risks and rewards of online self-expression. New Media & Society, 14(4), 634–651. https://doi.org/10.1177/1461444811422430
- Ellison, N., Heino, R., & Gibbs, J. (2006). Managing impressions online: Self-presentation processes in the online dating environment. Journal of Computer-Mediated Communication, 11(2), 415–441. https://doi.org/10.1111/j.1083-6101.2006.00020.x
- Forest, A. L., & Wood, J. V. (2012). When social networking is not working: Individuals with low self-esteem recognize but do not reap the benefits of self-disclosure on Facebook. Psychological Science, 23(3), 295–302. https://doi.org/10.1177/0956797611429709
- Goffman, E. (1959). The presentation of self in everyday life. Anchor Books. https://search.worldcat.org/title/The-presentation-of-self-in-everyday-life/oclc/733073728
- Gonzales, A. L., & Hancock, J. T. (2011). Mirror, mirror on my Facebook wall: Effects of exposure to Facebook on self-esteem. Cyberpsychology, Behavior, and Social Networking, 14(1–2), 79–83. https://doi.org/10.1089/cyber.2009.0411
- Haimson, O. L. (2018). Social media as social transition machinery. Proceedings of the ACM on Human-Computer Interaction, 2(CSCW), Article 63. https://doi.org/10.1145/3274332
- Hogan, B. (2010). The presentation of self in the age of social media: Distinguishing performances and exhibitions online. Bulletin of Science, Technology & Society, 30(6), 377–386. https://doi.org/10.1177/0270467610385893
- Leary, M. R., & Kowalski, R. M. (1990). Impression management: A literature review and two-component model. Psychological Bulletin, 107(1), 34–47. https://doi.org/10.1037/0033-2909.107.1.34
- Manago, A. M., Graham, M. B., Greenfield, P. M., & Salimkhan, G. (2008). Self-presentation and gender on MySpace. Journal of Applied Developmental Psychology, 29(6), 446–458. https://doi.org/10.1016/j.appdev.2008.07.001
- Marwick, A. E., & boyd, d. (2011). I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. New Media & Society, 13(1), 114–133. https://doi.org/10.1177/1461444810365313
- McKenna, K. Y. A., Green, A. S., & Gleason, M. E. J. (2002). Relationship formation on the Internet: What's the big attraction? Journal of Social Issues, 58(1), 9–31. https://doi.org/10.1111/1540-4560.00246
- Nesi, J., Choukas-Bradley, S., & Prinstein, M. J. (2018a). Transformation of adolescent peer relations in the social media context: Part 1—A theoretical framework and application to dyadic peer relationships. Clinical Child and Family Psychology Review, 21(3), 267–294. https://doi.org/10.1007/s10567-018-0261-x
- Nesi, J., Choukas-Bradley, S., & Prinstein, M. J. (2018b). Transformation of adolescent peer relations in the social media context: Part 2—Application to peer group processes and future directions for research. Clinical Child and Family Psychology Review, 21(3), 295–319. https://doi.org/10.1007/s10567-018-0262-9
- Toma, C. L., Hancock, J. T., & Ellison, N. B. (2008). Separating fact from fiction: An examination of deceptive self-presentation in online dating profiles. Personality and Social Psychology Bulletin, 34(8), 1023–1036. https://doi.org/10.1177/0146167208318067
- Valkenburg, P. M., Peter, J., & Schouten, A. P. (2006). Friend networking sites and their relationship to adolescents' well-being and social self-esteem. CyberPsychology & Behavior, 9(5), 584–590. https://doi.org/10.1089/cpb.2006.9.584
- Valkenburg, P. M., Meier, A., & Beyens, I. (2022). Social media use and its impact on adolescent mental health: An umbrella review of the evidence. Current Opinion in Psychology, 44, 58–68. https://doi.org/10.1016/j.copsyc.2021.08.017
- Vogel, E. A., Rose, J. P., Roberts, L. R., & Eckles, K. (2014). Social comparison, social media, and self-esteem. Psychology of Popular Media Culture, 3(4), 206–222. https://doi.org/10.1037/ppm0000047
- Walther, J. B. (1996). Computer-mediated communication: Impersonal, interpersonal, and hyperpersonal interaction. Communication Research, 23(1), 3–43. https://doi.org/10.1177/009365096023001001
- Zhao, S., Grasmuck, S., & Martin, J. (2008). Identity construction on Facebook: Digital empowerment in anchored relationships. Computers in Human Behavior, 24(5), 1816–1836. https://doi.org/10.1016/j.chb.2008.02.012
Social Comparison and the Recursive Production of Identity
Self-presentation and social comparison form a feedback loop. People observe curated representations of peers, evaluate their own standing, and may subsequently change what they choose to display. Chou and Edge documented links between Facebook use and beliefs about others' happiness; Vogel and colleagues investigated social comparison, platform use, and self-esteem through observational and experimental work. The appropriate account specifies whether comparison is upward or downward, which dimensions are salient, and whether an observed change affects private self-evaluation, public content, or both. Content exposure is not equivalent to producing content Chou & Edge (2012) Vogel et al. (2014).
A carefully arranged feed may distort the perceived base rate of successful experiences without containing any false statement. Users selectively report celebrations, accomplishments, and aesthetically favorable moments, leaving everyday difficulties less visible. Recipients can then infer that apparently common experiences are personally unavailable to them. These inferences depend on selection effects and audience interpretation. The research design should manipulate the content distribution to viewers and separately record whether subsequent presentation becomes more idealized, more withdrawn, or more expressive. Cross-sectional exposure counts cannot establish the direction of this reciprocal process Vogel et al. (2014) Hogan (2010).
Social comparison may motivate identity exploration as well as distress. The 2025 adolescent identity review linked comparison with both exploration and identity distress, suggesting that the same class of activity cannot be assigned one developmental meaning. Differences in motives, social support, comparison targets, self-concept clarity, and cultural expectations may account for heterogeneous outcomes. Analyses should therefore examine distributions and moderators rather than relying on average associations alone. Replication across platforms and cohorts is essential because mechanisms may persist while the social meanings of particular visual and textual conventions change Avci et al. (2025) Valkenburg et al. (2022).