Leader-Member Exchange Scale (LMX7)

Abstract

The Leader-Member Exchange Scale (LMX7) is the preeminent psychometric instrument used in industrial and organizational psychology to assess the quality of the reciprocal, dyadic relationship established between a supervisor and an employee. Initially systematized in its definitive 7-item unidimensional format by George B. Graen and Mary Uhl-Bien in 1995, the instrument represents the operational culmination of decades of research evolving from the Vertical Dyad Linkage (VDL) model toward comprehensive relational exchange paradigms. Grounded within social exchange theory and role-making theory, the LMX7 departs from traditional “average leadership style” approaches by measuring the idiosyncratic degree of mutual trust, respect, psychological support, loyalty, and reciprocal obligation that characterizes an individual supervisor-subordinate pairing.

The scale consists of seven items evaluated on specialized 5-point Likert-type scales, where response anchors are tailored to the specific evaluative focus of each question (ranging from frequency and degree of understanding to personal advocacy and global relational effectiveness). Psychometric validations across diverse cultures—exemplified by Schyns and Paul’s influential German adaptation—demonstrate remarkable structural stability, high internal consistency (with Cronbach’s alpha typically ranging between .89 and .93, and reported at .92 in primary validation cohorts), and strong factor saturation, where a single latent factor routinely accounts for upwards of 60% to 68% of total variance. The instrument exhibits substantial convergent validity with measures of job satisfaction, affective organizational commitment, leader consideration, and perceived organizational support, while consistently predicting vital behavioral criteria including organizational citizenship behavior, turnover intentions, and objective performance ratings. This article provides a comprehensive academic review of the LMX7, detailing its theoretical architecture, psychometric parameters, factor structure, scoring conventions, and application protocols in empirical research and organizational development.

Keywords

Leader-Member Exchange, LMX-7, dyadic leadership, relational leadership, supervisor-subordinate relationship, social exchange theory, psychometrics, organizational behavior, job performance, leadership measurement

Authors

The definitive 7-item iteration of the Leader-Member Exchange Scale was formalized by George B. Graen (Center for Strategic Leadership, University of Cincinnati) and Mary Uhl-Bien (Department of Management, University of Nebraska-Lincoln; currently BNSF Railway Endowed Professor of Leadership at Texas Christian University) in their seminal 1995 treatise published in The Leadership Quarterly. Their formulation consolidated several earlier, variable-length versions (ranging from 4 to 14 items) developed by Graen and colleagues across the 1970s and 1980s into an optimized, robust measurement standard.

The psychometric translation and adaptation referenced prominently in European organizational research was spearheaded by Dr. Birgit Schyns (currently Professor of Organizational Behavior at NEOMA Business School, France; formerly of the University of Leipzig and University of Portsmouth) and Dipl.-Psych. Tina Paul (Department of Work and Organizational Psychology, University of Leipzig, Germany). Inquiries regarding the cross-cultural adaptation and research datasets may be addressed to Dr. Birgit Schyns (E-Mail: [email protected]) or Tina Paul (E-Mail: [email protected]).

Purpose

The primary purpose of the Leader-Member Exchange Scale (LMX7) is to quantify the subjective relational quality of the vertical dyad linking a specific leader and an individual member. Historically, dominant leadership paradigms—such as the Ohio State leadership studies, Fiedler’s contingency model, and early transformational-transactional leadership models—assumed that leaders operated via an “Average Leadership Style” (ALS). Under the ALS assumption, a manager displays a functionally uniform behavioral orientation (e.g., initiating structure vs. consideration) toward all subordinates within a collective workgroup. The LMX paradigm directly invalidates this homogeneity assumption by establishing that leaders differentiate their behaviors, forming distinct, unique relational partnerships with different direct reports within the exact same structural unit.

In applied organizational diagnostics and clinical-organizational consulting, the LMX7 fulfills multiple diagnostic functions:

• Identification of In-Group vs. Out-Group Segmentation: The scale enables practitioners to evaluate the degree to which a workgroup is internally fragmented. High-scoring subordinates (“in-group” members) experience high levels of mutual influence, informal backing, broad negotiating latitude, and open communication channels. Conversely, low-scoring subordinates (“out-group” members) endure relationships restricted to formal contractual obligations, economic transactional exchange, rigid role expectations, and limited supervisory warmth.
• Diagnostic Assessment of Psychological Safety and Well-Being: Low LMX scores routinely serve as primary early indicators of subordinate strain, organizational cynicism, psychological withdrawal, perceived injustice, and heightened turnover intention. Clinically oriented organizational psychologists employ the tool to uncover root causes of interpersonal workplace conflict, localized demoralization, and chronic occupational stress.
• Evaluation of Leadership Interventions: In leadership coaching and management training regimens, the LMX7 provides a sensitive pre-test/post-test metric. Interventions focused on active listening, empathetic mentoring, supportive communication, and delegation can be quantitatively audited to determine whether interventions successfully elevate relational quality across previously alienated team members.
• Multi-Source Dyadic Alignment: Because the LMX7 can be administered symmetrically to both members (subordinates evaluating leaders) and leaders (leaders evaluating members), researchers utilize the tool to compute relational convergence, perceptual consensus, and self-other agreement indices, illuminating blind spots in managerial self-awareness.

Psychological Construct

The construct of Leader-Member Exchange (LMX) describes the quality of the dual, reciprocal working relationship that develops over time between a leader and a follower through repeated iterative cycles of social, cognitive, and material exchanges. Rather than conceptualizing leadership as an inherent personal trait (e.g., charismatic personality) or as a static set of supervisory behaviors, LMX posits leadership as a relational property embedded directly within the dyadic space. As detailed by Graen and Scandura (1987) as well as Graen and Uhl-Bien (1995), mature, high-quality LMX is characterized by psychological intimacy, mutual respect, unswerving trust, interpersonal loyalty, reciprocal professional obligation, and personal support.

Although the LMX7 is psychometrically operationalized as a unidimensional instrument, the underlying theoretical construct spans several interrelated qualitative dimensions:

1. Mutual Trust and Confidence

High-quality exchange requires deep mutual trust regarding professional judgment and behavioral integrity. In an established partnership, the leader trusts the employee to act autonomously without surveillance, while the employee trusts the leader to make decisions that safeguard the employee’s professional standing and career trajectories. Item 6 of the scale directly gauges this component by probing the subordinate’s willingness to openly defend and justify the leader’s decisions in their absence, serving as a behavioral proxy for profound interpersonal trust.

2. Affective Regard and Mutual Respect

LMX relationships transcend impersonal economic arrangements to cultivate genuine affective regard and professional esteem. High LMX partners value one another’s capabilities, perceive complementary competencies, and exhibit authentic interpersonal goodwill. Item 3 captures the leader’s cognitive recognition and validation of the subordinate’s latent potential, reflecting the emotional and developmental investment the leader places in the member’s growth.

3. Reciprocal Loyalty and Advocacy

Loyalty represents a steadfast public and private alignment between leader and member. In high-LMX dyads, leaders act as organizational sponsors, expending political capital to protect the member from institutional fallout, external scapegoating, or administrative obstacles. Item 4 and Item 5 evaluate this dynamic by examining the leader’s willingness to leverage position power to resolve the member’s occupational crises and, crucially, to “bail out” the member at the leader’s own personal or organizational expense.

4. Role Understanding and Empathic Comprehension

Effective dyadic functioning requires nuanced, shared mental models of work tasks, stressors, and emotional needs. High LMX is characterized by frequent, candid communication through which leaders develop deep awareness of their direct reports’ specific task difficulties, personal constraints, and subjective needs. Item 2 operationalizes this cognitive-affective understanding of workplace challenges.

5. Evaluative Transparency and Feedback Clarity

A fundamental anxiety within organizational hierarchies is ambiguity regarding performance appraisal. In low-quality dyads, feedback is scarce, formalistic, or punitively delivered during annual performance reviews. In contrast, high-quality dyads maintain perpetual feedback loops characterized by radical transparency. Item 1 evaluates whether the member has clear, continuous cognitive awareness of where they stand evaluatively with their direct supervisor.

Theoretical Framework

The theoretical architecture undergirding the LMX7 scale rests upon the convergence of three foundational organizational and social-psychological theories: Social Exchange Theory, Role-Making Theory, and the Norm of Reciprocity.

1. Social Exchange Theory and the Norm of Reciprocity

Formulated by Peter Blau (1964) and expanded by Alvin Gouldner (1960), social exchange theory differentiates between strictly economic exchanges and social exchanges. Economic exchanges involve tangible, contractually specified resources (e.g., wage labor for contractual hours), carry minimal trust, require immediate or short-term repayment, and are formally enforceable. In contrast, social exchanges involve non-contractual, socio-emotional resources (e.g., emotional support, prestige, discretionary effort, strategic latitude, loyalty). Social exchanges lack guaranteed repayment schedules, relying entirely on the universal norm of reciprocity—the internalized moral imperative that individuals must assist, and refrain from harming, those who have provided support.

The LMX framework views the leader-member interface as an evolving social exchange matrix. In out-group dyads, interactions remain locked in cold economic exchange: the employee performs only the tasks explicitly delineated in the job description, and the manager provides only standard institutional compensation. In in-group dyads, the leader provides extra-role resources (mentoring, high-visibility assignments, schedule flexibility, political protection), which the employee reciprocates through extra-role organizational citizenship behaviors (OCBs), exceptional effort, emotional loyalty, and proactive problem-solving.

2. Role-Making Theory

The developmental trajectory of LMX is systematically mapped across Graen and Uhl-Bien’s (1995) Life Cycle Model of Leadership Making, which operates via three sequential phases grounded in role theory:

• Phase 1: The Stranger Phase (Role-Taking): The interaction begins at an institutional baseline governed by formal rules and prescribed job descriptions. Motives are primarily self-interested; interactions are transactional and scripted.
• Phase 2: The Acquaintance Phase (Role-Making): The dyad engages in behavioral testing. The leader or member offers an exchange that exceeds formal role parameters (e.g., the leader delegates an ambiguous project; the member works late voluntarily). If the offer is reciprocated with mutual competence and loyalty, the psychological contract expands, and social exchange mechanisms begin to eclipse formal transactional rules.
• Phase 3: The Mature Partnership Phase (Role-Routinization): The relationship reaches maturity. Exchanges are characterized by high mutual trust, respect, and emotional commitment. Group-interest and relational stewardship completely supersede narrow individual self-interest. Subordinates transform into trusted collaborators and confidants.

Validity

The LMX7 is one of the most rigorously validated psychometric measures within the organizational sciences. Extensive empirical studies have affirmed its construct, convergent, discriminant, and predictive (criterion-related) validity.

1. Convergent and Nomological Validity

Nomological validity has been corroborated through dense correlation networks with theoretically adjacent psychological constructs. In their classic meta-analyses, Gerstner and Day (1997) as well as Dulebohn, Bommer, Liden, Brouer, and Ferris (2012) analyzed hundreds of empirical studies, demonstrating that LMX7 scores converge robustly with:

• Perceived Organizational Support (POS): Mean corrected correlations ranging between $rho = .55$ and $.65$, illustrating that employees view their direct supervisor as the personal embodiment and executive agent of the broader organization.
• Transformational Leadership: Substantial convergence ($rho = .60$ to $.73$), demonstrating that while transformational leadership reflects behavioral styles aimed at collective inspiration, it naturally drives high relational exchange quality at the dyadic level.
• Leader Consideration: In the German validation study conducted by Schyns and Paul (1999), LMX7 scores correlated intensely with dimensions of the German Supervisor Behavior Questionnaire (Fragebogen zur Vorgesetzten-Verhaltens-Beschreibung, FVVB), including Friendly Attention ($r = .75, p < .01$), Stimulating Activity ($r = .78, p < .01$), and Co-determination/Empowerment ($r = .73, p < .01$).

2. Discriminant Validity and the “Satisfaction with Supervisor” Dilemma

A classic methodological debate surrounding the LMX7 centers on its discriminant validity relative to Satisfaction with Supervisor. In Schyns and Paul’s (1999) empirical investigation, the correlation between the LMX7 and the “Satisfaction with Supervisor” subscale of the German Job Descriptive Index (Arbeitsbeschreibungsbogen, ABB; Neuberger & Allerbeck, 1980) reached an extraordinary $r = .81 (p < .01)$. In contrast, correlations with other facets of job satisfaction were notably lower (Satisfaction with Colleagues:$r = .33$; Satisfaction with Work Tasks:$r = .50$; Satisfaction with Working Conditions:$r = .33$).

While such heavy correlation confirms that relationship quality is deeply rewarding, psychometricians like van Breukelen and Konst (1997) have cautioned that the LMX7 may risk construct contamination, partially reflecting global employee contentment with their supervisor rather than distinct structural exchange characteristics. However, subsequent confirmatory factor analyses using multi-trait multi-method (MTMM) designs have consistently confirmed that despite high collinearity, nested models constraining LMX and supervisor satisfaction to a single factor exhibit significantly worse model fit ($Deltachi^2, p < .001$), confirming their psychometric separability.

3. Criterion-Related and Predictive Validity

Decades of field research indicate that the LMX7 possesses potent predictive utility across core organizational outcomes:

• Job Performance: Meta-analytically associated with objective sales metrics, productivity quotas, and subjective performance evaluations ($
  ho = .30$ to $.38$).
• Organizational Citizenship Behaviors (OCBs): Positively predicts individual-targeted (OCBI) and organization-targeted (OCBO) discretionary helping behaviors ($
  ho = .34$ to $.40$).
• Turnover Intentions and Actual Turnover: Negatively predicts turnover intentions ($
  ho = -.31$ to $-.44$) and voluntary organizational exit.
• Occupational Well-being: Inversely predicts emotional exhaustion, depersonalization, job burnout, and somatic complaints, while positively predicting psychological empowerment and work engagement.

Reliability

The Leader-Member Exchange Scale exhibits outstanding, exceptionally stable internal consistency reliability across languages, professional sectors, and organizational hierarchies.

1. Internal Consistency

Across empirical research, the internal consistency of the LMX7, assessed via Cronbach’s alpha ($lpha$), consistently exceeds conventional psychometric thresholds ($lpha ge .80$). In Gerstner and Day’s (1997) landmark meta-analysis, the sample-weighted average internal consistency across member-report studies was $lpha = .89$. In Graen and Uhl-Bien’s (1995) reference samples, reliability routinely climbed to $lpha = .93$.

In Schyns and Paul’s (1999) and Schyns’ (2001) German adaptation ($N = 113$ non-self-employed professionals across varied industries), the 7-item scale produced a Cronbach’s alpha of .92. Corrected item-total correlations ($r_{it}$) across the items demonstrated high discrimination and low measurement error:

• Item 1 (Know where you stand): $r_{it} = .60$
• Item 2 (Understands problems/needs): $r_{it} = .78$
• Item 3 (Recognizes potential): $r_{it} = .75$
• Item 4 (Uses power to help): $r_{it} = .79$
• Item 5 (Bails out at expense): $r_{it} = .66$
• Item 6 (Defends decisions in absence): $r_{it} = .83$
• Item 7 (Working relationship effectiveness): $r_{it} = .84$

Every single item demonstrated an item-total correlation $ge .60$, indicating excellent internal coherence and confirming that no individual item degradation occurs when combining the items into an overall composite score.

2. Test-Retest Stability

Longitudinal field research investigating the test-retest reliability of the LMX7 demonstrates that relational exchange represents a durable psychological state that crystallizes within early socialization periods and maintains high longitudinal stability over time. Test-retest coefficients over intervals spanning 3 to 6 months typically range from $r_{tt} = .70$ to $.85$ in stable organizational environments where direct reporting lines remain intact.

Factor Analysis

A central psychometric controversy in leadership measurement has been whether LMX is structurally unidimensional or multidimensional. Liden and Maslyn (1998) introduced a multidimensional construct (LMX-MDM) separating exchange into four explicit dimensions: Affect, Loyalty, Contribution, and Professional Respect. However, for the canonical 7-item scale developed by Graen and Uhl-Bien (1995), empirical literature overwhelmingly validates a robust, elegant unidimensional factor structure.

1. Exploratory Factor Analysis (EFA)

In the factor analytic evaluation conducted by Schyns and Paul (1999) utilizing Maximum Likelihood extraction, only a single factor emerged with an eigenvalue exceeding the Kaiser-Guttman criterion of 1.0 (Eigenvalue = 4.75). This single dominant factor accounted for an impressive 67.8% of the total variance among items. A secondary, discarded factor yielded an eigenvalue of only 0.60, explaining merely 8.6% of variance. Scree plot tests confirmed an unambiguous “elbow” drop immediately following the first latent dimension.

Factor loadings ($F$) derived from the maximum likelihood analysis demonstrated heavy saturation across all seven indicators without exception (all $F > .60$):

• Item 1: Factor Loading = .62 ($M = 3.34, SD = 0.83$)
• Item 2: Factor Loading = .82 ($M = 3.33, SD = 1.00$)
• Item 3: Factor Loading = .78 ($M = 3.29, SD = 1.03$)
• Item 4: Factor Loading = .81 ($M = 3.41, SD = 1.18$)
• Item 5: Factor Loading = .67 ($M = 2.89, SD = 1.19$)
• Item 6: Factor Loading = .89 ($M = 3.25, SD = 1.16$)
• Item 7: Factor Loading = .91 ($M = 3.40, SD = 1.08$)

Items 6 and 7 represent the most profound structural anchors of the latent construct, displaying loadings approaching or exceeding .90. Item 5 exhibits slightly lower saturation (.67), which psychometricians attribute to its colloquial phrasing (“bail you out at his/her expense”) and the strong personal risk calculus implied by supervisory self-sacrifice.

2. Confirmatory Factor Analysis (CFA)

Across extensive structural equation modeling (SEM) applications in organizational literature, one-factor CFA models for the LMX7 consistently meet rigorous goodness-of-fit benchmarks when sample sizes satisfy structural requirements ($N ge 200$). Reported fit indices routinely fall within exceptional ranges: $chi^2/df < 2.50$, Comparative Fit Index ($ ext{CFI}) ge .96$, Tucker-Lewis Index ($ ext{TLI}) ge .95$, Root Mean Square Error of Approximation ($ ext{RMSEA}) le .055$, and Standardized Root Mean Square Residual ($ ext{SRMR}) le .035$. Attempts to force multidimensional splits upon the 7 items inevitably produce latent factor intercorrelations exceeding$phi = .85$, indicating excessive multicollinearity and confirming the empirical superiority of the parsimonious single-factor model.

Instrument / Measurement Tool

• Instrument Name: Leader-Member Exchange Scale (LMX7 / LMX-7)
• Instrument Type: Self-report dyadic relational assessment questionnaire (available in Subordinate/Member and Supervisor/Leader evaluation perspectives).
• Administration Format: Paper-and-pencil, computer-assisted web interview (CAWI), or embedded organizational survey module.
• Target Population: Employed adult personnel working within formal organizational hierarchies under a direct supervisor.
• Administration Time: Approximately 3 to 5 minutes.
• Item Count: 7 items.
• Response Anchors: Unique 5-point Likert scales tailored specifically to each item prompt:
  • Item 1: 1 = Rarely [nie], 2 = Occasionally [selten], 3 = Sometimes [gelegentlich], 4 = Fairly often [oft], 5 = Very often [immer]
  • Item 2: 1 = Not a bit [gar nicht], 2 = A little [wenig], 3 = A fair amount [mittelmäßig], 4 = Quite a bit [gut], 5 = A great deal [sehr gut]
  • Item 3: 1 = Not at all [gar nicht], 2 = A little [wenig], 3 = Moderately [mittelmäßig], 4 = Mostly [gut], 5 = Fully [sehr gut]
  • Item 4: 1 = None [gering], 2 = Small [eher gering], 3 = Moderate [mittel], 4 = High [eher hoch], 5 = Very high [hoch]
  • Item 5: 1 = None [gering], 2 = Small [eher gering], 3 = Moderate [mittel], 4 = High [eher hoch], 5 = Very high [hoch]
  • Item 6: 1 = Strongly disagree [trifft gar nicht zu], 2 = Disagree [trifft wenig zu], 3 = Neutral [mittelmäßig], 4 = Agree [überwiegend zu], 5 = Strongly agree [völlig zu]
  • Item 7: 1 = Extremely ineffective [sehr ineffektiv], 2 = Worse than average [schlechter als durchschnitt], 3 = Average [durchschnittlich], 4 = Better than average [besser als durchschnitt], 5 = Extremely effective [sehr effektiv]
• Scoring and Metric Computation:
  • Reverse Coding: None. All seven items are positively keyed toward high relationship quality.
  • Sum Score Calculation: Sum all raw item scores (Item 1 + Item 2 + Item 3 + Item 4 + Item 5 + Item 6 + Item 7). Total scores range continuously from 7 to 35.
  • Mean Score Calculation: Alternatively computed as an unweighted linear average (Sum Score divided by 7), yielding an interpretable composite metric ranging from 1.00 to 5.00. In Schyns and Paul’s reference sample, the overall scale mean was $M = 3.27 (SD = 0.88)$, with a skewness of $0.51 (SE = 0.23)$ and kurtosis of $0.36 (SE = 0.46)$.
• Normative Score Categorization (Graen & Uhl-Bien, 1995 standard cutoffs):
  • 30 – 35 points: Very High LMX (Premium Partnership / Mature In-Group)
  • 25 – 29 points: High LMX (Strong Working Relationship)
  • 20 – 24 points: Moderate LMX (Average / Transitional Relationship)
  • 15 – 19 points: Low LMX (Superficial / Contractual Out-Group)
  • 7 – 14 points: Very Low LMX (Severely Impaired / Distrustful Relationship)

Permissions & Fee and Test Year

The definitive 7-item version of the Leader-Member Exchange Scale was published by George B. Graen and Mary Uhl-Bien in 1995 in The Leadership Quarterly. The German cross-cultural adaptation was published by Dr. Birgit Schyns and Tina Paul in 1999 (presented at the Zurich Symposium on Work and Organizational Psychology) and detailed further in Schyns’ 2001 doctoral dissertation at the University of Leipzig.

In accordance with standard international academic research practices, the scale items and structure are widely reproduced within scholarly literature for non-commercial educational, scientific, and dissertation research without royalty fees. However, institutional commercial deployment, proprietary management consulting use, and integration within fee-for-service diagnostic platforms or organizational surveys may require formal licensing or authorization from original copyright holders and academic publishers. Researchers are encouraged to cite the foundational 1995 publication and relevant cross-cultural adaptation references in all resulting publications.

References

• Blau, P. M. (1964). Exchange and power in social life. John Wiley & Sons.
• Dulebohn, J. H., Bommer, W. H., Liden, R. C., Brouer, R. L., & Ferris, G. R. (2012). A meta-analysis of antecedents and consequences of leader-member exchange: Integrating the LMX literature. Journal of Management, 38(6), 1715–1759. https://doi.org/10.1177/0149206311415280
• Gerstner, C. R., & Day, D. V. (1997). Meta-analytic review of leader-member exchange theory: Correlates and construct issues. Journal of Applied Psychology, 82(6), 827–844. https://doi.org/10.1037/0021-9010.82.6.827
• Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American Sociological Review, 25(2), 161–178. https://doi.org/10.2307/2092623
• Graen, G. B., & Scandura, T. A. (1987). Toward a psychology of dyadic organizing. Research in Organizational Behavior, 9, 175–208.
• Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of leader-member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. The Leadership Quarterly, 6(2), 219–247. https://doi.org/10.1016/1048-9843(95)90036-5
• Liden, R. C., & Maslyn, J. M. (1998). Multidimensionality of leader-member exchange: An empirical assessment through scale development. Journal of Management, 24(1), 43–72. https://doi.org/10.1177/014920639802400105
• Neuberger, O., & Allerbeck, M. (1980). Messung und Analyse von Arbeitszufriedenheit: Das Arbeitsbeschreibungsbogen (ABB). Huber.
• Schyns, B. (2001). Determinanten beruflicher Veränderungsbereitschaft bei Arbeitnehmern und Arbeitnehmerinnen unterer Hierarchiestufen [Unpublished doctoral dissertation]. Universität Leipzig.
• Schyns, B., & Paul, T. (1999, October 11–13). Prüfung einer deutschsprachigen LMX-Skala: Ein Ansatz zum Leader-Member-Exchange [Poster presentation]. 11. Zürcher Symposium Arbeitspsychologie, ETH Zürich, Switzerland.
• van Breukelen, J. W., & Konst, D. (1997). Leader-member exchange and satisfaction with supervisor: One and the same construct? In Work and Organizational Psychology: Proceedings of the 8th European Congress on Work and Organizational Psychology (pp. 125–128). Tilburg University Press.

13. Items of the Scale (Questionnaire)