Engaging Citizens on Government Social Media Page: The Role of Content Characteristics

Abstract

Assessing citizen engagement and exploring its influencing factors has become imperative to help the government in leveraging social media to improve its public service delivery. This study aims to examine the content characteristics posted by the government on social media and their impact on citizen engagement. Content analysis of 1264 Facebook posts from three Malaysian ministries were employed from May until June 2022 and coded into four content types, three media richness levels, and the percentage of comments replied to by the government. Findings revealed that information provision was the highest posted content type, yet symbolic presentation was the most engaging, across all ministries. On the other hand, moderate media richness, such as images, was the most attached in content and yielded the most engagement compared to low and high media richness. Responsiveness level was also low and had no impact on citizen engagement. This study presents a novel finding of the government's communicative practice on social media. The findings contribute significantly to understanding how citizen engagement is influenced by symbolic presentation content type and moderate media richness, while perceived responsiveness calls for immediate improvements.

Key points for practitioners:

While social media is currently common worldwide, governments in developing countries are still at the informational stage and have not yet fully realised its benefits.

This study contributes to the government's social media strategy and helps to maximise content posting using the right content and media richness.

The government should adopt a balanced mix of content types, instead of over-focusing on one content type, to cater to diverse audience needs by informing, involving, and emotionally engaging citizens.

It also contributes to the government's communication strategy as it provides evidence on the low quality of responsiveness level.

Keywords

government social media citizen engagement content type media richness perceived responsiveness

1. Introduction

Social media has revolutionised how private and public sectors communicate with their audience. For businesses, social media serves as a dynamic platform for expression, connection, and commerce, facilitating brands to drive sales and foster customer relationships. Moreover, social media has proven to be an effective tool for marketing and advertising, as well as product research and development (Heidemann et al., 2012). Similarly, governments worldwide have been employing social media to connect with citizens. Unlike business social media use, which is more profit-driven, government usage primarily focuses on broadening information access, enhancing citizens’ trust, and fostering transparency. While users typically follow businesses’ social media for their benefits, such as exclusive promotions, citizens follow the government for different reasons. Citizens are primarily motivated by the need to seek information and real-time updates on public issues (Islm et al., 2021), to observe the government's transparency in decision-making and resource allocation (Arshad & Khurram, 2020), as well as to keep track of the government's promises, public statements, or policy progress. Scholars posited that the desired outcome of the government's social media usage is to build a collaborative relationship with the citizens (Stone & Can, 2021) that can be achieved through meaningful citizen engagement. It allows the government to comprehend citizens’ concerns by assessing their responses on social media (Lee & Kwak, 2012), assisting them by responding appropriately (Chatfield & Reddick, 2018). Moreover, citizen engagement can induce public value creation (Alam, 2021), such as reduced corruption and citizen empowerment (Birghan et al., 2021).

Governments in developing countries are still at the informational stage and have not yet fully realised their benefits. Government's communication through social media remains at an early stage (Sandoval-Almazan et al., 2021) and is predominantly employed as a one-way communication tool (Al-Aufi et al., 2017; Islam & Bhuiyan, 2023) with limited or no expectations for citizen engagement (Arshad & Khurram, 2020; Neely & Collins, 2018). This limitation is reflected in the E-Participation Index (EPI) assessment, an index indicating how the government uses online platforms to enhance communication with citizens (United Nations, 2022). For instance, Malaysia's EPI ranking dropped from 29th in 2020 to 47th in 2022, suggesting a lack of strategic social media utilisation. While Malaysia strives to cultivate citizen engagement through social media as one of its primary communication platforms (Hasbullah et al., 2016), there has been limited research on how to leverage social media to advance EPI rankings (United Nations, 2020). Therefore, assessing citizen engagement and exploring its influencing factors has become imperative.

Social media are well-known for interactivity and user-generated content (Kietzmann et al., 2011) that serve as the backbone of governmental communication tools to encourage engagement (Hidayat et al., 2019). Reactions, such as likes and comments, are widely regarded as key indicators of citizen engagement, as they reflect compelling content and are easily measurable (Darwish, 2017). Previous studies have highlighted the critical role of content in driving engagement (Chen et al., 2021; Zhang et al., 2022), and it is becoming crucial to explore how specific content characteristics trigger such engagement. Thus, this study seeks to answer the question: How do the content characteristics influence citizen engagement in government social media?

Although there are various content characteristics (Henisa & Wilantika, 2021) in government social media, we examine three primary content characteristics: content type, media richness, and perceived responsiveness. Content type refers to the topic posted in the content (DePaula et al., 2018). Media richness is the content richness level in terms of the media presence in the post (Daft & Lengel, 1986). Perceived responsiveness denotes the government's proficiency in responding to citizens on social media (Al-Aufi et al., 2017) and is usually measured by the comment exchange between the government and citizens (Haro-de-Rosario et al., 2018). Citizen engagement is conceptualised as citizens’ responses to government social media in terms of reactions, comments, and shares. We hypothesise that content type, media richness, and perceived responsiveness influence citizen engagement. To test these hypotheses, we analysed data from the Facebook pages of Malaysian government agencies, given that Facebook is the most used social media platform in the country. Through content analysis, we articulate a better comprehension of the Malaysian government's communicative practice on Facebook and assess the individual impact of these factors on citizen engagement.

2. Content Characteristics in Government Social Media

Content characteristic is the core of effective government social media use (Karnowski et al., 2021), which refers to the content attributes that define its nature and how it engages the audience, including content type (Chen et al., 2020; DePaula et al., 2018), media richness (Chen et al., 2021; Rahim et al., 2019), and perceived responsiveness (Al-Aufi et al., 2017; Arshad & Khurram, 2020). As the government continues to embrace the advantages and challenges of integrating social media into its strategy, previous studies have tried to measure its effectiveness. For instance, Mergel (2013a) proposed a framework that can trace online engagement to support government missions using push, pull, and networking tactics. The push tactic refers to simple public information provision that supports transparency through informational and educational posts. The pull tactic refers to the government's interaction with citizens to acquire information and feedback, which fosters participation by promoting two-way interaction that can be measured through comments. Meanwhile, the networking tactic refers to government-citizen activities through online dialogue or physical activity that promotes collaborative goals, which can be measured through shares. The author concluded that the government was focusing too much on push tactics, missing out on the opportunities to improve citizen engagement. Later, DePaula et al. (2018) proposed a new category to update Mergel's framework, symbolic presentation, describing how the government uses social media to improve its reputation and promote public services.

Rahim et al. (2019) conducted a content analysis to study the health content characteristics that influence citizen engagement. Timely posts related to health education or risk communication with videos attached that induced emotional triggers were more appealing to the citizens. However, they concluded that the government has not been ideally leveraging social media, with a low average engagement level. Similarly, Alam and Gill (2020) conducted a content analysis, while Contri et al. (2024) adopted a mixed-methodology to investigate their local government's Facebook Pages and found that the government was more focused on information dissemination with minimal citizen participation. The researcher also emphasises the utility of content analysis for understanding government online engagement, which informs our methodology for assessing Malaysian government Facebook pages.

Apart from the governmental perspective, previous scholars also studied engagement from a citizen perspective. Hidayat et al. (2019) conducted surveys to explore the content type preferred by citizens, which ranged from healthcare and emergency to job-seeking information. Meanwhile, Arshad and Khurram (2020) found that information quality and trust are significant predictors of citizen engagement, while perceived responsiveness is a negative predictor. In summary, these studies collectively corroborated the importance of content quality and how the government strategically uses social media to foster meaningful engagement with citizens.

2.1 Hypotheses Development: Content Characteristics and Citizen Engagement

Citizen engagement denotes citizens’ participation in public affairs (Gil de Zúñiga et al., 2012) and is often employed as a metric to assess the government's social media use. Engagement is commonly measured through social media reactions, including likes, comments, and shares, as they are easily obtainable and quantifiable (Darwish, 2017). In this study, citizen engagement is defined as citizens’ responses to government social media posts regarding these reactions. Bonsón and Ratkai (2013) have proposed explicit metrics to measure citizen engagement by analysing reactions, comments, and shares, which include popularity, commitment, and virality. Popularity is measured by the number of reactions, representing how popular a post is. Commitment refers to the number of comments, representing how committed citizens are to interacting with the government. Virality refers to the number of shares, representing how viral a post is among the citizens. Table 1 shows the formulas used in the metrics adopted in this study.

Table 1.
Metrics for Citizen Engagement.

Name Sign Formula Measures

Popularity P1 Number of posts liked/total posts Percentage of posts that have been liked

P2 Total likes/total number of posts Average number of likes per post

P3 (P2/Number of followers)×1000 Average number of likes per post per 1000 fans

Commitment C1 Number of posts commented/total posts Percentage of posts that have been commented

C2 Total comments/total posts Average number of comments per post

C3 (C2/Number of followers)×1000 Average number of comments per post per 1000 followers

Virality V1 Number of posts shared/total posts Percentage of posts that have been shared

V2 Total shares/total posts Average number of shares per post

V3 (V2/Number of followers)×1000 Average number of shares per post per 1000 followers

Engagement E P3 + C3 + V3 Engagement index

Name	Sign	Formula	Measures
Popularity	P1	Number of posts liked/total posts	Percentage of posts that have been liked
P2	Total likes/total number of posts	Average number of likes per post
P3	(P2/Number of followers)×1000	Average number of likes per post per 1000 fans
Commitment	C1	Number of posts commented/total posts	Percentage of posts that have been commented
C2	Total comments/total posts	Average number of comments per post
C3	(C2/Number of followers)×1000	Average number of comments per post per 1000 followers
Virality	V1	Number of posts shared/total posts	Percentage of posts that have been shared
V2	Total shares/total posts	Average number of shares per post
V3	(V2/Number of followers)×1000	Average number of shares per post per 1000 followers
Engagement	E	P3 + C3 + V3	Engagement index

Source: (Bonsón & Ratkai, 2013).

Citizen engagement is crucial to assist the government in understanding the public concerns, especially in a crisis (Paul & Das, 2023). Recent studies have begun to focus on citizen engagement, emphasising that content characteristics have a significant role in influencing engagement. For instance, different content types impact citizen engagement differently based on their preferences (DePaula & Dincelli, 2016). Citizens in the United States are more inclined to interact with symbolic content posted by their government compared to other categories, which is a behaviour associated with the affective characteristics of social media use (DePaula & Dincelli, 2018). Media elements are also responsible for a notable part of the content posted by the government on social media, which can be influential on citizen engagement (Hussin et al., 2023). Scholar posited that image posts containing pictures gain more likes and comments compared to plain text posts (Hofmann et al., 2013) and are more inclined to evoke emotions than text-only posts, thereby fostering higher citizen engagement (Guidry et al., 2020). Additionally, responsiveness, which has been considered as standard for the government to offer a timely response to meet the citizens’ demands and needs, has been recognised to have a fundamental effect on citizens (Eom et al., 2018).

However, despite these insights, previous studies often treated these factors in isolation, focusing either on individual factors or on a single method. For instance, Chen et al. (2020) and Bonsón et al. (2019) explored citizen engagement in local government using content analysis and mixed methods, respectively. Yet, both studies only focus on content type and media richness. A comprehensive understanding of the implications of content characteristics from all its quality attributes remains an important frontier for future research.

2.2 Content Type and Citizen Engagement

The literature identifies content type as one of the most essential key factors in fostering citizen engagement (Chen et al., 2020; DePaula & Dincelli, 2018; Henisa & Wilantika, 2021; Li et al., 2021; Luo et al., 2021; Renshaw et al., 2021; Yang et al., 2021; Zhang et al., 2022). This study employs the typology proposed by DePaula et al. (2018) and Haq et al. (2022) for coding content type because of its generalizability and relevance across government departments, which includes four main categories as summarised in Table 2.

Table 2.
Metrics for Content Type.

Content Type Indicators

Information Provision Post that provides simple and accurate information to citizens

Input Seeking Post to acquire citizen feedback and information

Online Dialogue and Offline Interaction Post about live-streaming or direct invitations to participate or volunteer in government programmes.

Symbolic Presentation Post about impression management, positive imagery, and symbolic posts.

Content Type	Indicators
Information Provision	Post that provides simple and accurate information to citizens
Input Seeking	Post to acquire citizen feedback and information
Online Dialogue and Offline Interaction	Post about live-streaming or direct invitations to participate or volunteer in government programmes.
Symbolic Presentation	Post about impression management, positive imagery, and symbolic posts.

Source: (DePaula et al., 2018; Haq et al., 2022).

Information provision involves the factual data presentation designed to inform, educate, and raise awareness among the government and the citizens (DePaula et al., 2018). This category is akin to what Mergel (2013a) defines as push tactics, which support the government's transparency mission. It also represents one of the first social media strategies implemented by the government (Criado & Villodre, 2021), emphasising their important role to provide detailed and adequate information to the citizens (DePaula et al., 2018). Information provision includes the descriptive information regarding policy, government programmes, and the informative announcement aimed at educating, raising awareness, and advocating safety.

The second category, input seeking, refers to content that aims at obtaining information and feedback from the citizen (DePaula & Dincelli, 2018). This category resembles and may be part of the pull tactics proposed by Mergel (2013a), which encourage participation through inquiries. This type of content reflects the government's interest in getting information from the citizens, but it often lacks further intention to interact with them (Zavattaro & Sementelli, 2014). Input seeking content posted by the government on social media includes a call on citizens to fill out a survey or a poll (DePaula et al., 2018; Waters & Williams, 2011).

Next is online dialogue offline interaction, which focuses on the real-time dialogue and collaborative content regarding political and democratic interactions (DePaula et al., 2018; Haq et al., 2022). This category aligns with what Mergel (2013a) defines as a networking strategy and Haq et al. (2022) define as a discussion and press conference. While social media platform enables dialogue through features such as live streaming, they lack tools for collaborative project or programme establishment. However, social media could serve as a bridge for an offline collaboration between the government and citizens. This category involves content about live-streaming, an invitation for policy discussion or a meet and greet event with government personnel, and an invitation to collaborate or volunteer in the government's programme.

Symbolic presentation encompasses symbolic content and strategic self-presentation, indicating the government's social media use for self-promotion, symbolic exchange, and marketing (Bonsón et al., 2015). Governments often seek likability, showing their competency and worthiness particularly on social media (DePaula & Dincelli, 2018). Politicians affiliated with government agencies might also make use of social media to express their political perspective and desire to improve their public image (DePaula et al., 2018). Moreover, governments were also observed to post casual content, cultural exchanges, and little marketing. Therefore, in this study, symbolic presentations include self-presentation posts to improve reputation and increase self-likability, cultural themes posts, symbolic expressions, and posts showcasing government services to attract citizens to acquire.

The impact of content type on citizen engagement has been justified to varying degrees. Although Hofmann et al. (2013) argued that no content type can consistently guarantee citizen engagement, this may stem due to no one-size-fits-all circumstance. Citizens from different contexts might have different preferences over what information is valuable to them. For instance, Bonsón et al. (2019) found that city promotional content generates higher citizen engagement in Andalusia, which is one of the popular tourism destinations in Spain. On the other hand, Chen et al. (2020) found that information provision content is more popular among citizens during a crisis, while Metallo et al. (2020) found that it does not affect citizen engagement.

In contrast, Hidayat et al. (2019) explored the social media implementation by the Bahrain government and asserted that posting the right content is vital as citizens prefer to engage in content that piques their interest now, regardless of topics. The finding suggests that citizens are keener to interact with content that resonates personally or intrigues their immediate interest. Given that different content types serve distinct purposes during different times and contexts, governments need to adapt their content which reflect the latest issues. Therefore, it is crucial to examine the impact of content type on citizen engagement across various settings and contexts, leading to our first proposition:

H1: Different content type has different effects on citizen engagement in government social media.

2.3 Media Richness and Citizen Engagement

Media richness theory (MRT), founded by Daft and Lengel (1986), progressively gained prominence with the emergence of social media. MRT posits that communication platforms vary in their affordances of carrying information and delivering messages, with richer information being more effective in reducing ambiguity. On social media platforms, MRT has been employed to examine the effect of rich media on information dissemination. However, the shifting of communication from traditional to social media platforms requires a reinterpretation of the MRT application. While MRT proposed four types of richness based on the traditional affordances of communication, which vary from unprocessed documents (the lowest) to face-to-face communication (the richest), richness in the social media context is based on the social media affordances criteria (i.e., it's multimedia capabilities and content format). These affordances are able to improve the communication, varying from text content to pictures or video content. In this sense, low media richness refers solely to plain text content. Moderate richness refers to content with pictures, links, GIFs, or other equivalent media formats. High richness encompasses posts with video or other interactive media formats.

From the literature, media richness was discovered to have a significant impact on social media engagement (Chen et al., 2021; Henisa & Wilantika, 2021; Paul & Das, 2023; Rahim et al., 2019; Renshaw et al., 2021; Ross et al., 2018; Yang et al., 2021; Zhang et al., 2022). Chen et al. (2020) segregated media richness into three levels depending on the media type in the post: low richness (plain text post), moderate richness (picture post), and high richness (video post).

However, in academic society, media richness has been revealed to have a controversial discovery and remains inconclusive, as researchers have presented contradicting findings. For instance, scholars confirmed that image-based content generated more likes and comments compared to text posts (Hofmann et al., 2013), fostering higher citizen engagement (Bonsón et al., 2019). Conversely, Lee and Xu (2018) reported an adverse effect of media richness as it appears that citizens favoured plain text posts during a crisis, where citizens seek information to stay calm and well-informed. Similarly, Kim and Yang (2017) also supported the proposition that content with pictures reduces comments, while Denktaş-Şakar and Sürücü (2020) concluded that content with videos does not affect engagement at all. Given these mixed findings, it is essential to understand the effect of media richness on citizen engagement that can be adapted particularly to local government settings. This leads to the following proposition:

H2: Media richness has a significant effect on citizen engagement in government social media.

2.4 Perceived Responsiveness and Citizen Engagement

Dialogic Communication Theory (DCT) posits that dialogue is foundational to encourage engagement (Kent & Taylor, 1998). Dialogic loop, one of the core principles in DCT, implies that organisations must post content to stimulate dialogue and respond to it in a timely manner. Align with the dialogic loop, responsiveness acts as an essential key factor in fostering citizen engagement, enabled by social media features that afford real-time communication disregarding place and time barriers. A responsive government has been found to encourage citizens to be more inclined to engage with them (Arshad & Khurram, 2020; Sjoberg et al., 2017). Citizens often feel heard and valued when the government responds to their comments. This is because responsiveness has been widely recognised as a benchmark for meeting the citizens’ demand that affects trust and engagement (Eom et al., 2018).

Assessing the government's responsiveness and its impact on citizen engagement is crucial to determine the right timing to invest the resource due to the requirement for real-time human intervention. To measure perceived responsiveness, Haro-de-Rosario et al. (2018) and Nah and Saxton (2013) suggested that government replies to citizens’ comments need to be examined over a study period of one month minimum. As the requirement to be interactive with citizens is inevitable, further research is needed, leading to the third proposition:

H3: Perceived responsiveness has a significant effect on citizen engagement in government social media.

3. Research Method

We conducted content analysis to answer the research question as it is able to generate both quantitative data (e.g., post frequency and number of likes, comments, and shares) and qualitative data (e.g., content types), offering a rich and comprehensive understanding of the government's communicative practice. Content analysis is exceptionally valuable for exploratory analysis of large amounts of data and inferential study of content (Krippendorff, 2018; McKibben et al., 2022) and has been proven effective by previous studies (Bonsón et al., 2019).

Qualitative content analysis was carried out to identify the government communicative practice across content type, media richness, and perceived responsiveness. Meanwhile, quantitative content analysis was employed to assess the impact of these factors on citizen engagement. This study followed three phases of content analysis proposed by Elo and Kyngas (2008). First, in the preparation phase, the unit of analysis was defined, and initial data familiarisation took place. Second, in the organising phase, a deductive approach took place as the predefined analysis metrics in this study have been drawn from the literature. An unconstrained analysis matrix was employed as it permits flexibility in data collection and enables the left-over data to fit into the analysis matrix. During this phase, data were grouped, categorised, and coded before the hypothesis was tested. The last phase was the reporting phase. Findings from the previous phase were compiled and described in an organised format.

3.1 Data Preparation

The target population in this study comprised the official Facebook pages of Malaysia's Ministry of Defence (MOD), Ministry of Education (MOE), and Ministry of Health (MOH). Facebook was selected due to its wider interactive capabilities (Alam et al., 2022), allowing its users to express their concerns or opinions (Ononye & Igwe, 2017) and the government to respond to their comments (Reddick et al., 2017). Facebook is also the most prominent platform in Malaysia, with a user penetration of 91.7% (Malaysian Communication and Multimedia Commission, 2020). The ministries were chosen because they have the highest number of followers among Malaysian ministries. The unit of analysis is the Facebook post, which was manually collected from May to June 2022.

Year 2022 was the year after the peak of COVID-19; thus, it was selected to avoid bias from heightened activity on social media. Additionally, Malaysia's rank in EPI had dropped in 2022, making it significant to assess citizen engagement during this period. Meanwhile, May and June were chosen because there was no big event, such as elections, that could affect social media activity. According to Haro-de-Rosario et al. (2018), a minimum study duration of one month is an acceptable timeframe in conducting social media analysis. Therefore, a two-month-long period is considered adequate to gather the Facebook data for each ministry (Nah & Saxton, 2013).

Manual coding was employed as it allowed us to concentrate on nuanced data (Saldana, 2021), aligning with researchers’ time and expertise (Basit, 2003). Besides, manual coding is common in content analysis, often used by researchers in the social media field (e.g., (Bellucci et al., 2019; DePaula & Dincelli, 2018; Fissi et al., 2022). A total of 1264 posts, which were publicly available data, were collected and exported into MS Excel for analysis.

3.2 Data Coding and Analysis

The coding process adhered to metrics set from the literature review. The coding scheme for each content was as follows. Information Provision was coded as 1 for content posted or shared from another page by the government containing public service announcements, safety recommendations, or information on operation, policies, programme, or event, such as Covid-19 updates and vaccination awareness content. This also included when the government shared posts from other pages which is related to the government's role. Input Seeking was coded as 2 when the government posts content soliciting citizens’ opinions or feedback, such as surveys and polls. Online Dialogue and Offline Interaction was coded as 3 for live-streamed content, and invitation posts for citizens to government meet-and-greet events, policy discussion, or volunteer in government programmes. Lastly, Symbolic Presentation was coded as 4 when the government posted positive content aimed at enhancing its reputation or taking a firm stand on a political issue. This also included heartfelt expressions post such as congratulatory, condolences, cultural wishes, such as festival celebrations, promotional content for competitions held by the government, or services offered by the government.

Next, media richness was categorised based on the content's richness level following definitions by Chen et al. (2020). These media richness levels from low to high were coded as follows. Low richness was coded as 1 when the content posted a text-only post, which also included images that displayed text only. Moderate Richness was coded as 2 when the government included an image, GIF, or external links (e.g., YouTube link) that redirected to a different site in their post. High Richness was coded as 3 when the government included a video in their post. Perceived responsiveness measured the government's response and replies to citizens’ comments on their posts. Each comment was assessed to find comments that seek answers, direction, inquiries, and/or replies from the government. The response rate was evaluated as a percentage of the replies and was calculated by dividing the number of the government's replies by the number of comments asking for answers, direction, or inquiries. We input this variable as a continuous variable by entering the reply as a percentage. Citizen engagement was measured using the set of metrics proposed by Bonsón and Ratkai (2013) (See Table 1). Under this method, the interaction between the government and citizens was examined directly to determine the level of citizen engagement (Bonsón & Ratkai, 2013). The number of shares, comments, and reactions was collected for all posts to measure the aggregated engagement index (E).

To ensure reliability, a second coder was recruited and trained on a smaller subset of posts for practices after reviewing the coding metrics. After the first discussion and refinement of their mutual agreement, the second coder independently coded 10% of the coded data. Discrepancies between the two coders were settled through consensus meetings. Inter-rater reliability (IRR) was measured using Krippendorff's alpha (α) (Krippendorff, 2018). The alpha value is 0.8559, which shows a high IRR and acceptable reliability.

To test the proposition, engagement metrics were examined by content type, media richness, and perceived responsiveness. Normality tests were checked first to confirm the right statistical method for data analysis. Kolmogorov-Smirnov normality test showed a p-value less than the significance level of 0.05 for all independent variables, indicating that all factors were not normally distributed. Thus, non-parametric tests were employed. As a non-parametric data set with more than two category groups, the Kruskal-Wallis H test was conducted to check for any possible differences in citizen engagement by content type and media. On the other hand, Spearman's Rho Correlation was conducted to check for the correlation between perceived responsiveness and citizen engagement because both were continuous data.

4. Findings

The first part of the finding sections provides an overview of how the MOD, MOE, and MOH utilise social media based on content type, media richness, and perceived responsiveness. As shown in Table 3, MOH has the largest following, with over 5.7 million followers, significantly higher than MOE and MOD with 1.2 million and 781,229 followers, respectively. MOH also maintains consistent and substantial engagement, with 416 posts, while MOD has 546 posts, and MOE has the lowest at 272 posts. Posting activity is balanced between May and June for all ministries, with MOD and MOE posting slightly more in June (MOD = 51.5%; MOE = 54.4%), while MOH posted slightly more in May (51.7%).

Table 3.
Overview of Social Media Usage by Government Agencies.

Government Agencies MOD MOE MOH

Number of followers 781229 1213053 5719426

No. of Post Total 546 (100%) 272 (100%) 416 (100%)

May 265 (48.5%) 124 (45.6%) 215 (51.7%)

June 281 (51.5%) 148 (54.4%) 201 (48.3%)

Reactions Total reactions 31010 42988 809758

Total Comments 1521 12325 61644

Total Shares 1907 14261 176846

Government Agencies	MOD	MOE	MOH
Number of followers	781229	1213053	5719426
No. of Post	Total	546 (100%)	272 (100%)	416 (100%)
May	265 (48.5%)	124 (45.6%)	215 (51.7%)
June	281 (51.5%)	148 (54.4%)	201 (48.3%)
Reactions	Total reactions	31010	42988	809758
Total Comments	1521	12325	61644
Total Shares	1907	14261	176846

Overall, MOH stands out with 809,758 reactions, 61,644 comments, and 176,846 shares, significantly higher than MOD and MOE. This suggests that MOH's content may be relevant or resonating more with the public or simply reaching a wider audience. MOD and MOE have relatively lower numbers of total reactions, with MOD receiving 31,010 reactions, 1521 comments, and 1907 shares, while MOE garnered 42,988 reactions, 12,325 comments, and 14261 shares. Notably, MOE yielded more comments and shares compared to MOD despite posting fewer posts, indicating that MOE received better engagement per post.

All three ministries predominantly posted information Provision content, accounting for 85% of posts by MOD, 72.4% by MOE, and 75% by MOH, which aligns with their role in providing information to the public. Symbolic Presentation posts, which often involve patriotic or ceremonial content, are more common in MOD (10.6%) but generate notable interest in MOH (15.6%) and MOE (13.3%). This indicates that symbolic posts are well-received across ministries but may have different engagement effects depending on the ministry's audience and purpose. Online dialogue offline interaction contents are relatively rare, with MOE leading at 14.3%, suggesting a greater emphasis on interactive content compared to the other two ministries. On the other hand, input seeking is hardly ever used and almost absent in all ministries, with only MOH and MOD posting a low number of this content type, with 0.7%, which unveils the lack of efforts to seek public opinion. Table 4 summarises the content type posted by these ministries.

Table 4.

Overview of Content Type Posted by Government Agencies.

Content Type	MOD	MOE	MOH
Information Provision	464 (85.0%)	197 (72.4%)	312 (75.0%)
Input Seeking	4 (0.7%)	0 (0.0%)	3 (0.7%)
Online Dialogue Offline Interaction	20 (3.7%)	39 (14.3%)	36 (8.7%)
Symbolic Presentation	58 (10.6%)	36 (13.3%)	65 (15.6%)

Additionally, most posts fall under the moderate media richness category across all ministries, with MOD at 85.7%, MOE at 86.8%, and MOH at 77.2%, as shown in Table 5. This suggests a preference for media with a moderate level of interactivity, such as images or text with some visual elements. High media richness content (e.g., videos) is most common in MOH (15.4%) compared to MOD (14.1%) and MOE (11.4%). This could explain MOH's higher engagement rates, as richer media formats can attract more interaction.

Table 5.

Overview of Media Richness in Content Posted by Government Agencies.

Media Richness	MOD	MOE	MOH
Low	1 (0.2%)	5 (1.8%)	31 (7.5%)
Moderate	468 (85.7%)	236 (86.8%)	321 (77.2%)
High	77 (14.1%)	31 (11.4%)	64 (15.4%)

Lastly, all three ministries show a very low response rate to public comments, with MOD and MOE not replying to any inquiry comments, while MOH replies to only a small portion (1.2%). Additionally, a significant percentage of posts contained no inquiry comments: 98.5% for MOD, 89% for MOE, and 97.6% for MOH, suggesting that citizens may not frequently use social media to ask questions on government platforms or that there is limited opportunity for dialogue. Table 6 summarises these findings.

Table 6.

Overview of Perceived Responsiveness in Government Agencies.

Perceived Responsiveness	MOD	MOE	MOH
No inquiry comment	538 (98.5%)	242 (89.0%)	406 (97.6%)
Inquiry comment without reply	8 (1.5%)	30 (11.0%)	5 (1.2%)
Inquiry comments replied	0 (0.0%)	0 (0.0%)	5 (1.2%)

In summary, MOH leads in both followers and total reactions, potentially due to its use of high-media-rich content and relevance to public health issues. MOD and MOE have fewer total overall reactions but still emphasise information-heavy content to fulfil their public roles. The analysis shows a strong reliance on information provision and symbolic presentation across the ministries, with limited responsiveness to public inquiries, highlighting a predominantly one-way communication strategy.

4.1 Social Media Use by the Government

Content Type. Table 7 provides a detailed breakdown of content types used by government agencies to engage citizens, highlighting various strategies aimed at informing, involving, and fostering a positive image among the public.

Table 7.
Example of Content Types Coded across Government Agencies.

Content Type Content Description Examples

Information Provision Post about policy, operation, or event. “[…] Ministry of Defence has organised a Blood Donation Program […]” “Following the updated protocol for travellers on 1 May 2022, […]”

Post to raise public awareness. “Do you know your role and responsibility in defending the country? […]” “[…] An open call to you to work together to raise awareness about Thalassemia […]”

Input Seeking Post to get citizens’ feedback. “[…] You are invited to complete a survey […]” “HWCITF is conducting a survey […]”

Online Dialogue Offline Interactions live-streaming post. “[…] Media Conference by Senior Minister of Defence […]” “Watch the Malaysian Teachers Festival Appreciation Ceremony live, […]”

Invitation to government programmes or meet-and-greet. “[…] Everyone is invited to come and celebrate this Customer Day […]” “Visit the MOE booth around 2.30 pm for the #tanyaDRJ session.”

Post asking citizens to get involved in a government programme. “[…] Let's come together to the selected location on 3^rd May 2022 to help make the Blood Donation Campaign […]”

Symbolic Presentation Post to increase positive image and likability. “50 veteran home rehabilitation projects at the ATM have been completed […]” “Ministry of Health has won the National Protection Safety Premier Award […]”

Post to relay an expression related to a cultural or special occasion. “[…] the Ministry of Defence extends its condolences […]” “Happy Eid to all front liners and all Malaysians. […]”

Post to promote government services for citizens to acquire. “Application for admission to Form Six (6) of the Fully Boarding School […] “[…] Various activities will be held by the National Blood Centre, such as essay writing competition […]

Content Type	Content Description	Examples
Information Provision	Post about policy, operation, or event.	“[…] Ministry of Defence has organised a Blood Donation Program […]” “Following the updated protocol for travellers on 1 May 2022, […]”
Post to raise public awareness.	“Do you know your role and responsibility in defending the country? […]” “[…] An open call to you to work together to raise awareness about Thalassemia […]”
Input Seeking	Post to get citizens’ feedback.	“[…] You are invited to complete a survey […]” “HWCITF is conducting a survey […]”
Online Dialogue Offline Interactions	live-streaming post.	“[…] Media Conference by Senior Minister of Defence […]” “Watch the Malaysian Teachers Festival Appreciation Ceremony live, […]”
Invitation to government programmes or meet-and-greet.	“[…] Everyone is invited to come and celebrate this Customer Day […]” “Visit the MOE booth around 2.30 pm for the #tanyaDRJ session.”
Post asking citizens to get involved in a government programme.	“[…] Let's come together to the selected location on 3^rd May 2022 to help make the Blood Donation Campaign […]”
Symbolic Presentation	Post to increase positive image and likability.	“50 veteran home rehabilitation projects at the ATM have been completed […]” “Ministry of Health has won the National Protection Safety Premier Award […]”
Post to relay an expression related to a cultural or special occasion.	“[…] the Ministry of Defence extends its condolences […]” “Happy Eid to all front liners and all Malaysians. […]”
Post to promote government services for citizens to acquire.	“Application for admission to Form Six (6) of the Fully Boarding School […] “[…] Various activities will be held by the National Blood Centre, such as essay writing competition […]

Information Provision. Information provision contents are related to the government's operation, policy, events, programmes, and public service announcements. Despite the disparities in the ministries’ roles and number of followers, MOD, MOE, and MOH demonstrated a similar social media usage in terms of information dissemination. As expected, information provision was the most prevalent content type posted by MOD, MOE, and MOH with 85% (n = 464), 72.4% (n = 197), and 75% (n = 312), respectively.

Input Seeking. Input seeking content that was aimed at collecting citizens’ feedback through surveys or polls was scarce. Only a small percentage of input seeking content type (MOD = 0.7%, n = 4; MOE = 0; MOH = 0.7%, n = 4) was found in the collected posts, with a total absence in MOE. This suggests little effort in all ministries to seek input from the citizens using social media.

Online Dialogue and Offline Interactions. The third content type was the second most common post by MOE (14.3%, n = 39) and the third most common post by MOD (3.7%, n = 20), and MOH (8.7%, n = 36). This finding reflects the government's initiative in employing its networking strategy, which facilitates interactive engagement, both online and offline. However, it can be observed that MOD and MOE rarely post content with the intention to collaborate with the citizens.

Symbolic Presentations. Lastly, symbolic presentation contents were the second most common in MOD (10.6%, n = 58) and MOH (15.6%, n = 65), while it was the third most common in MOE (13.3%, n = 36). These posts often foster a sense of goodwill and national pride among the public by emphasising symbolic messages, the ministry's achievements and milestones, as well as cultural events.

Media Richness. A consistent pattern in media richness was found in all ministries. In general, MOD, MOE, and MOH posted most of their content with moderate richness level with 85.7% (n = 468), 86.8% (n = 236), and 77.2% (n =321), respectively, while content with high and low richness levels has a significantly lesser number. Notably, while moderate and high media richness were found included in all content types, it can be observed that all low richness content was under the information provision content type. This might be due to the need for longer elaboration on important information. Figure 1 illustrates the example of each media richness level, ranging from low to high.

Figure 1.

Media Richness: Example of Low (Left), Moderate (Middle), and High Richness (Right).

Perceived Responsiveness. Comments from citizens were assessed to find inquiry comments. As shown in Table 2, the number of posts that received inquiry comments was notably low with 1.5% (n = 8), 11% (n = 30), and 2.4% (n = 10) in MOD, MOE, and MOH, respectively. However, no reply was recorded in MOD and MOE. Over the two months, MOH received a total of 61644 comments, with only 10 comments classified as inquiries. From these, only 5 comments were replied to by MOH.

Figure 2 illustrates one example of the comments replied to by MOH. One citizen asked if the National Blood Centre is open for blood donation. The MOH responded by redirecting the citizen to another Facebook page for her to check for the updated blood donation schedule. In another post, a citizen commented, “If I read correctly, monkeypox has a vaccine. So, where can I get the vaccine?” with no response from the MOH. Interestingly, MOH responded to one comment, stating that he has donated his blood by expressing gratitude and congratulating the individual for being a hero.

Figure 2.

Perceived Responsiveness: Example of the Government's Reply to a Citizen's Comment.

While MOH established a selective response and interactions, it remains unclear what criteria guided the MOH to reply to inquiry comments. Although some inquiries received answers, many other important inquiries were still neglected, exhibiting the inconsistent responsiveness level in the ministry.

4.2 Differences in Citizen Engagement

Table 8 presents the analysis of the overall citizen engagement on the ministries’ Facebook pages. In comparison, MOH recorded the highest engagement (M = 0.421), surpassing MOD (M = 0.081) and MOE (M = 0.211). This result indicates that the citizens were participating more with MOH, potentially driven by lingering fear or habitual behaviour after the pandemic.

Table 8.
Citizen Engagement by Government Agencies.

Engagement MOD MOE MOH

Mean 0.081 0.211 0.421

Minimum 0.003 0.008 0.012

Median 0.046 0.045 0.175

Maximum 1.276 0.008 8.054

Standard Deviation 0.124 0.801 0.802

Engagement	MOD	MOE	MOH
Mean	0.081	0.211	0.421
Minimum	0.003	0.008	0.012
Median	0.046	0.045	0.175
Maximum	1.276	0.008	8.054
Standard Deviation	0.124	0.801	0.802

4.3 Citizen Engagement by Content Type

Kruskal-Wallis H test was conducted to determine whether content type influences citizen engagement across all ministries. Kruskal-Wallis H test showed that there was a statistically significant difference in citizen engagement between the content type (MOD: H(3) = 14.549, p = 0.002; MOE: H(2) = 38.783, p=<0.001; MOH: H(3) = 27.175, p=<0.001), indicating that at least one content type had a different median engagement score, as shown in Table 9. The effect size, as estimated by epsilon-squared (ε2), was 0.03 for MOD, 0.14 for MOE, and 0.07 for MOH, indicating that content type had a weak effect on citizen engagement in MOD, and a moderate effect in MOE and MOH.

Table 9.
Differences in Citizen Engagement by Content Type.

Engagement by Content Type Kruskal-Wallis Test

Information Input Online Dialogue Symbolic Asymp. Effect

Ministry Provision Seeking offline Interaction Presentation H value Df Sig. Size

MOD Mean Rank 275.14 80.25 184.98 304.20 14.549 3 0.002 0.03

Median 0.045 0.017 0.020 0.055

MOE Mean Rank 125.21 - 123.37 212.53 38.783 2 <0.001 0.14

Median 0.040 - 0.035 0.164

MOH Mean Rank 206.95 80.67 139.64 259.99 27.175 3 <0.001 0.07

Median 0.173 0.057 0.082 0.366

	Engagement by Content Type	Kruskal-Wallis Test
MOD	Mean Rank	275.14	80.25	184.98	304.20	14.549	3	0.002	0.03
Median	0.045	0.017	0.020	0.055
MOE	Mean Rank	125.21	-	123.37	212.53	38.783	2	<0.001	0.14
Median	0.040	-	0.035	0.164
MOH	Mean Rank	206.95	80.67	139.64	259.99	27.175	3	<0.001	0.07
Median	0.173	0.057	0.082	0.366

Dunn's pairwise post-Hoc tests were carried out for all pairs in the groups. In MOD, strong evidence of difference (p < 0.05, adjusted using the Bonferroni correction) was found between (1) Input seeking (mean rank = and symbolic presentations (adjusted p = 0.036); and (2) online dialogue offline interaction, and symbolic presentation (adjusted p = 0.021). Symbolic presentation has a higher mean rank engagement index compared to input seeking and online dialogue offline interactions. There was no evidence of a difference found significant between the other pairs in content type.

In MOE, strong evidence of difference (p < 0.05, adjusted using the Bonferroni correction) was found between (1) Online dialogue offline interactions and symbolic presentations (adjusted p = 0.000), and (2) information provision and symbolic presentation (adjusted p = 0.000). Symbolic presentation was found as the most engaging content type with a higher mean rank engagement index compared to information provision and online dialogue offline interaction. There was no evidence of a difference found significant between the other pairs in content type.

In MOH, strong evidence of difference (p < 0.05, adjusted using the Bonferroni correction) were found between (1) Online dialogue offline interactions and information provision (adjusted p = 0.009); (2) online dialogue offline interaction and symbolic presentation (adjusted p = 0.000); and (3) information provision and symbolic presentation (adjusted p = 0.007). Symbolic presentation has a higher mean rank engagement index compared to information provision and online dialogue offline interaction. There was no evidence of a difference found significant between the other pairs in content type.

Symbolic presentation was found as the most engaging content type across all ministries, suggesting that citizens preferred symbolic and cultural content. In contrast, the most frequently used content type, information provision, did not garner high citizen engagement and often has a lower median engagement index. On the other hand, online dialogue offline interaction, and input seeking have lower citizen engagement, indicating the one-way communication preferences among citizens. These results support Proposition 1.

4.4 Citizen Engagement by Media Richness

Kruskal-Wallis H test was also conducted to assess the effect of media richness on citizen engagement. The results indicate that only MOH has enough evidence to conclude that media richness influences citizen engagement (H(2) = 20.825, p=<0.001), suggesting that at least one richness level had a different median citizen engagement score, as shown in Table 10. The effect size, as estimated by epsilon-squared, was 0.00 for MOD, 0.01 for MOE, and 0.05 for MOH, indicating that media richness had a negligible effect on citizen engagement in MOD, a weak effect in MOE, and a moderate effect in MOH. Strong evidence of difference (p < 0.05, adjusted using the Bonferroni correction) was found between high and moderate richness (adjusted p = 0.000). Moderate richness generated the highest mean rank engagement index compared to high richness. There was no evidence of a difference found significant between the other pairs in media richness. These results partially support Proposition 2, indicating that media richness can influence citizen engagement but may vary depending on the richness level.

Table 10.
Differences in Citizen Engagement by Media Richness.

Engagement by Media Richness Kruskal-Wallis Test

Ministry Low Moderate High H value Df Asymp. Sig. Effect Size

MOD Mean Rank 270 275.34 262.34 0.449 2 0.799 0.00

Median 0.045 0.046 0.042

MOE Mean Rank 177.20 135.67 136.26 1.365 2 0.505 0.01

Median 0.072 0.045 0.041

MOH Mean Rank 187.39 222.37 149.15 20.825 2 <0.001 0.05

Median 0.113 0.227 0.097

	Engagement by Media Richness	Kruskal-Wallis Test
MOD	Mean Rank	270	275.34	262.34	0.449	2	0.799	0.00
Median	0.045	0.046	0.042
MOE	Mean Rank	177.20	135.67	136.26	1.365	2	0.505	0.01
Median	0.072	0.045	0.041
MOH	Mean Rank	187.39	222.37	149.15	20.825	2	<0.001	0.05
Median	0.113	0.227	0.097

4.5 Citizen Engagement by Perceived Responsiveness

Table 11 shows the results of Spearman's rho correlation analysis between perceived responsiveness and citizen engagement. The analysis showed that no significant correlation was found, indicating that perceived responsiveness has no relationship with citizen engagement. Therefore, the findings confirmed that the responsiveness level has no measurable impact on citizen engagement, thus failing to support Proposition 3.

Table 11.
Correlation among Perceived Responsiveness and Citizen Engagement.

Perceived Responsiveness - Engagement Spearman's Correlation Sig. (2-tailed)

0.036 0.461

Table 12 summarises the results of proposition testing in this study. The findings confirm the first hypothesis (H1), demonstrating that different content types have varied impacts on citizen engagement. This aligns with the notion that specific types of posts, such as information provision or symbolic presentation, elicit distinct responses from the citizen, emphasising the importance of tailored content strategies.

Table 12.

Summary of Proposition Testing.

Proposition	Decision
H1: Different content type has different effects on citizen engagement on government social media.	Supported
H2: Media richness significantly influences citizen engagement on government social media.	Partially Supported
H3: Perceived responsiveness has an effect on citizen engagement on government social media.	Not Supported

The second hypothesis (H2), which posits that media richness significantly influences engagement, is only partially supported. This suggests that while richer media formats generally enhance engagement, their effectiveness appears to depend on additional factors, such as the nature of the content or the platform's audience.

In contrast, the third hypothesis (H3) was not supported, indicating that the level of direct responsiveness from the government does not significantly impact overall engagement. Instead, this finding suggests that while responsiveness might be desirable and important, other aspects of social media use, such as content relevance and format, may play a more critical role in sustaining citizen interaction.

5. Discussion

Our findings addressed our first objective through the assessment of the communicative practice posted by MOD, MOE, and MOH on their Facebook page. In terms of content types, our results show that most of the content posted was information provision content type. This aligns with prior research indicating that social media use by the government was mainly to provide one-way communication (DePaula et al., 2018; Mergel, 2013b; Zheng & Zheng, 2014). While this practice fulfils the roles of government in information dissemination, the over-focus on this content type may reflect a lack of engagement-focused strategies. This trend of unidirectional communication strategy may be attributed to several factors, including the limited resources in human capital and training to handle bidirectional social media communication, and the need to curb public criticism or misinformation. In contrast, content encouraging citizen engagement was comparatively scarce, potentially contributing to lower engagement rates. Input seeking was rare across all ministries, indicating the lack of effort by these governments to solicit citizen input. Our finding corroborates Contri et al. (2024)'s finding that governments hardly use social media to seek public feedback, especially related to their concerns. Meanwhile, online dialogue offline interactions, and symbolic presentations were moderately posted by the governments, suggesting the government's attempt at networking and national-identity strategy.

Overall, information provisions yielded more reactions and shares across all ministries, reflecting the nature of a one-way push communication strategy that was simply meant for information dissemination. Citizens might appreciate the useful information without the need to comment and engage further. For posts coded as input seeking, reactions were significantly higher than comments and shares, likely because citizens are redirected to an external site without the need to engage further on social media. In online dialogue offline interactions, MOD and MOH received the most reactions, while MOE received the most comments. This discrepancy could stem from the content's relevance to citizens’ lives. MOE mainly focuses on posting content related to education, which impacts students, teachers, and parents daily, and often invites citizens to engage and discuss the content. Meanwhile, content related to defence and health may not always resonate personally and be relatable to the citizens’ daily lives, hence the lower number of comments than MOE. Symbolic presentations generated the most reactions in all government agencies. Since reaction is the easiest way to engage with a post, it could also show a generalised agreement and support for the post without requiring commenting and sharing. These findings underscore the importance of content strategy in government communication. To maximise citizen engagement, the government needs to construct a social media strategy to avoid over-focusing on one single content type and instead adopt a balanced mix of these content types to cater to diverse audience needs by informing, involving, and emotionally engaging citizens. Interestingly, while the frequency of content posting does not affect citizen engagement, which supports the previous results by Bonsón et al. (2019), who found that active posting on government social media does not affect citizen engagement, we found that different content types had different effects across the ministries that have different roles and functions.

In general, information provision was the highest number of contents posted in all government agencies, yet symbolic presentation was the most engaging across all ministries. These findings were consistent with DePaula and Dincelli (2018), who also found that symbolic presentation elicits higher reactions than other content types. High engagement in symbolic presentation content suggests that citizens resonate more with cultural and identity-centric content. Cultural and symbolic posts, such as public holidays or national celebrations, are usually emotionally evoking, as they tend to make people feel shared enjoyment, pride, or nostalgia. These feelings can motivate citizens to engage with the content to express their feelings. Moreover, symbolic presentations also strengthen citizens’ collective identity and sense of belonging by demonstrating shared values. Psychologically, symbolic presentation also fulfils human intrinsic needs for social connection, bonding, and attachment. As a result, citizens are more likely to engage with this content as it is more compelling compared to information provision content, which may lack emotional reinforcement. Our results further support the prior findings, which asserted that different content type serves the citizens in different ways and generate different reactions (Chen et al., 2020; Keib et al., 2018; Lee & Xu, 2018).

Second, posts with moderate media richness have emerged as the most commonly used by all ministries. This finding is consistent with a previous study (Yang et al., 2021), suggesting that moderate richness can effectively convey information through its visual and emotional appeal to the citizens. It is also more likely to capture the citizens’ attention than low and high richness content. Low richness posts are usually compact with information that often appeals to individuals who enjoy reading text content, while high richness content, though it can deliver complex information through visuals and audio, may trigger an individual's sensory levels.

We found that moderate richness was the most engaging content, which is consistent with prior research (Bonsón et al., 2014; Lappas et al., 2022). As this finding deepens our understanding of media richness impact, it also gives rise to a paradox: increasing media richness might hurt citizen engagement. Our results also showed that the low richness content received better engagement than high richness, which proved that while increasing richness in content can improve the message quality (Metallo et al., 2020), it does not necessarily have an impact on citizen engagement (Lee & Xu, 2018).

Third, the overall responsiveness level of Malaysian government agencies revealed significant room for improvement. Among all ministries, only MOH has been observed to have replied to some of the inquiry comments, even though it's only 1.2% of all collected posts. Prior studies have noted that responsiveness can be a struggle and challenging for the government when the levels of engagement increase (Panagiotopoulos et al., 2013). This limitation might be the case for these ministries. Due to many followings, it can be hard to keep up with notifications on questions from the citizens. As comments keep on showing up on their post, replying to so many comments can be time-consuming, thus, in some cases, near impossible.

Contrary to our expectations, our analysis revealed that the responsiveness level had no significant influence on citizen engagement. This finding contradicts the existing literature, which generally posits that a higher responsiveness level from the government on social media leads to improved citizen engagement. Studies by Haro-de-Rosario et al. (2018) have demonstrated that the government enjoyed greater citizen engagement when they were more responsive to the citizens’ comments on social media. However, our results suggest that this relationship may not be as straightforward in the Malaysian context. Several factors might explain this result. First, insufficient data might lead to this finding. Second, an insufficient or generic reply by the government might potentially fail to satisfy citizens’ expectations for meaningful interaction, but complete information could also give little reason for the citizen to engage further. Third, the cultural norms, particularly in the collectivist culture in Malaysia, might shape citizen engagement with government social media. According to Noelle-Neumann (1974)'s spiral of silence theory, individuals who fear community isolation are more inclined to remain silent when they are convinced that their opinions are the minority, thus pushing forward the dominant voice while silencing the minority perceptions. Aligning with this theory, the government's responsiveness level might not only reflect the government's communicative strategy but also be rooted in the cultural factor within the citizens, which deters self-expression or public questioning. As the citizens tend to silence their opinions, it may reinforce their belief that government responsiveness is not needed.

6. Limitations and Future Recommendations

Several limitations have arisen in this study and should be acknowledged. First, this study only focuses on Facebook, limiting its relevance to other platforms. As social media platform keeps evolving, these platforms have different features and affordances, which usually have users and audiences from distinct backgrounds. This may have a different impact on citizen engagement, thus limiting the generalizability of the findings of this study. Therefore, future studies should broaden and update the current findings using other social media applications such as TikTok or employ a cross-platform comparative study.

Second, as this study focused on Facebook due to its popularity, particularly in Malaysia, this study did not consider the influence of Facebook's architecture on citizen engagement. Facebook works through algorithms and content filtration rules, which may have an impact on content that appears on its users’ feeds. Considering this, government content may not have appeared on citizens’ feeds, thus hindering them from engaging with government posts. This may interfere with the full apprehension of citizen engagement behaviour. Therefore, future studies should acknowledge this limitation and consider the influence of social media dynamics on citizen engagement.

Third, the data collection was conducted in two months (May - June 2022), which can potentially increase bias in perceived responsiveness. Environmental factors such as shifting citizen concerns or political events might affect the government's responsiveness level, thus, this short study period might not be able to completely reflect perceived responsiveness. Furthermore, the sparse data with only five replies from the government may also weaken the strength of the analysis. The lack of interactions by the government may suggest either a general absence of engagement by the government or a possible underreporting caused by constraints in data collection. Therefore, future research could consider extending the data collection period or employ a mixed method for complementary findings, to reflect the government's responsiveness level in order to provide a more robust and reliable result to conclude the relationship.

Fourth, this study focuses only on observable interactions on social media, disregarding citizens’ underlying perceptions and motives. While social media engagement metrics can offer insights into the engagement level on a post, they may not fully capture citizens’ true intentions in their interactions. Therefore, future research could benefit from mixed-method research to triangulate the findings from the government perspective by taking the citizens’ perspective into account. Incorporating additional methods such as surveys or interviews can add a holistic overview of citizens’ experience and motivations beyond the visible interactions on social media. Furthermore, a mixed-method study also offers confirmation of findings from one method, which can enhance the validity and robustness of the results.

Finally, as our analysis indicates that perceived responsiveness has no immediate impact on citizen engagement, it could possibly stem from its long-term impact beyond engagement. A consistent and significant government's responsiveness might lead to fostering long-term trust in government, and this indirect effect may not be instantly visible through citizen engagement. Therefore, future research could consider this longitudinal aspect, exploring how perceived responsiveness shapes citizens’ trust in government.

7. Implications and Conclusions

This study provides valuable insights into how the Malaysian government can leverage social media to enhance citizen engagement, addressing a gap where there is little research in this field. The application of these findings can consequently improve Malaysia's ranking in EPI. This study offers several practical implications that can be implemented by other ministries as well.

First, the Malaysian government can benefit from the research findings by realising that posting relevant information can enhance citizen engagement, which leads to increased collaboration. Instead of overfocusing on one specific content type (i.e., information provision), the government needs to balance the content posted. On the other hand, while online dialogue offline interaction, and input seeking have the least citizen engagement, the government still needs to improve on posting these content types to encourage deeper interaction. For instance, the division of the content type posting can be in the form of 60% of information provision, 25% of symbolic presentations, 10% of online dialogue offline interactions, and 5% of input seeking. By doing it this way, the governments can maintain their role as the information provider while encouraging citizen engagement. Moreover, the government can leverage the dialogic affordances of social media and promote more meaningful engagement through a specific strategy. For example, the government can increase live-streaming content or conduct surveys and polls to gather public opinions. These interactive strategies can encourage two-way communication as they foster citizens’ sense of inclusion, consequently enhancing citizen engagement. In addition, media content also plays a vital role in stimulating citizens’ visualisation and fostering their engagement. Therefore, the Malaysian government can strategise and maximise content posting by integrating suitable topics with the right media richness. Balanced and citizen-tailored content could improve the usage of Malaysian government online communication, which could elevate the EPI ranking.

Secondly, our study underscores the need for the Malaysian government to take action in improving its responsiveness quality on social media. Even though it was found to have no impact on citizen engagement, it is essential for rebuilding trust and ensuring better public service delivery. As the government can now be reached online, citizens rarely need to go to the counter to ask questions. Being able to answer comments promptly on social media will satisfy the citizens’ needs and improve the government's commitment to serving the citizens. The government could allocate special resources to improve perceived responsiveness by, for instance, appointing social media teams to specifically handle citizens’ inquiries on social media.

In conclusion, Malaysian government agencies should establish a clear goal and strategy to address their shortcoming and leverage social media to foster citizen engagement. It is important to note that posting more content does not guarantee the expected engagement. A well-thought-out content combined with quality media and effective responsiveness will be far more impactful in encouraging citizen engagement.

Footnotes

ORCID iDs

Zainatun N. Hussin

Christie Pei-Yee Chin

Stephen L Sondoh Jr

Author Biographies

Zainatun N Hussin is currently a Ph.D. student in Management at Universiti Malaysia Sabah. She received the B.S degree in Biotechnology in 2011 and the M.S. Degree in Human Capital Management in 2015 from Universiti Malaysia Sabah.

Christie Pei-Yee Chin received the doctoral degree in Information Systems from the University of South Australia, South Australia, Australia, in 2018. She is currently a Senior Lecturer with the Faculty of Computing and Informatics, Universiti Malaysia Sabah. She has authored and co-authored various international journals, such as Information Systems Frontiers, Internet Research, and Journal of Organizational Computing and Electronic Commerce, as well as proceedings of conferences such as AMCIS and PACIS. Her research interests include enterprise social networks, public social networks, social analytics, knowledge management, fake news, entrepreneurship, ESG and technology adoption and usage.

Stephen L Sondoh Jr is currently working as an Associate Professor with the Faculty of Business, Economics and Accountancy (FBEA), Universiti Malaysia Sabah. His research direction is related to consumer behaviour, brand image, brand loyalty and psychology marketing. Currently, he is also conducting research related to social marketing, green marketing and services, green branding, and social media adoption. He has published in various journals including Journal of Cleaner Production, International Journal of Sustainable Development & World Ecology, Jurnal Pengurusan, Asian Academy of Management Journal, Journal of Islamic Accounting and Business Research, European Journal of Business and Social Sciences, Journal of Sustainable Development, Malaysia Management Review (MMR, International Journal of Arts and Commerce, International Journal of Global Management Studies Professional and have participated in numerous national and international conferences.

References

Alam

Meenar

Barraza

Khalil

M. B.

Knopp

(2022). Citizen engagement on local government Facebook pages: Experience from Aotearoa New Zealand. Cities, 123(103584), 1–12. https://doi.org/10.1016/j.cities.2022.103584

Alam

S. L.

(2021). Many hands make light work: Towards a framework of digital co-production to co-creation on social platforms. Information Technology & People, 34(3), 1087–1118. https://doi.org/10.1108/itp-05-2019-0231

Alam

S. L.

Gill

(2020). A social engagement framework for the government ecosystem: Insights from Australian government Facebook pages. Forty-first International Conference on Information Systems (ICIS), India.

Al-Aufi

A. S.

Al-Harthi

AlHinai

Al-Salti

Al-Badi

(2017). Citizens’ perceptions of government’s participatory use of social media. Transforming Government: People, Process and Policy, 11(2), 174–194. https://doi.org/10.1108/tg-09-2016-0056

Arshad

Khurram

(2020). Can government’s presence on social media stimulate citizens’ online political participation? Investigating the influence of transparency, trust, and responsiveness. Government Information Quarterly, 37(101486), 1–18. https://doi.org/10.1016/j.giq.2020.101486

Basit

(2003). Manual or electronic? The role of coding in qualitative data analysis. Educational Researches, 45(2), 143–154. https://doi.org/10.1080/0013188032000133548

Bellucci

Biagi

Manetti

(2019). Dialogic accounting and stakeholder engagement through social media: The case of top-ranked universities. The Review of Higher Education, 42(3), 1145–1184. https://doi.org/10.1353/rhe.2019.0032

Birghan

Robert

Meschede

(2021). Challenges, barriers, and approaches for providing digital citizen information. A Case Study in North Rhine-Westphalia, Germany. Proceedings of the 54th Hawaii internationnal conference on system sciences.

Bonsón

Perea

Bednárová

(2019). Twitter as a tool for citizen engagement: An empirical study of the Andalusian municipalities. Government Information Quarterly, 36(3), 480–489. https://doi.org/10.1016/j.giq.2019.03.001

10.

Bonsón

Ratkai

(2013). A set of metrics to assess stakeholder engagement and social legitimacy on a corporate Facebook page. Online Information Review, 37(5), 787–803. https://doi.org/10.1108/oir-03-2012-0054

11.

Bonsón

Royo

Ratkai

(2014). Facebook practices in Western European municipalities. Administration & Society, 49(3), 320–347. https://doi.org/10.1177/0095399714544945

12.

Bonsón

Royo

Ratkai

(2015). Citizens’ engagement on local governments’ Facebook sites. An empirical analysis: The impact of different media and content types in Western Europe. Government Information Quarterly, 32(1), 52–62. https://doi.org/10.1016/j.giq.2014.11.001

13.

Chatfield

A. T.

Reddick

C. G.

(2018). All hands on deck to tweet # sandy: Networked governance of citizen coproduction in turbulent times. Government Information Quarterly, 35(2), 259–272. https://doi.org/10.1016/j.giq.2017.09.004

14.

Chen

Min

Zhang

Evans

(2021). Factors driving citizen engagement with government TikTok accounts during the COVID-19 pandemic: Model development and analysis. Journal of Medical Internet Research, 23(2), e21463. https://doi.org/10.2196/21463

15.

Chen

Min

Zhang

Wang

Evans

(2020). Unpacking the black box: How to promote citizen engagement through government social media during the COVID-19 crisis. Computers in Human Behaviour, 110(106380), 1–11. https://doi.org/10.1016/j.chb.2020.106380

16.

Contri

Fissi

Gori

(2024). Citizen engagement and dialogic accounting through social media: A study of Italian regions. Accounting, Auditing & Accountability Journal, 37(3), 866–892. https://doi.org/10.1108/aaaj-02-2022-5663

17.

Criado

J. I.

Villodre

(2021). Delivering public services through social media in European local governments. An interpretative framework using semantic algorithms. Local Government Studies, 47(2), 253–275. https://doi.org/10.1080/03003930.2020.1729750

18.

Daft

R. L.

Lengel

R. H.

(1986). Organizational information requirements, media richness and structural design. Management Science, 32(5), 554–571. https://doi.org/10.1287/mnsc.32.5.554

19.

Darwish

E. B.

(2017). The effectiveness of the use of social media in government communication in the UAE. Journal of Arab & Muslim Media Research, 10(1), 41–63. https://doi.org/10.1386/jammr.10.1.41_1

20.

Denktaş-Şakar

Sürücü

(2020). Stakeholder engagement via social media: An analysis of third-party logistics companies. The Service Industries Journal, 40(11-12), 866–889. https://doi.org/10.1080/02642069.2018.1561874

21.

DePaula

Dincelli

(2016). An empirical analysis of local government social media communication. In Proceedings of the 17th international digital government research conference on digital government research, Shanghai, China.

22.

DePaula

Dincelli

(2018). Information strategies and affective reactions: How citizens interact with government social media content. First Monday. https://doi.org/10.5210/fm.v23i4.8414

23.

DePaula

Dincelli

Harrison

T. M.

(2018). Toward a typology of government social media communication: Democratic goals, symbolic acts and self-presentation. Government Information Quarterly, 35(1), 98–108. https://doi.org/10.1016/j.giq.2017.10.003

24.

Elo

Kyngas

(2008). The qualitative content analysis process. Journal of Advanced Nursing, 62(1), 107–115. https://doi.org/10.1111/j.1365-2648.2007.04569.x

25.

Eom

S.-J.

Hwang

Kim

J. H.

(2018). Can social media increase government responsiveness? A case study of Seoul, Korea. Government Information Quarterly, 35(1), 109–122. https://doi.org/10.1016/j.giq.2017.10.002

26.

Fissi

Gori

Romolini

(2022). Social media government communication and stakeholder engagement in the era of COVID-19: Evidence from Italy. International Journal of Public Sector Management, 35(3), 276–293. https://doi.org/10.1108/ijpsm-06-2021-0145

27.

Gil de Zúñiga

Jung

Valenzuela

(2012). Social media use for news and individuals’ social capital, civic engagement and political participation. Journal of Computer-Mediated Communication, 17(3), 319–336. https://doi.org/10.1111/j.1083-6101.2012.01574.x

28.

Guidry

J. P. D.

Meganck

S. L.

Lovari

Messner

Medina-Messner

Sherman

Adams

(2020). Tweeting about #Diseases and #Publichealth: Communicating global health issues across nations. Health Communication, 35(9), 1137–1145. https://doi.org/10.1080/10410236.2019.1620089

29.

Haq

E.-U.

Braud

Lee

L. H.

Mogavi

R. H.

Zhang

Hui

(2022). Tips, tidings, and tech: Governmental communication on Facebook during the COVID-19 pandemic DG.O 2022: The 23rd annual international conference on digital government research.

30.

Haro-de-Rosario

Saez-Martin

Caba-Perez

M. d. C.

(2018). Using social media to enhance citizen engagement with local government: Twitter or Facebook? New Media & Society, 20(1), 29–49. https://doi.org/10.1177/1461444816645652

31.

Hasbullah

S. S.

Maynard

Chik

R. Z. W.

Mohd

Noor

(2016). Automated content analysis. In Proceedings of the 10th international conference on ubiquitous information management and communication, Danang, Vietnam.

32.

Heidemann

Klier

Probst

(2012). Online social networks: A survey of a global phenomenon. Computer Networks, 56(18), 3866–3878. https://doi.org/10.1016/j.comnet.2012.08.009

33.

Henisa

Wilantika

(2021). Content characteristics of government social media and the impact on citizen engagement rate. In 2021 international conference on informatics, multimedia, cyber and information system (ICIMCIS).

34.

Hidayat

S. E.

Rafiki

Al Khalifa

M. H.

(2019). The social media adoption of public sector in the Kingdom of Bahrain. Journal of Advances in Management Research, 16(1), 23–37. https://doi.org/10.1108/jamr-11-2017-0105

35.

Hofmann

Beverungen

Räckers

Becker

(2013). What makes local governments’ online communications successful? Insights from a multi-method analysis of Facebook. Government Information Quarterly, 30(4), 387–396. https://doi.org/10.1016/j.giq.2013.05.013

36.

Hussin

Z. N.

Chin

C. P.-Y.

Sondoh

S. L.

Choo

K.-K. R.

(2023). How do governments leverage the use of social media? A systematic review. IEEE Transactions on Engineering Management, 71, 7242–7256. https://doi.org/10.1109/tem.2023.3239382

37.

Islam

M. A.

Bhuiyan

M. A.-F.

(2023). Social media adoption in the public sector of Bangladesh: Progress and efficiency. Information Polity, 28(3), 439–449. https://doi.org/10.3233/IP-239011

38.

Islm

Meng

Pitafi

A. H.

Ullah Zafar

Sheikh

Shujaat Mubarik

Liang

(2021). Why DO citizens engage in government social media accounts during COVID-19 pandemic? A comparative study. Telematics and Informatics, 62(101619), 1–20. https://doi.org/10.1016/j.tele.2021.101619

39.

Karnowski

Leiner

d. J.

Kümpel

A. S.

Leonhard

(2021). Worth to share? How content characteristics and article competitiveness influence news sharing on social network sites. Journalism & Mass Communication Quarterly, 98(1), 59–82. https://doi.org/10.1177/1077699020940340

40.

Keib

Himelboim

Han

J.-Y.

(2018). Important tweets matter: Predicting retweets in the #BlackLivesMatter talk on twitter. Computers in Human Behavior, 85, 106–115. https://doi.org/10.1016/j.chb.2018.03.025

41.

Kent

M. L.

Taylor

(1998). Building dialogic relationships through the world wide web. Public Relations Review, 24(3), 321–334. https://doi.org/10.1016/S0363-8111(99)80143-X

42.

Kietzmann

J. H.

Hermkens

McCarthy

I. P.

Silvestre

B. S.

(2011). Social media? Get serious! Understanding the functional building blocks of social media. Business Horizons, 54(3), 241–251. https://doi.org/10.1016/j.bushor.2011.01.005

43.

Kim

Yang

S.-U.

(2017). Like, comment, and share on Facebook: How each behavior differs from the other. Public Relations Review, 43(2), 441–449. https://doi.org/10.1016/j.pubrev.2017.02.006

44.

Krippendorff

(2018). Content analysis: An introduction to its methodology (4th ed.). Sage.

45.

Lappas

Triantafillidou

Kani

(2022). Harnessing the power of dialogue: Examining the impact of Facebook content on citizens’ engagement. Local Government Studies, 48(1), 87–106. https://doi.org/10.1080/03003930.2020.1870958

46.

Lee

Kwak

Y. H.

(2012). An Open Government Maturity Model for social media-based public engagement. Government Information Quarterly, 29(4), 492–503. https://doi.org/10.1016/j.giq.2012.06.001

47.

Lee

(2018). The more attacks, the more retweets: Trump’s and Clinton’s agenda setting on Twitter. Public Relations Review, 44(2), 201–213. https://doi.org/10.1016/j.pubrev.2017.10.002

48.

Chandra

Fan

(2021). Unpacking government social media messaging strategies during the COVID-19 pandemic in China. Policy & Internet, 14(3), 651–672. https://doi.org/10.1002/poi3.282

49.

Luo

Duan

Shang

Lyu

(2021). Understanding citizen engagement and concerns about the COVID-19 pandemic in China: A thematic analysis of government social media. Aslib Journal of Information Management, 73(6), 865–884. https://doi.org/10.1108/ajim-11-2020-0377

50.

Malaysian Communication and Multimedia Commission, M . (2020). Internet Users Survey 2020. https://www.mcmc.gov.my/skmmgovmy/media/General/pdf/IUS-2020-Report.pdf

51.

McKibben

W. B.

Cade

Purgason

L. L.

Wahesh

(2022). How to conduct a deductive content analysis in counseling research. Counseling Outcome Research and Evaluation, 13(2), 156–168. https://doi.org/10.1080/21501378.2020.1846992

52.

Mergel

(2013a). A framework for interpreting social media interactions in the public sector. Government Information Quarterly, 30(4), 327–334. https://doi.org/10.1016/j.giq.2013.05.015

53.

Mergel

(2013b). Social media adoption and resulting tactics in the U.S. federal government. Government Information Quarterly, 30(2), 123–130. https://doi.org/10.1016/j.giq.2012.12.004

54.

Metallo

Gesuele

Guillamón

M.-D.

Ríos

A.-M.

(2020). Determinants of public engagement on municipal Facebook pages. The Information Society, 36(3), 147–159. https://doi.org/10.1080/01972243.2020.1737605

55.

Nah

Saxton

G. D.

(2013). Modeling the adoption and use of social media by nonprofit organizations. New Media & Society, 15(2), 294–313. https://doi.org/10.1177/1461444812452411

56.

Neely

S. R.

Collins

(2018). Social media and crisis communications: A survey of local governments in Florida. Journal of Homeland Security and Emergency Management, 15(1), 1–13. https://doi.org/10.1515/jhsem-2016-0067

57.

Noelle-Neumann

(1974). The spiral of silence a theory of public opinion. Journal of Communication, 24, 43–51. https://doi.org/10.1111/j.1460-2466.1974.tb00367.x

58.

Ononye

U. H.

Igwe

(2017). The utilisation of Facebook for knowledge sharing in selected local government councils in Delta state, Nigeria. Interdisciplinary Journal of Information, Knowledge, and Management, 12, 269–282. https://doi.org/10.28945/3854

59.

Panagiotopoulos

Barnett

Brooks

(2013). Social media and government responsiveness: The case of the UK food standards agency. 12th International conference on electronic government (EGOV), Koblenz, Germany.

60.

Paul

Das

(2023). Investigating information dissemination and citizen engagement through government social media during the COVID-19 crisis. Online Information Review, 47(2), 316–332. https://doi.org/10.1108/oir-06-2021-0307

61.

Rahim

A. I. A.

Ibrahim

M. I.

A. S.

Ariffin

M. A. I.

(2019). Health information engagement factors in Malaysia: A content analysis of Facebook use by the ministry of health in 2016 and 2017. International Journal of Environmental Research and Public Health, 16(591), 1–11. https://doi.org/10.3390/ijerph16040591

62.

Reddick

Chatfield

Ojo

(2017). A social media text analytics framework for double-loop learning for citizen-centric public services: A case study of a local government Facebook use. Government Information Quarterly, 34(1), 110–125. https://doi.org/10.1016/j.giq.2016.11.001

63.

Renshaw

S. L.

Mai

Dubois

Sutton

Butts

C. T.

(2021). Cutting through the noise: Predictors of successful online message retransmission in the first 8 months of the COVID-19 pandemic. Health Security, 19(1), 31–43. https://doi.org/10.1089/hs.2020.0200

64.

Ross

Potthoff

Majchrzak

T. A.

Chakraborty

N. R.

Lazreg

M. B.

Stieglitz

(2018). The diffusion on crisis-related communication on social media: An empirical analysis of Facebook Reactions. Proceedings of the 51st Hawaii International Conference on System Sciences.

65.

Saldana

(2021). The coding manual for qualitative researches (4th ed.). Sage.

66.

Sandoval-Almazan

Valle-Cruz

Sandoval-Almazan

Kavanaugh

Ignacio Criado

(2021). Social media use in government health agencies: The COVID-19 impact. Information Polity, 26(4), 459–475. https://doi.org/10.3233/ip-210326

67.

Sjoberg

F. M.

Mellon

Peixoto

(2017). The effect of bureaucratic responsiveness on citizen participation. Public Administration Review, 77(3), 340–351. https://doi.org/10.1111/puar.12697

68.

Stone

J. A.

Can

S. H.

(2021). Factors influencing tweet purposes and citizen engagement with municipal Twitter accounts. Online Information Review, 45(3), 501–516. https://doi.org/10.1108/oir-05-2020-0175

69.

United Nations, T . (2020). E-Government Survey 2020. https://publicadministration.desa.un.org/publications/un-e-government-survey-2020

70.

United Nations, T . (2022). E-Government Survey 2022: The future of digital government.

71.

Waters

R. D.

Williams

J. M.

(2011). Squawking, tweeting, cooing, and hooting: Analyzing the communication patterns of government agencies on Twitter. Journal of Public Affairs, 11(4), 353–363. https://doi.org/10.1002/pa.385

72.

Yang

Deng

Zhang

Mao

(2021). Promoting public engagement during the COVID-19 crisis: How effective is the Wuhan local government's information release?. International Journal of Environmental Research and Public Health, 18(118), 1–17. https://doi.org/10.3390/ijerph18010118

73.

Zavattaro

S. M.

Sementelli

A. J.

(2014). A critical examination of social media adoption in government: Introducing omnipresence. Government Information Quarterly, 31(2), 257–264. https://doi.org/10.1016/j.giq.2013.10.007

74.

Zhang

Yuan

Zhu

Chen

Evans

(2022). Does citizen engagement with government social media accounts differ during the different stages of public health crises? An empirical examination of the COVID-19 pandemic. Frontiers in Public Health, 10(807459), 1–19. https://doi.org/10.3389/fpubh.2022.807459

75.

Zheng

(2014). Innovation through social media in the public sector: Information and interactions. Government Information Quarterly, 31, S106–S117. https://doi.org/10.1016/j.giq.2014.01.011

	Engagement by Content Type	Kruskal-Wallis Test
	Information	Input	Online Dialogue	Symbolic			Asymp.	Effect
Ministry		Provision	Seeking	offline Interaction	Presentation	H value	Df	Sig.	Size
MOD	Mean Rank	275.14	80.25	184.98	304.20	14.549	3	0.002	0.03
Median	0.045	0.017	0.020	0.055
MOE	Mean Rank	125.21	-	123.37	212.53	38.783	2	<0.001	0.14
Median	0.040	-	0.035	0.164
MOH	Mean Rank	206.95	80.67	139.64	259.99	27.175	3	<0.001	0.07
Median	0.173	0.057	0.082	0.366

Perceived Responsiveness - Engagement	Spearman's Correlation	Sig. (2-tailed)
0.036	0.461

Engaging Citizens on Government Social Media Page: The Role of Content Characteristics

Abstract

Keywords

1. Introduction

2. Content Characteristics in Government Social Media

2.1 Hypotheses Development: Content Characteristics and Citizen Engagement

2.4 Perceived Responsiveness and Citizen Engagement

3. Research Method

3.1 Data Preparation

3.2 Data Coding and Analysis

4. Findings

Table 8. Citizen Engagement by Government Agencies. Engagement MOD MOE MOH Mean 0.081 0.211 0.421 Minimum 0.003 0.008 0.012 Median 0.046 0.045 0.175 Maximum 1.276 0.008 8.054 Standard Deviation 0.124 0.801 0.802

Table 11. Correlation among Perceived Responsiveness and Citizen Engagement. Perceived Responsiveness - Engagement Spearman's Correlation Sig. (2-tailed) 0.036 0.461

6. Limitations and Future Recommendations

7. Implications and Conclusions

Footnotes

ORCID iDs

Author Biographies

References

Table 8.
Citizen Engagement by Government Agencies.

Engagement MOD MOE MOH

Mean 0.081 0.211 0.421

Minimum 0.003 0.008 0.012

Median 0.046 0.045 0.175

Maximum 1.276 0.008 8.054

Standard Deviation 0.124 0.801 0.802

Table 11.
Correlation among Perceived Responsiveness and Citizen Engagement.

Perceived Responsiveness - Engagement Spearman's Correlation Sig. (2-tailed)

0.036 0.461