Does dispositional awe promote customer citizenship behaviours? The multiple mediating effects of construal level and social connectedness

In today’s value co-creation era, customers often engage in behaviours such as recommending brands, helping other customers, and providing feedback. These customer citizenship behaviours (CCBs) are voluntary, extra-role actions that contribute to firm value creation, improving outcomes like customer happiness, repeat business, employee productivity, and competitive advantage. Among antecedents of CCBs, customer characteristics are especially influential. Prior research links CCBs with attitudes (e.g., identification, satisfaction, justice perceptions) and traits (e.g., self-efficacy, public self-awareness, agreeableness, extraversion, prosocial/proactive personality). However, many studied traits are highly stable and less actionable for firms. The authors argue trait emotions—dispositional tendencies in experiencing specific emotions—offer variability and are cultivable, making them promising predictors of CCBs. Dispositional awe, a trait emotion associated with prosocial outcomes (helping, socially responsible consumption, organizational citizenship, tourism helping), may similarly promote CCBs. The study asks whether people who frequently experience awe engage more in CCBs and through what mechanisms. Guided by the prototype model of awe (perceived vastness and need for accommodation) and the elaborated model of awe’s prosocial effects (emphasis on interdependence and internal transformation), the authors hypothesize that dispositional awe positively predicts CCBs and operates via social connectedness and, serially, via construal level then social connectedness. The study contributes by introducing trait emotions as antecedents of CCBs, extending awe’s prosocial implications to value co-creation, and disentangling intrapersonal (construal level) and interpersonal (social connectedness) mechanisms. It also offers practical ideas for motivating CCBs in marketing contexts.

Awe is an emotional response to vastness and a need for accommodation—stimuli that exceed current schemas and prompt cognitive restructuring. It can be state (transient) or dispositional (trait-like frequency/propensity). The prototype model (Keltner & Haidt) emphasizes perceived vastness and accommodative processing, underpinning awe’s transformative potential. The elaborated model of awe’s prosocial effects responds to critiques of the small-self hypothesis, proposing that awe fosters concern for an interdependent “we” through internal cognitive and motivational transformation rather than a mere attentional shift away from the self. Prior work documents awe’s prosocial outcomes (generosity, helping, OCBs, positive WOM, environmental responsibility) and suggests mechanisms including connectedness and abstract thinking. CCBs, conceptualized as recommendation, helping, and feedback (Groth’s three-dimensional model), are voluntary, positive, and non-tradable behaviours that benefit firms. The authors hypothesize: H1a–c, dispositional awe positively predicts recommendation, helping, and feedback; H2, dispositional awe increases social connectedness; H3a–c, social connectedness increases recommendation, helping, and feedback; H4a–c, social connectedness mediates awe → CCBs; H5, dispositional awe increases construal level; H6, higher construal level increases social connectedness; H7a–c, dispositional awe influences each CCB via serial mediation of construal level then social connectedness.

Design and measures: A questionnaire-based study was conducted in China. Dispositional awe was measured with the DPES awe subscale (Shiota et al., 2006; 6 items, e.g., “I often feel awe”). Construal level was measured with Vallacher and Wegner’s (1989) Behaviour Identification Form (25 dichotomous items; high-level choice=1, low-level=0; higher scores indicate higher construal level). Social connectedness was measured with the 20-item Social Connectedness Scale (Lee et al., 2001; 10 reverse, 10 forward items, e.g., “I feel disconnected from the world around me” [reverse], “I am in tune with the world”). CCBs were measured using Groth’s (2005) scale with three dimensions (recommendation, helping, feedback; 4 items each). Control variable: brand identification (Tuškej et al., 2013; 3 items). All except construal level used 7-point Likert scales (1=strongly disagree to 7=strongly agree). Scales were translated and back-translated, with expert review and minor item localization for clarity. Questionnaire composition: 66 scale items, one screening item, one brand fill-in, and four demographics. Structure: (1) consent; (2) dispositional awe, construal level, social connectedness, screening attention check; (3) a familiar brand fill-in, then brand identification and CCBs regarding that brand; (4) demographics (gender, age, education, occupation). Respondents could opt out. Power analysis: Using power4SEM and pwrSEM, the authors targeted SEM power ≥0.95. Measurement model df were 2058; structural model planned with item parceling (construal level 25→5 parcels; social connectedness 20→4 parcels) yielded 30 indicators and 75 free parameters (df=390). Close/not-close fit power suggested minimum N≈80–112; Monte Carlo power for mediations indicated N≈686 to achieve 0.95 power. Sampling and data collection: Data were collected via Credamo in September 2022 (professional Chinese online platform analogous to MTurk/Prolific). Of 776 responses, 701 valid questionnaires remained after excluding those failing the attention check, with extreme completion times, duplicate IPs, or straightlining (valid rate 90.3%; mean age 29.71). Brands covered diverse industries (e.g., smartphones/computers, apparel, cosmetics, home appliances, food/catering). Analysis: CFA and SEM were conducted in Mplus 8.3. To improve mediation model stability, construal level and social connectedness items were parceled using a balancing strategy based on factor loadings. Common method bias (CMB) was examined via Harman’s single-factor test, ULMC, and CFA marker variable (education) approach. Reliability assessed via AVE, CR, and Cronbach’s alpha. Structural models tested main effects and mediations (bootstrapping with 5000 resamples). Age and brand identification were included as controls; in later analyses age was also controlled due to correlations with CCBs.

Measurement model: Seven-factor CFA fit was good: χ²/df=2.478, RMSEA=0.046, SRMR=0.051, CFI=0.916, TLI=0.910. CMB tests indicated no serious bias: Harman’s first factor=28.231% (<50%); ULMC changes were small (ΔRMSEA=0.009, ΔSRMR=0.016, ΔCFI=0.052, ΔTLI=0.052); CFA marker method showed non-significant difference between Method-U and Method-R (p=0.256). Reliability/validity: All constructs exhibited acceptable convergent and discriminant validity (e.g., AVE≥0.50; CR≥0.80; alphas≥0.79). Correlations: Dispositional awe correlated positively with recommendation (r=0.499), helping (r=0.579), feedback (r=0.528), social connectedness (r=0.633), and construal level (r=0.336). Social connectedness correlated with recommendation (r=0.521), helping (r=0.578), feedback (r=0.551), and construal level (r=0.335). Main effects (controlling age, brand identification): Dispositional awe → recommendation β=0.608 (SE=0.035, t=17.422, p<0.001), helping β=0.704 (SE=0.029, t=24.278, p<0.001), feedback β=0.666 (SE=0.033, t=20.394, p<0.001). R²: recommendation=0.387, helping=0.508, feedback=0.485. Model comparisons supported the hypothesized mediation model (Model B) over alternatives (nested and non-nested). Mediation model (bootstrapped, 5000 resamples): Paths—Dispositional awe → social connectedness β=0.602 (SE=0.033, t=18.242, p<0.001); social connectedness → recommendation β=0.514 (SE=0.085, t=6.047, p<0.001), helping β=0.542 (SE=0.074, t=7.324, p<0.001), feedback β=0.530 (SE=0.083, t=6.386, p<0.001); dispositional awe → construal level β=0.410 (SE=0.040, t=10.178, p<0.001); construal level → social connectedness β=0.394 (SE=0.033, t=11.939, p<0.001). Direct effects attenuated but remained significant: awe → recommendation β=0.219 (SE=0.086, p=0.011), awe → helping β=0.295 (SE=0.074, p<0.001), awe → feedback β=0.265 (SE=0.084, p=0.002). Indirect effects: Awe → recommendation via social connectedness β=0.310 (95% CI [0.138, 0.474], PM=50.6%); via construal level→social connectedness β=0.083 (95% CI [0.038, 0.134], PM=13.6%). Awe → helping via social connectedness β=0.326 (95% CI [0.194, 0.460], PM=46.1%); via construal level→social connectedness β=0.088 (95% CI [0.049, 0.132], PM=12.3%). Awe → feedback via social connectedness β=0.319 (95% CI [0.161, 0.468], PM=47.6%); via construal level→social connectedness β=0.086 (95% CI [0.044, 0.133], PM=12.8%). R² in mediation model: construal level=0.380, social connectedness=0.620, recommendation=0.609, helping=0.669, feedback=0.649. Overall, H1a–c, H2, H3a–c, H5, H6, and H7a–c were supported.

The study addressed whether and how dispositional awe promotes customer citizenship behaviours. Results show that individuals higher in dispositional awe are more likely to recommend brands, help other customers, and provide feedback, aligning with and extending prior evidence on awe’s prosocial effects to the firm–customer value co-creation context. The independent mediation by social connectedness supports the elaborated model of awe’s prosocial effects: awe enhances concern for an interdependent “we,” reflecting a broadened sense of belonging and connectedness that encourages CCBs. The serial mediation via construal level and social connectedness indicates that awe’s accommodative processing fosters more abstract, holistic mental representations, which in turn heighten connectedness and prosocial engagement with firms and other customers. These findings refine theoretical accounts by emphasizing internal cognitive transformation (beyond mere attentional shifts) as a pathway from awe to prosocial consumer outcomes. Practically, the results suggest that firms can foster CCBs by targeting and nurturing awe-prone customers, embedding awe-inspiring elements in brand experiences and communications, and activating customers’ abstract thinking and social connectedness through narratives, values, community platforms, and collaborative initiatives.

Dispositional awe positively influences three dimensions of customer citizenship behaviours: recommendation, helping, and feedback. This effect operates through social connectedness and, serially, through construal level and social connectedness, underscoring awe’s role in promoting prosociality within value co-creation. The research contributes by introducing trait emotions as actionable antecedents of CCBs, extending the behavioural consequences of awe to co-creation, and clarifying the intrapersonal (construal level) and interpersonal (social connectedness) mechanisms. Future work should examine causal dynamics with longitudinal or experimental designs, test the effects of state awe on CCBs, explore additional mediators and moderators (e.g., values, brand symbolism, industry context), and validate findings across service sectors and diverse cultural settings.

Key limitations include: (1) cross-sectional design precludes strong causal inference; longitudinal and experimental studies are needed. (2) Focus on dispositional awe; it remains unclear whether induced state awe immediately affects CCBs. (3) Significant residual direct effects suggest additional mediators (e.g., value change, reduced territoriality) that warrant investigation. (4) Lack of tested moderators; customer attitudes and brand/firm characteristics may condition effects (e.g., symbolic status brands). (5) Sample skew toward product retail categories (e.g., mobile/IT, apparel) with fewer service brands, potentially limiting applicability across industries. (6) Single platform (Credamo) and solely Chinese respondents; cultural processing differences may limit generalizability, necessitating cross-cultural replication and broader sampling.