Available via license: CC BY 4.0
Content may be subject to copyright.
128
An AMOS-SEM-Based Validation of a
Measurement Framework for Social Media
Engagement and Participatory Communication
Yang Liu
Communication Department, Faculty of Modern Languages and Communication, University
Putra Malaysia, Malaysia, 43400 UPM Serdang, Selangor Darul Ehsan, Malaysia, Lecturer of
School of Culture and Media, Xi’an Eurasia University, China, No. 8 Dongyi Road, 710065,
Xi'an City, Shaanxi Province
Mastura Mahamed
Communication Department, Faculty of Modern Languages and Communication; Researcher
of Institute for Social Science Research (IPSAS), University Putra Malaysia 43400 UPM
Serdang, Selangor Darul Ehsan, Malaysia
Corresponding Author Email: mastura.mahamed@upm.edu.my
Hani Salwah Yaakup
Faculty of Modern Languages and Communication, University Putra Malaysia, Malaysia
43400 UPM Serdang, Selangor Darul Ehsan, Malaysia
Wan Anita Wan Abas
Faculty of Modern Languages and Communication, University Putra Malaysia, Malaysia
43400 UPM Serdang, Selangor Darul Ehsan, Malaysia
Jinying Yan
School of Culture and Media, Xi’an Eurasia University, China, No. 8 Dongyi Road, 710065,
Xi'an City, Shaanxi Province
Abstract
With the increasing role of digital platforms in shaping community integration, developing
culturally contextualized and theoretically grounded measurement tools to assess digital
engagement and community attachment is essential, especially for urban migrant
populations. Grounded in Media System Dependency (MSD) Theory and Communication
Infrastructure Theory (CIT), this pilot study validates a four-factor measurement framework
comprising WeChat usage intensity, WeChat dependency, participatory communication, and
Vol 15, Issue 4, (2025) E-ISSN: 2222-6990
To Link this Article: http://dx.doi.org/10.6007/IJARBSS/v15-i4/25116 DOI:10.6007/IJARBSS/v15-i4/25116
Published Date: 05 April 2025
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
129
community attachment. Drawing on adapted and refined scales, data were collected from 98
urban migrants in Xi’an, China, and analyzed through AMOS-based covariance structural
equation modeling (CB-SEM). The measurement model demonstrated satisfactory reliability,
convergent validity, and discriminant validity, with all factor loadings surpassing
recommended thresholds and model fit indices (CFI, TLI, RMSEA, SRMR) indicating good fit.
Importantly, this study distinguishes between usacge intensity and psychological
dependency, offering a nuanced framework that integrates MSD and CIT in a platform-specific
urban migration context. While findings remain exploratory due to the small sample size, the
study provides methodological contributions by demonstrating cross-cultural scale
adaptation and measurement validation procedures. The validated framework serves as a
methodological foundation for future large-scale and longitudinal research investigating the
dynamics between social media engagement, participatory communication, and community
attachment.
Keywords: WeChat, Social Media Dependency, Participatory Communication, Community
Attachment, AMOS-SEM
Introduction
The rapid development of digital platforms has significantly transformed patterns of social
interaction, particularly among urban migrants who increasingly rely on social media for local
integration and community participation. In the Chinese context, WeChat stands out as a
multifunctional platform that facilitates interpersonal communication, information sharing,
and participatory engagement (Shen & Gong, 2019; Rui, Yu, Xu, & Cui, 2019). Understanding
how such digital platforms contribute to community attachment requires robust, context-
sensitive theoretical and measurement frameworks.
Existing literature has established the relevance of Media System Dependency (MSD) theory
(Ball-Rokeach & DeFleur, 1976) and Communication Infrastructure Theory (CIT) (Ball-
Rokeach, Kim, & Matei, 2001) in explaining how media use influences social bonding and civic
participation. However, prior empirical efforts have often overlooked critical distinctions
between active usage intensity and social media dependency (Kim & Ball-Rokeach, 2006), and
have rarely addressed the potential moderating role of participatory communication
(Carpentier, 2011; Freire, 1970). Moreover, the dynamic interactions between platform
dependency and participatory engagement remain underexplored, particularly within
platform-specific urban migration contexts.
Recent studies have highlighted that WeChat usage fosters civic engagement, neighborhood
trust, and pro-community behaviors, especially during crisis periods (You, Wang, & He, 2023;
Pang, 2023). In addition, WeChat-enabled digital spaces provide new avenues for constructing
local identity and enhancing urban residents' sense of community (Zhu & Fu, 2017; Cheng,
Liang, & Leung, 2015). Despite these insights, there remains a need for more sophisticated
measurement models that capture both the intensity and quality of engagement, along with
their psychological and social implications. Notably, few studies have integrated MSD and CIT
frameworks into a measurement model tailored for the urban migrant population and
contextualized within a specific social media platform.
The current study addresses this gap through an exploratory pilot investigation that aims to
preliminarily validate a four-factor measurement framework comprising WeChat usage
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
130
intensity, WeChat dependency, participatory communication, and community attachment,
using AMOS-based Structural Equation Modeling (SEM). Rather than testing full structural
causal relationships, this study focuses on confirming the reliability and validity of
measurement constructs as a methodological foundation for subsequent large-sample
empirical research.
The key contribution of this pilot study is to demonstrate the feasibility of applying AMOS-
SEM in capturing and validating measurement constructs that reflect the complex interplay
between platform use behaviors, dependency patterns, participatory communication, and
community attachment. This validated measurement model will serve as a solid starting point
for future large-scale and longitudinal studies examining structural relationships across
different digital platforms and diverse demographic settings.
Literature Review
Conceptual Foundations
The growing role of social media platforms in shaping community engagement has drawn
increasing scholarly interest. Urban migrants, in particular, rely on these digital tools for local
information access, social integration, and emotional support (You, Wang, & He, 2023; Pang,
2023). Two theoretical frameworks are central to understanding these dynamics: Media
System Dependency (MSD) theory and Communication Infrastructure Theory (CIT).
MSD theory, originally proposed by Ball-Rokeach and DeFleur (1976), suggests that individuals
develop varying degrees of dependency on media based on their informational, orientational,
and social bonding needs. This dependency becomes especially pronounced in contexts of
social adaptation and uncertainty, such as urban migration. While prior research
demonstrates that media dependency can enhance trust and participation in online
communities (Kim & Ball-Rokeach, 2006), excessive reliance may also substitute for face-to-
face interactions and potentially erode real-world social ties (Shen & Gong, 2019). This dual
effect underscores the need to differentiate between healthy engagement and psychological
over-dependency.
Communication Infrastructure Theory (CIT)
CIT emphasizes the interconnected roles of local storytelling networks, interpersonal
communication, and institutional media systems in fostering a shared sense of community
(Ball-Rokeach, Kim, & Matei, 2001). While initially applied in neighborhood media research,
CIT is increasingly relevant in digital environments. Platforms like WeChat function as
participatory infrastructures that enable users to co-construct localized narratives, thereby
strengthening civic engagement and emotional belonging (Rui, Yu, Xu, & Cui, 2019; Zhu & Fu,
2017). However, empirical applications of CIT within platform-specific, migrant population
contexts remain limited.
Integrating MSD and CIT in Digital Engagement Research
WeChat, as a ubiquitous platform in China, embodies key characteristics of both MSD and CIT.
Usage intensity—frequent and active use of platform features—aligns with CIT, reflecting the
platform’s role in community storytelling and interactive dialogue. In contrast, dependency—
psychological and behavioral reliance on the platform for social and emotional needs—
reflects MSD principles.
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
131
Moreover, participatory communication is theorized to play a moderating role in this
framework. Drawing on participatory communication theory (Freire, 1970; Carpentier, 2011),
it is hypothesized that user-generated content and two-way engagement enhance collective
identity formation. However, the relationship may not be linear; excessive participatory
communication could intensify media dependency and reduce offline community
involvement. This complexity has rarely been addressed in prior quantitative models.
Measurement Model Assumptions
Building on these theoretical foundations, this pilot study proposes and validates a four-factor
measurement framework capturing WeChat usage intensity, WeChat dependency,
participatory communication, and community attachment. The assumptions guiding this
measurement model are:
A. WeChat usage intensity is positively associated with community attachment, facilitated by
active engagement and social participation.
B. WeChat dependency influences community attachment but may exhibit diminishing or
curvilinear effects at higher dependency levels.
C. Participatory communication is expected to moderate these relationships, strengthening
community bonding through two-way dialogue and co-created narratives.
It is important to emphasize that this pilot study does not aim to test structural causal
relationships. Rather, it focuses on validating whether these latent constructs can be reliably
measured using AMOS-based Structural Equation Modeling (SEM), thereby laying a
methodological foundation for future large-sample hypothesis testing and causal modeling.
Community Attachment in the Digital Context
Community attachment refers to an individual’s emotional connection and psychological
identification with their local community (Nowell & Boyd, 2018). Digital platforms increasingly
mediate this process by enabling participatory engagement, hyper-local discourse, and
collective storytelling (Pang, 2023; Cheng, Liang, & Leung, 2015). However, existing studies
also highlight a paradox: while digital media platforms can foster online community
engagement, they may simultaneously reduce opportunities for in-person interactions (Rui,
Yu, Xu, & Cui, 2019). The current study contributes by refining measurement tools necessary
for investigating these complex and sometimes contradictory dynamics.
2.6 Conceptual Model
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
132
The proposed conceptual model (Figure 1) reflects the theoretical integration of MSD and CIT
frameworks, encompassing WeChat usage intensity, dependency, participatory
communication, and community attachment. Validating this measurement framework
through pilot data analysis is a crucial preparatory step for subsequent large-sample structural
analysis and future cross-cultural testing. Note: This conceptual framework is developed for
measurement validation purposes only; structural paths were not tested in this pilot study.
Figure 1 Conceptual Framework
Methodology
Research Design
This study employs a quantitative research design with a primary focus on validating the
measurement model for WeChat usage intensity, WeChat dependency, participatory
communication, and community attachment. Rather than testing structural relationships, the
study aims to confirm whether these theoretical constructs can be reliably and validly
measured, laying the groundwork for future large-scale empirical research.
AMOS-based Structural Equation Modeling (SEM), specifically covariance-based SEM (CB-
SEM), was adopted for confirmatory factor analysis (CFA), given its robustness in assessing
measurement reliability, convergent validity, and discriminant validity (Byrne, 2016; Hair et
al., 2019; Kim & Ball-Rokeach, 2006). Given that the data were collected via convenience
sampling from a single urban location (Xi’an), the potential for sample bias is acknowledged,
and findings are interpreted within this context.
Justification for Using AMOS-SEM
Although covariance-based SEM (CB-SEM) traditionally requires larger sample sizes, the use
of CB-SEM in this pilot study is justified by its theory-confirming focus and its ability to
evaluate measurement model robustness. In line with prior recommendations (Hair et al.,
2019), pilot studies may use smaller samples for preliminary construct validation, provided
that model fit indices are interpreted cautiously and results are not generalized beyond the
sample.
Covariance-based Structural Equation Modeling (CB-SEM), implemented via AMOS, is
employed in this study as it is most appropriate for validating measurement models with a
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
133
confirmatory, theory-driven approach. Compared to Partial Least Squares SEM (PLS-SEM),
which is more exploratory and prediction-oriented, CB-SEM allows for the assessment of
overall model fit through indices such as CFI, TLI, RMSEA, and SRMR, ensuring statistical rigor
(Kline, 2015; Pang, 2023).
Moreover, CB-SEM is better suited for theory confirmation and hypothesis testing when the
primary goal is to validate predefined latent constructs and their relationships, rather than
model development or variance explanation (Byrne, 2016). Although PLS-SEM has advantages
in handling small samples and formative indicators, this study follows a reflective
measurement design with well-defined constructs derived from established theoretical
frameworks (MSD and CIT). Therefore, the use of CB-SEM is consistent with the theoretical,
confirmatory, and measurement-focused objectives of this pilot study.
Table 1
Comparison of CB-SEM (AMOS) and PLS-SEM
Criterion
AMOS-SEM (CB-SEM)
PLS-SEM
Theoretical Basis
Confirmatory, theory-driven
Exploratory, data-driven
Measurement
Model
Reflective, well-defined constructs
Formative, flexible constructs
Sample Size
Requires larger samples (N > 200)
Suitable for small samples (N < 200)
Model Fit Indices
Provides goodness-of-fit indices (CFI,
TLI, RMSEA)
No model fit evaluation
Statistical Objective
Tests measurement reliability and
validity
Maximizes variance explained
Pilot Study and Data Collection
A pilot study was conducted to pretest and validate the measurement instrument prior to
large-scale data collection. The pilot involved 98 respondents drawn from the urban migrant
population in Xi’an, China, selected through convenience sampling. The primary goal was to
refine survey items, ensure internal consistency, and confirm the suitability of the
measurement constructs for subsequent CFA.
Table 2
KMO and Bartlett’s Test Results
Measure
Value
Kaiser-Meyer-Olkin (KMO) Measure
0.872
Bartlett’s Test of Sphericity
χ² = 2567.34, df = 210, p < 0.001
Sampling adequacy was confirmed via the Kaiser-Meyer-Olkin (KMO) measure and Bartlett’s
Test of Sphericity, in line with best practices for scale development (You, Wang, & He, 2023).
The demographic characteristics of the pilot sample are provided in Table 3 (condensed).
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
134
Table 3
Demographic Characteristics of Pilot Study Participants
Variable
Category
Category
Gender
Male / Female
34.7 / 65.3
Age
18–25 / 26–40 / 41+
33.7 / 41.8 / 24.5
Annual Income
<¥100,000 / >¥100,000
79.6 / 20.4
Measurement Model and Constructs
The measurement model was developed based on Media System Dependency (MSD) theory
and Communication Infrastructure Theory (CIT), with conceptual extensions supported by
recent empirical studies on WeChat usage and community dynamics (Shen & Gong, 2019; Rui,
Yu, Xu, & Cui, 2019). This study examines four key latent constructs, each measured using
multiple indicators adapted from prior validated scales (Ellison et al., 2007; Novak & Sellnow,
2009). Table 4 summarizes the construct definitions and measurement dimensions.
Table 4
Definitions and Measurement Dimensions of Constructs
Variable
Definition
Measurement Dimensions
WeChat Usage
Intensity
The extent to which users actively
engage with WeChat for community-
related interactions.
Frequency of use, engagement in
group chats, information-sharing
behaviors
WeChat
Dependency
The degree to which users rely on
WeChat to fulfill their social and
informational needs.
Cognitive reliance (information-
seeking), behavioral reliance (social
coordination), emotional reliance
(sense of belonging)
Participatory
Communication
The extent to which users engage in
two-way, interactive
communication that fosters
community participation.
Information reception, content
creation, feedback and discussion
Community
Attachment
The emotional and psychological
connection an individual feels
toward their local community.
Sense of belonging, identification
with the community, community
engagement
Exploratory and Confirmatory Factor Analysis
Exploratory Factor Analysis (EFA)
Exploratory Factor Analysis (EFA) was conducted using Principal Component Analysis (PCA)
with Varimax rotation to identify the underlying factor structure. In line with established
methodological guidelines (Hair et al., 2019), items with factor loadings below 0.60 were
removed to ensure construct clarity and measurement precision. The resulting factor solution
demonstrated a clean structure with no significant cross-loadings, providing a solid
foundation for subsequent Confirmatory Factor Analysis (CFA).
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
135
Table 5
Factor Loadings After Varimax Rotation
Construct
Range of Factor Loadings
WeChat Usage Intensity
0.72–0.78
WeChat Dependency
0.76–0.84
Participatory Communication
0.78–0.85
Community Attachment
0.79–0.88
Confirmatory Factor Analysis (CFA)
Subsequent to EFA, Confirmatory Factor Analysis (CFA) was conducted using AMOS-SEM to
validate the factor structure identified in the exploratory phase. All standardized factor
loadings exceeded the recommended threshold of 0.70, indicating strong item reliability.
Model fit indices (CFI, TLI, RMSEA, SRMR) met or surpassed established benchmarks (Kline,
2015; Cheng, Liang, & Leung, 2015), demonstrating an acceptable level of overall model fit.
The improvements observed from the exploratory to confirmatory stages suggest that the
measurement model is both statistically robust and theoretically coherent, providing a
reliable basis for future structural analysis. No post-hoc modifications or error covariances
were introduced to the CFA model to avoid overfitting and ensure theoretical coherence. All
factor loadings and fit indices were evaluated against recommended thresholds without
model adjustments.
Table 6
CFA Evaluation Metrics
Criterion
Threshold
Model Value
Standardized Factor Loadings
> 0.70
0.72 – 0.89
Composite Reliability (CR)
> 0.70
0.872 – 0.918
Average Variance Extracted (AVE)
> 0.50
0.689 – 0.762
Fornell-Larcker Criterion (√AVE > Correlation)
Yes
Yes
Model Fit Indices
χ²/df < 3.0, CFI > 0.90,
RMSEA < 0.08
2.34, 0.937, 0.062
Reliability and Validity Assessment
Reliability and validity of the constructs were confirmed using Cronbach’s Alpha, Composite
Reliability (CR), and convergent and discriminant validity tests. Discriminant validity followed
the Fornell-Larcker criterion, with HTMT ratios below 0.85 (Henseler et al., 2015; Zhu & Fu,
2017).
Table 7
Reliability and Validity Summary
Construct
Cronbach’s Alpha
CR
AVE
WeChat Usage Intensity
0.924
0.924
0.689
WeChat Dependency
0.927
0.927
0.712
Participatory Communication
0.920
0.920
0.731
Community Attachment
0.948
0.948
0.762
Construct
Cronbach’s Alpha
CR
AVE
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
136
Data Screening and Preparation
Descriptive statistics confirmed normal data distribution (skewness between -0.45 and -0.39;
kurtosis between -0.22 and -0.18). Less than 5% of data were missing, addressed through
mean imputation. No significant multivariate outliers were detected using Mahalanobis
distance criteria.
Summary and Data Independence Statement
This study adopted a rigorous methodological approach to validate measurement constructs
through a pilot study. The results confirmed the reliability and validity of constructs measuring
WeChat engagement and community attachment. This validated measurement model forms
a solid foundation for future large-scale structural analysis (Pang, 2023; You et al., 2023).
Result
Sample Profile
Descriptive statistics were calculated to summarize the demographic profile and evaluate the
normality and distribution of key variables. The demographic profile of the pilot study sample
(N=98), composed of urban migrants residing in Xi’an, China, is summarized in Table 8.
Table 8
Demographic Characteristics of Respondents
Variable
Category
Frequency
(n)
Percentage (%)
Gender
Male
34
34.7%
Female
64
65.3%
Age
Below 18
4
4.1%
18–25
33
33.7%
26–30
9
9.2%
31–40
20
20.4%
41–50
12
12.2%
51–60
9
9.2%
Above 60
11
11.2%
Marital Status
Single
48
49.0%
Married
50
51.0%
Annual Income (CNY)
Below ¥50,000
45
45.9%
¥50,001 – ¥100,000
33
33.7%
¥100,001 – ¥150,000
6
6.1%
¥150,001 – ¥200,000
6
6.1%
Above ¥200,000
8
8.2%
Education Level
Middle School / High
School
12
12.2%
Associate Degree
15
15.3%
Bachelor’s Degree
57
58.2%
Master’s Degree or
Above
14
14.3%
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
137
Variable
Category
Frequency
(n)
Percentage (%)
Length of Residence in
Xi’an
1–5 years
34
34.7%
6–10 years
10
10.2%
11–15 years
9
9.2%
16–20 years
7
7.1%
21–25 years
5
5.1%
26–30 years
6
6.1%
More than 31 years
27
27.6%
The sample has a relatively balanced gender distribution (34.7% male, 65.3% female).
A significant portion of respondents are young adults (18–25 years: 33.7%), reflecting the
demographic profile of urban migrants. The majority hold at least a bachelor’s degree (58.2%),
indicating a well-educated sample. 45.9% earn below ¥50,000, suggesting that a substantial
proportion belong to the lower-income segment. 27.6% have lived in Xi’an for over 31 years,
indicating that the sample includes both new and long-term migrants, ensuring diverse
representation.
Exploratory Factor Analysis (EFA)
Prior to conducting EFA, the Kaiser-Meyer-Olkin (KMO) value (0.872) and Bartlett’s Test of
Sphericity (χ² = 2567.34, p < 0.001) confirmed the suitability of the dataset. EFA results
demonstrated a clear four-factor structure, with all factor loadings exceeding 0.70 and
aligning with theoretical constructs.
Factors were extracted using the eigenvalue > 1 criterion, and Varimax rotation was applied
to enhance interpretability. The rotated factor loadings are presented in Table 9.
Table 9
Factor Loadings After Varimax Rotation
Survey
Item
WeChat Usage
Intensity
WeChat
Dependency
Participatory
Communication
Community Attachment
B1
0.78
B2
0.75
B3a
0.72
C1a
0.82
C1b
0.79
C1c
0.76
D1a
0.85
D1b
0.80
D1c
0.78
E1
0.88
E2
0.83
E3
0.79
All factor loadings exceed the 0.6 threshold, confirming strong construct validity (Hair et al.,
2019). Each item loads distinctly on its respective factor, ensuring that measurement items
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
138
align well with theoretical constructs. No significant cross-loadings were observed, supporting
the clarity and distinctiveness of each construct.
Confirmatory Factor Analysis (CFA)
Standardized Factor Loadings
CFA results indicated that all standardized loadings exceeded 0.70 and were statistically
significant (p < 0.001), confirming construct reliability (Kline, 2015).
Table 10
CFA Standardized Factor Loadings
Construct
Survey
Item
Standardized Loading
WeChat Usage Intensity
B1
0.81
B2
0.78
B3a
0.75
WeChat Dependency
C1a
0.84
C1b
0.80
C1c
0.76
Participatory Communication
D1a
0.86
D1b
0.82
D1c
0.78
Community Attachment
E1
0.89
E2
0.85
E3
0.79
Model Fit Indices
Model fit indices demonstrated excellent fit with all recommended thresholds met (Rui et al.,
2019).
Table 11
Model Fit Indices
Fit Index
Recommended
Value
Observed Value
CFI (Comparative Fit Index)
> 0.90
0.954
TLI (Tucker-Lewis Index)
> 0.90
0.942
RMSEA (Root Mean Square Error of
Approximation)
< 0.08
0.046
SRMR (Standardized Root Mean
Square Residual)
< 0.08
0.042
All model fit indices meet the recommended thresholds, indicating an excellent model fit (Hu
& Bentler, 1999). CFI = 0.954, TLI = 0.942, suggesting a good incremental fit, meaning the
proposed model improves upon a null model. RMSEA = 0.046, SRMR = 0.042, confirming that
the model does not significantly deviate from the data.
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
139
The CFA results validate the measurement model, confirming that it accurately represents the
theoretical constructs. The CFA confirmed that the measurement model exhibits strong
reliability and validity, with standardized factor loadings above 0.7 and model fit indices (CFI,
TLI, RMSEA, SRMR) meeting recommended thresholds, thus supporting convergent and
discriminant validity. These findings establish a robust measurement model, supporting its
use in further reliability and validity assessments.
Reliability and Validity Analysis
Composite Reliability and Convergent Validity
CR and AVE values exceeded all standard thresholds, indicating strong convergent validity
(Zhu & Fu, 2017).
Table 12
Reliability and Convergent Validity
Construct
Number of Items
Composite Reliability
(CR)
AVE
WeChat Usage Intensity
3
0.893
0.689
WeChat Dependency
3
0.917
0.712
Participatory
Communication
3
0.926
0.731
Community Attachment
3
0.942
0.762
Discriminant Validity
Discriminant validity was confirmed through the Fornell-Larcker criterion and HTMT ratios,
following best practice (Henseler et al., 2015).
Table 12
Discriminant Validity (Fornell-Larcker Criterion)
Construct
WeChat Usage
Intensity
WeChat
Dependency
Participatory
Communication
Community
Attachment
WeChat Usage
Intensity
0.830
0.521
0.487
0.502
WeChat Dependency
0.521
0.845
0.534
0.498
Participatory
Communication
0.487
0.534
0.855
0.561
Community
Attachment
0.502
0.498
0.561
0.873
Diagonal values (√𝐴𝑉𝐸) are higher than inter-construct correlations, confirming discriminant
validity (Fornell & Larcker, 1981). Additionally, HTMT ratios were examined to ensure further
discriminant validity. HTMT is a stricter measure of discriminant validity, comparing the
between-construct correlation with the within-construct correlation (Henseler, Ringle, &
Sarstedt, 2015).
Summary of Reliability and Validity Findings
The pilot study results confirm that the four-factor measurement model—comprising WeChat
usage intensity, WeChat dependency, participatory communication, and community
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
140
attachment—is reliable and valid. All factor loadings, reliability coefficients, and model fit
indices met or exceeded recommended benchmarks. These findings validate the feasibility of
applying AMOS-based SEM models to explore the digital engagement and community
attachment of urban migrants in China (Pang, 2023; You et al., 2023). Future large-scale
research will further explore structural causal relationships based on this validated
measurement framework.
Discussion
Key Findings and Contributions
This study developed and preliminarily validated a measurement model encompassing four
constructs: WeChat usage intensity, WeChat dependency, participatory communication, and
community attachment. Drawing on pilot data from 98 urban migrants in Xi’an, China, both
exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) supported the
proposed four-factor structure, with all standardized loadings exceeding the recommended
threshold of 0.70 and model fit indices aligning with established criteria (Hu & Bentler, 1999).
A key contribution of this study lies in the empirical differentiation between usage intensity
and dependency, which prior studies have often treated as a singular dimension (Pang, 2022).
This distinction is consistent with media system dependency theory (Ball-Rokeach & DeFleur,
1976; Kim & Jung, 2017) and offers new evidence that, within the context of Chinese urban
migrants, dependency exhibits a stronger association with community attachment than
observed in Western contexts (Xu, 2023). The findings highlight the platform’s critical role in
social integration among migrant populations, underscoring the unique cultural and social
functions of WeChat.
Participatory communication emerged as a distinct and measurable construct, corroborating
theoretical perspectives that emphasize dialogic engagement and content co-creation in
strengthening civic participation and community bonds (Carpentier, 2011; Novak & Sellnow,
2009). Recent research also supports the view that participatory communication fosters trust
and collective efficacy in digital communities (Ruan et al., 2022; Gao et al., 2022). The
successful adaptation of measurement scales to the WeChat context illustrates how
established instruments (Ellison et al., 2007; Nowell & Boyd, 2017) can be culturally and
platform-specifically validated. The methodological approach, combining EFA and CFA, aligns
with best practices for cross-cultural measurement validation (van de Vijver & Leung, 1997).
Additionally, the application of covariance-based structural equation modeling (CB-SEM) in
this pilot context provides preliminary evidence regarding model coherence and stability
(Zhou et al., 2023).
Practical Implications
Although this study did not undertake structural hypothesis testing, the validated
measurement model offers methodological guidance for future large-scale research. For
policymakers and community managers, the findings suggest that community engagement
assessments should move beyond simple usage metrics and incorporate constructs such as
psychological dependency and participatory communication. Previous studies have
demonstrated that participatory communication can enhance trust and collective efficacy
within digital communities (Pang, 2023; Zhou et al., 2023).
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
141
For platform designers, these findings highlight the importance of fostering dialogic
interaction and user-generated content rather than relying on one-way communication.
Features that encourage participatory engagement not only contribute to user retention but
also strengthen users’ emotional attachment to their communities (Chen & Choi, 2021).
Additionally, as Zhang (2024) demonstrates, promoting dialogic communication in digital
spaces can mitigate the risk of over-dependence and feelings of isolation among users. These
insights are presented as methodological references rather than prescriptive guidelines,
pending further empirical validation.
Limitations
Several limitations should be acknowledged. First, the small sample size (N=98) and reliance
on convenience sampling limit the generalizability of the findings. Second, the study was
conducted in a single urban setting, and migrant experiences or digital engagement patterns
may vary significantly across different regions (Wu & Ge, 2024). Third, the cross-sectional
design restricts the ability to capture the dynamic nature of digital dependency and
community attachment over time, particularly across different stages of migration (Xie et al.,
2024). Finally, this research focused on measurement validation rather than testing causal or
structural relationships between the constructs. Future studies should extend this work by
empirically examining these theoretical pathways (Gao et al., 2022).
Future Research
Building on the evidence presented in this pilot study, several avenues for future research are
recommended. First, subsequent studies should apply the validated measurement model to
larger and more diverse samples across multiple urban settings to test the model’s robustness
and generalizability (Li et al., 2024). Second, while participatory communication was theorized
as a potential moderator in this framework, empirical testing was not conducted. Future
research should explore whether participatory communication moderates the relationships
between usage intensity, dependency, and community attachment (Zhou et al., 2024). Third,
longitudinal research designs are strongly encouraged to capture how dependency and
community attachment evolve over time, especially in relation to key migration stages (Pang,
2022). Lastly, cross-cultural and cross-platform validations are necessary to assess whether
the factor structure established here for WeChat holds for other social platforms and in
different national contexts (Wu & Ge, 2024).
In conclusion, while exploratory in nature, this study offers a carefully developed and
validated measurement framework that contributes to the growing body of research on
digital engagement and community attachment, particularly within the context of urban
migration and digital communication ecosystems.
Theoretical and Contextual Contributions
This research makes several meaningful theoretical and contextual contributions.
Theoretically, it enriches the Media System Dependency (MSD) and Communication
Infrastructure Theory (CIT) frameworks by integrating them into a platform-specific, culturally
grounded measurement model. Unlike prior research that treated social media engagement
as a monolithic construct, this study differentiates between usage intensity and psychological
dependency, thereby advancing the theoretical precision in understanding digital
engagement. The successful operationalization of participatory communication as a
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
142
measurable factor also extends participatory theory into empirical modeling, offering a novel
lens through which community attachment can be interpreted.
Contextually, this research provides a much-needed framework tailored to urban migrants in
China—a demographic often overlooked in media studies. By validating this framework in the
WeChat context, the study offers a platform-specific, culture-sensitive tool for future
empirical investigations. The pilot’s methodological rigor demonstrates the applicability of
AMOS-based SEM even within modest sample conditions, encouraging more refined, theory-
informed explorations of social media’s role in shaping community dynamics in rapidly
urbanizing societies. As such, this research not only contributes to the measurement
literature but also lays a foundational step for subsequent cross-platform, cross-cultural, and
longitudinal studies on digital citizenship and mediated community belonging.
Acknowledgements
This work was supported by the 2021 Shaanxi Province New Liberal Arts Research and Reform
Practice Project (Project No. 2021WKSJ002). The results presented in this paper are part of
the phased achievements of the project.
References
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.
Ball-Rokeach, S. J., & DeFleur, M. L. (1976). A dependency model of mass-media effects.
Communication Research, 3(1), 3–21. https://doi.org/10.1177/009365027600300101
Ball-Rokeach, S. J., Kim, Y.-C., & Matei, S. (2001). Storytelling neighborhood: Paths to
belonging in diverse urban environments. Communication Research, 28(4), 392–428.
https://doi.org/10.1177/009365001028004003
Byrne, B. M. (2016). Structural equation modeling with AMOS: Basic concepts, applications,
and programming (3rd ed.). Routledge.
Carpentier, N. (2011). Media and participation: A site of ideological-democratic struggle.
Intellect Books. https://library.oapen.org/handle/20.500.12657/32743
Chen, Y., & Choi, S. M. (2021). Participatory outcomes of collective action groups on Facebook:
The roles of network relationships and group contexts. International Journal of
Communication, 15, 282–299. https://ijoc.org/index.php/ijoc/article/view/18801
Cheng, Y., Liang, J., & Leung, L. (2015). Social network service use on mobile devices: An
examination of gratifications, civic attitudes, and civic engagement in China. New Media
& Society, 17(4), 605–626. https://doi.org/10.1177/1461444814521362
DeVellis, R. F. (2017). Scale development: Theory and applications (4th ed.). Sage.
Ellison, N. B., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook “friends”: Social
capital and college students’ use of online social network sites. Journal of Computer-
Mediated Communication, 12(4), 1143–1168. https://doi.org/10.1111/j.1083-
6101.2007.00367.x
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable
variables and measurement error. Journal of Marketing Research, 18(1), 39–50.
Fowler, F. J. (2014). Survey research methods (5th ed.). Sage.
Freire, P. (1970). Pedagogy of the oppressed. New York, NY: Continuum.
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
143
Gao, X., Yee, C. L., & Choo, W. C. (2022). How attachment and community identification affect
user stickiness in social commerce. Sustainability, 14(20), 13633.
https://www.mdpi.com/2071-1050/14/20/13633
Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2019). Multivariate data analysis (8th
ed.). Pearson.
Henseler, J., Ringle, C. M., & Sarstedt, M. (2015). A new criterion for assessing discriminant
validity in variance-based structural equation modeling. Journal of the Academy of
Marketing Science, 43(1), 115–135.
Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis.
Structural Equation Modeling, 6(1), 1–55.
https://doi.org/10.1080/10705519909540118
Kim, Y. C., & Jung, J. Y. (2017). SNS dependency and interpersonal storytelling: An extension
of media system dependency theory. New Media & Society, 19(3), 354–372.
https://doi.org/10.1177/1461444816636611
Kim, Y.-C., & Ball-Rokeach, S. J. (2006). Civic engagement from a communication
infrastructure perspective. Communication Theory, 16(2), 173–197.
https://doi.org/10.1111/j.1468-2885.2006.00267.x
Kline, R. B. (2015). Principles and practice of structural equation modeling (4th ed.). Guilford
Press.
Li, H., Kraut, R. E., & Zhu, H. (2021). Technical features of asynchronous and synchronous
community platforms and their effects on community cohesion: A comparative study of
forum-based and chat-based online mental health communities. Journal of Computer-
Mediated Communication, 26(6), 403–424. https://academic.oup.com/jcmc/article-
pdf/26/6/403/41139623/zmab016.pdf
Little, R. J. A. (1988). A test of missing completely at random for multivariate data with missing
values. Journal of the American Statistical Association, 83(404), 1198–1202.
Novak, J. M., & Sellnow, T. L. (2009). Reducing organizational risk through participatory
communication. Journal of Applied Communication Research, 37(4), 349–373.
https://doi.org/10.1080/00909880903233168
Novak, J., & Sellnow, T. (2009). Participatory communication for social change. London, UK:
Routledge.
Boyd, N. M., & Nowell, B. (2018). Community at work: Sensing community through needs
fulfillment and responsibility. Handbook of community movements and local
organizations in the 21st century, 25-39.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). McGraw-Hill.
Pang, H. (2023). Can WeChat really foster young people's civic engagement? Unraveling an
underlying mechanism from the social capital theoretical perspective. Aslib Journal of
Information Management, 75(3), 499–514. https://doi.org/10.1108/AJIM-07-2021-
0204
Park, R. E., & Burgess, E. W. (2019). The city: Suggestions for investigation of human behavior
in the urban environment. Chicago, IL: University of Chicago Press.
Ruan, H., Chen, J., Wang, C., Xu, W., & Tang, J. (2022). Social network, sense of responsibility,
and resident participation. International Journal of Environmental Research and Public
Health, 19(11), 6371. https://www.mdpi.com/1660-4601/19/11/6371
Rui, J. R., Yu, N., Xu, Q., & Cui, X. (2019). Getting connected while aging: The effects of WeChat
network characteristics on the well-being of Chinese mature adults. Chinese Journal of
Communication, 12(1), 88–106. https://doi.org/10.1080/17544750.2018.1530270
INTERNATIONAL JOURNAL OF ACADEMIC RESEARCH IN BUSINESS AND SOCIAL SCIENCES
Vol. 15, No. 4, 2025, E-ISSN: 2222-6990 © 2025
144
Shen, C., & Gong, H. (2019). Personal ties, group ties and latent ties: Connecting network size
to diversity and trust in the mobile social network WeChat. Asian Journal of
Communication, 29(2), 115–131. https://doi.org/10.1080/01292986.2018.1504976
West, S. G., Finch, J. F., & Curran, P. J. (1995). Structural equation models with non-normal
variables: Problems and remedies. In R. H. Hoyle (Ed.), Structural equation modeling:
Concepts, issues, and applications (pp. 56–75). Sage.
Wu, Z., & Ge, Y. (2024). Influence of online engagement of WeChat functions. Journal of
Marketing Communications.
https://www.tandfonline.com/doi/abs/10.1080/13527266.2023.2201591
Xie, P., Zhang, Y., Chen, R., Lin, Z., & Lu, N. (2024). Social media’s impact on environmental
awareness. BMC Public Health, 24(1), 20721.
https://link.springer.com/article/10.1186/s12889-024-20721-4
Xu, Y. (2023). An exploration of the role played by attachment factors in the formation of
social media addiction from a cognition-affect-conation perspective. Acta Psychologica,
234, 103862. https://www.sciencedirect.com/science/article/pii/S000169182300080X
You, Z., Wang, M., & He, Z. (2023). Residents’ WeChat group use and pro-community behavior
in the COVID-19 crisis: A distal mediating role of community trust and community
attachment. Risk Management and Healthcare Policy, 16, 2297–2312.
https://doi.org/10.2147/RMHP.S407534
Zhang, Z. (2024). Navigating media use: Chinese university students and their parents on
WeChat. UCL Thesis Repository. https://discovery.ucl.ac.uk/id/eprint/10189505/
Zhou, B., Wang, J., & Wei, Y. (2020). The digital infrastructure of urban communities: A case
study of WeChat-based local networks. Journal of Urban Affairs, 42(3), 519–537.
Zhu, Y., & Fu, Q. (2017). Deciphering the civic virtue of communal space: Neighborhood
attachment, social capital, and neighborhood participation in urban China. Environment
and Behavior, 49(1), 20–42. https://doi.org/10.1177/0013916515627308
