Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Parasocial Interaction and Mediated Social Presence

Abstract. Parasocial interaction refers to the felt experience of social exchange with a mediated figure, parasocial relationship to an enduring representation of that figure as socially meaningful, and mediated social presence to the perception of another's social availability through a communication environment. This fundamental research monograph analyzes their historical origins, definitions, perceptual mechanisms, relationship dynamics, measurement strategies, experimental findings, longitudinal challenges, and outcomes from broadcasting to social media, livestreaming, and artificially mediated personae. The critical narrative synthesis draws on original conceptual and empirical work, controlled addressing experiments, psychometric evaluations, social-presence theory, research mapping, and meta-analyses published through 2026. It identifies experimentally supported short-term effects of direct address, substantial evidence for distinguishing momentary interaction from durable relationship, and heterogeneity in the relationships among parasocial engagement, well-being, persuasion, health judgments, and social resources. The cumulative literature does not support a universal loneliness-deficiency explanation. Major unresolved questions concern causal direction, measurement invariance, long-term relationship development, platform affordances, cultural transportability, and the boundary between perceived and actual reciprocity. The work is a critical scholarly narrative monograph; it is not an independently conducted systematic review, meta-analysis, clinical trial, or original experiment.

Keywords: parasocial interaction; parasocial relationship; mediated social presence; media psychology; intimacy at a distance; social cognition; Experience of Parasocial Interaction; direct address; mediated familiarity; livestreaming; social media influencers; perceived reciprocity; attachment; loneliness; media persuasion; mental health; human–artificial relations; Artificial Era

Research Object and Scientific Questions

Parasocial interaction is an audience member's subjective experience of a social encounter with a person presented through media. The person may look into a camera, address viewers, disclose personal details, or enact a recognizable social role. The viewer can experience being attended to even when the communication is broadcast and no individualized response occurs. A parasocial relationship is the more durable psychological representation of a mediated person as familiar, significant, emotionally meaningful, or part of one's recurrent social world. Mediated social presence is the perception that another socially responsive being is available or psychologically present through a communication environment. The three phenomena can coincide, but they are different explanatory targets: episodic encounter, continuing relationship, and perceived social availability. This distinction follows the historical account of Horton, D., & Wohl, R. R (1956), the measurement critique of Dibble, J. L., Hartmann, T., & Rosaen, S. F (2016), and the social-presence framework of Biocca, F., Harms, C., & Burgoon, J. K (2003).

The research object is a specific form of mediated social perception and relational experience involving human public figures, television performers, fictional characters, journalists, podcast hosts, influencers, livestreamers, and other recognizable media personae. Its units of analysis include the exposure episode, the viewer–persona relationship, the creator–audience communication arrangement, and the platform features governing visibility and feedback. Its outcomes include subjective mutual awareness, familiarity, trust, attachment-like feelings, social learning, persuasion, relational disappointment, and health-related judgments. Affective intensity, frequency of exposure, following behavior, and platform engagement are possible correlates rather than interchangeable definitions.

Five questions organize this monograph. How did the original television-era concept become a general theory of mediated social experience? Which observable cues generate momentary perceived interaction, and which support persistent bonds? How do measures distinguish parasocial interaction, relationship strength, identification, and social presence? What do experiments, longitudinal observations, reviews, and meta-analyses actually establish about consequences? Finally, how does technological change—from broadcasts to creators, livestreaming, and increasingly artificial social interfaces—alter the conditions under which human audiences experience social responsiveness? The method is a critical scholarly narrative synthesis of historically foundational studies, experimental evidence, psychometric investigations, systematic reviews, and meta-analyses. No new data were collected, and no systematic-search completeness is claimed.

Historical Origin: Intimacy at a Distance

Donald Horton and R. Richard Wohl introduced the language of para-social interaction in their 1956 essay on mass communication and apparent intimacy with performers. Their historical context matters. In a broadcast environment, performers could address an indefinite audience as though a personal relationship were unfolding; programs could cultivate a recurrent figure whose verbal style, bodily orientation, and social role supplied stable expectations. The analytic problem was how a one-to-many medium could produce a distinctly interpersonal quality of experience. The article was published in Psychiatry in 1956, volume 19, pages 215–229, and its original bibliographic identity is confirmed by the Horton, D., & Wohl, R. R (1956). The journal's present online posting date is not the date of the original scientific contribution.

Early parasocial theorizing addressed the dramatic organization of media encounters. A performer repeatedly appearing as a recognizable person can establish a sense of continuity across episodes, even though the audience has little personal access to the performer. Performed friendliness, direct address, the promise of familiarity, and an audience's learned knowledge of the persona together make the encounter intelligible as social. The model highlights a feature of media design: communication can display the grammatical form of interpersonal attention while allocating actual communication rights asymmetrically. Audiences can interpret a mediated address as meaningful without possessing a comparable capacity to influence the original performance.

This historical idea persists because asymmetry is a property of the distribution of communicative opportunities, rather than a judgment about the audience's rationality or health. A viewer may know perfectly well that a broadcast anchor is speaking to thousands and still respond to the encounter with social perception, felt warmth, or anticipation. Repetition can increase knowledge of an individual's style and biography; it does not by itself prove a durable emotional relationship. Later scholarship separated the immediate episode from the longer-term bond, an essential transition for cumulative research Giles, D. C (2002); Dibble, J. L., Hartmann, T., & Rosaen, S. F (2016).

From Broadcast Reception to Relational Research

Research after the foundational essay progressively operationalized parasocial experience through surveys, controlled stimuli, and longitudinal accounts of audience involvement. The important transition was from interpreting media personae as an illustrative cultural phenomenon to asking which measurable psychological variables predict the experience and which consequences follow. An influential early investigation by Rubin, A. M., Perse, E. M., & Powell, R. A (1985) collected questionnaires from 329 respondents about local television news, viewing motives, loneliness, and favorite news personalities. They reported that parasocial interaction and loneliness were not correlated in that sample, while news affinity, perceived realism, and instrumental motives were important predictors. The study was observational: its regression patterns cannot establish that increasing perceived realism causes parasocial closeness.

The early literature also encountered a problem of terminology. Some investigators used parasocial interaction to label the felt exchange during a program; others used it for a stable attachment or favorable disposition toward a media person. This accumulation of partly overlapping measures created apparent theoretical disagreement that could arise from different operational targets. Giles, D. C (2002) reviewed the literature and outlined ways to understand its development; a much later review by Schramm, H., Liebers, N., Biniak, L., & Dettmar, F (2024) examined 281 English- and German-language studies published between 2016 and 2020. The latter review demonstrates the breadth of contemporary applications but does not convert heterogeneous study designs into one unified causal estimate.

The field's expansion tracked the historical diversification of media. Radio and television foregrounded a recognizable speaker within a scheduled program; fictional television and film introduced character attachment and possible narrative loss; social media added a persistent persona feed, self-disclosure, and visible engagement; livestreaming introduced asynchronous and real-time signals that sometimes permit actual reply. These environments can all generate parasocial experience, while the architecture of possible response varies substantially. A coherent contemporary theory therefore requires attention to the technical distribution of visibility and agency as well as to the human viewer's psychological response.

Parasocial Interaction as an Episode of Perceived Reciprocity

The central phenomenological feature of parasocial interaction is experienced mutual orientation: the mediated figure seems to notice, address, include, or share a social moment with the audience member. Such an experience can occur during a brief exposure and does not require a preexisting relationship. A television presenter looking into the camera may elicit this response during a first encounter; an admired creator may fail to elicit it during a specific impersonal post. Momentary interaction and cumulative relationship strength must therefore be measured separately.

Hartmann, T., & Goldhoorn, C (2011) tested this distinction through a two-by-two experiment in which a television performer varied bodily and verbal addressing. Their study included 198 viewers and introduced an Experience of Parasocial Interaction scale. The published abstract reports greater perceived parasocial experience when viewers received direct bodily and verbal address. This design isolates features of the stimulus and supports a causal inference about short-term perceived interaction under the tested conditions. It does not establish enduring relationship change or generalized population effects from one exposure.

Direct address is neither the only cue nor a universal guarantee. A performer may convey attentiveness through pacing, turn-like speech, close framing, emotionally responsive facial expression, or a narration that invites the recipient to imagine being included. These features interact with viewers' prior expectations, familiarity, perspective-taking, and evaluation of the persona. Some encounters elicit pleasure or comfort, others embarrassment, resistance, or distrust. Describing all attention as parasocial bonding would flatten this functional diversity. The appropriate empirical outcome is the viewer's report of experienced social exchange during the actual exposure.

Parasocial Relationships as Enduring Social Representations

A parasocial relationship concerns an enduring social representation of a mediated figure, which may survive intervals without media exposure. The viewer can know the person's recurring habits, anticipate reactions, evaluate their reliability, discuss them with others, and experience their absence as meaningful. A strong parasocial relationship may contain affection, identification, admiration, imagined friendship, romantic interest, moral disappointment, or ambivalence. These are possible attributes of a relational representation; no one emotion exhausts the construct.

The distinction between a bond and a transient encounter has empirical support. In a meta-analysis of 120 studies, Tukachinsky, R., Walter, N., & Saucier, C. J (2020) found that parasocial relationships and interactions were moderately related, indicating meaningful overlap without identity. The same synthesis connected parasocial relationships with perceived similarity, identification, transportation, and persuasive outcomes. A meta-analytic association does not imply that every audience member follows the same developmental pathway, nor that repeated interaction invariably becomes a stable attachment.

Continuity arises through both exposure history and meaning attribution. Repeated encounters supply information, while the viewer integrates that information into social memory and future expectation. The mediated figure may become a reference point for routines, tastes, aspirations, or identity narratives. A relationship can continue psychologically even after the individual stops posting or a character's series ends. Researchers should therefore distinguish time spent with content, remembered familiarity, emotional importance, and perceived reciprocity in separate measures. Treating follower counts or watch time as a direct index of relationship strength would assume what ought to be measured.

Mediated Social Presence: A Distinct but Intersecting Construct

Social presence concerns the felt social reality or availability of others through communication media. It applies to many settings beyond parasocial media encounters: a video call, a multiplayer environment, a classroom stream, or an avatar encounter can produce different forms of awareness of another person. Biocca, F., Harms, C., & Burgoon, J. K (2003) reviewed problems of theoretical definition and measurement and advanced criteria for more robust social-presence research. Their approach shows why subjective presence must be analyzed as a multidimensional construct rather than equated with technical image quality.

In the broader tradition of mediated interpersonal communication, Walther, J. B (1996) developed an account of impersonal, interpersonal, and hyperpersonal processes in computer-mediated communication. That theory helps explain why a communication channel's limited cues do not preclude strong relational inference. Its original emphasis on communicating partners should be distinguished from the absence of individualized reply that can characterize parasocial encounters.

Presence may be facilitated by vivid cues, behavioral contingency, shared attention, spatial arrangement, and the ability to coordinate responses. Yet technical richness and social presence are not equivalent: a simple voice message from a known person can create an intense sense of social presence, while elaborate imagery may convey little interpersonal engagement. The experimental work of Nowak, K. L., & Biocca, F (2003) on agency and anthropomorphism in virtual environments illustrates that perceived presence depends on interpreted social properties and design arrangements, with complex rather than mechanically linear effects.

Parasocial interaction and social presence often converge when a mediated personality appears socially available. The analytical separation remains decisive. Social presence can occur in fully reciprocal contact without a parasocial relationship, and a parasocial bond can persist during the absence of a momentary feeling of copresence. Conversely, a viewer can feel a streamer is present in a shared event while having no attachment to that streamer. Future experiments should test these constructs with distinct latent variables and manipulation checks, preventing a high-presence interface from being mistaken automatically for a strong parasocial relationship.

Levels of Explanation and a Method for Evidence Synthesis

A complete account operates at several levels. At the stimulus level, camera direction, self-disclosure, verbal address, narrative continuity, and timing provide social cues. At the encounter level, viewers interpret these cues as a sign of attention, agency, realism, or shared experience. At the relationship level, repetition and appraisal contribute to durable familiarity, trust, expectation, and emotional investment. At the platform level, algorithms, notification systems, channel memberships, comment moderation, and creator access distribute opportunities for encounters and feedback. At the social level, peer discussion, community norms, age, culture, and interpersonal resources shape how the mediated figure becomes meaningful.

The evidentiary hierarchy depends on the claim. Historical original publications establish the emergence of the theory. Controlled stimulus experiments can identify the local effects of specific addressing or response features. Psychometric studies assess whether items reflect the proposed construct. Cross-sectional surveys identify covariation but leave substantial ambiguity about direction and selection. Longitudinal panels improve temporal ordering but remain vulnerable to time-varying confounding. Meta-analyses estimate patterns across operationally heterogeneous studies; their pooled results inherit the strengths and weaknesses of their input evidence. This monograph evaluates these traditions together while preserving distinctions among descriptive observation, causal identification, and theoretical explanation.

A useful working rule follows: claims about the existence of a reported social experience require validated descriptions; claims about its causes require controlled or otherwise credible causal variation; claims about long-term outcomes require repeated follow-up and attention to confounders. A psychologically plausible pathway is a hypothesis until supported by an appropriate design. The field advances when it tests which contextual variables change the encounter, rather than taking any technological platform or audience behavior as proof of a particular relational mechanism.

Social Address, Gaze, and the Organization of a Mediated Encounter

Media performances become socially legible through familiar interactional cues. A gaze directed toward the lens approximates an interlocutor's gaze, a second-person pronoun establishes an addressee, and a pause can resemble the conversational opportunity to respond. These signs operate through ordinary social interpretation: perceivers attribute attention, communicative intention, and emotional orientation using partial evidence. Their operation is context-dependent. A direct gaze in an advertisement may be interpreted as manipulative, whereas a similarly framed gaze by a trusted creator may feel welcoming. The causal question is what the same performer and message would evoke under different patterns of social address.

The controlled study by Hartmann, T., & Goldhoorn, C (2011) provides a particularly clear test because bodily and verbal addressing were manipulated separately. It shows that a conventional broadcast display can alter momentary parasocial experience even though the performer remains unavailable for a personal conversation. The implication is theoretical as well as methodological: perceived reciprocity can be evoked by a structured representation of communicative attention without actual reciprocal turn-taking. The relevant dependent measure is the viewer's state during exposure, not a subsequent claim that a relationship has necessarily formed.

A sophisticated causal design would further manipulate the congruence of gaze and wording, the temporal placement of pauses, and the audience's prior relationship with the presenter. Such a design could determine whether already-familiar presenters generate larger immediate effects or whether unfamiliar figures benefit more from highly explicit address. It should include checks for discomfort, perceived sincerity, and cultural expectations about gaze. Treating every increment of viewer-directed gaze as an increment of intimacy would impose a linearity that the empirical record does not warrant.

Familiarity, Social Memory, and Expectation

Repeated exposure gives a media figure informational continuity. Viewers learn stable traits, habits, speech patterns, values, relationships, and preferred narratives. This information can support a social representation that is automatically retrieved when the performer reappears. Familiarity facilitates prediction: audiences anticipate jokes, reactions, or points of view and can experience continuity across otherwise disconnected episodes. The mechanism is similar in one respect to how interpersonal knowledge supports expectations, although the informational channels and opportunities for correction are markedly different.

A familiar voice can become an environmental cue in daily routines, such as commuting or evening listening, and repeated content can acquire autobiographical associations. Those associations make the figure personally relevant even when the viewer has never had an individualized exchange. This is a theoretically plausible route to parasocial relationship strength; evidence about a particular route must still be derived from designs that separate exposure history from selective preference. People who already like a figure may choose to watch more, making frequency a consequence of affinity as well as a possible contributor.

Rubin, A. M., Perse, E. M., & Powell, R. A (1985) identified news affinity, perceived realism, and information-oriented motives as salient correlates of parasocial interaction with local news personalities. Their findings argue for differentiated models: the same exposure quantity can serve information acquisition, leisure, habit, social connection, or emotional regulation. A useful longitudinal approach would obtain within-person time series of preference, actual exposure, and subsequent relationship ratings, testing whether changes in exposure predict later changes in closeness after accounting for prior closeness. Measurement of creator availability and changes in production style would improve identification.

Self-Disclosure, Perceived Authenticity, and Trust

Creators may disclose personal experiences, everyday challenges, family routines, or emotional reactions to create a recognizable interpersonal narrative. Self-disclosure can convey vulnerability and distinctiveness, thereby providing material that viewers use to infer character and trustworthiness. Yet mediated disclosure is also an edited performance: a creator can curate what is shown, repeat recurring intimate motifs, or present staged spontaneity. The psychological variable of interest is perceived authenticity and the basis on which audiences infer it, rather than an investigator's unverified assumption that the performer is transparently revealing private life.

Perceived authenticity may affect parasocial bonds through credibility, similarity, and predictability. A creator's apparent willingness to admit error or describe ordinary limitations could elicit identification and confidence; a mismatch between commercial incentives and intimate presentation might instead produce skepticism. The field needs experiments that hold content information constant while changing disclosure framing, commercial sponsorship signals, and prior reputational cues. Such studies can assess whether any observed change is attributable to apparent intimacy, epistemic trust, social similarity, or general liking.

Influencer evidence exemplifies this overlap. Lou, C., & Yuan, S (2019) studied content value, credibility, trust, and branded-content responses in influencer marketing. These associations provide relevant evidence about the persuasive environment but do not isolate a universal mechanism by which disclosure causes relationship strength. A rigorous synthesis preserves the distinct constructs of credibility, message evaluation, perceived similarity, emotional closeness, and intention to act. Commercial disclosure experiments should include sponsor knowledge and advertising literacy as candidate moderators rather than treating influencer trust as inherently warranted.

Similarity, Identity, and Perspective-Taking

Perceived similarity supplies a plausible foundation for affinity with media figures. Shared values, language, interests, identity claims, experiences, or age-relevant concerns make the figure legible as someone with whom an audience member could potentially relate. Similarity may encourage attention and a favorable evaluation before parasocial processes become measurable. It may also result from selective interpretation: highly invested viewers can magnify commonalities and discount differences. This bidirectionality is central to theories of homophily and media preference.

The meta-analysis by Tukachinsky, R., Walter, N., & Saucier, C. J (2020) connected parasocial relationship strength to homophily and other involvement constructs. A causal interpretation would require randomized similarity information or credible natural variation, ideally complemented by preexposure measurement of viewer characteristics. Randomizing a fictional character's stated preferences may be easier than randomizing similarity to an established celebrity. However, experimental simplification may sacrifice ecological validity because real audiences often have extensive prior knowledge of familiar performers.

Perspective-taking is another candidate process. In the Hartmann, T., & Goldhoorn, C (2011), greater perspective-taking ability was associated with stronger reported parasocial experience in the tested viewing situation. This result supports a role for interpreters' social-cognitive capacities but should not be mistaken for a universal causal effect of a stable trait. An important research question is whether perspective-taking helps viewers perceive mutual awareness, infer the performer's circumstances, identify with narrative choices, or simply produce richer reports about an otherwise identical encounter.

Attachment, Belonging, and Relational Substitution Hypotheses

Parasocial bonds can contribute to people's felt social worlds by offering predictable familiarity, admiration, comfort, and an object of social thought. This creates a question about their relationship with interpersonal needs. One hypothesis proposes compensation, whereby people lacking satisfying offline relationships form stronger parasocial attachments. Another proposes extension, whereby parasocial connections coexist with and elaborate an already socially active life. These are empirical hypotheses with different predictions, not mutually exclusive categories of people.

The early Rubin, A. M., Perse, E. M., & Powell, R. A (1985) study found no correlation between loneliness and parasocial interaction in its local-news sample. More broadly, the Tukachinsky, R., Walter, N., & Saucier, C. J (2020) meta-analysis did not find a significant relationship between social deficiencies and parasocial relationship strength. These results undermine claims that parasocial attachment is generally reducible to loneliness or deficient interpersonal resources. They do not imply that no lonely individual ever uses parasocial content for compensation, nor that all contexts produce equivalent patterns.

A better model includes the function of a given bond. The same viewer may follow an admired performer for entertainment, use another creator's content for companionship during isolation, and rely on a third as an information source. The association between belonging needs and attachment may depend on personality, developmental transition, life stress, accessibility constraints, and the type of media figure. Longitudinal studies should analyze change in perceived isolation and parasocial involvement within individuals, distinguishing stable between-person differences from time-varying states. Interventions that alter social opportunity ethically could test substitution more directly, although randomizing loneliness would be unacceptable and conceptually imprecise.

Media Genres, Persona Roles, and Relational Expectations

Different media genres establish different rules for how a figure can address an audience. A local news presenter derives apparent familiarity partly from recurrent contact and a trustworthy institutional role. A fictional protagonist supplies narrative continuity but cannot be expected to answer an audience message. A podcast host may become integrated into daily life through an intimate acoustic format. A beauty creator can share routines and recommendations while simultaneously conducting a commercial enterprise. A livestreamer may visibly react to selected messages among thousands, creating intermittent actual reciprocity within a larger asymmetric arrangement.

These differences complicate pooled estimates across media genres. A high parasocial rating about a fictional character may arise from narrative identification and emotional investment, whereas a similarly high rating about a real political commentator may involve ideological agreement and source credibility. The same question—"I feel as though this person is my friend"—may have different pragmatic meanings across contexts. The Schramm, H., Liebers, N., Biniak, L., & Dettmar, F (2024) mapping of 281 studies underscores how quickly media characters, types, and research purposes diversified during 2016–2020. Comparability requires explicit measurement of persona ontology and genre.

Researchers should model genre as an affordance system rather than a decorative categorical label. A fictional character is constrained by a narrative production process; a livestream performer has variable but real opportunities for audience response; a public official's perceived warmth may have consequences for political credibility; an educator's mediated address may affect perceived instructional support. These affordances can change which mechanisms are possible and which outcomes are ethically salient. A common conceptual vocabulary allows comparison while preserving substantive differences.

Livestreaming and the Partial Reciprocity Problem

Livestreaming creates one of the strongest tests of traditional parasocial concepts because some viewers can actually receive responses. Chat messages, usernames, donations, audience votes, and repeated interaction can make a creator aware of particular participants. Yet an individual viewer's experience may still depend primarily on collective address or the possibility of being noticed rather than on a personal exchange. The social arrangement often combines truly reciprocal microepisodes with an enduring many-to-one audience relationship.

Wulf, T., Schneider, F. M., & Queck, J (2021) examined Twitch experiences in an online experiment with 251 participants. They manipulated forms of addressing and a streamer's attention to chat in a three-by-two design. This is valuable because it moves beyond counting platform features and tests distinct communicative arrangements. The study's stimulus-level manipulations can support causal claims about immediate viewing outcomes in that experiment; they do not establish the long-term psychological consequences of being an active paying subscriber or participating in a large streamer community.

Kowert, R., & Daniel, E., Jr (2021) described the conceptual challenge of a partially reciprocal livestream environment. A research design should independently code whether a viewer's message was seen, whether the creator replied, how many other people were addressed, and whether the viewer recognized a response as directed specifically to them. This would enable causal and measurement distinctions between direct interaction, anticipated interaction, generalized social presence, and parasocial relationship strength. The language of reciprocity should track actual response contingencies and perceived reciprocity as separate variables.

The Relational Economy of Influencer Platforms

Influencer platforms turn visibility, trust, apparent accessibility, and repeat audience attention into economic resources. Parasocial closeness can coexist with sponsorship, subscription, merchandise promotion, affiliate links, advertising, and fundraising. The commercial setting changes the opportunity structure of a relationship: creators may cultivate routine access, intimate self-presentation, and frequent audience acknowledgment while depending financially on engagement. A viewer's perceived intimacy therefore intersects with media business models and disclosure obligations.

The research-trends review by Schramm, H., Liebers, N., Biniak, L., & Dettmar, F (2024) shows that the growing research corpus includes multiple relationship contexts that make universal conclusions hazardous. Sokolova, K., & Kefi, H (2020) investigated blogger credibility, parasocial processes, and purchase intentions. These constructs can covary because audiences choose whom to follow, perceive similarity, and evaluate product suitability. Self-reported purchase intention is an outcome distinct from observed purchase behavior; observational associations cannot isolate a causal effect of parasocial closeness from all other source attributes.

The strongest research designs would randomize clear sponsorship disclosure and message framing while preserving the informational content of the recommendation. Researchers could then separately assess attention, perceived credibility, recognition of persuasive intent, relational feeling, and independently measured behavior. A central ethical issue is whether the presentation of intimate connection creates a mistaken expectation that creators are personally invested in each viewer's welfare. Disclosure and media literacy become scientifically relevant when they alter perceived obligations, trust calibration, and behavioral decisions.

Character Loss, Creator Scandal, and Parasocial Rupture

A persistent social representation can be disrupted when a program ends, a fictional character leaves, a creator withdraws, or credible new information contradicts prior expectations. The resulting experience may include sadness, disappointment, anger, grief-like reactions, and reinterpretation of earlier content. The scientific object is a change in a mediated bond's meaning over time. The observation that distress can occur does not turn every parasocial rupture into a psychiatric disorder, and intensity must be assessed in relation to duration, functioning, and context.

Cohen, J (2003) developed measures of individual differences in reactions to the dissolution of parasocial relationships, while Cohen, J (2004) investigated anticipated parasocial breakup involving favorite television characters and the role of attachment styles and relationship intensity. These studies establish a recognizable research tradition around media-character loss. Their specific samples and imagined or anticipated loss designs limit any inference about how all audiences respond to actual creator disappearance, sudden death, or public scandal.

The temporal structure of rupture matters. An expected series ending may permit preparation; a sudden loss may interrupt routine and identity references; evidence of deception may generate moral evaluation and loss of trust rather than simple separation distress. Digital archives can preserve the figure's image after the relationship has changed, making continued exposure psychologically ambivalent. Researchers should distinguish severed access, changed evaluation, and changed perceived reciprocity. Repeated assessments before and after naturally occurring transitions, with attention to ethical consent, could clarify which consequences are transient and which involve durable reorganization of audience routines and expectations.

The Measurement Problem: What Does a Parasocial Scale Measure?

Parasocial research depends upon the match between a proposed construct and the questions used to observe it. A scale asking whether a television host seems likable may index evaluation; one asking whether the viewer imagines friendship may index relationship representation; another asking whether the host seemed to speak directly to the viewer during a particular clip may index momentary perceived interaction. A high correlation among these instruments is possible without equivalence. If researchers label all three "parasocial interaction," an apparent theoretical regularity could reflect a mixture of psychological processes.

Dibble, J. L., Hartmann, T., & Rosaen, S. F (2016) directly compared a widely used parasocial interaction scale with the Experience of Parasocial Interaction scale. In an experiment including 383 participants, a woman in a short video either addressed the viewer bodily or did not. The newer experience scale more clearly differentiated the manipulated state of mutual awareness, whereas the older scale potentially captured liking or longer-term dispositions. The contribution is not a universal rejection of historical instruments; it is a demonstration that construct validity must be judged against the causal or theoretical property researchers intend to measure.

Scale selection requires attention to wording, time reference, persona familiarity, modality, and dimensionality. Asking "during the video, it felt as if she noticed me" identifies a distinct reference window from "I have come to regard her as someone important in my life." Reliability coefficients alone cannot resolve this difference because internally consistent items may still measure the wrong concept. Validation should include convergent and discriminant measures, known-groups comparisons, experimental responsiveness, test–retest evidence when stability is predicted, and qualitative interpretation of items. Translation and age adaptation require explicit tests of conceptual and psychometric equivalence.

The Experience of Parasocial Interaction and Stimulus Manipulation

The Experience of Parasocial Interaction approach treats perceived interaction as a state generated within a defined viewing event. Its methodological advantage is that conditions can be experimentally varied while an audience member's immediate social experience is assessed. Hartmann, T., & Goldhoorn, C (2011) introduced the approach with a 2 × 2 manipulation of bodily and verbal address and a sample of 198. The result connects a specific design feature to a plausible perceptual mechanism, avoiding the inference that television exposure alone produces strong interpersonal attachment.

Experimental interpretation still depends on the treatment contrast. If a performer both makes eye contact and uses personal language, changes in perceived warmth, realism, attention, or credibility can accompany the intended parasocial manipulation. The study design should therefore specify whether addressing is conceptualized as the treatment or as an instrument for a latent social state. If the objective is to show that direct address causes stronger perceived interaction, randomization can establish the average response difference. If the objective is to show that perceived interaction itself causes persuasion, the mediator–outcome relationship requires additional identification assumptions or direct manipulation.

An ideal replication would preregister primary outcomes, report all manipulations and exclusions, use multiple performer identities, and assess whether effects survive across cultural and linguistic contexts. Independent replications could add objective eye tracking or physiological responses, but those measures would index attention and arousal, not parasocial experience automatically. Such multimethod evidence is strongest when paired with psychometrically appropriate self-reports and behaviorally informative choices.

Persistent Relationship Measures and Temporal Validity

Durable parasocial relationships require a different measurement frame. Respondents should report an established figure with whom they have prior exposure, describe familiarity and emotional significance over time, and distinguish relationship strength from enjoyment of a recent piece of content. Measuring relationship intensity immediately after one highly engaging video risks treating state emotion as stable attachment. Conversely, asking only about a favorite persona creates selection bias because the sample excludes weaker or negative relationships.

Panel designs can assess stability through repeated measurements before and after media events. Within-person change may reflect exposure, new knowledge about a creator, changes in personal circumstances, or shifts in platform recommendation. A two-wave association is insufficient to untangle dynamic reciprocal causation; additional waves, exposure records, and appropriate time scales improve the analysis. Researchers should register the unit of a relationship—a particular performer, a fictional character, a channel persona, or a creator collective—because these identities may not align.

The meta-analytic separation of Tukachinsky, R., Walter, N., & Saucier, C. J (2020) establishes a strong reason to protect measurement distinctions in both primary work and synthesis. Measurement models should also recognize negative relationships and ambivalence. A critical viewer may experience acute presence or strong familiarity while rejecting the persona morally. If instruments measure only liking, they conflate positive valence with social significance and cannot explain controversial or antagonistic forms of media involvement.

Social Presence Instruments and Multidimensional Experience

Social presence research asks whether a communication environment conveys the awareness and responsiveness of others. This can include perceived copresence, psychological involvement, attentional allocation, mutual understanding, and inferred agency. Biocca, F., Harms, C., & Burgoon, J. K (2003) argued that theory and measurement had to capture more than superficial technological realism. Their critique remains pertinent in video-mediated and immersive environments, where designers may equate higher resolution or anthropomorphic imagery with humanlike social contact.

Researchers should decide whether social presence is an antecedent of parasocial interaction, a concurrent experiential dimension, a mediator between communication features and parasocial response, or an outcome in its own right. Each alternative generates different design and analysis requirements. If presence and PSI are measured by nearly identical items, their statistical association becomes a tautological measure of shared wording. If presence is measured by perceived copresence and PSI by felt personal address, their relationship becomes a testable hypothesis.

Nowak, K. L., & Biocca, F (2003) showed that manipulated agency and anthropomorphism in virtual environments do not map straightforwardly onto perceived copresence or social presence. This suggests a design-sensitive rather than purely visual account. An experiment could compare visually realistic avatars with behaviorally contingent low-fidelity agents and ask which features most strongly predict perceived social agency. Such a study should also distinguish the presence of a real remote human partner, a programmed representation, and a fictional persona.

Survey Evidence, Selection, and the Limits of Correlation

Surveys have been foundational because parasocial ties are embedded in voluntary media use and relationships that can last for years. They can document experiences that laboratory exposures rarely reproduce, including strong familiarity, routine integration, and high emotional investment. They are also susceptible to selection effects. A viewer who follows a creator may already share their interests, support their ideology, or appreciate their personality. Similarity and perceived trust can precede media exposure, emerge during it, or be retrospectively intensified by existing attachment.

The Rubin, A. M., Perse, E. M., & Powell, R. A (1985) is a useful example of why explanatory factors should be modeled jointly: instrumental information motives and perceived realism explained relevant variation, whereas loneliness did not show the hypothesized simple correlation with parasocial interaction. The study's 329 questionnaires clarify relationships among measured variables within that sample, but its observational design cannot identify the effect of randomly increasing viewing time or changing a person's social opportunities.

Contemporary creator surveys share these risks. Bond, B. J (2016) studied adolescents' relationships with media figures in a social-media context, while Lee, H., Sun, Y., & Rojas, H (2024) compared audience involvement with livestreamers across South Korea and the United States. Such studies are valuable for ecological and comparative questions. Without experimental allocation or sufficiently rich repeated measures, however, a platform-use coefficient should be interpreted as an association rather than an independently established platform effect.

Experiments, Mechanisms, and External Validity

Experiments can isolate specific perceptual cues because randomly assigned viewers can encounter otherwise comparable material differing in direct address, visible chat response, or persona characteristics. Hartmann, T., & Goldhoorn, C (2011), Dibble, J. L., Hartmann, T., & Rosaen, S. F (2016), and Wulf, T., Schneider, F. M., & Queck, J (2021) represent a coherent experimental tradition. Their findings support the conclusion that interactive presentation and addressing structures can change immediate parasocial reports under specified conditions.

Experimental stimuli are nevertheless limited representations of an ecological relationship. A laboratory subject watching a short unknown presenter has neither the months of exposure nor the accumulated personal meanings of a devoted follower. If researchers infer changes in long-term attachment from an immediate EPSI score, they cross a construct boundary. A stronger design combines random assignment to recurring formats with repeated measures of familiarity, relational expectation, behavior, and well-being. It also samples multiple creators, genres, and audience populations so that the estimated effect is not hostage to a single unusually charismatic presenter.

The mechanism linking experimental manipulation to outcome must be specified independently of the effect. A camera-directed address might increase perceived social presence, but it could also increase attention or processing fluency. Mediation analysis requires assumptions about confounding and measurement error beyond randomization of the initial cue. Factorial experiments separating address, source credibility, and opportunities for real response would allow more discriminating tests. Reporting null results and confidence intervals is essential, particularly when field expectations encourage selective celebration of strong effects.

Longitudinal Development, Repeated Exposure, and Reverse Causality

Parasocial development unfolds through repeated encounters, making time central to explanation. An individual's growing closeness might follow repeated exposure, but the desire for closeness may itself produce additional viewing and subscription. A creator's change of subject may simultaneously alter exposure and trust; audience life transitions may change both social need and media habits. Longitudinal inquiry must account for this feedback rather than simply measure an early predictor and a late outcome.

A useful design would sample viewers before following a creator, then record incoming recommendations, voluntary video starts, active engagement, recognition of response, and recurring bond measures. The model should differentiate between-person association ("frequent viewers tend to report stronger bonds") from within-person association ("this viewer reports stronger connection after an unusual increase in exposure"). These estimands answer different questions, and neither independently proves that exposure causes closeness. Exogenous shocks, such as transparent changes to display rules, can offer natural experiments if alternative explanations and interference are examined.

Relationship persistence can also be tested through interruption. A program ending, creator hiatus, change of platform, or removal of a subscription feature may disrupt exposure while the viewer's remembered attachment continues. Such events can reveal whether social significance persists beyond the immediate media stimulus. Cohen, J (2004) provides important starting points for studying loss and attachment, although more direct prospective field measurements would strengthen causal inference.

Meta-Analysis, Construct Heterogeneity, and Evidence Aggregation

Meta-analysis is particularly valuable in a field where individual studies vary by genre, population, measurement scale, and media technology. Tukachinsky, R., Walter, N., & Saucier, C. J (2020) synthesized 120 studies and reported no significant general association between social deficiencies and parasocial relationship strength. Their findings also supported relationships with homophily, identification, and persuasion, while differentiating PSI from PSR. This is stronger cumulative evidence against an overly simple deficit model than any single survey, but it remains constrained by the conceptual and design heterogeneity of the included literature.

The Schramm, H., Liebers, N., Biniak, L., & Dettmar, F (2024) research-trends review of 281 studies is different in purpose. It maps the characteristics and development of a rapidly expanding research area across a defined publication period rather than estimating one psychological effect across all studies. Its contribution is coverage and methodological diagnosis. Combining this mapping evidence with effect-size synthesis can reveal where the field has abundant descriptive observations but relatively few prospective causal investigations.

Researchers should report which construct and outcome each quantitative synthesis actually aggregates. A meta-analysis of influencer-related well-being cannot be generalized automatically to fictional-character attachment or historical TV news without examining moderators and transportability. Statistical techniques can model dependence among multiple effect sizes and heterogeneity, but they do not resolve construct mismatch by themselves. Robust systematic procedures also demand transparent inclusion criteria, risk-of-bias assessment, and access to sufficient study metadata. This monograph is a narrative critical synthesis of existing sources and makes no claim to reproduce these systematic review procedures.

Replication, Preregistration, and Open Measurement

The strongest next phase of parasocial science will be distinguished by reproducible manipulations, open psychometric documentation, and coordinated studies across media types. Replication requires more than a second sample using a slightly different scale. Investigators should specify whether they are replicating the effect of direct verbal address, the factor structure of a relationship instrument, or an association with a health outcome. Each target has a different success criterion.

For experimental work, preregistration should distinguish the manipulation, primary outcome, time window, attention checks, exclusions, and intended estimand. For observational work, transparent model specifications should describe exposure measurement, confounding adjustment, missingness, platform selection, and the treatment of algorithmic recommendation. For translated instruments, researchers should test measurement invariance instead of equating a back-translation with psychometric equivalence. Longitudinal data-sharing must protect sensitive follow relationships, emotional disclosures, and potentially identifying creator information.

A cumulative research program also benefits from negative and boundary findings. An absence of loneliness association in a properly powered sample, a failed effect of a direct-address manipulation in an unfamiliar genre, or a lack of longitudinal change after exposure would revise theoretical scope. The goal is a model capable of making falsifiable predictions about when an apparent relationship emerges, what it does, and the conditions under which it dissipates. Preserving uncertainty where the data cannot resolve mechanism is a component of scientific precision.

Positive and Negative Psychological Outcomes

Parasocial involvement has several plausible consequences, and their interpretation depends on the outcome. A sense of familiarity may be experienced as comforting; a creator's success may inspire learning; an admired person's disclosure may reduce stigma. The same connection can coexist with jealousy, social comparison, disappointment, compulsive checking, or susceptibility to misleading advice. A scientific outcome model must therefore analyze distinct response domains rather than divide media relationships into uniformly healthy and unhealthy kinds.

Li, Y., Liu, Z., & Liu, F (2026) conducted a systematic review and three-level random-effects meta-analysis specifically addressing parasocial engagement with social media influencers and mental health outcomes. They included 17 studies and 52 effect sizes. Associations were positive for both adaptive outcomes (28 effects; pooled r = .38, 95% CI [.18, .55]) and maladaptive outcomes (24 effects; r = .24, 95% CI [.08, .38]). These coefficients show that higher parasocial engagement covaried with more favorable measures such as well-being and with less favorable measures such as problematic engagement or distress. They are correlational syntheses, not estimates of the consequences of experimentally creating a relationship.

The review identified substantial variation by outcome domain, parasocial construct, age, and region. It warned that heterogeneous psychological outcomes should not be treated as interchangeable components of one general mental-health variable. The study included English-language empirical work and concentrated on social media influencers, so its estimates should not be generalized mechanically to radio presenters, historical television characters, or artificial conversational systems. The associations are scientifically significant precisely because they resist a one-dimensional causal story: mechanisms, baseline vulnerabilities, media practices, and platform conditions matter.

Loneliness, Social Resources, and Everyday Companionship

The substitution hypothesis assumes that viewers turn toward a parasocial bond because interpersonal companionship is missing. In some circumstances this may be a plausible individual-level explanation, particularly during isolation or stress. As a general empirical law, however, the proposition conflicts with important accumulated evidence. Rubin, A. M., Perse, E. M., & Powell, R. A (1985) reported no significant direct correlation between loneliness and parasocial interaction in their local-news sample. Tukachinsky, R., Walter, N., & Saucier, C. J (2020) likewise found no significant meta-analytic relationship between social deficiencies and parasocial relationship strength across the studies they synthesized.

This evidence supports treating parasocial experience as an ordinary dimension of mediated social life whose functions vary. Someone with abundant friendships can develop an intense admiration for an author, actor, athlete, or commentator. Someone with limited social access may find reliable emotional comfort in a familiar voice. A viewer can also devote so much attention to media that the balance of daily activities changes. These situations must be distinguished empirically through measures of interpersonal network quality, subjective loneliness, time allocation, distress, personal goals, and individual trajectories.

A strong study of companionship would record the participant's social circumstances before and after meaningful changes in exposure, assess what the relationship subjectively supplies, and separate intentional comfort-seeking from automatic habitual viewing. Cross-sectional correlations between media use and loneliness cannot answer whether following creators changes loneliness, whether loneliness changes following behavior, or whether unobserved circumstances affect both. The source of causal variation must be specified.

Adolescence, Developmental Needs, and Identity

Adolescence brings intensified identity exploration, social comparison, autonomy development, and sensitivity to peer recognition. Media figures can become sources of vocational models, stylistic cues, emotional narratives, social norms, or imagined advice. The developmental role of parasocial experience depends on the person followed, the content represented, the adolescent's circumstances, and the opportunities to evaluate claims critically. Researchers should avoid treating all adolescent parasocial engagement as either a social deficit or an educational benefit.

Bond, B. J (2016) investigated adolescents' parasocial relationships with media personae in connection with social-media following, making it relevant to exposure pathways and the ecology of youth relationships. The 2026 Li, Y., Liu, Z., & Liu, F (2026) reported stronger associations for adolescent samples than for some other age groups, but subgroup comparisons may reflect uneven study availability, different instruments, and different platform practices. Such comparisons do not by themselves show that adolescents are uniquely vulnerable to a fixed causal mechanism.

Research should distinguish normal developmental experimentation from functional impairment. Attention to a celebrity, the wish to imitate a creator, and discomfort at the end of a series can occur in ordinary development. Clinically important questions concern distress severity, displacement of necessary activity, exposure to harmful advice, coercive monetization, and the absence of protective support. Longitudinal sampling, age-appropriate psychometrics, and ethical safeguards are required to identify developmental changes rather than impose adult interpretations on adolescent media practices.

Cross-Cultural Variation and Generalizability

Social norms organize what counts as familiarity, warmth, respectful distance, authenticity, celebrity, and appropriate public disclosure. These norms can shape both exposure practices and how questionnaire items are understood. A direct gaze that conveys warmth in one communicative setting may convey pressure or intrusion in another. Likewise, public figures can occupy different moral and institutional roles across countries. Cross-cultural research must measure the underlying experience, not merely compare translations of English-language scales.

Lee, H., Sun, Y., & Rojas, H (2024) studied livestreamer relationships in South Korea and the United States, showing the value of comparative design for examining contextual variation. The 2026 Li, Y., Liu, Z., & Liu, F (2026) found region-based differences in pooled associations with mental-health outcomes, including stronger results in some Eastern samples. Its authors cautioned about interpretation across heterogeneous studies. A regional pooled coefficient could reflect differences in sampled platforms, age distributions, relationship scales, or reporting practices, and cannot be taken as a timeless trait of entire cultural populations.

A comparative program should test measurement invariance across language versions, recruit multiple age and socioeconomic groups, describe platform access and creator types, and predefine meaningful contrasts. It should also analyze media figures who address transnational audiences, multilingual creators, and diaspora communities. Cultural differences become scientifically interpretable when researchers specify which norms or affordances vary and identify mechanisms linking them to observed experience.

Persuasion, Source Credibility, and Behavioral Consequences

Parasocial relationship strength can be associated with the persuasiveness of messages from familiar sources. A viewer who feels close to a creator may attend more closely to recommendations, interpret uncertain claims charitably, or regard disclosed experiences as relevant evidence. Yet persuasion also depends on argument quality, prior agreement, perceived expertise, commercial incentives, and the behavioral costs of complying. Treating relationship intensity as a sufficient cause of belief change obscures these competing pathways.

The Tukachinsky, R., Walter, N., & Saucier, C. J (2020) meta-analysis connects parasocial relationships with persuasive outcomes. Sokolova, K., & Kefi, H (2020) studied influencer credibility and purchase intentions, while Lou, C., & Yuan, S (2019) examined the trust dynamics of branded influencer content. These findings identify relevant associations and theoretical mediators. They do not justify a general claim that strong parasocial bonds compel purchase, increase susceptibility in all audiences, or override message quality.

Testing causality requires manipulation of the message and its source attributes while measuring preexisting bonds before exposure. Field experiments could randomize credible disclosure, informative qualifications, or promotional framing, then measure observed choices rather than intention alone. Behavioral outcomes may be delayed, constrained by financial resources, or affected by social discussion. An ethically sound research design should detect persuasion without deceiving participants about serious health risks or exposing them to uncorrected misinformation.

Health Communication, Trust, and Misinformation

Media relationships matter in health communication because public figures can normalize help-seeking, illustrate personal coping experiences, promote interventions, or give audiences a reason to attend to otherwise ignored information. The same perceived trust may magnify unreliable treatment claims, distorted risk impressions, or commercial promotion of ineffective interventions. A relational cue is not equivalent to medical expertise. The scientific problem is the interaction between social trust, message accuracy, and the audience's capacity to evaluate evidence.

A 2026 meta-analysis by Saucier, C. J., Andrews, E. A., Walter, N., & Tukachinsky Fo (2026) synthesized 58 studies on parasocial relationships and health. The published abstract reports positive associations with beneficial as well as harmful health attitudes, with health information seeking and sharing, and with mediating constructs including source credibility, perceived risk, and self-efficacy. Moderator analyses identified differences according to health condition, media personality, and use of experimental stimuli. These results suggest differentiated opportunities and risks; they do not establish that parasocial attachment itself produces a net health benefit.

Health researchers should measure actual advice quality and independently assess whether a message is consistent with authoritative guidance. Outcomes should distinguish information seeking, stated attitudes, knowledge retention, behavior, and objectively verifiable health endpoints. Intervention studies need informed consent, clear disclosures, and appropriate scientific or clinical oversight. Translating an observational relationship coefficient into a recommendation to cultivate influencer attachment would exceed what the evidence presently identifies.

Well-Being, Distress, and Problematic Engagement

Psychological outcomes need construct-specific interpretation. Subjective well-being, self-esteem, loneliness, anxiety symptoms, problematic media engagement, and distress following a creator scandal measure different experiences. Each can relate to parasocial involvement through distinct mechanisms. For example, inspiration and belonging might predict higher life satisfaction, while relentless self-comparison or repeated checking might predict distress. The same participant could report both experiences during different periods or in relation to different media figures.

Li, Y., Liu, Z., & Liu, F (2026) found associations of parasocial engagement with positive and negative mental-health domains in the influencer literature. The authors' subgroup analyses produced different magnitudes across categories and reported high heterogeneity. Their pooled estimates should be read as descriptions of the available studies rather than as predictions for a specific individual. In particular, a correlation between parasocial engagement and problematic use cannot establish that relational feelings caused compulsive media behavior. Habitual exposure might itself contribute to both the measured bond and difficulties disengaging.

A productive clinical-adjacent research design would separate degree of involvement from its functioning-related costs, account for co-occurring symptoms, and follow changes over time. It should examine adverse experiences without pathologizing ordinary fandom. This dossier concerns psychological processes and evidence appraisal; it offers no individualized diagnosis, treatment plan, or pharmaceutical recommendation. Clinical claims should be evaluated using disorder-specific diagnostic standards and independent specialist review where warranted.

Fictional Characters, Loss, and Continuity of Emotional Meaning

Parasocial attachment to fictional characters adds a distinct problem: the represented person may never have existed outside a narrative, yet the emotional significance of the character can be durable. A viewer can acquire knowledge of the character's traits, anticipate future actions, compare personal experiences with the character's story, and experience disappointment when a plot changes. The psychological reality of that experience is measured through the viewer's cognition and affect; the ontology of the fictional figure remains a separate question.

Cohen, J (2003) examined parasocial breakups, and Cohen, J (2004) considered relationship intensity and attachment patterns in responses to anticipated loss. These findings show why researchers must investigate event-specific reactions and prior bonds separately. A favorite character's disappearance may be anticipated, narratively explained, reversed through sequels, or preserved in accessible media archives. Such circumstances shape the meaning of a reported breakup.

The broader research significance is continuity. Media artifacts can remain available after the author, performer, or fictional universe changes, and viewers can revisit earlier episodes while holding altered evaluations. This makes a parasocial relationship a dynamic social representation rather than a fixed amount of exposure or a binary active/inactive status. Prospective repeated measures and contextual interviews can reveal whether media-based familiarity is reconstructed, withdrawn, redirected, or preserved during change.

Artificially Mediated Figures and the Boundary of the Research Object

Artificially generated representations introduce a related but distinct explanatory domain. A human influencer may use filters, virtual animation, generated speech, or editorial assistance without changing the human origin of the public persona. An entirely synthetic character may be presented as an influencer; a conversational agent may generate contingent responses to individual users. These cases vary in the source of agency, the possibility of actual individual feedback, and the continuity of the system's behavior. The mere presence of a humanlike display does not establish humanlike interiority or a human social relationship.

Stein, J.-P., Breves, P. L., & Anders, N (2024) experimentally investigated real and virtual influencers, including the role of perceived similarity and human likeness. Their findings indicate why a categorical opposition between "real" and "virtual" must be supplemented by specific tested properties. A viewer's social response can arise from representation and cue interpretation, whereas the technical source of the persona's behavior is separately observable. Effects concerning a virtual influencer should not automatically be generalized to a conversational assistant capable of generating responses in context.

In the Artificial Era, research on mediated social presence can contribute operational questions to human–artificial relations: what response-contingency patterns cause perceived attention, how do users infer stable identity from repeated interaction, and when does perceived reciprocity correspond to actual adaptive behavior? These hypotheses can be tested without stipulating unmeasured psychological states in an artificial system. This dossier retains human media figures as its primary canonical object. Reciprocal human–artificial interaction, AI companionship, and attachment to artificial agents belong to related research programs with their own definitions, samples, methods, and canonical ownership.

A Causal Framework for Mediated Social Experience

A scientific explanation of parasocial phenomena should specify the sequence connecting environmental features to social interpretation and longer-term psychological organization. A useful synthesis begins with audience selection and prior expectations: people approach a message with preferences, social histories, and knowledge of a persona. They encounter identifiable social cues such as gaze, address, personal disclosure, voice, and response-contingency. They interpret those cues in terms of interpersonal attention, agency, familiarity, or credibility. The immediate result may include experienced parasocial interaction and social presence. Across repeated occasions, selected experiences can become integrated into a parasocial relationship with expectations that persist outside the episode.

This sequence is a research-organizing synthesis, not an empirically certified universal path. It contains feedback loops. A strong existing relationship may increase selective exposure and favorable interpretations of new cues; recommendation systems may respond to engagement by increasing subsequent encounters; changes in personal circumstances may alter the meaning of the same performer. Experimental evidence for the cue-to-experience relation is relatively direct in studies such as Hartmann, T., & Goldhoorn, C (2011), while evidence for transition to enduring relationships and downstream health outcomes is considerably more dependent on observational designs. The distinction determines which causal arrows can be claimed with confidence.

A full model must also include social and material consequences. Continued involvement might support learning, comfort, and social discussion, or contribute to problematic checking, persuasion, and disappointment. Whether these effects occur depends upon content quality, viewer circumstances, platform organization, and competing resources. Tukachinsky, R., Walter, N., & Saucier, C. J (2020) and Li, Y., Liu, Z., & Liu, F (2026) provide evidence of heterogeneous associations rather than a singular outcome law. Future interventions should randomize manipulable features and observe explicitly defined endpoints.

Evidence Appraisal Across Major Research Traditions

The original Horton, D., & Wohl, R. R (1956) essay has foundational conceptual importance. Its status as an original theoretical observation is different from that of a contemporary preregistered experiment. The Rubin, A. M., Perse, E. M., & Powell, R. A (1985) survey contributes ecologically situated associations and a directly relevant null result concerning loneliness. Its sampling and cross-sectional method limit generalization and causal inference. Hartmann, T., & Goldhoorn, C (2011) and Wulf, T., Schneider, F. M., & Queck, J (2021) improve local causal identification for stimulus features; their brief exposure settings cannot substitute for a prospective study of persistent relationships.

The psychometric work of Dibble, J. L., Hartmann, T., & Rosaen, S. F (2016) is decisive for interpretation because measurement mismatch can compromise even well-randomized experiments. The corpus map by Schramm, H., Liebers, N., Biniak, L., & Dettmar, F (2024) documents research diversity and changing emphases. The 120-study meta-analysis by Tukachinsky, R., Walter, N., & Saucier, C. J (2020) is especially informative about the correlational structure of relationships and the limits of a deficiency model. The 2026 influencer and health syntheses by Li, Y., Liu, Z., & Liu, F (2026) and Saucier, C. J., Andrews, E. A., Walter, N., & Tukachinsky Fo (2026) extend the evidence to outcomes with substantial practical significance, while leaving causal direction and study heterogeneity as central questions.

Across traditions, recurring methodological risks include convenience samples, inconsistent instruments, limited cultural diversity, selective exposure, single-person stimuli, unstable platform designs, common-method self-report, unreported null outcomes, and insufficient longitudinal follow-up. Some articles summarized here were evaluated through publisher abstracts or accessible full text; the narrative does not imply that every paywalled page has been independently inspected. These limitations inform both the interpretation of individual findings and the present dossier's source-review status.

Ethical Organization of Creator–Audience Relations

Parasocial connections produce practical ethical questions when an audience experiences intimacy within a structurally unequal communication arrangement. An influencer can address a large audience using the language of friendship while retaining unilateral control over disclosure, availability, sponsorship, and the terms of participation. This arrangement may be entirely acceptable when its limits are understood, but problems arise when commercial or institutional conduct exploits an expectation of personal care. Investigators should distinguish what audiences believe was promised, what creators actually represented, and what can be observed about communicative access.

Research ethics in this domain therefore concerns both the study process and the mediated environment. Researchers should obtain appropriate consent for participation, avoid unnecessary collection of private messages or follower identities, and respect contextual expectations around ostensibly public posts. When studying adolescents, distressed audiences, or relational loss, procedures should minimize exposure to harmful material and provide appropriate safeguards. Claims about health interventions or medical products require independent evidence assessment; personal familiarity with a creator is not a clinical qualification.

Platform designers can clarify sponsorship, create accessible mechanisms for reporting abuse, and make the difference between broadcast messages and individualized replies intelligible. These design choices are possible intervention targets. Their psychological effects should be evaluated empirically rather than inferred from intentions. A major unanswered question is whether transparency about creator incentives and response limitations changes perceived intimacy, trust calibration, or informed consumer behavior without diminishing valuable social engagement.

Research Priorities and Falsifiable Hypotheses

A priority for the field is a coordinated test of the relation among parasocial interaction, parasocial relationships, and social presence using separate, validated measures. Researchers can randomize direct address and response-contingency, then assess immediate PSI, perceived copresence, and later relationship development across repeated sessions. A key prediction is that address changes state PSI more readily than it changes stable relationship strength after one exposure. This prediction follows the distinction examined by Dibble, J. L., Hartmann, T., & Rosaen, S. F (2016) and could be falsified through appropriately sensitive measures.

A second priority concerns selection and relational development. Cohorts recruited before beginning to follow a new creator could provide exposure logs, repeated measures of affiliation, and records of genuine responses. Statistical models should distinguish who selects content from what its exposure changes. Randomized recommendation or invitation conditions, where ethical and feasible, could sharpen causal identification. The major test is whether shifts in social address or access predict later bond strength after accounting for prior preference.

Third, outcome-specific field studies should test the pathways linking parasocial involvement to well-being, information learning, purchasing, health beliefs, and problematic engagement. The divergent associations summarized in Li, Y., Liu, Z., & Liu, F (2026) justify separate preregistered outcome families. Fourth, cross-cultural work should evaluate whether apparently different results reflect true social mechanisms or measurement differences. Fifth, research on AI-generated persona features should distinguish visible human likeness, actual algorithmic contingency, and perceived personhood, without transferring results from human celebrity engagement directly to agent-centered relationships.

Integrated Scientific Synthesis

The accumulated evidence supports five interdependent conclusions. First, a mediated communication environment can organize recognizable experiences of interpersonal attention without a direct individual exchange. This is the central historical insight of Horton, D., & Wohl, R. R (1956), given sharper experimental definition by Hartmann, T., & Goldhoorn, C (2011). Second, a momentary interaction-like experience and a persistent relationship are empirically related but separable, as indicated by Dibble, J. L., Hartmann, T., & Rosaen, S. F (2016) and Tukachinsky, R., Walter, N., & Saucier, C. J (2020).

Third, media relationships draw upon ordinary social-cognitive capacities such as recognition, similarity judgment, expectation, and perspective-taking. Such mechanisms can serve a diverse range of functions, and a general social-deficiency explanation is unsupported by the cited early survey and meta-analysis. Fourth, the consequences of parasocial involvement depend on the outcome: stronger engagement can correlate with favorable and unfavorable responses, as shown in the 2026 Li, Y., Liu, Z., & Liu, F (2026) and Saucier, C. J., Andrews, E. A., Walter, N., & Tukachinsky Fo (2026). Fifth, interactive media alter the opportunities for actual response but do not eliminate the analytic difference between perceived and observed reciprocity.

These conclusions place parasocial interaction within the wider psychology of mediated social perception. The central scientific task is to measure how representation becomes experienced as encounter, how repeated encounters become organized as social memory, and how platforms and audience conditions influence the consequences of that organization. Each link is investigable through distinct methods and can be supported, revised, or rejected by evidence.

Conclusion

Parasocial Interaction and Mediated Social Presence names a coherent research field concerning the organization of social experience across mediated environments. Its foundational phenomenon is the viewer's perception of interpersonal contact with a represented figure; its durable relational form is the parasocial relationship; its environmental dimension is the sense of social presence afforded by communication media. Their distinctions are essential for scientific explanation and measurement. The development of the field from 1950s broadcasts to digital creator platforms has increased the observable diversity of social cues, relational expectations, and opportunities for actual response.

Experimental work establishes that forms of direct address and mediated responsiveness can alter immediate parasocial experience under defined conditions. Psychometric studies show that some historical scales confound state interaction with liking and relationship strength. Survey findings and meta-analyses establish associations with homophily, identification, persuasion, and both beneficial and adverse outcome domains, while weakening the assumption that parasocial relationships generally substitute for deficient offline ties. The growing evidence base requires better causal designs, longitudinal observation, measurement invariance, independent replications, and domain-specific outcome analysis.

Within the Artificial Era, the research extends toward questions about responsiveness and perceived agency in artificial communication environments. The explanatory pathway must remain empirically explicit: measurable system behavior, perceived social cues, audience interpretation, relational history, and observed consequences. The scientific contribution of this monograph is the systematic organization of these existing traditions and evidence boundaries around one stable research object, suitable for cumulative inquiry across Media and Technology Psychology and its adjoining fields.

References

  • Biocca, F., Harms, C., & Burgoon, J. K. (2003). Toward a more robust theory and measure of social presence: Review and suggested criteria. Presence: Teleoperators and Virtual Environments, 12(5), 456–480. https://doi.org/10.1162/105474603322761270
  • Bond, B. J. (2016). Following your "friend": Social media and the strength of adolescents' parasocial relationships with media personae. Cyberpsychology, Behavior, and Social Networking, 19(11), 656–660. https://doi.org/10.1089/cyber.2016.0355
  • Cohen, J. (2003). Parasocial breakups: Measuring individual differences in responses to the dissolution of parasocial relationships. Mass Communication & Society, 6(2), 191–202. https://doi.org/10.1207/S15327825MCS0602_5
  • Cohen, J. (2004). Parasocial break-up from favorite television characters: The role of attachment styles and relationship intensity. Journal of Social and Personal Relationships, 21(2), 187–202. https://doi.org/10.1177/0265407504041374
  • Dibble, J. L., Hartmann, T., & Rosaen, S. F. (2016). Parasocial interaction and parasocial relationship: Conceptual clarification and a critical assessment of measures. Human Communication Research, 42(1), 21–44. https://doi.org/10.1111/hcre.12063
  • Giles, D. C. (2002). Parasocial interaction: A review of the literature and a model for future research. Media Psychology, 4(3), 279–305. https://doi.org/10.1207/S1532785XMEP0403_04
  • Hartmann, T., & Goldhoorn, C. (2011). Horton and Wohl revisited: Exploring viewers' experience of parasocial interaction. Journal of Communication, 61(6), 1104–1121. https://doi.org/10.1111/j.1460-2466.2011.01595.x
  • Horton, D., & Wohl, R. R. (1956). Mass communication and para-social interaction: Observations on intimacy at a distance. Psychiatry, 19(3), 215–229. https://doi.org/10.1080/00332747.1956.11023049
  • Kowert, R., & Daniel, E., Jr. (2021). The one-and-a-half sided parasocial relationship: The curious case of live streaming. Computers in Human Behavior Reports, 4, 100150. https://doi.org/10.1016/j.chbr.2021.100150
  • Lee, H., Sun, Y., & Rojas, H. (2024). Parasocial relationships with live streamers: Evidence from South Korea and the United States. Asian Journal for Public Opinion Research, 12(3), 166–183. https://doi.org/10.15206/ajpor.2024.12.3.166
  • Li, Y., Liu, Z., & Liu, F. (2026). Parasocial engagement with social media influencers and mental health outcomes: Systematic review and meta-analysis. JMIR Mental Health, 13, e96331. https://doi.org/10.2196/96331
  • Lou, C., & Yuan, S. (2019). Influencer marketing: How message value and credibility affect consumer trust of branded content on social media. Journal of Interactive Advertising, 19(1), 58–73. https://doi.org/10.1080/15252019.2018.1533501
  • Nowak, K. L., & Biocca, F. (2003). The effect of agency and anthropomorphism on users' sense of telepresence, copresence, and social presence in virtual environments. Presence: Teleoperators and Virtual Environments, 12(5), 481–494. https://doi.org/10.1162/105474603322761289
  • Rubin, A. M., Perse, E. M., & Powell, R. A. (1985). Loneliness, parasocial interaction, and local television news viewing. Human Communication Research, 12(2), 155–180. https://doi.org/10.1111/j.1468-2958.1985.tb00071.x
  • Saucier, C. J., Andrews, E. A., Walter, N., & Tukachinsky Forster, R. (2026). A meta-analysis of parasocial relationships and health: Examining theoretical mechanisms, health outcomes, and moderating factors. Health Communication. Advance online publication. https://doi.org/10.1080/10410236.2026.2641695
  • Schramm, H., Liebers, N., Biniak, L., & Dettmar, F. (2024). Research trends on parasocial interactions and relationships with media characters: A review of 281 English and German-language studies from 2016 to 2020. Frontiers in Psychology, 15, 1418564. https://doi.org/10.3389/fpsyg.2024.1418564
  • Sokolova, K., & Kefi, H. (2020). Instagram and YouTube bloggers promote it, why should I buy? How credibility and parasocial interaction influence purchase intentions. Journal of Retailing and Consumer Services, 53, 101742. https://doi.org/10.1016/j.jretconser.2019.01.011
  • Stein, J.-P., Breves, P. L., & Anders, N. (2024). Parasocial interactions with real and virtual influencers: The role of perceived similarity and human-likeness. New Media & Society, 26(6), 3433–3453. https://doi.org/10.1177/14614448221102900
  • Tukachinsky, R., Walter, N., & Saucier, C. J. (2020). Antecedents and effects of parasocial relationships: A meta-analysis. Journal of Communication, 70(6), 868–894. https://doi.org/10.1093/joc/jqaa034
  • Walther, J. B. (1996). Computer-mediated communication: Impersonal, interpersonal, and hyperpersonal interaction. Communication Research, 23(1), 3–43. https://doi.org/10.1177/009365096023001001
  • Wulf, T., Schneider, F. M., & Queck, J. (2021). Exploring viewers' experiences of parasocial interactions with videogame streamers on Twitch. Cyberpsychology, Behavior, and Social Networking, 24(10), 648–653. https://doi.org/10.1089/cyber.2020.0546