From Expectation and Participation to Satisfaction: The Moderating Role of Perceived Government Responsiveness in Digital Government

Abstract

This study examines the mechanisms shaping citizen satisfaction in the context of digital government, taking Guangdong Province’s highly centralized “Yue Sheng Shi” platform as a case study. Building on the American Customer Satisfaction Index (ACSI) framework, a structural model was tested with survey data from 647 respondents and variance-based structural equation modeling. The results indicate that digital service expectations and citizen participation both enhance perceptions of service quality, with participation showing the stronger influence. Higher perceived service quality leads to greater citizen satisfaction, while government responsiveness strengthens this relationship. These research findings enrich the theoretical understanding of how satisfaction with e-government services is formed and extend the application of the ACSI framework to the Chinese digital governance context, while offering practical implications for governments on managing expectations, promoting citizen participation, and enhancing responsiveness in building citizen-centered digital platforms.

1. Introduction

With the continuous development of information and communication technologies (ICTs) and the transformation of digital public service models, governments around the world are actively advancing e-government initiatives aimed at improving the accessibility, efficiency, and quality of public services (Eom & Lee, 2022; Latupeirissa et al., 2024). By integrating online service platforms and mobile applications, governments are able to go beyond traditional administrative processes, providing intelligent approval systems, automated responses, enhanced data governance, and personalized services, thus improving administrative efficiency, shortening service delivery times, and making public services more accessible (van Noordt et al., 2023; Vrabie, 2025; Valackiene & Giedraitiene, 2024). These technological advancements not only optimize the service delivery model but also provide more channels for citizen participation, changing the interaction model between government and citizens (Shin et al., 2024).

However, despite these technological upgrades providing citizens with more opportunities to engage in public affairs and changing the interaction model between government and citizens, the practice of digital government has still not fully met citizens’ core expectations, especially in terms of improving public satisfaction. Specifically, these technological advancements have provided more channels for interaction, but if the government fails to respond effectively to citizens’ needs, the advantages of technology cannot be transformed into improved service quality and increased citizen satisfaction. Through these platforms, citizens not only have the opportunity to express their opinions and provide feedback, but also actively participate in policy formulation, service design, and implementation processes, thereby enhancing transparency and accuracy (Asimakopoulos et al., 2025; Randma-Liiv, 2022; Scupola & Mergel, 2022). However, some services still do not align with citizens’ needs, particularly in terms of service supply, quality, and response speed. For example, disparities in service accessibility across regions, mismatches between platform design and user needs, and insufficient timeliness in government responses to public concerns all directly impact citizen satisfaction (Beh et al., 2022; Y. Cheng & Zheng, 2023; Ma & Wu, 2020). These issues are not solely due to inadequate technological capacity but reflect a deeper misalignment between digital service design and citizens’ needs, making the improvement of citizen satisfaction an urgent core issue to address.

Therefore, to bridge this gap and improve the effectiveness of digital public services, not only are technological upgrades required, but also the implementation of a citizen-centered service philosophy. In this process, citizen participation and government responsiveness are crucial in bridging the gap between expectations and actual experiences (Mo & Beh, 2025; Sjoberg et al., 2017; Vrabie, 2025). Currently, despite the unprecedented opportunities digital technologies offer for citizen participation, there is still a lack of a systematic framework that can comprehensively explore the relationships between citizens’ expectations, participation behaviors, service quality perceptions, government responsiveness, and satisfaction.

At the theoretical level, various standardized assessment systems have been developed internationally to evaluate the effectiveness of public services. Among them, the American Customer Satisfaction Index (ACSI) was introduced into the U.S. federal government’s service performance evaluation system as early as 1999, serving as an important instrument for measuring user satisfaction with digital services (Fornell et al., 1996; Morgeson et al., 2023). ACSI, together with a series of service quality evaluation models, has provided a significant theoretical foundation for subsequent research, particularly in exploring the causal pathway from service expectation to perceived quality and ultimately to satisfaction (T. Li & Wang, 2021; Tian et al., 2025). Although originally developed in the U.S., the ACSI framework has also been applied in the Chinese public service context, including studies on government service delivery and local governance, demonstrating its cross-cultural applicability (D. Cheng, 2021; Q. Wang et al., 2020). Building on this foundation, the present study applies the ACSI model to digital governance in China, extending it by incorporating citizen participation and government responsiveness, and thereby constructing a citizen-centered analytical framework.

The contributions of this study are twofold. First, it advances e-government service satisfaction research by integrating the service quality perspective from classical satisfaction models (e.g., ACSI) with the dimensions of citizen participation and government responsiveness, thereby constructing a citizen-centered analytical framework tailored to the digital government context. This integration has narrowed the gap between traditional service evaluation methods and the participatory and interactive nature of contemporary digital governance. Secondly, through empirical analysis, this study explores how digital service expectations, citizens’ digital participation, perceived service quality, and government responsiveness jointly influence citizen satisfaction, and also identifies the direct correlations and moderating effects among these variables. These findings not only deepen our theoretical understanding of the mechanism of satisfaction formation in digital public services but also provide practical guidance for improving service design, refining response strategies, and optimizing the overall citizen experience.

2. Literature Review

2.1. Digital Service Expectations

In the performance evaluation of e-government, one metric that garners significant attention is citizen satisfaction, which is widely regarded as a core criterion for assessing the effectiveness of service delivery. The formation of this satisfaction is not arbitrary; it is influenced by antecedent variables such as service expectations and perceived service quality (Morgeson et al., 2023). Digital service expectations refer to users’ preconceived perceptions of public digital services prior to their actual use. These expectations play a crucial role, not only serving as a psychological benchmark during service evaluation but also subtly influencing the cognitive process through which people interpret service outcomes (Badri et al., 2015; Chatterjee & Suy, 2019). Examining the American Customer Satisfaction Index (ACSI) model reveals that expectations are positioned at the inception of the causal chain, directly influencing perceptions of service quality and judgments of service value (Morgeson, 2013; T. Li & Wang, 2021). A substantial body of prior research has consistently demonstrated a significant pattern: when actual service performance meets or exceeds customer expectations, evaluations of service quality tend to be markedly more positive. Conversely, when there is a substantial gap between actual performance and expectations, perceived service quality experiences a notable decline (Qin et al., 2025; Tian et al., 2025).

In recent years, the increasing application of emerging technologies such as artificial intelligence, big data, and cloud computing in e-government platforms has significantly elevated public expectations for service quality, particularly in terms of convenience, intelligence, and personalization (OECD, 2024). Consequently, for governments, the effective establishment, timely communication, and proper management of these expectations can substantially reduce the gap between public anticipation and actual experience, thereby naturally enhancing the perceived quality of services (Guo et al., 2025; Z. Li & Xu, 2017).

In summary, both theoretical and empirical studies suggest that digital service expectations play a crucial role in shaping citizens’ perceptions of service quality. Therefore, the following hypothesis is proposed:

H1. 

Digital service expectations have a significant positive effect on perceived service quality.

2.2. Citizen Digital Participation

Citizen digital participation has emerged as a critical factor in enhancing the performance of digital governance (Fledderus et al., 2015; Sorrentino et al., 2018). Just like in the “user-driven service co-production” model, citizen participation no longer remains at the stage of service consumption as before, but extends to the entire process, including policy-making, service design, and post-implementation feedback (Scupola & Mergel, 2022; Cordella & Paletti, 2017). Digital platforms have played a significant role in this regard, offering a wide range of participation tools such as public consultations, various surveys, and open feedback channels. With these tools, the depth and breadth of citizen participation have been expanded. Gradually, the role of citizens has also changed, from being mere “information receivers” to “cooperators” and even “co-decision makers” (Clifton et al., 2020; Scupola & Mergel, 2022).

Moreover, the benefits of high participation are obvious. It enables more frequent interaction between the government and citizens, makes information more symmetrical, and reduces uncertainties in the service process. As a result, people’s perception of service quality can be significantly enhanced (Meijer & Bolívar, 2016; Porumbescu, 2015). Especially in digital environments such as intelligent approval systems, online customer support, and real-time feedback platforms, those who actively participate tend to more easily perceive high levels of service in terms of accessibility, responsiveness, and personalization (Xin et al., 2022). However, a critical factor lies in the depth of participation, which directly determines the efficacy of engagement. As Lee and Kwak (2012) have elucidated in their open government maturity model, participation ranges from basic information access (data transparency and open participation) to deeper collaborative engagement (open collaboration). Superficial information-oriented participation, such as merely acquiring basic information, often fails to significantly enhance public perception of service quality. In contrast, when citizens engage deeply in the co-design and feedback processes of services, the scenario transforms markedly, enabling service design to better align with public needs. In line with this distinction, the present study measured citizen participation through items reflecting both informational participation (e.g., frequency of platform use) and co-design participation (e.g., provision of suggestions or feedback).

Taken together, the evidence indicates that citizen participation, particularly deeper forms of engagement such as co-design and feedback, plays a vital role in shaping perceptions of service quality. Accordingly, the following hypothesis is proposed:

H2. 

Citizen digital participation has a significant positive effect on perceived service quality.

2.3. The Impact of Perceived Service Quality on Satisfaction

In the evaluation of e-government service performance, citizens’ subjective assessments of digital service quality are generally regarded as a key direct factor influencing satisfaction (Špaček & Špačková, 2022). Perceived digital service quality refers to the overall evaluation of the services and information provided by digital government platforms by users, especially in terms of system functionality, information reliability, and service responsiveness (Chatterjee & Suy, 2019). Specifically, it encompasses multiple dimensions: information quality (whether the information is comprehensive and timely), credibility (whether the information is perceived as authoritative and reliable), usability (whether the platform is user-friendly and easy to navigate), and efficiency (whether the services are concise and delivered quickly) (Chan et al., 2021; T. Cheng et al., 2021; Nookhao & Kiattisin, 2023). These dimensions form the core of user experience and directly affect public satisfaction with digital government services (Špaček & Špačková, 2022). Classic models such as the American Customer Satisfaction Index (ACSI) regard perceived service quality as a critical antecedent factor driving user satisfaction. This indicates that higher perceived digital service quality typically correlates with elevated satisfaction levels.

Numerous empirical studies have confirmed that there is a significant positive relationship between the perceived quality of e-government services and citizen satisfaction (Badri et al., 2015; Qin et al., 2025; C. Wang & Ma, 2022). For instance, research has shown that enhancing the service quality of e-government platforms can significantly increase user satisfaction. The better the service quality, the higher the level of citizen satisfaction naturally becomes (Chatterjee & Suy, 2019; Nookhao & Kiattisin, 2023). Further research has investigated the role of mediating variables, such as trust, in the relationship between perceived quality and satisfaction. The findings reveal that digital service quality significantly enhances both trust and satisfaction; however, trust does not exhibit a notable mediating effect. This indicates that improving service quality alone can directly elevate satisfaction levels (Lanin & Hermanto, 2018; Hu et al., 2020; Xin et al., 2022).

Over the years, the private sector has continuously optimized the user experience of digital services, raising citizens’ expectations for efficient and intuitive service delivery. This has also extended to the public sector, where there are higher demands for information integrity, system usability, and service efficiency of government platforms. However, research shows that the satisfaction rate of digital public services is sometimes about 21 percentage points lower than that of the private sector (Global Government Forum, 2025), indicating that there is still room for improvement in platform design and service provision. To address these challenges, governments around the world have adopted digital strategies such as integrating platforms, optimizing interfaces, and enhancing system response speeds, with the aim of improving the user experience of e-government services and strengthening the public’s positive perception of service quality (Mao & Zhu, 2025; Latupeirissa et al., 2024).

In summary, perceived digital service quality is a critical antecedent influencing citizen satisfaction, occupying a central position in the causal pathways of service satisfaction models. Digital government platforms that provide more comprehensive and authoritative information, offer more user-friendly interfaces, and deliver services more efficiently are better able to meet citizens’ needs, thereby increasing satisfaction. This factor reflects the process through which digital government enhances citizens’ sense of gain by improving the quality of service delivery. Based on this, the following hypothesis is proposed:

H3. 

Perceived service quality has a significant positive effect on citizen satisfaction.

2.4. Government Responsiveness as a Moderating Variable

In e-government research, government responsiveness is widely regarded as one of the key factors influencing citizen satisfaction. It generally refers to the speed and quality with which governments respond to public needs, feedback, and opinions (Su & Meng, 2016). Traditional service management theories also consider responsiveness a core dimension of service quality, emphasizing whether service providers can promptly, accurately, and effectively meet user needs (Parasuraman et al., 1988). Within the context of digital public service, government responsiveness is reflected not only in the efficiency of responding to online inquiries, complaints, and suggestions from the public but also in the transparency, professionalism, and credibility of the response content (Mishra, 2020).

Existing research suggests that the mere presence of a government response is not sufficient to enhance citizen satisfaction; rather, it is the way in which the response is delivered—particularly whether it effectively addresses the public’s core concerns—that serves as a critical mechanism shaping perceptions of service performance (Su & Meng, 2016). This underscores the importance of value congruence, meaning that the greater the alignment between the content of government responses and public expectations, the stronger the relationship between perceived service quality and citizen satisfaction. In this regard, a study by Nie and Wang (2023) in the context of environmental governance found that when governments responded to public concerns with concrete actions or reasonable explanations, satisfaction levels significantly increased.

Conversely, if the government does not respond or merely gives perfunctory replies, citizens’ overall evaluation of government services will be discounted. Following this line of research, some scholars have further proposed that the government’s responsiveness not only directly affects satisfaction but may also play a moderating role in other key factors in scenarios such as digital public services.

If citizens perceive the government’s response as timely, transparent, and professionally reliable, they are more likely to appreciate the value of public services. In this case, the positive impact of perceived service quality on satisfaction will be stronger (Jiang & Fan, 2024). This mechanism can be understood as a “value congruence” path: that is, when the government’s response closely aligns with citizens’ core needs and expectations, people are more willing to convert their positive impression of service quality into a satisfied evaluation (Shen et al., 2023). Therefore, in some circumstances, government responsiveness may act as a background factor, making the connection between service quality and satisfaction more robust.

Moreover, in the absence of an effective response mechanism or when response quality is poor, objectively good services may not translate into user satisfaction. This may stem from citizens’ doubts regarding the government’s attitude and problem-solving capability. This phenomenon highlights that in the context of digital public service, the effect of service quality on satisfaction is not isolated; instead, it may be significantly moderated by the interactive characteristics of government behavior. Therefore, based on both theoretical reasoning and empirical evidence regarding the moderating role of government responsiveness, this study proposes the following hypothesis:

H4. 

Government responsiveness positively moderates the relationship between perceived service quality and citizen satisfaction.

The relationships between the independent, dependent, and moderating variables in this research are represented in Figure 1.

3. Research Design and Methodology

3.1. Background and Research Context

The “Yue Sheng Shi” platform was officially launched on 21 May 2018, as China’s first digital public service WeChat mini-program. It integrates multiple government services from 24 departments in Guangdong Province, providing a unified access point, standardized procedures, and a 24 h intelligent assistant (Cai, 2023). In December 2024, the “Yue Sheng Shi” platform processed an average of 3.236 million transactions and inquiries per day, with the total number of real-name registered users reaching 190.5 million (Guangdong Provincial Government Affairs Service and Data Management Bureau, 2024). As a typical representative of provincial-level digital government platforms in China, “Yue Sheng Shi” has high research representativeness; therefore, it was chosen as the subject of this study.

3.2. Measures

Government responsiveness was measured based on the “responsiveness” dimension of the SERVQUAL model developed by Parasuraman et al. (1988) and the operationalization of responsiveness in digital government settings proposed by Su and Meng (2016), capturing perceptions of transparency in policies and service guidelines, service standardization and professionalism, trustworthiness, and the timeliness of governmental reactions to public needs (efficiency at the response level). Digital service expectations were assessed following the American Customer Satisfaction Index (ACSI) framework (Fornell et al., 1996) and e-government service expectation research by T. Li and Wang (2021), covering anticipated overall quality, reliability, convenience, and personalization before using the platform. Citizen digital participation was measured with reference to co-production behavior scales from Scupola and Mergel (2022) and Xin et al. (2022), and aligned with Lee and Kwak’s (2012) model distinguishing basic information use from deeper collaborative engagement, thus capturing both informational participation (e.g., frequency of platform use) and co-creative participation (e.g., feedback provision, interactive exchanges, and participation opportunities). Perceived service quality was measured following dimensions commonly applied in e-government studies (Chan et al., 2021; T. Cheng et al., 2021; Nookhao & Kiattisin, 2023; Špaček & Špačková, 2022), assessing information completeness and timeliness, authority and reliability (credibility), user-friendliness (usability), and the operational speed and simplicity of services (efficiency at the service delivery level). Citizen satisfaction was measured based on the ACSI framework (Fornell et al., 1996) and e-government satisfaction research by Morgeson et al. (2023) and Badri et al. (2015), evaluating overall satisfaction, comparisons with expected and ideal services, and satisfaction with handling of complaints and suggestions. All constructs were modeled as reflective indicators.

The questionnaire underwent expert review and a pilot test before the formal survey to ensure reliability and validity. A total of 97 valid responses were collected in the pilot test. The reliability analysis showed that all constructs had Cronbach’s α values ranging from 0.843 to 0.893, with all CITC values above 0.50, indicating strong internal consistency. The Kaiser–Meyer–Olkin (KMO) value was 0.823, and Bartlett’s test of sphericity was significant (χ2 = 1293.864, df = 190, p < 0.001), confirming the suitability of the data for factor analysis. These results demonstrate that the measurement scale had good reliability and construct validity, and thus the original items were retained for the main survey.

3.3. Sample and Procedure

This study adopted a convenience sampling method within the framework of random sampling, collecting data through an online questionnaire. Convenience sampling is a non-probability sampling method suitable when the target population is easily accessible, resources are limited, and the research purpose is exploratory (Etikan et al., 2015). In e-government platform research, this method enables the rapid collection of authentic feedback from a large number of actual users. The questionnaire was distributed from 1 July to 1 August 2023, via the “Yue Sheng Shi” platform and related online channels. Respondents were registered users from Guangzhou, Shenzhen, and Zhuhai. Before answering, participants were informed of the research purpose and confidentiality principles and provided informed consent. A total of 750 questionnaires were collected. The first question served as a screening item, and only responses from participants who had previously used the “Yue Sheng Shi” digital public service platform were retained, resulting in 647 valid responses and an effective response rate of 86.27%.

While convenience sampling may limit the strict generalization of results, it is a widely adopted approach in e-government research to capture actual platform users in a cost-effective and timely manner, especially when the population is large and resources are constrained (Etikan et al., 2015). By covering three major cities in Guangdong and obtaining a large and diverse sample, this study provides a reasonable empirical basis to reflect user experiences at the provincial level, though province-wide extrapolation should be made with caution.

3.4. Data Analysis Technique

Partial Least Squares Structural Equation Modeling (PLS-SEM) was employed for data analysis. Compared to covariance-based structural equation modeling (CB-SEM), PLS-SEM is more suitable for exploratory research and theories that are not yet fully developed (Rigdon, 2012), and it offers advantages in handling small sample sizes, non-normal data distributions, and models containing both reflective and formative constructs (Hair et al., 2011; Civelek, 2018). The data analysis was conducted in two stages: (1) Measurement model assessment: internal consistency was examined using Cronbach’s α and composite reliability (CR); convergent validity was assessed via average variance extracted (AVE); and discriminant validity was evaluated using the Fornell–Larcker criterion. (2) Structural model assessment: model fit and hypothesis significance were examined through path coefficients, coefficient of determination (R²) and predictive relevance (Q²), with significance tested using bootstrapping with 5000 resamples. All statistical analyses were performed using SPSS 22.0 and SmartPLS 4.0.

4. Results

4.1. Sample Characteristics

Table 1. Descriptive statistical analysis of respondents.

Frequencies (N = 647)
	Items	Frequency	Percent (%)	Cumulative Percent (%)
Gender	Male	306	47.3	47.3
	Female	341	52.7	100.0
Age	18 years old and below	77	11.9	11.9
	19–30 years old	130	20.1	32.0
	31–40 years old	148	22.9	54.9
	41–50 years old	128	19.8	74.7
	51–60 years old	113	17.5	92.1
	61 years old and above	51	7.9	100.0
Occupation	Farmer	62	9.6	9.6
	Laborer	72	11.1	20.7
	Individual businessman	96	14.8	35.5
	Retired	45	7.0	42.5
	Student	140	21.6	64.1
	Enterprise employee	75	11.6	75.7
	Organization and institution employee	123	19.0	94.7
	Other	34	5.3	100.0
Education level	Junior high school (junior college) and below	46	7.1	7.1
	High school	87	13.4	20.6
	Vocational/technical school	144	22.3	42.8
	Undergraduate	209	32.3	75.1
	Graduate and above	161	24.9	100.0
Household income range (RMB)	Less than 5000	48	7.4	7.4
	5001–10,000	56	8.7	16.1
	10,001–15,000	97	15.0	31.1
	15,001–20,000	128	19.8	50.9
	20,001–25,000	159	24.6	75.4
	25,001–30,000	104	16.1	91.5
	More than 30,000	55	8.5	100.0
Total		647	100.0	100.0

presents the socio-demographic characteristics of the respondents. Of the 647 participants, 52.7% were female and 47.3% were male. The age distribution was relatively balanced, with the largest proportions falling within the 31–40 (22.9%), 19–30 (20.1%), and 41–50 (19.8%) age groups. In terms of occupation, students (21.6%) and employees of public institutions/government agencies (19.0%) accounted for the largest shares, followed by individual business owners (14.8%), enterprise employees (11.6%), and laborers (11.1%). Regarding educational attainment, more than half of the respondents held a bachelor’s degree (32.3%) or a postgraduate degree (24.9%). Monthly household income was concentrated in the RMB 20,001–25,000 (24.6%) and RMB 15,001–20,000 (19.8%) brackets. Overall, the demographic distribution of the sample is relatively balanced and broadly representative of the primary user base of the “Yue Sheng Shi” platform. However, it should be noted that the sample as a whole is relatively more educated and has higher household income levels compared to the general population, which may introduce a potential upward bias in satisfaction evaluations.

Table 2. Descriptive statistics of the main constructs (N = 647).

Construct	Min	Max	Mean	SD
Digital Service Expectation (DSE)	1	5	2.93	1.01
Citizen Participation (CDP)	1	5	3.52	1.04
Perceived Service Quality (PSQ)	1	5	3.54	1.04
Citizen Satisfaction (CS)	1	5	3.18	0.98
Government Responsiveness (GR)	1	5	2.85	1.06

presents the descriptive statistics of the main constructs. On a five-point scale, perceived service quality (M = 3.54, SD = 1.04) and citizen participation (M = 3.52, SD = 1.04) were both above the scale midpoint, indicating generally positive evaluations of the digital government platform. Citizen satisfaction was moderate (M = 3.18, SD = 0.98), suggesting that respondents were somewhat satisfied overall but not highly enthusiastic. In contrast, digital service expectation (M = 2.93, SD = 1.01) and government responsiveness (M = 2.85, SD = 1.06) were slightly below the midpoint, implying that citizens had relatively cautious expectations and perceived limited responsiveness from the government. These findings provide a broader context for the subsequent structural equation modeling results by offering insights into the general tendencies of respondents’ opinions.

4.2. Measurement Analysis

Before assessing reliability and validity, potential collinearity and common method bias (CMB) were examined. As presented in

Table 3. Inner VIF values for structural paths.

Structural Path	VIF Value
CDP → PSQ	1.115
DSE → PSQ	1.115
PSQ → CS	1.172
GR × PSQ → CS	1.031

, the variance inflation factor (VIF) values for the structural paths in the inner model ranged from 1.031 to 1.172, all well below the conservative threshold of 3.3 (Kock, 2015). This indicates that multicollinearity is not a concern and that CMB is unlikely to bias the results.

As presented in

Table 4. Measurement model results: factor loadings, reliability, and AVE.

Construct	Item	Loading	Cronbach’s Alpha (α)	Composite Reliability	AVE
Citizen Digital Participation	CDP1	0.893	0.89	0.924	0.753
	CDP2	0.866
	CDP3	0.851
	CDP4	0.86
Digital Service Expectation	DSE1	0.876	0.863	0.907	0.709
	DSE2	0.813
	DSE3	0.842
	DSE4	0.836
Perceived Service Quality	PSQ1	0.897	0.895	0.927	0.761
	PSQ2	0.872
	PSQ3	0.853
	PSQ4	0.868
Government Responsiveness	GR1	0.886	0.879	0.917	0.734
	GR2	0.836
	GR3	0.851
	GR4	0.854
Citizen Satisfaction	CS1	0.882	0.848	0.898	0.688
	CS2	0.816
	CS3	0.81
	CS4	0.807

, the PLS path model analysis shows that all measures meet the commonly suggested criteria for measurement model assessment (Chin, 1998; Henseler et al., 2009; Hair et al., 2011). Convergent validity was first assessed by examining the standardized factor loadings, composite reliability (CR), and average variance extracted (AVE). As shown in Table 2, all standardized loadings are above the recommended threshold of 0.70 (p < 0.001), indicating strong indicator reliability. The AVE values for all constructs range from 0.688 to 0.761, exceeding the minimum criterion of 0.50 (Henseler et al., 2009; Bagozzi & Yi, 2012), thereby supporting convergent validity. Additionally, Cronbach’s alpha (α) values for all constructs are above 0.84, and CR values range from 0.898 to 0.927, surpassing the commonly accepted threshold of 0.80, thus demonstrating satisfactory internal consistency.

Discriminant validity was then assessed using the Fornell–Larcker criterion (Fornell & Larcker, 1981). As presented in

Table 5. Discriminant validity assessment.

	Citizen Digital Participation	Citizen Satisfaction	Digital Service Expectation	Government Responsiveness	Perceived Service Quality
Citizen Digital Participation	0.868
Citizen Satisfaction	0.419	0.829
Digital Service Expectation	0.322	0.42	0.842
Government Responsiveness	0.385	0.392	0.46	0.857
Perceived Service Quality	0.565	0.395	0.346	0.347	0.872

, the square root of each construct’s AVE (shown in bold on the diagonal) is greater than its correlations with any other construct (off-diagonal elements), indicating adequate discriminant validity. Taken together, the measurement model results confirm the reliability, convergent validity, and discriminant validity of all constructs in this study.

4.3. Evaluation of the Structural Model

To test the proposed hypotheses, this study employed the PLS-SEM method to evaluate the structural model and applied a bootstrapping procedure with 5000 resamples at a 5% significance level to examine the significance of the path coefficients. The explanatory power of the model was assessed using the coefficient of determination (R²). The results show that the adjusted R² value for perceived service quality (PSQ) is 0.347, and that for citizen satisfaction (CS) is 0.235, indicating that the model effectively explains the variance of the two endogenous variables.

The hypothesis testing results are presented in

Table 6. Results of hypothesis testing.

Hypothesized Path	Path Coefficient (β)	t-Value	p-Value	Supported?
CDP → PSQ	0.506	15.612	<0.001	Yes
DSE → PSQ	0.184	5.283	<0.001	Yes
PSQ → CS	0.312	8.801	<0.001	Yes
GR × PSQ → CS	0.095	2.859	0.004	Yes

and illustrated in Figure 2. Citizen digital participation (CDP) has a significant positive effect on perceived service quality (PSQ) (β = 0.506, t = 15.612, p < 0.001), supporting H1. Digital service expectation (DSE) also exerts a significant positive effect on PSQ (β = 0.184, t = 5.283, p < 0.001), supporting H2. Perceived service quality (PSQ) has a significant positive effect on citizen satisfaction (CS) (β = 0.312, t = 8.801, p < 0.001), supporting H3. In addition, the moderating effect of government responsiveness (GR) on the relationship between PSQ and CS is significant (β = 0.095, t = 2.859, p = 0.004), supporting H4.

Overall, all four hypotheses are supported, indicating that the proposed model can effectively explain the formation mechanism of citizen satisfaction in the context of digital government. The findings provide empirical evidence for a deeper understanding of the relationships among digital government service quality, government responsiveness, and citizen satisfaction.

4.4. Predictive Relevance (Q²)

This study assessed the predictive relevance of the structural model using the Q² statistic proposed by Stone and Geisser, calculated through the blindfolding procedure (Geisser, 1974). This method systematically omits every dth data point and predicts the omitted part, while ensuring that the chosen d results in a non-integer ratio of valid observations (Hair et al., 2019). Following established recommendations, the cross-validated redundancy measure was adopted, as it takes into account both the structural and measurement models (Chin, 1998; Henseler et al., 2009).

A Q² value greater than zero indicates that the model has predictive relevance for the given endogenous construct. The results of this study show that the Q² value for perceived service quality (PSQ) is 0.261 and for citizen satisfaction (CS) is 0.160, both exceeding zero, thereby confirming the predictive capability of the model.

5. Discussion

This study focuses on the influence mechanism of citizen digital participation (CDP) and digital service expectations (DSEs) on perceived service quality (PSQ), and further explores the moderating role of government responsiveness (GR) in the relationship between perceived service quality and citizen satisfaction (CS). We conducted a PLS-SEM analysis on 647 valid questionnaires from three cities. The results show that the research model has excellent model fit and predictive power, and all proposed hypotheses are supported by statistical data. Next, we will conduct an in-depth discussion of the empirical results of each hypothesis and conduct a comparative analysis with existing research to clarify the theoretical value and practical significance of this study.

5.1. The Effect of Digital Service Expectations on Perceived Service Quality (H1)

Hypothesis H1 posits that digital service expectation (DSE) has a significant positive effect on perceived service quality (PSQ). The empirical results show that the path coefficient for DSE → PSQ is β = 0.184 (t = 5.283, p < 0.001), indicating that the higher the service expectations citizens form before using a digital government platform, the more likely they are to perceive higher service quality during actual use.

This finding aligns with the conclusions of Morgeson (2013), based on the American Customer Satisfaction Index (ACSI) model, which suggest that expectations are not only an antecedent of satisfaction but also an important psychological benchmark for shaping perceptions of quality. When service performance meets or exceeds users’ psychological expectations, perceived quality tends to improve; conversely, when actual experiences fall short of expectations, perceived quality decreases significantly (Badri et al., 2015; T. Li & Wang, 2021). In the context of e-government, service expectations are shaped not only by users’ prior online government service experiences but also by the government’s signaling behaviors in platform development, promotion, and functional positioning (Chatterjee & Suy, 2019).

Existing studies provide further empirical support. Qin et al. (2025) and Tian et al. (2025) found that clear and reasonable service expectations help enhance users’ overall perceptions of digital public service quality. This is consistent with the OECD (2024) international comparative report, which notes that, in the context of widespread adoption of artificial intelligence, big data, and cloud computing, public expectations for convenience, intelligence, and personalization in e-government have risen significantly. E-government platforms that can effectively set and manage public expectations are more likely to receive positive quality evaluations during service delivery (Guo et al., 2025; Z. Li & Xu, 2017).

It is worth noting that the strength of the relationship between expectations and perceived quality may vary across institutional and technological contexts. In some developed countries, where government platforms are relatively mature and service experiences are stable, public service expectations have reached near-saturation, and the marginal gains of expectation increases on perceived quality are limited (Meijer & Bolívar, 2016). In contrast, in China, e-government is still in a phase of rapid iteration and upgrading, and public expectations for new features and service models remain highly dynamic. For example, the “Yue Sheng Shi” platform leverages cross-departmental data sharing, integrated workflows, and intelligent approval mechanisms to continuously introduce “beyond-expectation” services—such as integrated service processing, intelligent customer service, and personalized notifications (Cai, 2023), which not only optimize the user experience but also significantly amplify the positive impact of service expectations on perceived quality.

Therefore, this study provides both theoretical and empirical support for H1, confirming that digital service expectation is an important antecedent for improving perceived service quality. This conclusion not only responds to the ACSI-based theoretical hypothesis discussed in the literature review but also offers new empirical evidence for understanding how governments, in the context of rapid technological evolution and centralized governance, can enhance perceived quality by effectively managing public expectations.

Nevertheless, beyond convenience and personalization, which have been central to much of the existing discourse, broader contextual factors such as institutional arrangements (Randma-Liiv et al., 2022), policy environments (Breaugh et al., 2023), and technological trends (Ashaye & Irani, 2019) may also play a critical role in shaping how citizens form and adjust their expectations. Future research could further investigate these external influences to complement the present findings.

5.2. The Effect of Citizen Digital Participation on Perceived Service Quality (H2)

Hypothesis H2 posits that citizen digital participation (CDP) has a significant positive effect on perceived service quality (PSQ). The empirical analysis shows that the path coefficient for CDP → PSQ is β = 0.506 (t = 15.612, p < 0.001), which is not only statistically significant but also indicates a substantial effect size. This suggests that, within e-government platforms, the more actively citizens engage—by submitting opinions, providing feedback, participating in surveys, or collaborating online—the higher their perceived quality of government services.

This research finding is highly consistent with existing achievements. Morgeson et al. (2023) found in the context of local government research in the United States that citizen participation can significantly enhance the public’s perception of service efficiency, professionalism, and transparency. Similarly, Fledderus et al. (2015) and Scupola and Mergel (2022) also emphasized that active citizen participation can improve the alignment of services with public needs, reduce information asymmetry, and thereby enhance the responsiveness and personalization of services. The results of this study further validate the applicability of the “user-driven service co-production” theory in the context of China’s digital government (Cordella & Paletti, 2017; Clifton et al., 2020), that is, the quality of public services is not solely provided by the government but is co-created through the interaction between the government and users (Osborne et al., 2016).

Compared with existing studies, this research finds that citizen digital participation has a significant and relatively strong positive impact on perceived service quality. The previous literature indicates that differences in institutional environments and technological conditions can affect the efficiency of public participation in transforming into perceived service quality (Lee-Geiller & Lee, 2019; Meijer & Bolívar, 2016). In some Western contexts, issues such as insufficient platform integration, limited cross-departmental collaboration, and imperfect feedback mechanisms may restrict the enhancing effect of participation on perceived quality (Porumbescu, 2015; Thorsby et al., 2017). In contrast, China’s integrated digital government platforms (such as “Yue Shengshi”) demonstrate a high degree of integration and synergy in both institutional design and technical architecture. As Luo (2022) pointed out, these platforms break down traditional departmental information barriers through cross-departmental data sharing, process optimization, and intelligent approval mechanisms, significantly shortening the transmission chain from public feedback to policy adjustment. This system enables citizens’ opinions and collaborations provided through digital participation to be more directly embedded in service optimization and process improvement, thereby reducing information asymmetry and enhancing the responsiveness and personalization of public services.

Therefore, this study provides both theoretical and empirical support for H2, indicating that in the context of digital government, citizen participation is a key driver of improvements in perceived service quality. This conclusion not only confirms the hypothesized “digital participation–service quality perception” pathway outlined in the literature review but also offers empirical evidence for understanding how highly centralized digital government platforms can achieve effective value co-creation.

At the same time, it should be noted that the relationship between participation and perceived quality may not always be linear. External contextual factors, such as incentive structures shaping engagement (Omar et al., 2016) or overly optimistic socio-technical visions in technology adoption (Meijer & Bolívar, 2016), can distort participation patterns, suggesting that future research should explore these contingencies more deeply.

5.3. Perceived Service Quality and Citizen Satisfaction (H3)

Hypothesis H3 posits that perceived service quality has a significant positive effect on citizen satisfaction. The empirical results indicate that the path coefficient for PSQ → CS is β = 0.312 (t = 8.801, p < 0.001), suggesting that the higher the public’s evaluation of a e-government platform’s quality, the greater their overall satisfaction. This finding is consistent with multiple empirical studies, which demonstrate that in both developed countries and emerging economies, improvements in information completeness and authority, system usability, and service responsiveness in digital services can significantly enhance user satisfaction (Badri et al., 2015; Qin et al., 2025; C. Wang & Ma, 2022).

From a theoretical perspective, perceived service quality not only reflects the maturity of technical functions in e-government but also embodies the service’s overall performance in meeting public needs, reducing transactional frictions, and improving interaction efficiency (Chatterjee & Suy, 2019; Nookhao & Kiattisin, 2023). For example, when a platform provides authoritative, accurate, and timely information, complemented by a clear and intuitive interface and efficient service processes, users are more likely to experience smooth and appropriate handling of their affairs, thereby increasing their satisfaction.

It is important to note that the strength of this positive relationship may vary depending on institutional and environmental contexts. Existing research shows that in regions with higher degrees of fiscal decentralization, local governments enjoy greater autonomy in resource allocation and service design, enabling them to adjust information provision, interface design, and response mechanisms according to local needs—thereby further strengthening the influence of perceived quality on satisfaction (Y. Wang et al., 2023). By contrast, in highly centralized governance systems, public services often operate under uniform standards and policy frameworks, with more limited scope for local adaptation in service details, which may lead to different patterns in how perceived quality translates into satisfaction.

Nevertheless, the empirical evidence from this study confirms the significant positive impact of perceived service quality on citizen satisfaction within the Chinese context. This result suggests that even in a highly standardized and unified public service system, e-government can still effectively meet public needs and enhance satisfaction through improvements in information completeness, interface friendliness, and responsiveness. These findings not only provide empirical support for understanding service performance in a centralized institutional environment but also highlight the importance of continuously optimizing service quality details to better align with citizens’ core expectations.

5.4. The Moderating Role of Government Responsiveness (H4)

Hypothesis H4 posits that government responsiveness (GR) positively moderates the relationship between perceived service quality (PSQ) and citizen satisfaction (CS). The empirical analysis shows that the interaction term PSQ × GR → CS has a path coefficient of β = 0.095 (t = 2.859, p = 0.004). Although statistically significant, the effect size is relatively small, suggesting that the moderating influence of responsiveness should be interpreted with caution. Nevertheless, the results indicate that in the context of e-government, citizens’ positive perceptions of service quality, when coupled with a high level of government responsiveness, can still amplify their overall satisfaction.

This finding aligns closely with recent research in public administration and e-government. Chen et al. (2023) conducted an empirical analysis on digital platforms of local governments in China and found that the practicality of response content has a greater impact on enhancing public satisfaction than the timeliness of response. Moreover, high-quality responses can significantly strengthen the association between perceived service quality and satisfaction. Similarly, Nie and Wang (2023) also found in their research on environmental governance that when government responses directly address citizens’ core concerns and provide specific solutions, positive evaluations of service quality are more likely to translate into higher satisfaction. Ma and Wu (2020) further observed that through online collaborative production mechanisms, when citizens perceive that their feedback has been adopted by the government and implemented as actual improvement measures, their recognition of the overall service experience will significantly increase.

Furthermore, the moderating effect identified in this study reflects the “feedback loop” characteristic of value co-creation in digital government platforms (Sorrentino et al., 2018). When responses are not only prompt but also targeted and solution-oriented, the pathway from perceived quality to satisfaction becomes smoother. This mechanism echoes the “digital interaction quality” concept proposed by Wirtz and Kurtz (2017), which posits that the quality of responses largely determines whether citizens evaluate their service experience positively.

Unlike some prior studies, this research reaches this conclusion within a highly centralized governance system. While centralized systems may impose certain constraints on the diversity and differentiation of responses (Ibrahim, 2024), they also shorten the policy adjustment chain, allowing platform feedback to be more swiftly institutionalized. This, to some extent, strengthens the moderating effect. At the same time, because responses in such systems are often highly standardized with limited room for differentiation, this may partly explain why the moderating effect, although statistically significant, is relatively weak in magnitude. Future research could further examine whether this moderating relationship becomes stronger in more decentralized or participatory governance environments, where responses are likely to be more diverse and tailored to citizens’ needs. Thus, government responsiveness emerges not only as an independent driver of satisfaction but also as a critical contextual factor that amplifies the impact of perceived quality on satisfaction.

6. Conclusions

This study takes the highly centralized and integrated digital government platform “Yue Sheng Shi” in Guangdong Province, China, as the research context, constructing and validating a comprehensive model incorporating digital service expectation (DSE), citizen digital participation (CDP), perceived service quality (PSQ), government responsiveness (GR), and citizen satisfaction (CS). It particularly examines the moderating role of GR between PSQ and CS. The results indicate that both DSE and CDP have significant positive effects on PSQ, with CDP exerting a stronger influence. PSQ significantly enhances CS, and GR amplifies the effect of PSQ on CS through its moderating role. These findings deepen the understanding of the performance formation mechanism in digital government and hold important theoretical and practical implications.

Theoretically, this study responds to the need in public administration and e-government research for a comprehensive explanation of the mechanisms at play. While previous studies often focused on technology adoption models or single quality dimensions, rarely considering the joint effects of service expectations and participation behavior, the findings here suggest that service expectations provide a psychological benchmark for quality perception, whereas participation behavior embeds itself in the service improvement process through feedback and collaboration. The two complement each other to shape quality perception. Furthermore, by positioning GR as a moderating variable, this study reveals how the timeliness, specificity, and implementability of government responses under different institutional conditions influence the efficiency of translating perceived quality into satisfaction. Compared with most research conducted in decentralized governance contexts in the West, this study validates the mechanism under a centralized system, noting that while centralized structures may limit response diversity, they can leverage cross-departmental data sharing and integrated workflows to shorten policy adjustment chains, thereby strengthening the moderating effect under certain conditions. It should be noted that the theoretical contribution of this study still has room for further extension. For example, future research could incorporate dimensions such as trust, system quality, and usability to enhance the generalizability of the model across different institutional contexts.

Practically, the findings offer several implications for the design and operation of e-government platforms. First, in terms of expectation management, governments should not only guide citizens’ overall expectations of services through policy communication but also adopt differentiated strategies for different groups. For example, they can implement age-friendly interface designs (such as large fonts and voice interaction), create differentiated service entry points (e.g., customized modules for youth, elderly, or vulnerable groups), and utilize intelligent recommendation mechanisms (pushing relevant services based on usage patterns) to meet diverse needs and achieve more precise expectation management. Second, in terms of citizen participation, governments should establish convenient digital feedback channels that allow citizens to submit issues and suggestions with minimal barriers. Specifically, mobile applications or government mini-programs can incorporate “snap-and-report” features that enable instant submissions, which are then linked to back-end processing systems to ensure rapid assignment, handling, and feedback. At the same time, a public tracking mechanism should be created so that citizens can view the progress and outcomes of their submissions, thereby enhancing the effectiveness, transparency, and credibility of participation. Third, regarding government responsiveness, it is essential not only to ensure timeliness but also to enhance the usefulness and verifiability of responses. For example, governments can establish standardized knowledge bases to ensure consistency and professionalism in replies; include relevant legal provisions or service guidelines in responses to facilitate citizen verification; and introduce intelligent routing and precise matching mechanisms that leverage artificial intelligence and big data to automatically categorize issues and assign them to the most appropriate departments or personnel. These measures can improve the relevance and professionalism of responses while avoiding empty or purely formalistic replies.

Despite the robustness of the findings, this study has several limitations. First, the data were collected through a cross-sectional survey conducted in Guangzhou, Shenzhen, and Zhuhai, which does not capture the dynamic evolution of variable relationships. Future research could adopt longitudinal designs to examine the temporal effects in the formation of satisfaction. Second, as one of the earliest government mini-programs, “Yue Sheng Shi” has certain regional limitations in terms of functional design and user characteristics, and thus, the generalizability of the findings to the national level should be approached with caution. In addition, the sample in this study is relatively more educated and has higher income levels, which, to some extent, reflects the economic development of the selected cities but may also introduce representativeness bias. Prior research has shown that socioeconomic factors such as education and income can significantly predict individual performance outcomes (Sangsawang & Yang, 2025), implying that citizen satisfaction may also vary across different socioeconomic and institutional contexts. Therefore, future research should expand the sample coverage to include a wider variety of regions and groups and incorporate socioeconomic variables into the analysis, in order to enhance both the explanatory power and generalizability of the conclusions.

Another limitation concerns the explanatory power of the model. The structural model yielded relatively modest R² values (0.347 for PSQ and 0.235 for CS), suggesting that additional unobserved factors may influence the formation of citizen satisfaction. In particular, this study did not explicitly model constructs such as trust, perceived risk, or system quality, even though certain aspects (e.g., usability and credibility) were already captured within the measurement of perceived service quality. Prior e-government research has highlighted that trust and system-related attributes play a crucial role in shaping adoption, satisfaction, and long-term platform engagement (Porumbescu, 2015; W. Li, 2021). Evidence from adjacent digital service domains further confirms this view: Yadulla et al. (2024), for example, demonstrate in the context of digital finance platforms that system design and perceived trust strongly influence both behavioral intention and platform stability. Moreover, this study did not examine other emerging security mechanisms, such as blockchain-based identity systems, which may also shape user expectations and satisfaction in digital government. Future research could therefore enhance explanatory power by incorporating these dimensions more explicitly, thereby providing a more comprehensive understanding of the multidimensional evaluation of digital government services.

In conclusion, this study not only enriches the theoretical explanation of the formation mechanism of citizen satisfaction in the context of e-government but also offers actionable recommendations for governments to optimize platform construction, enhance user experience, and build trust during the process of digital public service transformation. By effectively managing public expectations, activating citizen participation, and strengthening high-quality responsiveness mechanisms, digital public service platforms can improve service performance while achieving collaborative governance between governments and citizens.