RESEARCH DOSSIER
Social Comparison and Self-Evaluation
Abstract. Social comparison and self-evaluation concern how information about other people becomes a standard for judgments of one's own standing, competence, worth, aspirations, and circumstances. This fundamental critical narrative research monograph examines the historical foundations of social comparison theory, the selection of upward, downward, and lateral targets, motivational and information-processing mechanisms, individual differences, relationship-specific effects, and methodological questions. It evaluates experimental findings, the accumulated meta-analytic evidence on comparison choice and reaction, classroom and developmental research, and studies of digitally mediated appearance and achievement comparisons. The synthesis distinguishes immediate evaluative contrast from aspiration, informational learning, longer-term behavioral consequences, and correlational associations with well-being. It examines measurement validity, target-selection bias, cultural and institutional context, and unanswered questions about replication and causality. Human comparison with artificial performance standards is treated as a research agenda requiring explicit tests. The central conclusion is that effects depend on the comparator, domain, appraisal, motive, time scale, and social environment.
Keywords: social comparison; self-evaluation; upward social comparison; downward comparison; lateral comparison; assimilation and contrast; social comparison orientation; self-esteem; academic self-concept; social media; body image; social psychology; Artificial Era
The Historical Formation of Social Comparison Theory
Leon Festinger's 1954 article, Festinger (1954), established a research program around a drive to evaluate opinions and abilities, the circumstances under which objective standards are insufficient, and the informational value of other people. His proposal placed social standards within psychological explanation at a time when group influence and interpersonal reference points were prominent experimental questions. Similarity occupied a central position: individuals were expected to prefer other people who supplied informative comparisons, while the relevant dimension and the presence of objective evaluation could modify that preference. Festinger also proposed an upward tendency in ability evaluation. The theory's historical achievement was to make standards of self-appraisal a tractable topic for hypotheses about target choice, group membership, and changes in evaluations. Its propositions became research questions rather than permanent empirical axioms.
Subsequent work expanded the motivating functions of comparisons. Wills (1981) formulated downward-comparison principles to explain how people under stress or negative affect might regulate well-being by considering others who were worse off. Taylor & Lobel (1989) showed why explicit evaluation and information seeking must be differentiated: under threat, individuals may evaluate downward while seeking knowledge or affiliation upward. Wood (1989) then integrated social cognition and self-related motives, emphasizing that comparison may be imposed by the social setting and is shaped by motives beyond accuracy. These developments changed the presumed process from a simple selection of similar peers into an interaction between motivation, exposure, interpretive framing, and circumstance. The historical sequence also widened the empirical contexts beyond isolated ability tasks to coping, close relationships, health, and institutional settings.
Research in the 1990s and 2000s connected comparisons to self-maintenance, automatic evaluation, accessibility of attributes, and differences among observers. Tesser (1988) modeled how a close other's achievement can either support or threaten self-evaluation depending on the personal relevance of the domain. Collins (1996) investigated why encounters with superior performers can have constructive as well as deflating consequences. Gilbert et al. (1995) tested the spontaneous emergence of comparisons, and Mussweiler (2003) presented a selective-accessibility account of the cognitive basis of assimilation and contrast. Gibbons & Buunk (1999) gave individual differences a measurable instrument, while Gerber et al. (2018) assessed decades of accumulated selection and reaction findings using meta-analysis. The field's development thus reflects increasing precision about what is chosen, what is processed, what is felt, and what endures.
The Anatomy of a Comparison Episode
A comparison begins with a possible dimension: intelligence, appearance, income, generosity, popularity, performance, or another attribute. The observer must identify an actual or imagined comparator and decide whether the comparator's observed characteristic is commensurate with the self. This is an inferential step. A coworker's salary may be evaluated as evidence of institutional recognition, personal merit, unfair advantage, or a misleading difference in responsibilities. These interpretations produce distinct judgments even when the numerical salary gap is identical. Wood (1996) cautions that methods which ask participants merely to rate themselves relative to others cannot always establish that a natural comparison process occurred. A rigorous study separates exposure to the standard, comparative processing, attribution of meaning, and the measured end point.
Comparison direction describes the comparator's perceived standing relative to the observer on the focal dimension. An upward target appears better off or more capable; a downward target appears worse off; a lateral target appears similar. Direction is relative to a selected attribute, rather than a fixed characteristic of the other person. One classmate can be upward for academic performance and downward for physical endurance. Direction must also be separated from valence: a downward comparison can evoke sympathy, anxiety, or anticipated vulnerability, and an upward comparison can evoke admiration, hope, resentment, or discouragement. The informational content of these reactions depends on perceived similarity, attainability, moral significance, and interpersonal connection. Findings synthesized by Collins (1996) and Gerber et al. (2018) explain why a directional label alone has limited predictive power.
Comparison unfolds through at least five analytically distinct operations: attention to a standard, construal of the comparison dimension, assessment of relative standing, integration with an existing self-view, and possible behavioral adjustment. This decomposition is a research framework rather than a claim that every episode involves conscious sequential steps. Automaticity studies Gilbert et al. (1995) show that a preliminary comparative response can arise before reflective correction; the selective-accessibility model Mussweiler (2003) indicates that the information retrieved during judgment can change the result. Repeated exposure may alter which standards are salient and remembered, while social environments constrain which standards can be encountered at all. Research must therefore measure both the distribution of available comparison opportunities and the conditional responses to each opportunity.
Comparison Standards and Target Selection
Target selection is an observable choice with several possible motives. For accurate performance assessment, a similar peer may be more diagnostically useful than an expert with vastly different experience. For aspiration, a more successful person may illustrate an attainable pathway. For reassurance, someone worse off can provide a temporarily favorable reference point. For identity, members of a valued group may carry greater significance than strangers. Festinger's Festinger (1954) original argument about similarity anticipated one prominent informational motive, but empirical work has repeatedly shown that similarity competes with the search for aspiration, affiliation, emotional regulation, and salient environmental standards. These motives can be simultaneous. The chosen target may reflect both the observer's interests and a menu determined by a school, organization, or digital feed.
The experimental literature does not justify treating downward selection as the dominant response to negative feeling. The meta-analysis by Gerber et al. (2018), which explicitly distinguished target-selection designs from reaction designs, reported a pronounced upward selection preference when no threat was present and did not find downward comparison to be the dominant selection even under threat. Offering a lateral alternative made choices less sharply differentiated. These findings challenge a simple translation from downward-comparison motives to expected selection frequencies. They also demonstrate that results depend on available options. A forced choice between a better-off and a worse-off peer differs from an open social environment where similarity, familiarity, and accessibility affect the distribution of encountered targets.
Target choice in everyday life is subject to selection bias. People may follow high-achieving peers, leave threatening networks, and seek communities in which their own standing feels legible. Consequently, the observed frequency of upward comparison cannot be interpreted without knowing which targets were available and why they were selected. Diary and experience-sampling studies can distinguish voluntary choice from unbidden exposure, but repeated self-reports require careful definitions because respondents may recognize intentional comparisons more readily than automatic ones. Experimental randomization of comparator standing can identify short-run effects of information, whereas observational network analyses can show naturally occurring exposure patterns but have difficulty separating selection from influence. Wood (1996) supplies the foundational measurement warning, and Gerber et al. (2018) shows how design differences change aggregate conclusions.
Upward Comparison, Inspiration, and Threat
Upward comparisons place the observer beside a person who appears superior on a valued dimension. Their effects depend on the degree to which the standard is construed as an attainable future self, an informative model, a close rival, or evidence of one's own inadequacy. In Lockwood & Kunda (1997), the impact of high-achieving exemplars on self-views depended on relevance and perceived attainability. A successful target can help organize learning when the observer sees a path to similar success, but may undermine self-evaluation when the same standard is interpreted as unreachable. These mechanisms invite separate measurements of perceived attainability, identification, affect, and later effort rather than a single question about how the comparison 'felt.'
The broader literature reviewed by Collins (1996) documented conditions under which upward comparison improves or worsens subjective self-evaluation. One influential interpretation distinguishes assimilation toward an upward target from contrast away from it. Assimilation can involve seeing the target's achievements as evidence of what the self could become or what the shared group represents. Contrast highlights a gap between the observed target and the current self. These outcomes need not correspond to stable personality types. They are responses to the combined characteristics of a target, task, framing, and existing self-knowledge. The importance of relational closeness and relevance is made explicit in Tesser (1988), which predicts different self-maintenance pressures when achievement belongs to a close person in a personally important domain.
The quantitative synthesis by Gerber et al. (2018) found that contrast was substantially more common than assimilation in the reaction studies it analyzed, with particularly pronounced effects on ability estimates under some experimental manipulations. This result places limits on the popular claim that high-performing models generally elevate self-perception. It also leaves a significant explanatory puzzle: upward targets may be chosen even when a relatively deflating response is likely. Choice can be motivated by information, learning, or aspiration despite a cost to immediate affect. Future tests should measure both concurrent self-evaluation and prospective behavior, distinguish perceived similarity from objective standing, and include no-comparison controls. An inspiring interpretation in an interview is useful evidence of meaning, but by itself it does not establish a lasting behavioral benefit.
Downward Comparison, Coping, and Defensive Evaluation
Downward comparison occurs when an observer takes a lower-standing person as a reference point on the relevant dimension. Wills (1981) argued that such comparisons can protect or elevate subjective well-being under unfavorable circumstances. A patient coping with illness may evaluate their own situation against that of someone with a more severe prognosis; a student may regain a sense of competence by comparing performance with a less successful peer. These examples illustrate potential short-term emotion regulation, not a universal therapeutic effect. A downward standard may also communicate future risk, activate empathy, or expose a gap between moral ideals and behavior. Its consequences depend on whether the observer interprets the lower-standing other as distant, similar, deserving of care, or representative of a possible future.
Taylor & Lobel (1989) made an essential distinction between downward evaluation and upward contact. Under threat, information gathering can orient toward people who have navigated the problem successfully even as favorable self-evaluation draws on less fortunate cases. The same person can thus use different comparison directions for different goals within one situation. Studies that ask only which person an individual would like to meet risk conflating advice seeking with comparative self-appraisal. Likewise, a downward ranking selected in a laboratory may reflect task expectations or the wish to avoid humiliation rather than a general preference for unfavorable others. Methodologically, observed contact, explicit comparative ratings, perceived similarity, and affect need to be treated as separate indicators.
Defensive comparison may temporarily reduce distress yet be costly when it blocks accurate feedback or encourages contempt for those with fewer resources. Research should therefore examine change over time and consequences for relationships, not simply pretest–posttest mood. Wood (1989) emphasized that social comparison serves several self-related objectives and can be elicited by circumstances beyond deliberate control. Gerber et al. (2018) further cautions against inferring universal selection preferences from a motivational theory. A credible causal model distinguishes a trigger such as threat, a comparison opportunity, the appraisal that follows, and the broader ecology in which alternative coping strategies are available. Such a model accommodates reassurance, avoidance, social learning, and compassion without assigning one inevitable emotional response to downward comparison.
Lateral Comparison, Similarity, and Diagnosticity
Lateral comparisons use others perceived as similar in current standing. Similarity can be defined on the focal variable, on background features, or on presumed causes of performance; these forms of similarity carry different informational weight. A novice musician assessing progress may learn more from similarly trained players than from an international soloist. Yet similarity on age or group membership may be psychologically salient without improving objective diagnosticity. Festinger's Festinger (1954) proposal concerned the informational use of relevant others, and later work questioned which similarity criteria observers actually apply. Studies must distinguish perceived similarity from documented equivalence and determine whether observers select peers because they are informative, socially available, or emotionally comfortable.
A comparison with a comparable peer can make a small difference feel decisive. If another person shares one's preparation, opportunities, and experience, a minor performance gap can be interpreted as evidence of competence or effort. At the same time, similar others can offer more credible learning strategies, reveal normative variation, or normalize temporary difficulty. The reaction depends on the precision of the information and its personal stakes. Wood (1989) shows why the comparison relation must be considered alongside goals and context, while Mussweiler (2003) describes how selective attention to similarity or difference changes judgment. Lateral comparisons may have little effect on global esteem while changing the perceived fairness of an outcome or the expected rate of improvement.
Operational definitions should not classify all peer comparisons as lateral. A 'classmate' or 'friend' is a social category, not an empirical indicator that the target is equally skilled or similarly placed. Likewise, similar school grades can conceal different task conditions or evaluation criteria. Designs that manipulate percentile standing while holding background characteristics constant help to isolate direction, whereas factorial designs can vary similarity and direction independently. Gerber et al. (2018) found that including lateral options materially altered target-selection patterns in the pooled experimental literature. This supports making the choice set an explicit property of a study. Observed comparison behavior is always conditional on what the observer could have compared.
Self-Evaluation Maintenance and Relational Closeness
Self-evaluation is maintained in relationships as well as within isolated judgments. Tesser (1988) proposed that a close person's strong performance can benefit self-evaluation through a reflection process when the domain has low relevance to one's own identity, but can threaten self-evaluation through a comparison process when achievement occurs in a highly self-relevant domain. The model makes predictions from the interaction of relational closeness, personal relevance, and relative performance. For example, admiration of a friend's success in an unfamiliar activity can coexist with discomfort about the same friend's superiority in one's core professional specialization. The central phenomenon is a context-sensitive relation between an interpersonal bond and self-evaluation.
The model also suggests that people may regulate exposure, reinterpret domain importance, change performance investment, or adjust closeness. Such possibilities require ethical and empirical precision: a theoretical adaptation strategy is not necessarily a consciously selected behavior or an established universal outcome. Tests should measure the independently manipulated performance relation and relevance of the domain, rather than infer both from self-reported jealousy. Relational comparisons can affect generosity, cooperation, feedback, and willingness to celebrate another person's achievement. Their consequences may differ across intimate friendships, teams, sibling relationships, and professional networks because each setting carries distinct expectations of mutual support and competition.
Self-evaluation maintenance complements broader comparison theories by specifying the relational meaning of a standard. Collins (1996) and Mussweiler (2003) concentrate on how superior standards are interpreted and how activated information shapes judgments. Tesser (1988) locates these processes within closeness and domain relevance. A strong empirical synthesis should test whether relational closeness changes perceived similarity, attainable pathways, and defensive motivation, and whether the predicted interactions replicate in diverse populations. Relationships provide repeated exposure to the same comparator, making it possible to investigate cumulative consequences and recovery after a threatening event. The mechanism can therefore be assessed at the episode, dyad, and developmental time scale.
Assimilation, Contrast, and Selective Accessibility
Assimilation is a shift in self-judgment toward the standard: after considering a high performer, the observer feels more capable or identifies with that performer's standing. Contrast is a shift away from the standard: the same encounter lowers the observer's relative competence assessment. These are outcomes, not synonyms for upward and downward direction. Mussweiler (2003) proposed that selective accessibility provides an informational mechanism. Similarity testing activates information consistent with resembling the target, while dissimilarity testing activates information that differentiates the self. The selectively accessible evidence can then bias the final judgment. This model explains why the same target may produce different self-evaluations depending on the processing question, the context, or a prior cue.
Experiments reported by Mussweiler et al. (2004) tested the predicted relation between similarity-focused and dissimilarity-focused processing and the resulting assimilation or contrast. The theoretical claim is about accessibility of self-relevant information at judgment time, which can be measured or manipulated, rather than the assumption that an observer consciously chooses optimism or pessimism. When studies manipulate comparison standards, instructions, and target attributes, a causal estimate is possible for that specific operation. It does not automatically generalize to years of repeated social exposure or to all dimensions of identity. The research also invites separating early automatic comparisons from later correction: Gilbert et al. (1995) found evidence that participants can generate an initial comparison even when it is of little diagnostic value.
Meta-analytic evidence provides a crucial constraint on mechanism-focused theories. Gerber et al. (2018) found robust evidence of contrastive reactions across the body of experiments they assessed, with substantially weaker support for assimilation under several conditions. The contrastive tendency varied with domain, experimental manipulation, and the handling of publication bias. A selective-accessibility account can explain conditions that support either response, but a general synthesis must accurately represent their observed prevalence. Future studies should measure what information was retrieved, test whether that retrieval mediates change, and consider alternatives such as demand characteristics, response-scale anchoring, and preexisting self-beliefs. Mechanistic success requires experimental differentiation between the proposed processing route and rival explanations.
Motives: Accuracy, Enhancement, Improvement, and Belonging
Social comparison has been connected to several motives that are analytically separable. Self-assessment concerns accurate information about current standing. Self-enhancement concerns a favorable evaluation or protection of self-worth. Self-improvement concerns identifying information and strategies that support future performance. Affiliation concerns learning from or approaching other people, and belonging concerns locating oneself within a valued social community. An upward standard may satisfy assessment and improvement while reducing immediate self-esteem; a downward standard may protect mood while offering little practical guidance. Taylor & Lobel (1989) provides an influential demonstration of diverging evaluation and contact preferences under threat, while Wood (1989) emphasizes that comparisons can serve multiple purposes beyond neutral accuracy.
To identify a motive, a study must do more than infer it from comparison direction. Choosing an outstanding classmate could indicate aspiration, curiosity, the desire to understand a grade, or simple familiarity. The same choice may also reflect an externally assigned menu of comparators. Strong designs manipulate motivational state and measure multiple dependent variables: selected standard, perceived informativeness, expected attainability, immediate self-evaluation, and subsequent behavior. Comparisons should be interpreted within participants' goals and available alternatives. Gerber et al. (2018) is especially useful because its aggregation separates selections from reactions and demonstrates how the assumed link between threat and downward selection can fail under different experimental arrangements.
Motivational plurality explains apparent inconsistencies across contexts. A person preparing for a professional examination may seek highly competent models for learning, compare with same-level peers for calibration, and recall worse performers when recovering from a setback. These actions need not reflect enduring traits or one consciously formulated strategy. Institutions also steer motives by publishing rankings, awards, or achievement distributions. The psychological outcome depends on which information those mechanisms make salient and what the observer is trying to accomplish. Research should specify the motivational manipulation and avoid assigning a motive simply because the observed behavior fits a familiar theoretical narrative. The measurement recommendations in Wood (1996) remain directly relevant to this distinction.
Individual Differences and Comparison Orientation
People differ in how often they notice differences in others' achievements, seek comparative information, or report using other people's circumstances to evaluate themselves. Gibbons & Buunk (1999) developed and evaluated the Iowa–Netherlands Comparison Orientation Measure (INCOM), an instrument designed to measure such individual differences in American and Dutch samples. The research reported psychometric analyses and tests of the relation between questionnaire scores and comparison behavior. The scale offers a structured indicator of habitual comparison orientation, but a scale score is not an independent count of naturally occurring comparison episodes and does not identify the causal origin of a comparison tendency.
A high comparison-orientation score may reflect increased social attention, uncertainty about self-assessment, performance concerns, interpersonal sensitivity, or a combination of traits and environments. Distinguishing these possibilities requires convergent data. Questionnaire responses are vulnerable to memory, self-presentation, and interpretation of what counts as comparison. Experience sampling, behavior during comparator-choice experiments, and passive observation of actual information exposure can supply complementary indicators. Wood (1996) warned that frequently used measures of comparison do not all capture the same operation. Researchers should examine predictive validity and whether instrument meaning remains stable across ages, languages, cultures, and platform settings rather than treating a successful initial validation as universal.
Individual differences are scientifically useful chiefly as potential moderators. A given feed or educational ranking may produce stronger short-term self-evaluation changes for some participants than for others, but the proposed moderator should be tested through prespecified interaction analyses with adequate power. Associations between comparison orientation and distress cannot establish that comparison causes distress: negative mood or uncertainty might increase the tendency to compare, and both may be influenced by broader social adversity. The research strategy is to combine trait measures with within-person observations and random variation in comparison information. Gibbons & Buunk (1999), Nesi & Prinstein (2015), and Meier & Johnson (2022) supply complementary psychometric, developmental, and critical-review perspectives on the need to distinguish person-level tendencies from episode-level effects.
Experimental Paradigms and Causal Identification
Controlled studies can randomize participants to upward, lateral, downward, or no-comparison standards while keeping the focal task and available information comparable. The outcome may be a domain-specific self-rating, immediate affect, goal selection, persistence, or observed performance. Such designs identify the effect of experimentally presented standards under study conditions. Vogel et al. (2014) combined an observational study of Facebook use and self-esteem with an experiment varying profiles containing comparatively favorable or unfavorable information. The two components answer different questions: the association in natural use is not causal, whereas the randomized profile contrast estimates an immediate effect of that particular exposure and accompanying comparative cues.
Strong laboratory research must separate several manipulations that commonly travel together. An upward target may also be more attractive, socially admired, competent, wealthy, or familiar. A negative emotional reaction might follow general exposure to a negative stimulus rather than comparison itself. Factorial designs can hold presentation style constant while varying standing, domain relevance, relational proximity, and the perceived attainability of success. Repeated measurement can test persistence beyond the immediate stimulus. A no-comparison condition allows researchers to distinguish the consequences of upward and downward standards from changes that would have occurred after any social information. Gerber et al. (2018) noted that the relatively limited availability of no-comparison controls constrained conclusions about absolute changes.
Experiments face their own validity problems. Demands about 'evaluating yourself' can create an artificial comparative frame, short interventions can exaggerate salience, and student samples restrict transportability. Failure to manipulate-check perceived direction may leave the causal treatment ambiguous. The study should specify sample size rationale, randomized assignment and exclusions, primary outcomes, timing, multiple-testing handling, and measurement reliability. Where possible, independent replications should test effects in different tasks and populations. Experimental causality should be stated narrowly: exposure caused a short-run shift in a measured outcome within the study's conditions. Wood (1996) and Gerber et al. (2018) together show why conceptually precise manipulation and design-aware evidence synthesis are indispensable.
Surveys, Longitudinal Designs, and Experience Sampling
Cross-sectional surveys reveal covariation among comparison habits, self-esteem, affect, and contexts, while leaving direction of influence unresolved. Chou & Edge (2012) studied how university students' Facebook use related to perceptions that other people had happier or better lives; its associations were measured in one population using self-report indicators and cannot identify which individuals would change after an intervention. Fardouly & Vartanian (2015) similarly examined associations involving appearance comparison and body image. Such studies are informative when the question concerns the distribution of beliefs and reported behavior. They become less informative when descriptive associations are presented as evidence of a universal psychological mechanism.
Longitudinal studies permit tests of temporal ordering, within-person change, and reciprocal associations when data quality and spacing are appropriate. In a study of adolescents, Nesi & Prinstein (2015) used measurements across two time points and identified associations between technology-based comparison or feedback seeking and depressive symptoms, with subgroup differences. Such evidence strengthens description of developmentally situated associations but cannot eliminate unmeasured confounding or guarantee that exposures preceded every reported process. Experience sampling improves temporal resolution by asking about recent comparison episodes, moods, and contexts repeatedly in everyday life. Its own challenges include reactivity, compliance, missed episodes, and dependence on retrospective identification of an event.
An improved analytic design would distinguish whether a person who compares more often than usual subsequently reports a deviation in self-evaluation from their own baseline, while also estimating stable differences between people. Person-mean centering and multilevel models can separate these components, but parameter interpretation depends on timing, measurement error, and omitted variables. The causal target should be declared before data collection: an immediate response to exposure, an accumulated consequence of repeated comparisons, or a reciprocal developmental process. Meier & Johnson (2022) emphasizes that digital comparison associations may be heterogeneous and reciprocal. Longitudinal research can test this complexity only when outcome and exposure are measured often enough to represent the postulated processes.
Psychometrics, Construct Validity, and Statistical Interpretation
The scientific validity of a comparison study begins with the correspondence between a concept and its indicator. Social comparison orientation, frequency of observed episodes, propensity to select upward targets, negative affect, self-esteem, body dissatisfaction, and academic self-concept are different constructs. Gibbons & Buunk (1999) measures a disposition to compare, while experiments such as Vogel et al. (2014) manipulate comparative information and measure state responses. Mixing these constructs in a pooled interpretation can manufacture apparent consistency or contradiction. Instrument items should be evaluated for internal structure, temporal stability when appropriate, measurement invariance, and relation to observable behavior. A domain-specific outcome should not be silently relabeled as global self-worth.
Self-reported comparison direction also demands caution. Participants may remember upward episodes that were emotionally intense, recognize only deliberate evaluation as comparison, or infer a direction after describing the emotional result. An observed rank offers another indicator but does not reveal subjective meaning; a student ranked twentieth can compare favorably with a particular peer while comparing unfavorably with a class average. Researchers can combine choice tasks, recorded exposures, self-reports, and direct observation while treating disagreements among measures as informative. Wood (1996) remains an important conceptual review of how different measurement procedures support different inferences about comparative processing.
Statistical evaluation should report effect sizes and uncertainty intervals as well as significance tests, distinguish planned hypotheses from exploratory analyses, and consider dependence of observations collected within schools, networks, or repeated episodes. Mediation models are particularly vulnerable to overinterpretation: an indirect association through comparison reports does not by itself establish a causal pathway when the mediator is unrandomized or confounded. The many plausible analytic choices in technology research make transparent specifications important. Orben & Przybylski (2019) demonstrates how specification-curve analysis can reassess broad digital-technology associations, and Meier & Johnson (2022) applies a critical lens to claims about social comparison, envy, and well-being. The methodological lesson is to match the strength of each claim to its design.
Meta-Analytic Evidence and the Selection–Reaction Puzzle
The review by Gerber et al. (2018) is a central cumulative assessment because it separates two outcomes often mistakenly combined: which target people choose and how they react after a comparison. In binary choice studies without threat, upward target selection was common; giving a lateral alternative reduced differentiation in preferences. Reaction analyses, however, supported contrast more consistently than assimilation, particularly for ability estimates. A person may therefore pursue information from a high-performing peer even when that encounter makes their current performance look relatively poor. This is a genuine theoretical problem for any model that assumes target choice always maximizes immediate self-esteem.
The meta-analysis also provides a lesson in interpreting pooled evidence. The meaning of a summary effect depends on comparison tasks, domains, response measures, and whether baseline or no-comparison conditions exist. Gerber et al. (2018) evaluated moderators and publication bias and identified the limited number of no-comparison control designs among the literature's shortcomings. Directional effects should not be turned into prescriptions or universal psychological laws. A contrast effect may be strong for a novel ability estimate while an aspiration measure is unaffected, and the same participant may later change actual behavior. Sampling and study selection remain relevant even when an effect is statistically precise within the analyzed corpus.
A mature synthesis integrates motivation and processing. Taylor & Lobel (1989) shows that evaluation and information seeking can diverge under threat; Mussweiler (2003) identifies a cognitive route through similarity or difference testing; Lockwood & Kunda (1997) emphasizes attainability in responses to outstanding role models. Together these findings explain why the terms upward, downward, and lateral identify a feature of the standard rather than a deterministic emotional outcome. Further progress requires studies that jointly randomize target opportunity, measure the information activated during appraisal, distinguish immediate from delayed outcomes, and replicate across populations. The resulting model should predict choice and reaction separately, then specify when their coupling is expected to be strong or weak.
Academic Comparison and the Big-Fish–Little-Pond Effect
Schools create especially visible comparison ecologies because achievement is measured repeatedly and shared with peers, teachers, and families. Marsh & Hau (2003) tested the big-fish–little-pond effect in samples from 26 countries. The core claim is that students of comparable individual achievement can report lower academic self-concept when embedded in higher-achieving school contexts. The phenomenon concerns the relation between individual standing and contextual achievement rather than simple differences in actual learning. Its interpretation requires control for a student's own performance and appropriate modeling of clustered educational data. School-average achievement may supply a locally salient standard, but its observed relation to self-concept does not imply that selective schools are uniformly harmful in every respect.
The review by Dijkstra et al. (2008) describes motives, comparison dimensions, direction, and consequences in educational settings. Students frequently compare upward with somewhat similar, higher-performing peers. Such comparisons may support effort while also provoking unfavorable feelings or lower academic self-concept. This combination mirrors the general selection–reaction distinction. Grades, public rankings, classroom tracking, competition, and peer familiarity can each determine which standard becomes salient. A fair evaluation of an educational intervention must measure learning, belonging, performance goals, stress, and self-concept separately rather than equating positive academic achievement with positive psychological experience.
Educational comparisons have implications for design of feedback and classroom environments. Reporting a single rank may communicate something different from giving individual progress, criterion-referenced mastery feedback, or constructive peer models. The appropriate research comparison is between explicitly defined feedback conditions with repeated assessments, while accounting for teacher practices and student background. Observational cross-national regularities are valuable tests of generalization but cannot substitute for direct tests of a proposed intervention. Marsh & Hau (2003) provides evidence about broad contextual patterns; Dijkstra et al. (2008) documents considerable variation in educational studies and their outcomes. Both are necessary for explaining how social structure and individual appraisal intersect in the classroom.
Development, Adolescence, and Life-Course Change
Social comparison changes as people acquire more complex abilities to represent peers, infer social standing, and differentiate performance domains. Classrooms, families, friendships, and extracurricular activities supply developmentally specific standards. Dijkstra et al. (2008) reviews educational studies in which pupils' motives and targets differ by task, achievement context, and age. Developmental explanation must distinguish a change in comparison competence from a change in opportunities to compare. A younger child and an adolescent may face different ranking regimes even when their underlying sensitivity to competence feedback is similar. Longitudinal measurement is essential for separating maturation, institutional transitions, and cohort-specific exposure to digital platforms.
Adolescence intensifies the relevance of peer groups and social visibility in many environments, but claims about a uniform developmental vulnerability require evidence. Nesi & Prinstein (2015) studied adolescent social-media comparison and feedback seeking in relation to depressive symptoms, reporting moderation by gender and peer popularity in the observed sample. These results concern a specific set of self-reported behaviors and outcomes, not an inevitable clinical consequence of online life. Valkenburg et al. (2006) also found that positive and negative feedback received through early networking sites related differently to adolescents' social self-esteem and well-being. Taken together, the studies show why feedback quality, perceived audience, and interpersonal context should be represented alongside comparison frequency.
A life-course account also asks how standards shift with employment, family change, health, aging, or retirement. A person may compare income during early career, parenting decisions in middle adulthood, and functional independence during later life. These are plausible domain changes, but their specific developmental pattern must be demonstrated rather than deduced from common experience. The research challenge is to retain measurement comparability while allowing the relevant standard to vary. Wood (1989) describes the multidimensional motivational architecture that such work must capture, and Taylor & Lobel (1989) demonstrates the distinction between evaluative reassurance and information seeking under threat. Future cohorts should be followed across transitions to test whether particular comparisons predict later self-evaluation or adapt to circumstances.
Culture, Institutions, and Inequality in Comparison
Comparison is shaped by cultural understandings of achievement, obligation, fairness, modesty, and belonging. A ranking can be interpreted as personal merit, collective family reputation, or institutional allocation of opportunity depending on local practices. Yet broad labels such as individualist and collectivist cannot substitute for measured mechanisms within populations. Gibbons & Buunk (1999) constructed its comparison-orientation scale with American and Dutch participants, illustrating both the possibility of cross-context measurement and the need to verify equivalence. Marsh & Hau (2003) reported the big-fish–little-pond association in educational data from 26 countries, offering stronger cross-context evidence for a specified academic mechanism than a single-site survey could provide.
Institutions structure which social differences become publicly legible. Exams, salaries, occupational titles, popularity rankings, athletic records, and social-media metrics are not interchangeable standards. Some are tied to objective performance criteria; others reflect resource distribution, selection into opportunity, or culturally preferred visibility. A comparison can therefore transmit information about inequality without accurately identifying individual ability. Studies should record whether targets are objectively matched on relevant opportunities and whether perceived fairness mediates evaluative reactions. This prevents the observer's relative judgment from being treated as a transparent measure of underlying merit. Wood (1996) emphasizes the conceptual specificity required to interpret comparative ratings.
The intersection of culture and inequality also matters for external validity. A student in a selective school, a worker in a precarious labor market, and an online creator evaluated through engagement metrics encounter structurally different comparison sets. Evidence from one context cannot simply be transplanted into another. Cross-cultural designs should test measurement invariance, sampling coverage, translated item meaning, within-country diversity, and institutional-level variation. Research should also consider who is excluded from visible rankings and whether downward targets are socially stigmatized. Marsh & Hau (2003) demonstrates the utility of hierarchical cross-national analysis, while Gerber et al. (2018) cautions that the design of comparator alternatives changes the apparent distribution of choices.
Social Media, Algorithmic Exposure, and Self-Evaluation
Social media expands the number, recency, and visual richness of available comparison targets. Curated profiles can display achievements, relationships, appearance, and leisure while omitting ordinary constraints. These features may alter perceived normative standing through selective presentation and repeated exposure. Chou & Edge (2012) found associations between Facebook involvement and beliefs that others enjoyed happier or better lives in a university sample. Vogel et al. (2014) reported both survey associations and experimental differences in immediate self-evaluation after viewing relatively high- versus low-standing profiles. The evidence supports studying comparisons as a mechanism, but it does not show that all platform use is harmful or that a particular design feature has the same effect across users.
Digital behavior needs finer categories than a binary active-versus-passive distinction. Viewing a supportive friend's update, comparing oneself with a celebrity, seeking reassurance after a setback, and watching instructional content can all count as consumption while providing different social information. Verduyn et al. (2017) reviewed evidence associating passive social-network use with poorer well-being and active use with some social-connection benefits, although the latter pattern was less robust. Meier & Johnson (2022) subsequently emphasized newer results indicating heterogeneity, reciprocal associations, and sometimes positive links between comparison and well-being. The scientific synthesis therefore treats activity labels as preliminary proxies. The relevant exposure is the particular comparison opportunity, its meaning, and its relation to the person's prior state.
Algorithmic ranking and recommendation may make certain standards unusually available, but an empirical account needs data about the actual content shown to users and the reasons for its selection. Researchers should distinguish the effect of seeing high-performing peers from an effect of platform-specific visual formatting or metrics. Randomized feed manipulation, preregistered natural experiments, platform-provided exposure records, and ethically collected momentary assessments can address parts of this problem. Privacy limits and selection into platforms constrain inference. Orben & Przybylski (2019) shows that broad digital-technology effects may be small and sensitive to analysis decisions; it does not isolate social comparison itself. This separation prevents overall screen time from standing in for the distinct mechanism under investigation.
Appearance Comparison, Body Image, and Self-Worth
Appearance constitutes a highly salient comparison domain because images are easily encountered and are routinely evaluated against culturally defined standards. A photograph can become an upward standard even when its production conditions, editing, and selection are unknown to the viewer. Fardouly et al. (2015) experimentally assigned young women to brief exposure to Facebook, an appearance-related magazine site, or a neutral control; Facebook exposure was associated with worse mood than the neutral condition, with appearance-comparison tendency moderating some facial or appearance-discrepancy outcomes. The experiment supports claims about a short-term effect in its sampled population and task. It does not establish long-term body-image change, and the conditions varied in several content characteristics beyond simple target rank.
In a related correlational study, Fardouly & Vartanian (2015) reported that negative appearance comparisons statistically mediated the link between Facebook use and body-image concerns among young women. Such mediation maps a plausible explanatory pathway but remains vulnerable to common-method variance and reverse causation. Existing dissatisfaction may increase exposure to appearance-focused content and make appearance differences more noticeable. The observer may also compare against a digitally altered image without recognizing its construction. This creates a difference between objectively presented material and subjectively inferred standards. Rigorous analysis must record stimulus properties and individual interpretation rather than attribute the full effect to the platform brand.
Body-image outcomes should be distinguished from global self-esteem, mood, disordered-eating symptoms, and general psychological distress. A brief appearance discrepancy may cause transient concern without implying a psychiatric diagnosis; repeated distress and impairment require separate validated assessments. The same image may be used for inspiration, aesthetic appreciation, comparison, or identity expression. Meier & Johnson (2022) warns against assuming a uniform negative pathway from digital exposure through comparison and envy to poor well-being. To advance the evidence, multiwave studies could test prior appearance concerns, actual exposure, momentary comparison episodes, and later outcomes, while experiments isolate image editing, target similarity, and feedback metrics. Clinical implications must follow established clinical evidence rather than being inferred from nonclinical social-psychological experiments.
Emotional Consequences, Envy, Admiration, and Agency
Comparative self-evaluation can produce pride, relief, embarrassment, admiration, envy, anxiety, hope, or indifference. These emotions are differentiated by interpretation, not exclusively by direction. An upward target can elicit admiration when success appears deserved and attainable, or envy when the gap is self-relevant and threatening. A downward target can elicit relief or concern if the observer imagines a similar future. Lockwood & Kunda (1997) demonstrates how personal relevance and attainability shape responses to high-achieving role models. Tesser (1988) adds that relational closeness and domain importance can reverse the implications of another person's success. A mechanism-focused study needs validated measures of distinct emotion categories and the antecedent appraisals that organize them.
Emotional responses can regulate behavior. A contrastive loss in self-confidence may be followed by disengagement, or it may lead to practice when a clear path to improvement is available. Admiration may encourage learning without elevating an immediate ability rating. Such cases underscore the difference between state self-evaluation and performance outcomes. Gerber et al. (2018) documents the prevalence of contrastive reactions in experimental samples, but its scope does not authorize an assumption about persistent motivation. Dijkstra et al. (2008) suggests that academically relevant upward comparisons can support performance-related striving while also evoking unpleasant feelings. Researchers should measure changes in effort and task selection independently from postexposure mood.
Digital research often treats envy as an intermediary between comparison and distress. Appel et al. (2016) reviewed relationships among Facebook use, comparison, envy, and depression, while Meier & Johnson (2022) describes evidence that is conditional, mixed, and sometimes reciprocal. Because envy and dissatisfaction may influence which content people seek, a one-time survey cannot order the proposed pathway causally. Experimental manipulation of comparison material with subsequent affective measurement can test short-term effects; prospective designs can investigate stability, while carefully controlled interventions are needed to estimate benefits from changing exposure or appraisal. A sound synthesis recognizes emotions as components of an evolving process involving targets, goals, self-concepts, and environmental conditions.
Human–Artificial Comparison in the Artificial Era
The Artificial Era introduces new comparison standards in which people encounter machine-generated achievements, automated rankings, synthetic social representations, and artificial interlocutors. A human observer may compare writing speed, knowledge recall, creativity, conversational responsiveness, or task performance with an artificial system. The psychological comparison remains an empirically testable relation between a human self-evaluation and perceived attributes of a comparator. The central design question is whether the artificial target is construed as a social agent, a tool, an institutional benchmark, or a representation of human performance. Each construal predicts a different informational value and potential reaction. These propositions extend the established comparison framework rather than claiming that equivalent mechanisms have already been established across biological and artificial systems.
An experimental program can separately manipulate comparator provenance, performance level, anthropomorphic presentation, task relevance, and perceived attainability. Human and artificial comparator conditions should be matched on displayed achievements before investigators attribute an effect to artificial identity. Outcome measures can include competence estimates, aspiration, task persistence, trust calibration, willingness to collaborate, and perceived agency. The mechanisms analyzed by Mussweiler (2003), Lockwood & Kunda (1997), and Tesser (1988) provide hypotheses about similarity testing, possible-self identification, and relational significance. Yet it remains an open empirical question whether people experience closeness to artificial systems in the same way that the self-evaluation maintenance model defines closeness to human relations.
Comparisons involving artificial systems also raise measurement questions. A machine's performance is sensitive to software version, prompt conditions, access to tools, and benchmark design, while human judgments may be influenced by disclosure and presentation. Claims about artificial experience or psychological interiority require their own operational definitions and evidence. The immediate subject of this dossier is how humans evaluate themselves relative to perceived artificial standards, with the reciprocal psychological organization of artificial systems belonging to adjacent research fields. This distinction supports collaboration among Social Psychology, Media and Technology Psychology, and Psychology of Human–Artificial Relations. Within the Artificial Era, comparison becomes a measurable element of changing human environments and requires transparent rather than purely metaphorical empirical study.
Critical Scientific Discussion and Competing Explanations
The principal achievement of social comparison research is to show that relative social information can shape self-appraisals, with effects mediated or moderated by the comparison set, domain, perceived similarity, personal relevance, and interpretive processing. Important uncertainties concern the frequency of spontaneous comparisons, the proportion that lead to lasting change, and the settings in which immediate contrast can coexist with later improvement. Gerber et al. (2018) provides relatively strong pooled evidence for particular experimental selection and reaction patterns, while Meier & Johnson (2022) presents a mixed account of digital comparison and well-being. These studies address different estimands and therefore should not be forced into a single universal answer about whether social comparison is beneficial.
Several rival mechanisms can produce similar observations. Low self-esteem may increase upward comparison rather than result from it; envy may arise from perceived injustice rather than relative standing; an artificial feed may change mood through emotionally charged content without activating self-evaluation; a selective school's resources may improve achievement while its comparison context lowers academic self-concept. To discriminate these possibilities, researchers need clear baseline measures, exogenous changes in exposure, careful mediator timing, and analysis of sample selection. Wood (1996) emphasizes whether tasks genuinely capture comparative processing, while Orben & Przybylski (2019) shows the importance of analytic flexibility in broad digital-well-being associations. Published correlations, even with impressive sample sizes, do not resolve a causal process by themselves.
Several recurrent problems remain across domains: convenience sampling, short follow-up, inconsistent labels, post hoc subgroup interpretation, common-source measurement, insufficient no-comparison controls, and publication or selection bias. Researchers should register primary questions, publish detailed stimulus materials, report complete model specifications, and include negative and null results. Replication should test whether a predicted effect survives changes in language, culture, socioeconomic setting, life stage, and comparison medium. Scientific progress will come from specifying when a given type of standard produces a given appraisal under definable conditions. A credible theory should also predict occasions on which comparison exposure produces no measurable effect, and it should reconcile the divergence between target selection and immediate reaction identified by Gerber et al. (2018).
Conclusion
The empirical and theoretical record supports a multi-component account: social comparison is an encounter with a possible reference standard, interpreted in relation to a particular self-relevant dimension and embedded in a structured environment. It can serve accuracy, aspiration, reassurance, or social belonging, and the motive behind selecting a standard may differ from the effect of encountering it. Festinger (1954) identified the fundamental dependence of appraisal on social standards; Wood (1989) expanded motives and forms of comparison; Tesser (1988) located self-maintenance within relationships; Mussweiler (2003) specified information-processing conditions; and Gerber et al. (2018) tested accumulated predictions across experimental studies. Together these sources establish the field's core explanatory architecture while retaining its unresolved tensions.
The strongest warranted conclusions are conditional. Experimental comparisons can change immediate self-evaluation, and contrast frequently dominates responses under studied conditions. School contexts provide evidence of contextual effects on academic self-concept, while social-media studies demonstrate that platform-mediated information can enter self-assessment and affective experience. Marsh & Hau (2003), Vogel et al. (2014), and Fardouly et al. (2015) document different parts of this evidence. The totality of social-media exposure cannot be treated as a direct measure of social comparison, and negative cross-sectional associations cannot establish lasting harm. Meier & Johnson (2022) underscores mixed, heterogeneous, and reciprocal outcomes. Research quality improves when comparison direction, interpretation, and time scale are specified separately.
The next stage of inquiry requires preregistered comparison-choice experiments with neutral controls, dense longitudinal measures of actual episodes, valid cross-cultural instruments, transparent exposure data, and explicit models of individual and institutional selection. A major theoretical target is to explain why upward comparison is often chosen even when immediate contrast is likely. The expansion of artificial performance standards creates further tests of whether perceived source, attainability, and agency transform evaluation dynamics. The established scientific conclusion is that comparative self-evaluation is a context-dependent psychological process with measurable consequences for judgment, motivation, and affect. Its future investigation should preserve construct precision and the separation of observed effects from speculative explanation.
References
- Appel, H., Gerlach, A. L., & Crusius, J. (2016). The interplay between Facebook use, social comparison, envy, and depression. Current Opinion in Psychology, 9, 44–49. https://doi.org/10.1016/j.copsyc.2015.10.006
- 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
- Collins, R. L. (1996). For better or worse: The impact of upward social comparison on self-evaluations. Psychological Bulletin, 119(1), 51–69. https://doi.org/10.1037/0033-2909.119.1.51
- Dijkstra, P., Kuyper, H., van der Werf, G., Buunk, A. P., & van der Zee, Y. G. (2008). Social comparison in the classroom: A review. Review of Educational Research, 78(4), 828–879. https://doi.org/10.3102/0034654308321210
- Fardouly, J., Diedrichs, P. C., Vartanian, L. R., & Halliwell, E. (2015). Social comparisons on social media: The impact of Facebook on young women's body image concerns and mood. Body Image, 13, 38–45. https://doi.org/10.1016/j.bodyim.2014.12.002
- Fardouly, J., & Vartanian, L. R. (2015). Negative comparisons about one's appearance mediate the relationship between Facebook usage and body image concerns. Body Image, 12, 82–88. https://doi.org/10.1016/j.bodyim.2014.10.004
- Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7(2), 117–140. https://doi.org/10.1177/001872675400700202
- Gerber, J. P., Wheeler, L., & Suls, J. (2018). A social comparison theory meta-analysis 60+ years on. Psychological Bulletin, 144(2), 177–197. https://doi.org/10.1037/bul0000127
- Gibbons, F. X., & Buunk, B. P. (1999). Individual differences in social comparison: Development of a scale of social comparison orientation. Journal of Personality and Social Psychology, 76(1), 129–142. https://doi.org/10.1037/0022-3514.76.1.129
- Gilbert, D. T., Giesler, R. B., & Morris, K. A. (1995). When comparisons arise. Journal of Personality and Social Psychology, 69(2), 227–236. https://doi.org/10.1037/0022-3514.69.2.227
- Lockwood, P., & Kunda, Z. (1997). Superstars and me: Predicting the impact of role models on the self. Journal of Personality and Social Psychology, 73(1), 91–103. https://doi.org/10.1037/0022-3514.73.1.91
- Marsh, H. W., & Hau, K.-T. (2003). Big-fish-little-pond effect on academic self-concept: A cross-cultural (26-country) test of the negative effects of academically selective schools. American Psychologist, 58(5), 364–376. https://doi.org/10.1037/0003-066X.58.5.364
- Meier, A., & Johnson, B. K. (2022). Social comparison and envy on social media: A critical review. Current Opinion in Psychology, 45, Article 101302. https://doi.org/10.1016/j.copsyc.2022.101302
- Mussweiler, T. (2003). Comparison processes in social judgment: Mechanisms and consequences. Psychological Review, 110(3), 472–489. https://doi.org/10.1037/0033-295X.110.3.472
- Mussweiler, T., Rüter, K., & Epstude, K. (2004). The ups and downs of social comparison: Mechanisms of assimilation and contrast. Journal of Personality and Social Psychology, 87(6), 832–844. https://doi.org/10.1037/0022-3514.87.6.832
- Nesi, J., & Prinstein, M. J. (2015). Using social media for social comparison and feedback-seeking: Gender and popularity moderate associations with depressive symptoms. Journal of Abnormal Child Psychology, 43(8), 1427–1438. https://doi.org/10.1007/s10802-015-0020-0
- Orben, A., & Przybylski, A. K. (2019). The association between adolescent well-being and digital technology use. Nature Human Behaviour, 3(2), 173–182. https://doi.org/10.1038/s41562-018-0506-1
- Taylor, S. E., & Lobel, M. (1989). Social comparison activity under threat: Downward evaluation and upward contacts. Psychological Review, 96(4), 569–575. https://doi.org/10.1037/0033-295X.96.4.569
- Tesser, A. (1988). Toward a self-evaluation maintenance model of social behavior. Advances in Experimental Social Psychology, 21, 181–227. https://doi.org/10.1016/S0065-2601(08)60227-0
- 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
- Verduyn, P., Ybarra, O., Résibois, M., Jonides, J., & Kross, E. (2017). Do social network sites enhance or undermine subjective well-being? A critical review. Social Issues and Policy Review, 11(1), 274–302. https://doi.org/10.1111/sipr.12033
- 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
- Wills, T. A. (1981). Downward comparison principles in social psychology. Psychological Bulletin, 90(2), 245–271. https://doi.org/10.1037/0033-2909.90.2.245
- Wood, J. V. (1989). Theory and research concerning social comparisons of personal attributes. Psychological Bulletin, 106(2), 231–248. https://doi.org/10.1037/0033-2909.106.2.231
- Wood, J. V. (1996). What is social comparison and how should we study it? Personality and Social Psychology Bulletin, 22(5), 520–537. https://doi.org/10.1177/0146167296225009
Social Comparison as a Psychological Process
Social comparison is the cognitive and motivational use of information about another person's attributes, performance, circumstances, or opinions to locate, interpret, or regulate one's own standing. The comparison may be deliberate, as when a student studies classmates' scores, or relatively spontaneous, as when a colleague's success changes an assessment of one's competence. Its object is a relation between a perceived self and a selected or encountered standard along a particular dimension. This definition locates the scientific phenomenon in target identification, attribute matching, interpretation, and consequences for self-evaluation. It excludes treating every experience of social influence as a comparison. The distinction matters because being praised, following a rule, or noticing someone's behavior can change the self without producing an explicit comparative judgment. Conceptual refinements by Wood (1989) and Wood (1996) make the operations of comparison, its antecedents, and its possible effects independently investigable.
Self-evaluation is the judgment a person makes about their own ability, worth, adequacy, progress, or position on a dimension. A comparison can alter a transient judgment of mathematical skill, a broader academic self-concept, or momentary affect, yet these are distinguishable outcomes. Self-esteem concerns a more global evaluation of personal worth; perceived competence concerns a specific capability; social status concerns a relational or institutional position; aspiration concerns a desired future level. The same comparison episode may influence these variables differently and at different times. For example, observing a proficient peer can reduce today's relative competence rating while increasing tomorrow's willingness to practice. The central task of this dossier is to explain how this divergence arises, rather than to assume that comparison has a single affective direction. The accumulated meta-analytic evidence in Gerber et al. (2018) makes this distinction between target selection and postcomparison reaction especially consequential.
The appropriate unit of analysis is an episode embedded in a history of repeated encounters. A target's apparent success becomes informative only when the observer notices it, attributes it to relevant causes, evaluates whether the domain matters, and infers what the observation implies for the self. Social proximity, institutional ranking, cultural norms, baseline mood, goals, and platform design each alter these steps. Observable outcomes therefore range from comparison choices and reaction times to self-ratings, affective reports, behavioral persistence, and changes in interpersonal behavior. The scientific program examines their joint organization through controlled experiments, repeated measures, school and workplace field designs, and longitudinal research. It must preserve the boundary between measured effects and claims about why those effects occurred. Gilbert et al. (1995) demonstrates why even knowing that a target is diagnostically inappropriate does not necessarily prevent an immediate comparative response.