Customer Engagement With Online Restaurant Community During COVID-19–Model Inventory

1. Abstract

The Customer Engagement With Online Restaurant Community During COVID-19–Model Inventory is an empirical measurement instrument developed by Liu, Xu, Li, and Wei (2023) to assess the multifaceted socio-psychological dynamics governing customer engagement behaviors within virtual brand communities during acute public health crises. Designed in the context of restaurant-hosted mobile social networking groups (specifically WeChat enterprise and customer groups) in the People’s Republic of China during the COVID-19 pandemic, this instrument operationalizes a complex nomological network comprising six core latent constructs: Server Disclosure, Customer Disclosure, Customer Trust, Swift Guanxi, Customers’ Social Influence Engagement, and Customers’ Knowledge-Sharing Engagement. The measurement inventory consists of 23 target items rated on a 7-point Likert response format ranging from 1 (“extremely disagree”) to 7 (“extremely agree”).

Grounded in the synergistic frameworks of social penetration theory and social exchange theory, the scale evaluates how reciprocal, multi-layered disclosures between boundary-spanning hospitality service personnel and patrons foster relational capital (trust and swift guanxi), which in turn drives high-value customer voluntary performance behaviors. The psychometric properties of the instrument were empirically validated through a rigorous methodological design involving back-translation procedures, assessment of common method variance using Harman’s single-factor criterion (accounting for 40.3% of variance), and structural equation modeling (SEM) via confirmatory factor analysis (CFA). The measurement model demonstrated robust construct validity, with standardized factor loadings exceeding 0.60 (p < 0.001), satisfactory average variance extracted (AVE) values exceeding or approaching recommended benchmarks, and discriminant validity supported by the Fornell-Larcker criterion. Reliability evaluations confirmed high internal consistency, with construct Cronbach’s alpha coefficients ranging between 0.670 and 0.924 and composite reliability (CR) indices spanning 0.762 to 0.938. The instrument provides hospitality researchers, organizational psychologists, and service marketing practitioners with a statistically validated diagnostic tool for evaluating digital relationship-building and customer value co-creation under conditions of environmental disruption and physical confinement.

2. Keywords

Customer Engagement, Online Restaurant Community, Server Disclosure, Customer Disclosure, Customer Trust, Swift Guanxi, Social Influence Engagement, Knowledge-Sharing Engagement, Social Penetration Theory, Social Exchange Theory, COVID-19 Pandemic, Psychometrics

3. Authors

The scale was developed and validated by a research team specializing in hospitality management, consumer behavior, and tourism geography from Xiamen University and Qingdao University:

• Min Liu — School of Management, Xiamen University, Xiamen, Fujian, China. ORCID: 0000-0002-5696-5405. Email: [email protected].
• Jie Xu (Corresponding Author) — School of Management, Xiamen University, 422 South Siming Road, Siming District, Xiamen, Fujian, China (Postal Code: 361005). ORCID: 0000-0002-1964-0820. Email: [email protected].
• Shuhao Li — School of Tourism and Geography Science, Qingdao University, Qingdao, Shandong, China. Email: [email protected].
• Min Wei — School of Management, Xiamen University, Xiamen, Fujian, China. ORCID: 0000-0002-8441-4600. Email: [email protected].

4. Purpose

The overarching purpose of the Customer Engagement With Online Restaurant Community During COVID-19–Model Inventory is to assess, diagnose, and quantify consumer interactive behavior and relationship-building mechanisms within digital hospitality communities established during extraordinary environmental stress. The onset of the COVID-19 pandemic precipitated unprecedented systemic shocks across global service industries, most acutely within physical food service and hospitality operations. Government-mandated dining room closures, social distancing regulations, and consumer contamination fears severely curtailed traditional face-to-face service encounters. In response, food service organizations rapidly shifted toward digital platforms and closed-loop social networks—predominantly proprietary WeChat groups—to preserve operational viability, maintain touchpoints with loyal patrons, and organize direct-to-consumer delivery or curbside pickups.

Despite this operational migration, conventional metrics of digital interaction (such as page views, click-through rates, or transactional frequency) failed to capture the deep psychological dynamics and relational mechanisms occurring between frontline service employees and quarantined consumers. The authors identified a critical need for an instrument that could examine the socio-psychological architecture of these micro-communities. Specifically, the instrument seeks to capture how reciprocal personal disclosures between boundary-spanning employees (servers) and consumers create interpersonal safety and psychological proximity, thereby generating cognitive trust and context-specific relational bonds (termed “swift guanxi” within Chinese cultural environments).

In applied research and organizational practice, the scale serves multiple key functions:

• Relational Diagnostic Tool: It enables hospitality managers and community moderators to evaluate whether communication strategies utilized by frontline staff transcend transactional updates and effectively cultivate perceived integrity, reliability, and emotional resonance.
• Predictor of Value Co-Creation: The instrument measures distinct downstream manifestations of customer engagement—specifically differentiating between outward-facing advocacy (social influence engagement) and collaborative organizational development (knowledge-sharing engagement). This distinction is vital for understanding how customer engagement translates into both brand equity diffusion and operational service improvements.
• Cross-Cultural Relational Modeling: By formalizing the measurement of “swift guanxi” alongside generalized “customer trust,” the tool addresses a longstanding void in psychometrics concerning how rapid-onset, digitally mediated interpersonal connections function outside Western individualistic paradigms.
• Crisis Resilience Evaluation: The inventory offers public health communicators, service design theorists, and crisis management scholars a systematic framework to analyze the digital preservation of organizational social capital when physical infrastructure is disabled.

5. Psychological Construct

The measurement inventory is structured across six interrelated socio-psychological constructs that reflect an input-mediator-output operational sequence. Each construct captures a discrete dimension of social interaction, relational cognition, or behavioral manifestation:

5.1. Server Disclosure

Server disclosure represents the customer’s subjective perception of the extent to which frontline restaurant personnel proactively, authentically, and transparently reveal operational details, professional opinions, and menu-related advisory insights within the virtual group. Grounded in the communicative dimensions of interpersonal disclosure, this construct does not measure generic advertising broadcasts; rather, it assesses high-value, vulnerable, or tailored communicative acts. Such acts include promptly admitting delivery or kitchen service delays, offering candid evaluations regarding product quality or portion sizes, and delivering personalized culinary recommendations tailored to specific client needs.

5.2. Customer Disclosure

Customer disclosure conceptualizes the degree to which patrons reciprocate communicative openness by sharing private, identity-relevant, or preference-based information within the community space. This includes expressing genuine emotional appreciation toward service staff, articulating dietary idiosyncratic preferences (such as culinary restrictions, allergies, or vegetarian lifestyle commitments), and voluntarily signaling customer loyalty (such as declaring oneself an established regular patron). Within relational psychology, customer disclosure serves as a primary marker of psychological safety and community embeddedness.

5.3. Customer Trust

Customer trust reflects the consumer’s cognitive and affective confidence in the reliability, benevolence, capability, and moral integrity of the restaurant’s personnel. In computer-mediated communication (CMC) environments characterized by heightened risk and physical detachment—such as purchasing prepared food amidst a viral contagion—trust represents an indispensable psychological buffer. The construct operationalizes customer expectations that frontline staff will fulfill implicit service promises, adhere to stringent hygiene standards, and maintain authentic concern for the customer’s welfare.

5.4. Swift Guanxi

Derived from indigenous Chinese relational psychology and adapted to rapid online interactions, guanxi represents a deeply rooted network of interpersonal connections underpinned by mutual obligation, reciprocal favor exchange, and affective empathy. While traditional guanxi requires longitudinal, face-to-face cultivation over decades, swift guanxi operationalizes a dynamic, digitally compressed variant. It captures immediate mutual understanding, pseudo-friendship framing, and interpersonal harmony forged swiftly between transacting parties within virtual environments through frequent, informal, and empathetic communicative exchanges.

5.5. Customers’ Social Influence Engagement

Customers’ social influence engagement constitutes an externally directed, behavioral manifestation of customer engagement. It assesses the customer’s proactive willingness to act as an evangelist and brand ambassador across personal and external social networks. This construct captures behaviors such as discussing favorable dining experiences with acquaintances, broadcasting the functional and economic benefits realized from the restaurant, and actively generating organic word-of-mouth (WOM) across external social media platforms.

5.6. Customers’ Knowledge-Sharing Engagement

Customers’ knowledge-sharing engagement operationalizes an internally directed, organizational-learning behavior. Unlike public advocacy, knowledge-sharing engagement captures the customer’s willingness to invest intellectual and cognitive resources into the restaurant’s continuous service improvement. This dimension measures constructive feedback behaviors, including suggesting targeted modifications to food recipes or packaging, proposing operational improvements to service delivery, and collaborating on the ideation of future menu offerings.

6. Theoretical Framework

The theoretical architecture of the model integrates two foundational theories from social psychology and sociology: Social Penetration Theory (SPT) and Social Exchange Theory (SET).

6.1. Social Penetration Theory

Formulated by Altman and Taylor (1973), Social Penetration Theory posits that interpersonal relationship development progresses systematically from superficial, non-intimate behavioral layers to deeper, highly intimate psychological strata through reciprocal self-disclosure. The progression of relational closeness is governed by two essential vectors: breadth (the range of topics discussed) and depth (the degree of personal intimacy, vulnerability, and private disclosure involved).

Liu et al. (2023) mapped this interpersonal progression onto digital commercial communities during the COVID-19 pandemic. Under normal operating conditions, commercial exchanges between food service employees and patrons are predominantly transactional and confined to superficial “outer-layer” interactions (e.g., ordering items, settling bills). However, the confinement and health anxieties generated by the pandemic created a psychological context wherein frontline servers initiated “deeper-layer” communicative disclosures—openly acknowledging operational vulnerabilities, service errors, and personal assessments of menu items. According to the core tenets of SPT, reciprocal disclosure is driven by a psychological pressure toward equilibrium: when one party reveals deeper layers of their personal reality, the interacting counterpart experiences a normative impetus to reciprocate with personal preferences, dietary restrictions, and expressions of gratitude.

6.2. Social Exchange Theory and Reciprocity Dynamics

While Social Penetration Theory explains the structural depth of communication, Social Exchange Theory (Blau, 1964; Homans, 1958) clarifies the motivational mechanisms that sustain these interactions. SET asserts that human relationships are maintained through an ongoing cost-benefit calculus governed by the universal norm of reciprocity (Gouldner, 1960). In digital service groups, the psychological and informational resources provided by servers (e.g., tailored guidance, honest error disclosure) generate a psychological debt or relational equity.

Consumers resolve this imbalance not merely through monetary transactions, but through relational and behavioral dividends. First, the mutual exchange of disclosures enhances relational mediators—building cognitive and affective Customer Trust and establishing Swift Guanxi. Second, these robust relational states motivate consumers to engage in extra-role, non-transactional customer behaviors, manifested as voluntary advocacy (Social Influence Engagement) and constructive feedback (Knowledge-Sharing Engagement). By uniting SPT and SET, the measurement inventory models how communicative inputs transform into cognitive-affective relational bonds, which ultimately materialize as behavioral co-creation.

7. Validity

The construct validity, convergent validity, and discriminant validity of the instrument were systematically verified using structural equation modeling (SEM) and confirmatory factor analysis (CFA) within a consumer sample in mainland China (Liu et al., 2023).

7.1. Content and Translation Validity

Items were adapted from validated international psychometric literature—specifically drawing upon the empirical foundations of Hwang et al. (2015) for employee disclosure, Fu et al. (2022) for customer disclosure and community engagement, Lin et al. (2018) for trust paradigms, and Li et al. (2021) for swift guanxi operationalizations. To ensure cross-linguistic equivalence and conceptual integrity, the authors implemented Brislin’s (1980) classical forward- and back-translation protocol. The items were translated from English into Simplified Chinese by bilingual hospitality scholars, subsequently back-translated by an independent researcher blind to the original English texts, and scrutinized for semantic parity, cultural resonance, and contextual accuracy in relation to WeChat group dynamics.

7.2. Convergent Validity

Convergent validity evaluates the extent to which individual operational items converge to measure their theoretical construct. According to established statistical thresholds (Anderson & Gerbing, 1988; Hair et al., 2010), convergent validity is evidenced when standardized factor loadings are statistically significant and exceed 0.60, and the Average Variance Extracted (AVE) exceeds 0.50. In the validation study:

• All standardized item factor loadings demonstrated statistical significance at the p < 0.001 level and comfortably exceeded the minimum cutoff of 0.60.
• The calculated Average Variance Extracted (AVE) values across the constructs met or closely approximated the benchmark of 0.50 (Fornell & Larcker, 1981; Lam, 2012), confirming that the majority of variance in the observed indicators was accounted for by the underlying latent constructs rather than measurement error.

7.3. Discriminant Validity

Discriminant validity was established using the classical Fornell-Larcker criterion (1981). According to this standard, discriminant validity is verified when the square root of the AVE for each latent construct is demonstrably greater than the correlation coefficients between that construct and any other construct in the measurement model. The empirical results confirmed that the square roots of the AVEs exceeded all corresponding inter-construct correlations, demonstrating that the six constructs—while theoretically integrated—represent empirically distinct phenomena. Specifically, this validated the empirical separability between generalized Customer Trust and culturally contextualized Swift Guanxi.

8. Reliability

The reliability of the measurement model was assessed by evaluating internal consistency via both Cronbach’s alpha ($lpha$) and Composite Reliability (CR) indices for each of the six latent dimensions.

8.1. Internal Consistency Statistics

In accordance with psychometric standards (Nunnally & Bernstein, 1994; Bagozzi & Yi, 1988), an alpha coefficient and composite reliability score exceeding 0.70 reflect satisfactory to excellent internal consistency, while values above 0.65 are acceptable in exploratory and specialized crisis contexts. The constructs demonstrated the following empirical profiles:

• Cronbach’s Alpha ($lpha$): Ranged from 0.670 to 0.924 across the six dimensions, indicating robust internal item consistency across all subscales.
• Composite Reliability (CR): Ranged from 0.762 to 0.938 across the six dimensions, exceeding the recommended 0.70 benchmark and establishing that the latent variables are reliably measured by their respective manifest indicators.

8.2. Error Variance and Indicator Consistency

The combined evaluation of high CR values alongside substantial factor loadings indicates that random measurement error was tightly constrained. The scale demonstrates reliable structural consistency, making it suitable for both structural equation modeling and aggregate descriptive scoring.

9. Factor Analysis

9.1. Common Method Variance (CMV)

Because all measures were collected via self-report questionnaires administered to consumer respondents within cross-sectional settings, the authors systematically evaluated the risk of common method variance (CMV). Using SPSS 26.0, Harman’s single-factor test was executed via unrotated exploratory factor analysis. The extraction produced six distinct factors with eigenvalues greater than 1.0, reflecting the multidimensional design of the scale. The first unrotated factor accounted for 40.3% of the total variance, safely below the conventional critical threshold of 50.0% (Podsakoff et al., 2003). Consequently, common method bias was confirmed not to be a confounding threat to the study’s structural interpretations.

9.2. Confirmatory Factor Analysis (CFA)

A confirmatory factor analysis was conducted to assess the overall structural fit of the 23-item, six-factor measurement model using maximum likelihood estimation. Model fit evaluation followed standard criteria established by Hu and Bentler (1999) and Kline (2015). The empirical goodness-of-fit indices demonstrated acceptable to excellent alignment with the empirical data:

• Chi-Square / Degrees of Freedom ($\chi^2/df$): $\chi^2 = 528.682$, $df = 215$, yielding $\chi^2/df = 2.459$. This ratio falls comfortably below the recommended upper limit of 3.0, indicating satisfactory parsimonious fit.
• Comparative Fit Index (CFI): 0.933 (Benchmark > 0.90 indicates good fit).
• Incremental Fit Index (IFI): 0.934 (Benchmark > 0.90 indicates good fit).
• Tucker-Lewis Index (TLI): 0.921 (Benchmark > 0.90 indicates good fit).
• Normed Fit Index (NFI): 0.893 (Approaching the ideal 0.90 threshold).
• Goodness-of-Fit Index (GFI): 0.869 (Acceptable within complex field-based SEM models).
• Root Mean Square Error of Approximation (RMSEA): 0.066 (Benchmark < 0.08 reflects good approximation error; 90% confidence intervals aligned with robust fit).
• Root Mean Square Residual (RMR): 0.114.

Collectively, the CFA indices confirm that the hypothesized six-factor structure accurately reflects the underlying multidimensional data matrix without requiring post-hoc cross-loadings or arbitrary error-covariance adjustments.

10. Instrument / Measurement Tool

• Instrument Name: Customer Engagement With Online Restaurant Community During COVID-19–Model Inventory
• Constructs / Subscales: 6 dimensions: (1) Server Disclosure, (2) Customer Disclosure, (3) Customer Trust, (4) Swift Guanxi, (5) Customers’ Social Influence Engagement, (6) Customers’ Knowledge-Sharing Engagement.
• Item Count: 23 target questionnaire items across the 6 subscales.
• Response Scale: 7-point Likert scale ranging from 1 = “Extremely Disagree” to 7 = “Extremely Agree”. Intermediate anchor points are typically labeled: 2 = Disagree, 3 = Somewhat Disagree, 4 = Neutral / Neither Agree nor Disagree, 5 = Somewhat Agree, 6 = Agree, 7 = Extremely Agree.
• Administration Format: Standardized self-report inventory suitable for paper-and-pencil, computer-assisted, or mobile digital survey administration (e.g., embedded within WeChat mini-programs or online survey suites).
• Target Population: Adult consumers (aged 18 and older) who are members of online restaurant communities, brand communication groups, or mobile messaging circles (such as WeChat enterprise groups) operated by food service establishments.
• Language: Validated in Simplified Chinese; original theoretical design adapted from English-language measurement literature.
• Scoring Protocol: Dimensional scores are generated by calculating the mean or sum of the items comprising each subscale. Higher mean scores indicate greater perceived disclosure, stronger relational bonds, or elevated levels of voluntary customer engagement. No reverse-scored items are utilized. In structural equation modeling, latent construct scores should be extracted via maximum likelihood estimation using the item loadings derived from the CFA.

11. Permissions & Fee and Test Year

• Year of Development: 2023.
• Copyright & Intellectual Ownership: Copyright © 2023 is retained by the authors (Min Liu, Jie Xu, Shuhao Li, and Min Wei) and the publisher (Elsevier Ltd / Journal of Hospitality and Tourism Management).
• Permissible Usage: The scale is published in an academic peer-reviewed journal and is available for academic research, non-profit educational instruction, and scholarly replication without royalty fees.
• Commercial Licensing: Commercial implementation, incorporation into proprietary consumer analytics platforms, or monetization within corporate consulting frameworks requires explicit written authorization and licensing from the corresponding author and copyright holders.
• Contact for Permissions: Dr. Jie Xu, School of Management, Xiamen University, 422 South Siming Road, Xiamen, China (Email: [email protected]).

12. References

• Altman, I., & Taylor, D. A. (1973). Social penetration: The development of interpersonal relationships. Holt, Rinehart & Winston.
• Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103(3), 411–423. https://doi.org/10.1037/0033-2909.103.3.411
• Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16(1), 74–94. https://doi.org/10.1007/BF02723327
• Blau, P. M. (1964). Exchange and power in social life. John Wiley & Sons.
• Brislin, R. W. (1980). Translation and content analysis of oral and written materials. In H. C. Triandis & J. W. Berry (Eds.), Handbook of cross-cultural psychology: Methodology (Vol. 2, pp. 389–444). Allyn & Bacon.
• Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39–50. https://doi.org/10.1177/002224378101800104
• Fu, S., Chen, X., & Zheng, H. (2022). Exploring an adverse impact of customer disclosure on community engagement: A social penetration perspective. Information & Management, 59(2), Article 103590. https://doi.org/10.1016/j.im.2021.103590
• Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American Sociological Review, 25(2), 161–178. https://doi.org/10.2307/2092623
• Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate data analysis (7th ed.). Prentice Hall.
• Homans, G. C. (1958). Social behavior as exchange. American Journal of Sociology, 63(6), 597–606. https://doi.org/10.1086/222355
• Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. https://doi.org/10.1080/10705519909540118
• Hwang, J., Ekinci, Y., & Yoon, S. (2015). Frontline employees’ self-disclosure: Its impacts on customer trust, loyalty, and satisfaction. International Journal of Hospitality Management, 49, 137–146. https://doi.org/10.1016/j.ijhm.2015.06.009
• Kline, R. B. (2015). Principles and practice of structural equation modeling (4th ed.). Guilford Press.
• Lam, L. W. (2012). Impact of employee competition on front-line employee service performance: The role of socialization. Journal of Applied Psychology, 97(5), 1084–1094. https://doi.org/10.1037/a0028885
• Li, X., Hess, T. J., & Valacich, J. S. (2021). Building swift guanxi in online marketplaces: The role of interactive virtual agents. Electronic Commerce Research and Applications, 47, Article 101046. https://doi.org/10.1016/j.elerap.2021.101046
• Lin, J., Wang, B., Wang, N., & Lu, Y. (2018). Understanding the evolution of consumer trust in mobile commerce: A longitudinal study. Information Technology & People, 31(1), 162–184. https://doi.org/10.1108/ITP-08-2016-0182
• Liu, M., Xu, J., Li, S., & Wei, M. (2023). Engaging customers with online restaurant community through mutual disclosure amid the COVID-19 pandemic: The roles of customer trust and swift guanxi. Journal of Hospitality and Tourism Management, 56, 124–134. https://doi.org/10.1016/j.jhtm.2023.06.019
• Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). McGraw-Hill.
• Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903. https://doi.org/10.1037/0021-9010.88.5.879

13. Items of the Scale