Introduction

Social media platforms, particularly WeChat, have become essential tools for fostering public engagement in China’s national parks (NPs). As the most widely used social media platform in China, WeChat provides unique opportunities to raise environmental awareness, promote community involvement, and support conservation efforts. The Chinese government has recognised the importance of ‘new media for government affairs,’ emphasising digital platforms like WeChat in modernising public services1. Despite global attention on the role of social media in environmental protection, research on the specific environmental factors—such as policies, news, and administrative strategies—that shape public interaction on WeChat, particularly in the context of China’s NPs, remains limited. NPs worldwide face significant challenges, including overtourism, which jeopardises ecological integrity and visitor experiences2. In China, NPs face similar pressures, demanding innovative strategies to manage visitor flows and enhance public participation.

While existing literature primarily focuses on traditional forms of public participation, such as community meetings and educational programmes, digital platforms have received less attention3. Although studies have examined the role of social media in environmental communication4, research specifically addressing China’s NPs, especially within the WeChat remains sparse. Studies indicate that social media plays a significant role in raising awareness of protected areas and influencing tourism patterns. For example, research on US NPs has shown a positive correlation between social media exposure and increased visitation5. Furthermore, social media is used to track public sentiment toward NPs and identify critical issues such as park management and environmental impacts6. However, engagement strategies in protected areas often focus on information dissemination rather than encouraging active public involvement in policy decisions or conservation efforts7. While WeChat is widely used in China’s NPs, its potential to foster public participation in environmental protection and policymaking remains underexplored due to its limited interaction features. Moreover, existing studies generally lack a socio-cognitive perspective, which is essential for understanding how environmental factors influence public behaviour.

Based on this, this study aims to investigate how environmental factors, such as policies, news, and academic conferences, shape public interaction on WeChat in China’s NPs. Using monthly panel data from official WeChat accounts from July 2018 to December 2023, this study employs advanced econometric models, including two-way fixed effects, Least Squares Dummy Variable (LSDV), and Difference Generalized Method of Moments (D-GMM). The study seeks to answer the following specific questions:

  1. (1)

    Do and how do environmental factors such as policies, news, and academic conferences influence public interaction behavior on China’s NPs WeChat account?

  2. (2)

    What role does WeChat play in the mediating effect of these environmental factors on public behaviour?

  3. (3)

    How does the administrative level of NPs moderate the relationship between environmental factors and public engagement?

By addressing these questions, the study provides valuable insights into optimising public engagement strategies in the management and conservation of NPs. This study introduces several innovative dimensions. First, it extends Social Cognitive Theory (SCT) to social media and NP governance, categorises environmental factors into policy, news and academic environment, and reveals how these three factors influence public behaviors through WeChat. Second, the study also provides a new empirical benchmark for understanding public perceptions and identifying effective policy strategies, highlighting the mediating role of WeChat and the moderating role of the administrative levels in public engagement.

Literature review

Environmental factors shaping public engagement: policy, news, and academic conferences

Drawing on existing research base, this paper explores how policy, news and academic conferences serve as environmental factors that influence public engagement with NPs via social media platforms such as WeChat8,9,10. These dynamics align with the ‘information’ and ‘community building’ stage in the layered model of social media engagement11, where users are initially exposed to issues and policies relevant to NPs.

The policy environment includes regulations, guidelines, and government decisions communicated through official channels, including social media. Digital platforms have become essential tools for informing the public about NP policies, encouraging advocacy, information sharing, and participation in management initiatives12. Social media’s broad reach facilitates the rapid dissemination of national policy messages13. Policies shared via WeChat shape how individuals understand and engage with NP-related issues, directly influencing public engagement with these matters14. The news environment also plays a pivotal role in shaping public attitudes towards NPs, as media coverage influences discourse and societal perceptions15. Traditional media (e.g. newspapers) still hold sway in public cognition, as exemplified by the rock art controversy in the Pedra Furada Reserve in Brazil16. Additionally, news stories related to environmental protection, biodiversity conservation, or wildlife management often trigger increased engagement on platforms like WeChat, sparking public interest and stimulating discussions17. Local media coverage, such as that of the Tuscan Archipelago NP, often reflects political tensions, while platforms like Facebook amplify political content18,19.

Academic conferences and research also influence public engagement by disseminating scientific knowledge related to NP management20. The growing role of social media in science communication has enabled broader public engagement on ecological issues such as climate change and species distribution studies21,22,23. For instance, the “A Year in the Life of Canada’s Climate Facebook” study demonstrates how social media platforms can foster action and digital engagement24, broadening public attention to ecological and environmental issues and encouraging more diverse voices in scientific discourse25. Research on China’s NP management, shared through WeChat since 2017, has increased public involvement and enhanced the credibility of conservation efforts.

Public engagement in the context of social media

Social media plays a central role in shaping public interactions with NPs. Research has shown a strong relationship between public interest in protected areas and their visibility on the internet26. Specifically, online reviews and discussions about NPs not only enhance public recognition of their ecological value but also create a positive feedback loop that encourages greater engagement in conservation activities27. In China, all five officially established national parks currently use WeChat as their primary social media platform. The Northeast Tiger and Leopard NP and Three River Source NP were the first to adopt WeChat in August and October 2017, respectively. This was followed by Wuyishan NP in July 2019, National Park of Hainan Tropical Rainforest in August 2020, and Giant Panda NP in September 2020. Since WeChat was introduced as a platform for disseminating information about NPs in China, it has significantly increased public engagement, promoted policy acceptance, and facilitated innovation in public services28,29.

On NP WeChat accounts, public engagement is defined as interactions with NP-related content, such as likes, watching, and reads. The ‘likes’ function represents one-way communication30. and can indicate deeper engagement when users resonate with content, prompting actions like sharing or commenting31,32. ‘Watching’ signifies reciprocal engagement, where users actively interact with content of tweets, potentially leading to further participation24. ‘reads’ measure passive engagement, marking the reach and visibility of NP-related content30. This initial engagement can eventually lead to more active participation. These various forms of engagement offer valuable insights into the public’s awareness, interest, and involvement in NPs management.

Public engagement within the social cognitive framework

Social Cognitive Theory (SCT) provides a robust framework for understanding how environmental factors shape public engagement. In the context of NP engagement on WeChat, external stimuli (such as policy, news, and academic content) influence individuals’ cognitive processes and behaviours. For example, exposure to policy information and news shapes individuals’ attitudes toward NP management, affecting their engagement on WeChat through actions like comments, likes, or shares.

WeChat facilitates interactions between the public, policymakers, the media, and the academic community, enabling dynamic engagement with topics like ecological research, climate change, and biodiversity conservation23,33. As social media grows in importance for environmental communication, WeChat plays a crucial role in raising awareness and promoting long-term participation in NP management and conservation efforts. This continuous interaction between environmental stimuli and personal factors reflects the concept of ‘connected action,’ where engagement behaviours are actively shaped by both cognitive and external influences.

From an SCT perspective, the ‘likes,’ watching,’ and ‘reads’ on WeChat represent different levels of engagement. The ‘likes’ align with cognitive processes, reflecting the public’s endorsement of content that resonates with their values. ‘Watching’ suggests deeper interaction, where individuals engage with content in various ways, and ‘reads’ signify initial, passive consumption of information. These engagement metrics indicate how individuals process and act on information related to NPs.

Theoretical basis and research hypotheses

Social cognitive theory (SCT)

Social Cognitive Theory (SCT), developed by Albert Bandura, posits that learning and behavioural change occur through the reciprocal interaction between personal factors, behaviour, and the environment34. In the context of NPs, SCT helps explain how individuals perceive, access, and act on information related to environmental conservation. Environmental factors play a crucial role in SCT by influencing both personal cognition and behavioural processes. This study further divides the environment into three key dimensions:

The policy environment includes regulations, guidelines, and decisions related to NPs, which shape public understanding and perception of NP management and conservation. Policy content disseminated via platforms such as WeChat not only informs but also influences public attitudes and behaviours regarding participation in NP-related activities. Social media platforms play a pivotal role in shaping how the public engages with NP policies.

The news environment involves media coverage that frames environmental issues and influences public perceptions of NPs and their conservation challenges. Media, by reporting on environmental protection, Media, by reporting on environmental protection, biodiversity, and ecological management, shapes public social cognition and spurs engagement on social media platforms.

The academic environment includes research findings, academic conferences, and scholarly discussions on NP management and conservation. These academic channels provide professional perspectives and deeper public understanding. When research is disseminated via social media, it enhances the public’s focus on NP conservation issues.

These three environmental factors not only independently influence public cognition and behaviour but also interact to shape public views and engagement in NP-related matters. Social media platforms, especially WeChat, facilitate the dissemination of policy information, news reports, and academic research, enabling the public to participate in conservation efforts.

Hierarchy of social media engagement

This study draws on the ‘information-community-action’ model11 to examine social media engagement on China’s NPs WeChat account. The model outlines a gradual progression of public participation across three stages:

Information Dissemination Stage: In this initial stage, the public primarily engages with and consumes information. For NPs, this often involves ‘reads’ content related to policies, news, or academic reports, shaping the public’s understanding of key issues.

Community Building Stage: Here, the public shifts from passive information consumption to active engagement. Actions such as ‘likes’ and ‘watching’ content signal approval and encourage further participation through sharing or commenting. This stage marks a transition from passive reception to social interaction.

Action Stage: In the final stage, the public moves beyond engagement to taking actual conservation actions, such as online advocacy, participating in discussions, or even taking part in offline NP management efforts.

Due to WeChat’s limited comment function, interactions are primarily through ‘likes’ and ‘watching’, restricting deeper discussion. Nonetheless, these actions allow the public to express interest in NP issues, laying the groundwork for future participation. Thus, social media serves as both an information dissemination tool and a catalyst for community building and behavioural change.

Research hypothesis

Direct influence from policy, news, and academic conferences

The government’s dissemination of NP policy information is crucial in shaping the public perceptions, with a significant impact across social media platforms35. Social media has proven effective in distributing policy content and supporting its implementation36. Policy communication through social media not only spreads information but also promotes public discussions and interactions, further increasing engagement31. Clear and effective policy communication enhances public awareness of conservation efforts, making individuals more likely to participate online and offline37. Thus:

  1. H1:

    The policy environment is positively associated with NP social media ‘likes’ (a), ‘watching’ (b), ‘reads’ (c).

The news environment serves as a powerful tool in framing environmental issues and influencing how the public views NPs and their conservation challenges. The rise of social media has revolutionised news dissemination, allowing rapid shifts in public perception38. High-quality news reporting increases public knowledge and stimulates engagement39. Media coverage of conservation issues, such as wildlife management or pollution in NPs, often drives significant social media interaction17. Therefore, the following hypotheses are proposed:

  1. H2:

    The news environment is positively associated with NP social media ‘likes’ (a), ‘watching’ (b), ‘reads’ (c).

Academic conferences facilitate the development and dissemination of knowledge through critical analysis of societal trends40. These platforms promote scientific dialogue and contribute to public knowledge41. Academic content, especially when shared through social media, engages a broader audience42. Social media helps extend academic discussions beyond conferences, encouraging public engagement in conservation issues43. Therefore:

  1. H3:

    The academic environment is positively associated with NP social media ‘likes’ (a), ‘watching’ (b), and ‘reads’ (c).

However, differing technology, interest, and knowledge levels may lead to varying public perceptions of issues in the news media, academia, and the public44. Traditional information dissemination may not always stimulate public engagement45. Scholars using platforms like Twitter during academic conferences have utilised multimodal strategies to engage discussions46. Despite social media’s accessibility, it may not always promote meaningful interactions, particularly when content is highly specialised42. This limitation often leads to fewer interactions on NP-related topics. Thus:

  1. H4:

    The academic environment is negatively associated with NP social media ‘likes’ (a), ‘watching’ (b), and ‘reads’ (c).

The mediating role of social media posts

Social media platforms are essential for promoting interpersonal communication and disseminating information from other sources47. For NPs, platforms like WeChat are essential for disseminating information and engaging the public in real-time government developments48. Social media bridges the gap between scientists, policymakers, and the public, enabling collaborative participation49. Policy-related information distributed through social media channels significantly influences public discourse and engagement metrics15. Posts aligned with users’ preferences tend to generate higher engagement, such as more likes and shares19. News coverage of environmental issues shared on social media often generates substantial public interaction50. Moreover, academic research shared via social media increases engagement, extending the reach of academic discussions and fostering deeper dialogues12. Therefore:

  1. H5:

    Posts positively mediate the relationship between the policy environment and ‘likes’ (a), ‘watching’ (b), and ‘reads’ (c).

  2. H6:

    Posts positively mediate the relationship between the news environment and ‘likes’ (a), ‘watching’ (b), and ‘reads’ (c).

  3. H7:

    Posts positively mediate the relationship between the academic environment and ‘likes’ (a), ‘watching’ (b), and ‘reads’ (c).

The moderating role of the administrative level

To explore the influence of administrative levels on NP governance and public engagement, we examine the hierarchical diffusion model common in Chinese public policy51,52. This top-down approach significantly influences the allocation of political resources, innovation diffusion, and government responsiveness, particularly for higher-level NPs, which benefit from stronger organisational capacity53,54,55. For instance, the Northeast Tiger and Leopard NP, with its higher administrative status, was the first to operate an official WeChat account.

Moreover, comparative studies highlight the importance of administrative coordination and power distribution for enhancing governance efficiency and policy diffusion. Examples include a ‘convention check’ of Austrian NPs revealing discrepancies in implementation measures across governance levels56, Germany’s ‘Biodiversity Explorer’ program, and analyses of marine NPs in Sweden and Norway57,58. The dissemination of big data policies by the Chinese government underscores the significance of considering administrative hierarchies in national policies for optimal outcomes59. Therefore:

  1. H8:

    The positive relationship between policy environment and ‘likes’ (a), ‘watching’ (b), and ‘reads’ (c) is amplified in NPs with higher administrative levels.

  2. H9:

    The positive relationship between news environment and ‘likes’ (a), ‘watching’ (b), and ‘reads’ (c) is amplified in NPs with higher administrative levels.

  3. H10:

    The positive relationship between academic environment and ‘likes’ (a), ‘watching’ (b), and ‘reads’ (c) is amplified in NPs with higher administrative levels.

Building on the above theories and assumptions, we propose the conceptual framework presented in Fig. 1.

Fig. 1
Fig. 1
Full size image

Theoretical Model of the Impact of Social-Cognitive Environment on Public Engagement Behavior.

Empirical research design

Research objectives

This study examines public interaction with the official WeChat accounts of five NPs in China. Managed by their respective NP authorities, these accounts ensure content standardisation, diversity, and authority while remaining publicly accessible. Users engage with NP content through likes, watching, and reads, offering insights into public interest and interaction patterns. To achieve comprehensive data coverage, we collected 561 policies, 1,495 news reports, and 161 academic conferences related to the five NPs. Additionally, using Python web scraping, we gathered 9,193 posts from NP WeChat accounts, along with engagement data on likes, watching, and reads. Table 1 provides details of the data sources.

The dataset includes all available policies, news reports, academic conferences, and social media posts to minimise bias. Data was collected from August 2017 to December 2023, with the final collection conducted on 16 April 2024. Among the five NPs, Giant Panda NP had 27 local management branch accounts at the time of collection. To ensure data consistency, this study focuses on data from the Chengdu Giant Panda NP WeChat account, selected based on its account activity and institutional validation. A data integrity check revealed no missing values, duplicates, or outliers. After processing, the final sample covers 287 observations.

Modelling

The study utilises an unbalanced panel dataset, where different entities (the five NPs) are observed over varying time periods, with some data points missing for certain periods. To mitigate potential biases, we conducted a Hausman test, which indicated that a fixed effects model is the most appropriate approach (\(\:{\chi}^{2}\)=131.53, p = 0.0000). We then tested individual fixed effects, time fixed effects, and two-way fixed effects separately. The results showed that the two-way fixed effects model provides the most comprehensive control, accounting for both entity-specific and time-specific factors, reducing omitted variable bias and correcting for heteroskedasticity. The baseline model is specified as follows:

$$\:{Y}_{it}={\beta\:}_{0}+{\beta\:}_{1}{Policy}_{it}+{\beta\:}_{2}{News}_{it}+{\beta\:}_{3}{Conference}_{it}+\gamma\:{Controls}_{it}+{\alpha\:}_{i}+{\lambda\:}_{t}+{\epsilon}_{it}$$
(1)

Where \(\:i\) represents the NP, \(\:t\) represents the month, and \(\:{Y}_{it}\) represents NP WeChat interactions (likes, watching, reads) for NP \(\:i\) in month \(\:t\). \(\:{\beta\:}_{0}\) is the intercept, \(\:{\beta\:}_{1}\), \(\:{\beta\:}_{2}\), \(\:{\beta\:}_{3}\) are the impact coefficients of policy, news, and academic conferences, respectively. \(\:{Policy}_{it}\), \(\:{News}_{it}\), and \(\:{Conference}_{it}\) represent the number of policies, news reports, and academic conferences for NP \(\:i\) in month \(\:t\). \(\:\gamma\:\) is the vector of control variable coefficients, \(\:{Controls}_{it}\) represents control variables, \(\:{\alpha\:}_{i}\) accounts for fixed effects for each NP that do not vary over time, \(\:{\lambda\:}_{t}\) accounts for time fixed effects, and \(\:{\epsilon}_{it}\) is the random error term.

Description of variables and data sources

Dependent variables: likes, watching, reads

Public engagement with NP WeChat accounts is captured through ‘likes’, ‘watching’, and ‘reads’, as three dependent variables. These engagement metrics align with existing social media impact research60. To mitigate heteroskedasticity, we applied natural logarithmic transformations to the data. Data collection was conducted using Python web scraping.

Independent variables: policy, news, academic conferences

The Policy Environment variable is measured by the monthly number of policies published by central and provincial government departments concerning five NPs, following the approach used in government WeChat studies61. The data is sourced from Pkulaw (https://pkulaw.com), which provides comprehensive legal documentation. This variable represents the formal regulations and guidelines that shape public perceptions of NPs and their activities.

The News Environment variable is measured by the monthly number of news reports on the five NPs published in major Chinese newspapers, inspired by research on the social media of EU institutions62. The data is sourced from a comprehensive database of major Chinese newspapers. This variable reflects public discourse and societal attitudes toward NPs.

The Academic Environment variable is measured by the monthly number of academic conferences featuring reports on NPs at domestic academic events20. The data is drawn from the China Knowledge Network (CNKI) conference database, which contains extensive records of academic events. This variable captures scholarly discourse and knowledge production related to NPs.

Mechanism variables: WeChat posts, administrative level

Mediator Variable – WeChat Posts: Inspired by Twitter engagement studies63, we selected NP WeChat posts as the mediator variable. Data was collected via Python web scraping, and logarithmic transformation was applied to account for missing values. This variable represents how environmental factors influence public engagement via social media.

Moderator Variable – NP Administrative Level: Based on research on big data and governance64, administrative levels were assigned as follows1:Northeast China Tiger and Leopard NP: 5, Three River Source NP: 4, and other NPs: 3.This variable moderates the relationship between environmental factors and public engagement.

Controlled variables

Consumer Price Index for Transport and Communication: Measures economic conditions affecting public online engagement65. Adjusted by multiplying by 0.01, data sourced from the National Bureau of Statistics.

Duration of NP WeChat Account Existence: Measured in logarithmic form to mitigate correlation with time effects62. Controls for differences in platform maturity and engagement trends.

Social Media Platform Competition & Comment Features: Assesses competitive presence of other social media platforms. Data was sourced from official NP websites, with a binary coding system (1 = feature used, 0 = not used), following the UN 2022 e-Government Survey criteria. The main variable indicator measurements are shown in Table 1.

Table 1 Indicator measures of variables and data sources.
Full size table

Descriptive statistics

Table 2 presents descriptive statistics for key variables. The variables— ‘likes’, ‘watching’, and ‘reads’—exhibit volatility, suggesting varying levels of engagement across content types, necessitate individual regression analysis. ‘reads’ dominate, indicating high content consumption but low active engagement (likes, watching), possibly due to content style or audience preferences. Policy data shows a standard deviation of 4.43, while news data has a standard deviation of 6.46, highlighting uneven monthly coverage. The variability in policy and news coverage suggests fluctuating public engagement, where more frequent and positive exposure may drive greater interaction.

Table 2 Results of descriptive statistics of variables.
Full size table

Empirical analyses

Test for multicollinearity of variables

Figure 2 shows the correlation matrix of the variables, with darker shades indicating stronger correlations. The analysis shows that all coefficients are within an acceptable range, peaking at 0.72 between ‘posts’ and ‘reads’ on NP WeChat. The VIF test confirms that there is no problem with multicollinearity, as the highest VIF value is 2.32, well below the threshold of 10. This ensures the stability of the subsequent regression analysis.

Fig. 2
Fig. 2
Full size image

Heat map of the matrix of correlation coefficients between variables.

Baseline regression results

Our empirical analysis employs a two-way fixed effects model to examine the direct impact of policies, news, and academic conferences on public interaction behaviors, controlling for WeChat posts, consumption indices, account duration, and media types. Table 3 presents the regression outcomes. Each additional policy correlates with a 1.50% rise in ‘likes’ (p < 0.05) and a 1.40% increase in ‘watching’ (p < 0.10), affirming H1a and H1b, but not H1c. Additional news items boost ‘likes’ by 2.60% (p < 0.10), confirming H2a but not H2b or H2c. Conversely, each academic conference is linked to a 7.80% decrease in ‘likes’ (p < 0.05) and a 7.10% drop in ‘watching’ (p < 0.10), verifying H4a and H4b, while hypotheses H3a, H3b, H3c, and H4c are not supported. Policies and news seem to bolster ‘community building’ on NP WeChat, whereas academic conferences may detract from it. The environmental factors’ influence on ‘reads’ remains insignificant, indicating no significant effect on messaging. Figure 3 is the path diagram visualizing the direct impact of policies, news, and academic conferences on public interaction behaviors.

Table 3 Two-way fixed effects baseline regression results.
Full size table
Fig. 3
Fig. 3
Full size image

Path diagram visualizing the direct impact of policies, news, and academic conferences on public interaction behaviors.

Endogeneity test

SCT suggests that there is a dynamic continuity in public engagement, implying persistent effects from changes in policy, news, and academic conferences, with the possibility of reverse causality. The public’s perception and behavior on WeChat can be shaped by these environmental factors, and vice versa. The Durbin-Wu-Hausmann test reveals a significant negative residual coefficient (β = −10.46, 1% level), indicating the need to account for endogeneity in the model. To address this issue, lagged ‘likes’, ‘watching’, and ‘reads’ are used as instrumental variables in a dynamic panel data model. To further mitigate the endogeneity arising from the use of lagged predictors, the least squares dummy variable (LSDV) regression is applied, utilizing Difference Generalized Method of Moments (D-GMM) estimators as initial values, with standard precision (1/T), and a 40-fold bootstrap method for robust estimation. Table 4 presents the LSDV results, with AR(1) tests close to 0 and AR(2) and Sargan tests above 0.05, validating the model. The signs and significance of the coefficients remain consistent with the baseline regression, confirming the robustness of the results.

Table 4 Regression results of endogeneity test for LSDV.
Full size table

Robustness tests

Robustness test for lagged independent variables

This paper conducts robust checks on the lagged explanatory variables, as detailed in Table 5. The one-period lagged NP policy retains a significant positive impact on public interactions, while the impact of news changes from positive to insignificant, and the negative impact of academic conferences at one lag becomes insignificant. These results further confirm the robustness of the results. After a two-period lag, the effects of policies, news, and conferences on interactions are no longer significant, suggesting that the influence of these environmental factors on public behaviors and perceptions diminishes over time.

Table 5 Robustness tests: lags of independent variables.
Full size table

Replacing the dependent variables

This paper reassesses the robustness of the results by changing the calculation of the dependent variables. We chose the monthly averages of ‘likes’, ‘watching’ and ‘reads’ as alternative dependent variables and applied a logarithmic transformation. Additionally, we also chose the Qingbo index to calculate the total and average communication power of NP WeChat as alternative dependent variables for further robustness testing. As shown in Table 6, the results confirm the empirical findings’ robustness. However, it is observed that the impact of policies and academic conferences is diminished within the dissemination power model, suggesting that the influence of these environmental factors on public behaviors and perceptions may wane in the context of communication strategies.

Table 6 Robustness tests: replacing dependent variables.
Full size table

Analysis of impact mechanisms

Mediating effects of WeChat posts

Policies, news, and academic conferences do not significantly influence ‘reads’, but WeChat posts significantly increase ‘likes’, ‘watching’, and ‘reads’, prompting an exploration of the influence mechanism. We investigated WeChat posts as a mediating variable and constructed mediation models as follows:

$$\:{Lnposts}_{it}={\pi\:}_{0}+{\pi\:}_{1}{Policy}_{it}+{\pi\:}_{2}{News}_{it}+{\pi\:}_{3}{Conference}_{it}+\delta\:{Controls}_{it}+{\alpha\:}_{i}+{\lambda\:}_{t}+{\varsigma\:}_{it}$$
(2)
$$\:{Y}_{it}={\beta\:}_{0}{+{\beta\:}_{1}}^{{\prime\:}}{Policy}_{it}+{\beta\:}_{2}{\prime\:}{News}_{it}+{\beta\:}_{3}{\prime\:}{Conference}_{it}+{\beta\:}_{4}{Lnposts}_{it}+\gamma\:{{\prime\:}Controls}_{it}+{\alpha\:}_{i}+{\lambda\:}_{t}+{\epsilon}_{it}$$
(3)

In model (2), \(\:{Lnposts}_{it}\) mediates the effects of policies, news, and conferences, with \(\:{\pi\:}_{1}\),\(\:{\pi\:}_{2}\),\(\:{\pi\:}_{3}\) as their coefficients. Model (3) shows \(\:{{\beta\:}_{1}}^{{\prime\:}}\),\(\:{{\beta\:}_{2}}^{{\prime\:}}\),\(\:{{\beta\:}_{3}}^{{\prime\:}}\) reflecting the influences on interaction behaviors post-mediation, and \(\:\gamma\:{\prime\:}\) as control variable coefficients.

The Sobel-Goodman test and the bootstrap method (using 1000 replicate samples) were used to analyze mediation effects. The results for ‘likes’ show that news, transmitted via WeChat posts, has a significant positive effect. Specifically, the Sobel test p = 0.000, bs_1 p = 0.000 and bs_2 p = 0.004, which are all less than 0.05, confirming a partial positive mediation effect, which accounts for 40.59% of the total effect. For ‘watching’, the Sobel p-value and bs_1 are both less than 0.05, and p > 0.05 for bs_2, indicating a positive mediating effect. The positive mediating effect is 2.69 times larger than the direct effect, which causes the direct effect of news on ‘watching’ to become insignificant. In the case of ‘reads’, both the Sobel p-value and bs_1 are less than 0.05, while bs_2 p = 0.167, indicating a positive mediating effect. The mediating effect is 1.59 times larger than the direct effect, making the direct effect of news on “reading” insignificant. This confirms that WeChat posts mediate the effect of news on NP WeChat interactions, thereby supporting hypotheses H6a, H6b, and H6c. However, the mediation of policies and academic conferences by posts was not supported, leading to the rejection of hypotheses H5a, H5b, H5c, H7a, H7b, and H7c.

Moderating effects of the administrative level

We examined the moderating influence of the NP administrative level on the relationship between policy, news, academic conferences, and public interaction behaviour using the following model:

$$\:\begin{array}{c}{Y}_{it}={\beta\:}_{0}+{\beta\:}_{1}{Policy}_{it}+{\beta\:}_{2}{News}_{it}+{\beta\:}_{3}{Conference}_{it}\\\:+{\beta\:}_{4}\left({Level}_{it}*{Policy}_{it})+{\beta\:}_{5}({Level}_{it}*{News}_{it}\right)\\\:+{\beta\:}_{6}\left({Level}_{it}*{Conference}_{it}\right)+\gamma\:{Controls}_{it}+{\alpha\:}_{i}+{\lambda\:}_{t}+{\epsilon}_{it}\end{array}$$
(4)
Fig. 4
Fig. 4
Full size image

Path diagram visualizing the moderating effects of the NP administrative level.

Where \(\:{Level}_{it}\) represents the administrative level of NP \(\:i\) at time \(\:t\), and \(\:{\beta\:}_{4}\), \(\:{\beta\:}_{5}\), \(\:{\beta\:}_{6}\) are the coefficients of the interaction terms for policy, news, and academic conferences, respectively. The results in Table 7 indicate a significant positive interaction effect between NP levels and policies for ‘reads’, significant at the 5% level (β = 0.72), confirming H8c, while H8a and H8b were not supported. This suggests that higher NP levels enhance information dissemination on WeChat. Conversely, the interaction terms for news exhibit significant negative coefficients at the 1%, 5%, and 1% levels, implying that higher NP levels reduce the positive impact of news on interactions. This contradicts H9a, H9b, and H9c. Regarding academic conferences, the interaction effects were not significant, leaving H10a, H10b, and H10c unconfirmed. Figure 4 presents a path diagram visualising the moderating effects of the NP administrative level on public interaction behaviour.

Table 7 Moderating effects: administrative level.
Full size table

Heterogeneity analysis: regional and developmental heterogeneity

Given the regional differences, NP social media interactions may vary due to factors such as online attention66. To explore this, the five NPs were divided into eastern and western groups based on their geographical location. The Three River Source NP and Chengdu Giant Panda NP are located in the west, while the remaining three NPs are located in the east. After excluding 58 unique observations, a multivariate analysis of variance (MANOVA), with p-values all approaching 0, revealed significant regional differences in the impact of environmental factors (such as policies, news, and academic conferences) on public interaction behaviour.

As shown in Table 8, the direct and moderating effects of policies and news differ by region. In the eastern group, policy significantly increases ‘likes’ (p < 0.01), ‘watching’ (p < 0.05), and ‘reads’ (p < 0.10), while in the west, no significant effects were observed. The interaction between administrative level and policies in the east has a positive effect on ‘likes’ (β = 0.34, p< 0.05), while in the west, the effect on ‘reads’ is significantly negative (β=−0.07, p < 0.10), suggesting that policy localisation plays a role in shaping public interaction behaviour67. The interaction between administrative level and news in the east negatively affects public interaction behaviour (p < 0.01), while no effect is found in the west. This suggests that the NP administrative level does not moderate the relationship between the academic environment and public interaction, reflecting the academic independence in NP research fields. Figure 5 presents the path diagram visualising the direct and moderated effects by region.

Fig. 5
Fig. 5
Full size image

Path diagram visualizing the direct and moderated effects by region. (a) Path diagram for western region; (b) Path diagram for eastern region.

Regarding the temporal dimension, the data were split into two periods: before (prior to October 2021) and after the establishment of the NP. Two observations were dropped due to uniqueness. The MANOVA revealed significant differences in the impact of policy, news and academic conferences across these periods. The regression results in Table 9 show that news significantly increased ‘likes’ (p < 0.05), ‘watching’ (p < 0.10) and ‘reads’ (p< 0.05) in the post-establishment period, while no significant effects were found in the pre-establishment period. Academic conferences had a negative effect on ‘likes’ (p < 0.05) and ‘watching’ (p < 0.05) in the pre-establishment period, with no effect in the post-establishment period, indicating the evolution of NP policy across developmental stages68.

In the moderated effects model, the interaction of administrative level with policy positively influenced ‘reads’ (β = 0.07, p < 0.05), while its effect on ‘watching’ in the pre-establishment period was significantly negative (β = −0.43, p < 0.10). The interaction between administrative level and news negatively affected public interaction behaviour in the pre-establishment period (p < 0.01), with no significant effect on ‘likes’ and ‘watching’ after the NP establishment. The lack of a moderating effect of administrative level between academic conferences and WeChat interactions highlights the academic autonomy within NP research fields. Figure 6 visualises the direct and moderating effects by temporal dimension.

Fig. 6
Fig. 6
Full size image

Path diagram visualizing the direct and moderate effects by region. (a) Path diagram for the pre-establishment period; (b) Path diagram for the post-establishment period.

Table 8 Heterogeneity analyses: regressions of area heterogeneity groupings.
Full size table
Table 9 Heterogeneity analysis: regression of heterogeneous subgroups by stage of development.
Full size table

Discussion and conclusions

Discussion

This study examines the influence of policy, news, and academic environments on public interaction behaviours (likes, watching, and reads) on NP social media platforms, focusing on the mediating role of WeChat posts and the moderating effect of administrative levels. The key findings are summarised below.

Policy environment

Policies positively influence ‘likes’ and ‘watching’, but their effect on ‘reads’ is not significant. The moderating role of administrative levels amplifies the effect of policies, particularly in enhancing information dissemination for ‘reads’. This finding aligns with policy diffusion theory, which suggests that well-communicated policies stimulate community engagement69. Furthermore, institutional theory70suggests that policies issued by higher administrative levels are perceived as more authoritative, leading to increased public engagement71,72. This reinforces the idea that institutional legitimacy plays a crucial role in shaping public participation. The effectiveness of policies in driving engagement, particularly in initial interactions such as ‘likes’ and ‘watching’, supports the argument that communication strategies are essential in managing overtourism2.

News environment

News significantly increases ‘likes’, but its effect on ‘watching’ and ‘reads’ is not significant. The mediation analysis confirms that WeChat posts play a pivotal role in facilitating news engagement, particularly in enhancing ‘likes’. However, higher administrative levels reduce the effectiveness of news in driving interactions. News is more engaging when disseminated through trusted social media platforms17. However, formal news from higher governance levels may be perceived as too authoritative or distant, reducing its immediacy and relatability73. The framing of news content is crucial, as centralised information may not resonate well with the public if it fails to align with audience expectations74. From the perspective of social cognitive theory, WeChat news provides vicarious experiences, allowing users to stay informed about NP development without physical visitation. Positive or inspiring news stories enhance self-efficacy, which in turn promotes public interaction75.

Academic environment

Academic conferences negatively affect ‘likes’ and ‘watching’, with no significant effect on ‘reads’. Additionally, NP administrative level does not moderate the relationship between the academic environment and interaction behaviours. This challenges the assumption that academic activities inherently foster public interest. Social media has provided valuable opportunities for academic dissemination, particularly during the pandemic, when virtual events became an alternative to in-person conferences76. However, with the return of face-to-face conferences and changes in social media algorithms, the effectiveness of these channels for academic engagement has diminished. Academic content is often highly specialised, making it less engaging for the public77. To increase public engagement, academic communication should prioritise accessibility and emotional resonance rather than focusing solely on technical details4,46.

Regional and Temporal heterogeneity

Significant regional differences are observed, with policies and news having stronger effects in the eastern region. In contrast, the western region exhibits a more localised response, particularly in policy-related engagement. The temporal analysis shows that news had a stronger impact post-establishment of NPs, while academic conferences negatively affected interactions before NP establishment. These findings highlight the importance of contextual factors in shaping public engagement, reinforcing the argument that policy implementation and public response are highly localised78. Furthermore, public cognition of policies is heterogeneous79. Understanding local migration patterns and demographic factors is essential for effective NP conservation management80.

Conclusions

This paper applies social cognitive theory and the hierarchy of social media participation to examine the role and mechanisms of environmental factors—particularly the policy, news, and academic environments—in influencing public interaction behaviours on WeChat. Using monthly panel data from July 2018 to December 2023, along with econometric methods including two-way fixed effects, LSDV, and D-GMM estimators, we reveal the complex dynamics of public interaction behaviours and messaging influenced by these environmental factors.

Key findings

Our findings highlight the direct impact of environmental factors on public interaction behaviour on WeChat. The policy and news environment, characterised by government initiatives aimed at modernising public services, significantly shaped public interest in NPs. In contrast, the academic environment has a negative effect on public engagement. These factors primarily influence the ‘community building’ level of the social media participation hierarchy, rather than directly affecting the ‘information’ layer. This suggests that while policies and news content are essential for shaping community engagement, they are less effective in driving the basic transmission of information.

The study also highlights the mediating role of WeChat in amplifying the effects of policy and news on public interaction behaviours. This underscores the platform’s role in translating environmental factors into observable public behaviours, thereby serving as a conduit for the influence of policy, news, and academic discussions. But due to the lack of a commenting feature, the application of the WeChat platform in the context of NPs remains largely at the stage of information dissemination and community building, and it has not effectively promoted the public to take concrete actions to participate in environmental protection or the formulation of relevant policies.

The study also reveals the moderating effect of NP administrative levels on the relationship between environmental factors and public behaviours. Higher administrative level seems to intensify the impact of policy-related messages, potentially overshadowing the influence of news. This implies that while news can stimulate engagement, higher administrative levels may reduce the immediacy and relatability of the news content. Additionally, the objectivity and scientific nature of academic environments, while valuable, may not directly translate into increased social media engagement, suggesting a need for more accessible and engaging communication strategies in academic settings.

Finally, the study identifies significant regional and developmental heterogeneity in NP engagement. The regional differences highlight the importance of tailoring public service strategies to account for local contexts and specific developmental stages of NPs. This underscores the need for context-sensitive strategies that adapt to the evolving needs and circumstances of different NP locations.

Theoretical and managerial implications

From a theoretical perspective, the findings contribute to environmental studies and social media research, advancing our understanding of how digital communication strategies can enhance public engagement in environmental conservation. The study reinforces the relevance of social cognitive theory in explaining how external stimuli such as policy, news, and academic content shape public cognition and behaviour. Additionally, the results suggests that social media posts amplify public engagement by acting as intermediaries, filtering content before it reaches the public.

From a managerial perspective, the study offers several actionable recommendations for policymakers and NP authorities. First, policymakers should create a supportive policy and news environment by continuing to modernise public services and leverage platforms like WeChat to foster community engagement and shared responsibility23,81. Second, academic content should be made more accessible to the public through collaborations with influencers and content creators, enhancing its engagement and relevance20. Third, social media interactivity should be enhanced, with features like commenting added to WeChat, enabling direct public feedback and participation in conservation efforts82. Finally, public service strategies should be tailored to the regional and developmental contexts of each NP, considering factors such as culture, economy, and environment to improve public engagement effectiveness.

Limitations

This study has several limitations. Although WeChat is the most widely used platform in China’s NPs, relying solely on WeChat may overlook interactions on other social media platforms. Future research could benefit from multi-platform analysis. Additionally, while this study applies social cognitive theory, it does not explore individual cognitive factors such as beliefs and expectations. A more comprehensive approach integrating individual, environmental, and behavioural factors would be beneficial. Additionally, future communication strategies should consider the emotional resonance of content, local contexts, and institutional legitimacy to foster greater public participation in conservation efforts.