Diversity of information refers to the heterogeneity of media content available to individuals within a society, which is an essential part of a pluralistic society and is crucial for democratic deliberation and governance (Van Cuilenburg, 1999; Rössler, 2007; Entman, 2007; Stroud, 2011). Diversity of information has been studied in media policy circles on different levels and has been the object of media investigation by regulatory bodies (Iselin, 1988; Van Dijk, 1995; McQuail, 2002; Carpenter, 2010). The research on how diversity of information has affected our society can be categorized from a societal level to an individual level. On a societal level, the diversity of information is an indicator of freedom for democracy. Scholars have reached an agreement that the more diversity of information, the greater expressive freedom for democracy (Himelboim and Limor, 2008; Ananny, 2014) and a country’s political ecology (Himelboim and Limor, 2008). For example, Entman (Entman, 2007) argued that diversity of information is important for ensuring a well-informed public, which helps prevent media monopolization. Bozdag et al. (Bozdag et al., 2014) analyzed Twitter user data to investigate whether users isolate themselves in ideological bubbles by following individuals solely from one end of the political spectrum, thereby exhibiting biased sharing behaviors. Beyond serving as an indicator of democratic vibrancy within society, Scholars have pointed out that diversity of information exerts a significant influence at the individual level, shaping information freedom perception (Himelboim and Limor, 2008; Ananny, 2014; Sætra, 2019), media trustworthiness (Aarts et al., 2012; Ceron, 2015), and political trust (Avery, 2009; Ariely, 2015; Ceron, 2015; An et al., 2012).

Political trust has been defined as a basic evaluative orientation toward the government (Green, 1962). When a government is perceived as trustworthy, citizens are more likely to comply with demand (Levi and Stoker, 2000), which will benefit both elected officials and political institutions (Hetherington, 1998). Political trust is an objective that has been widely studied in government studies (Hoffmann and Putnam, 2003; Stolle and Hooghe, 2005) because it is closely related to how the government changes its related regulation and policy. However, rather than focusing on the impact of political trust, scholars in media studies have been paying attention to determinants of political trust, such as social attitudes (Schoon and Cheng, 2011), media usage (Avery, 2009), social media and online media outlets (Ceron, 2015), etc. The purpose of studying determinants of political trust in media world is to leverage media as a tool to influence citizens’ thoughts and behaviors (Iselin, 1988; De Almeida et al., 2014; Norris, 2011; Ceron and Memoli, 2015; Merkley et al., 2020; Sætra, 2019). As we entered into the Web 3.0 era, scholars have been studying how digital media affects political trust, particularly in terms of diversity of information and media use (Brosius et al., 2019; Strömbäck et al., 2016; Shen and Guo, 2013)). Existing research confirms the relevance of information diversity for both democratic participation and individual-level trust formation, yet it rarely explains the mechanisms through which diverse information becomes translated into political trust, nor does it examine these mechanisms under regulated or hybrid media conditions. The following hypothesis is proposed for research drawing on the insights of previous studies:

Information freedom perception, reflecting to which people see their information choices limited by others (Verme, 2014), is highly related to diversity of information. When individuals have access to a wide range of information sources, they feel less restricted in their choices and perceive a greater sense of freedom in their ability to seek and evaluate information (Zhao et al., 2023). The availability of diverse viewpoints helps mitigate feelings of censorship and control, fostering a sense of autonomy and freedom (Ju et al., 2023). Studies have shown that an increase in information diversity leads to higher levels of perceived information freedom because individuals can compare different perspectives and make more informed decisions (Xiao, 2013; De Almeida et al., 2014). Especially in the Era of Big Data (Sivarajah et al., 2017; Uche, 2017), the effects of the diversity of information on information freedom perception are widely discussed in both communication and political studies. The proliferation of digital platforms and user-generated content contributes to a richer information environment, enhancing perceptions of freedom. For example, De Almeida et al. (2014) adopted survey data of 555 active members of two XBOX online P3 communities in Brazil and found that greater freedom of information perceived within a community will positively affect their trust in the media platform and social identification. What is more, Ariely (Ariely, 2015) analyzed survey data from 32 European countries and found that media trust and political trust are positively related, also, when there is higher diversity of information, the higher the information freedom perception and higher the political trust, the magnitude of the relation differs across countries. Guided by previous research, we formulate the following hypothesis:

Media trust is the extent to which citizens evaluate media content as credible, reliable, and meeting the public interest (Tsfati and Ariely, 2014), which has long been a key variable in media studies from its early days, dating back to the pioneering work of Carl Hovland (Hovland et al., 1953). The significance of media trust extends beyond news consumption and media behavior in communication studies (Tsfati and Cohen, 2005; Ladd, 2011). Media trust influences how individuals process information and form attitudes about news and media outlets, affecting their overall media consumption patterns and behaviors (Shen and Guo, 2013; Brosius et al., 2019). Higher media trust leads to greater reliance on media for information, while lower trust may drive individuals to seek alternative sources or become skeptical of all media content. Furthermore, media trust is a valuable mediator to political trust in political studies (Levi and Stoker, 2000; Shen and Guo, 2013). This is because the media acts as a primary conduit for political information (Brosius et al., 2019). When citizens trust the media, they are more likely to trust the political institutions it covers; conversely, low media trust can lead to political cynicism and decreased trust in the government. In the digital era, some scholars pointed out that information freedom perception has influenced media trust in different countries with different ideologies. Generally, information freedom perception is affected by the diversity of media sources, which can be typically categorized as traditional media outlets and alternative media outlets (Xu et al., 2022; Schranz et al., 2018). In the Western world, scholars have reached the argument that people who are more regularly use alternative media sources have less trust in media. However, the Chinese media environment is hugely different from the Western world, with its characteristic of a combination of rapid technological advancement and strict government oversight, which makes it a unique landscape within the global media ecosystem (Tai and Sun, 2007; Yao, 2008). Chinese citizens tend to consume alternative media to traditional media to seek diversity of information as well as the cross-validation of certain information types. Thus, it can be assumed that information freedom perception has a positive impact on media trust.

Though political trust and media trust are customarily studied independently, these two are regularly examined together to explore the influence of media on political issues (Jones, 2004; Gronke and Cook, 2007). The relationship between media trust and political trust is grounded in the role of media as a primary source of political information. When citizens perceive the media as trustworthy, they are more likely to accept and believe the information presented, which can enhance their trust in political institutions and leaders (Schoon and Cheng, 2011; Brosius et al., 2019). Conversely, when media trust is low, citizens may view the media as biased or unreliable, leading to political cynicism and decreased trust in government (Tsfati and Ariely, 2014; Kalogeropoulos et al., 2019). This interconnection highlights the media’s significant role in shaping political perceptions and trust. Some scholars have agreed that media trust is an important indicator of political trust in democracies (Ariely, 2015; Brosius et al., 2019). For example, Tsfati and Ariely (Tsfati and Ariely, 2014) used the World Value Survey of 44 countries to determine the predictors of media trust, which includes media utilization behaviors, intensity of media use, diversity of media categories, etc. In addition, Kalogeropoulos et al. (2019) based on a survey of news users in 35 countries found that diversity of information, especially alternative news sources, is associated with higher levels of media trust. Some researchers, considering the political trust and media trust, explore the influence of media on political issues (Jones, 2004; Gronke and Cook, 2007). For example, Schranz et al. (2018) adopted survey data on media usage and media trust in 2016 for 13 selected countries and applied regression analysis to find that higher news consumption behavior in digital media world promotes media trust and strengthens political trust. Lu et al. (2019) adopted two surveys conducted in China in 2013 and 2018, and they found that media use and media trust shape citizens’ political trust, especially different types of media have various effects on political trust. While information freedom perception and media trust have been acknowledged as predictors of political trust, existing studies rarely integrate them into a sequential mediating process, nor do they contextualize their interaction under differential governance conditions. Thus, we propose the following hypothesis:

Need for cognition is a psychological term that describes an individual’s tendency to engage in and enjoy effortful cognitive activities (Cacioppo and Petty, 1982), which reflect the extent to which individuals prefer critical thinking, complex problem solving, information process, etc. (De Choudhury et al., 2011; Vraga and Tully, 2021). In communication studies, need for cognition is seen as a dimension to measure media literacy (Daneels and Vanwynsberghe, 2017; Craft et al., 2017). Multiple prior studies have investigated how diversity of information in media would trigger people’s need for cognition and stir certain information behavior in marketing and information system (Frewer et al., 1999; Lu et al., 2019; Lambillotte et al., 2022). For example, Lu et al. (2019) adopted data of the World Values Survey with 51,960 residents in 36 countries to find that diversity of information has positively influenced residents’ need for cognition and political trust. Also, Lambillotte et al. (2022) based on signaling theory proposed that not only people need for cognition is affected by a diversity of information in their media environment, but also information freedom perception and information transparency are influenced by people’s need for cognition.

As a mediator, the need for cognition will adjust the relationship between the diversity of information and information freedom perception, media trust, and political trust (Maksl et al., 2015). Scholars have found that the need for cognition positively affects information freedom perception, as people with higher need for cognition are more likely to be motivated to seek information, which can help assume that they are more likely to be exposed to a variety of information (Xiao, 2013; Lambillotte et al., 2022). For example, Kessler and Zillich (2019) conducted an experimental study with 56 undergraduate students from a German university that consisted of a survey and eye-tracking while browsing the internet to find that users with higher need for cognition have a positive effect on information freedom perception cause they are more likely to spend time to expose themselves to information. What is more, Flanagin et al. (2020) conducted an experiment using a representative sample of the US Internet users in online encyclopedias and found that need for cognition adjust information freedom perception and how people process information to generate thoughts and trust over media.

In terms of media trust and political trust, scholars have agreed that need for cognition have positively shaped media trust and political trust (Kalogeropoulos et al., 2019; Lu et al., 2019; De Choudhury et al., 2011). For example, Briñol and Petty (2015) found in their research that need for cognition, including validation processes, is important for understanding whether thoughts are reliable and whether media is trustworthy. Also, Schreurs and Vandenbosch (2021) adopted social media literacy model in their study and found that need for cognition has positive effect on media trust as a construct of media literacy. Taken together, the literature confirms that need for cognition plays a moderating or mediating role in shaping perceptions of information freedom, media credibility, and political trust, yet it is still unclear how this psychological mechanism operates within hybrid media systems where information diversity is simultaneously expanded and regulated. Therefore, we posit the subsequent hypothesis based on earlier studies, as shown in Figure 1:

Research on political trust has historically been grounded in liberal-democratic media systems, where information diversity, media autonomy, and institutional accountability are assumed to operate in mutually reinforcing ways. However, this assumption does not fully apply to hybrid or authoritarian media ecologies, where digital pluralism coexists with state-oriented information governance. In such systems, political trust is not solely a product of institutional performance or individual media use, but also reflects the negotiated relationship between state-led communication, platform infrastructure, and citizens’ cognitive strategies for evaluating information. Moreover, while digitalization has expanded the channels through which citizens can access and circulate information, this expansion does not necessarily weaken state influence; instead, it reshapes how credibility and authority are negotiated between government, media institutions, and users (Vraga and Tully, 2021). Therefore, studying political trust in hybrid media environments requires moving beyond Western-origin theories that equate media pluralism with institutional voluntarism. By situating the Chinese post-pandemic case within this broader theoretical shift, the present study contributes to an emerging strand of research that re-examines the cognitive, perceptual, and institutional mechanisms of political trust under conditions of both digital diversification and communicative governance.

The objective of this research is to advance the knowledge of on what level people trust media and politics in the digital media era in Chinese Characteristic settings. To this purpose, we investigated the participants nationwide to explore the mechanism of the influencing factors affecting political trust in China based on our proposed theoretical model. The questionnaire is structured into three parts. The first section collects demographic information of the participants, including gender, age, educational level, major, and political affiliation. The second section probes the participants’ connections with Hubei province—a region severely affected early in the pandemic—inquiring about their residence, employment, or educational status in Hubei during the outbreak, as well as their personal or familial connections to the province and any direct relations to infected individuals. This part also asks participants about their media usage. The third section assesses key constructs of our study using established scales to ensure validity and reliability. Measures include the diversity of information, need for cognition, information freedom perception, media trust, and political trust. 6 items measuring the diversity of information were adapted from McDonald and Dimmick (2003). The need for cognition was measured using a scale developed by Lord and Putrevu (2006). Perception of information freedom was evaluated with items from Breskaya and Giordan (2019). Media trust was measured with 3 items from Zeng and Dou (2019), while political trust was assessed with seven items adapted from previous studies by Craig et al. (1990) and Zhang and Sha (2024).

A pilot study with 53 participants was undertaken to ensure the clarity of the questionnaire and its appropriateness for the post-pandemic context and Chinese Characteristic settings. The pilot study yielded a Cronbach’s Alpha value exceeding the minimum acceptable threshold of 0.70 (Bland and Altman, 1997), and the questionnaire was then slightly modified based on participants’ feedback. As a result, 27 questions were specifically tailored to suit the Chinese context. The finalized survey was distributed through a simple random sampling method. Announcements were made online, providing an introduction to the research, a link to the survey, and participation instructions. A total of 435 questionnaires were collected between October, 20th to December, 25th in 2023. Given the context of recent public health developments, it is particularly timely to conduct this survey in this period. In December 2022, China lifted all COVID-19 restrictions, and throughout 2023, the population has been dealing with various epidemics such as influenza A, influenza B, and Mycoplasma pneumonia. This period provides a unique opportunity to examine the formation and evolution of political trust in the post-pandemic era, as it encompasses the public’s reaction to the lifting of restrictions and their ongoing experiences with new health crises. During this period, public trust in the government has significantly eroded, with many people equating these new epidemics with COVID-19, reflecting a pervasive skepticism rooted in the government’s actions during the pandemic (Zhai, 2023; You et al., 2024). This distrust underscores the need to investigate the current state of public opinion and the lingering effects of the pandemic on political trust. By exploring these dynamics, we can better understand how the handling of COVID-19 and subsequent health crises has impacted public confidence in government. After excluding incomplete responses and those failing an attention check (the instruction was “Please choose ‘disagree’“), 370 valid questionnaires were retained, resulting in an 85.01% response rate.

Common Method Bias (CMB) is a potential issue in studies that use self-reported data due to the possible influence of the survey method on observed variances (Kock, 2015a). As this study is collected using a single data source, a common method bias test was employed to evaluate the potential presence of CMB in our data collection. Although, Podsakoff et al. (2003) proposed that Harman’s single-factor should be used to eliminate the common method bias, Guide and Ketokivi (2015) argued that it is no longer adequate for assessing the CMB. Consequently, a fully collinearity test by evaluating the Variance Inflation Factor (VIF) is opted in the study. Our result offered reassurance, indicating that the model was free from CMB issues since the VIF for all constructs is lower than 3.3, which is the criterion value suggested by Kock (2015b). This outcome confirmed that CMB was not a significant concern in the study. To further establish the reliability and validity of our scales, we utilized SPSS 23.0 for factor analysis (Carmines and Zeller, 1979). The calculated Cronbach’s alpha for all items is 0.893, and the Kaiser-Meyer-Olkin (KMO) measure for all items is 0.878. These values indicate the reliability and validity of the scale employed in the research (Eisinga et al., 2013).

This study applied the Partial Least Squares – Structural Equation Modelling (PLS-SEM) approach to analyze the data and test the hypotheses. This method is suitable for the exploratory nature of our study and the complexity of the proposed conceptual framework (Cheah et al., 2018; Hair et al., 2019). In addition, PLM-SEM is suitable for studies involving data with a non-normal distribution (Hair et al., 2019), and has been widely used in communication and media studies (Yap et al., 2018; Wang et al., 2019). This study used SmartPLS 3.0 software to conduct path analysis.

To establish the reliability of each research construct, we evaluated [factor loadings] composite reliability (CR) and average variance extracted (AVE). Generally, good convergent validity is typically indicated when the standardized factor loadings of measurement variables on their corresponding latent variables exceed 0.7, and the AVE for each latent variable is above 0.5 (Chiu and Wang, 2008). As presented in Table 1, [the factor loading for each variable surpasses 0.7]. Most CR values exceed 0.7 and most AVE values surpass 0.5, signifying a measurement model with a relative good reliability and validity (Henseler et al., 2015; Hair et al., 2016).

Referring to Table 1, diversity of information is characterized by a range of indicators related to variety of information outlets, such as information resources, opinions, and contents, etc. Need for cognition is closely related to citizens’ media literacy, such as enjoyment of cognitive stimulation, preference for complexity, commitment to cognitive effort, and desire for understanding (Lord and Putrevu, 2006). This underscores the significant contributions of elements such as commitment to cognitive effort. Furthermore, media trust is closely related to national-level mainstream media, local-level mainstream media, and We-media, in which We-media is of the most importance. Moreover, political trust is composed of the political system, government officials, and policies.

Table 2 affirms the discriminant validity analysis of the scale evaluated through AVE (Fornell and Larcker, 1981; Davison, 1983). A construct must share more variance with its indicators than other constructs of the research model, which occurs when the square root of AVE between each pair of constructs is higher than the estimated correlation between those factors, thus ratifying its discriminant validity. The square foot values of all constructs in this study are greater than the maximum value of the estimated inter-factor correlation coefficients, demonstrating the constructs possess good discriminant validity.

Among 370 valid responses, there were 166 male and 204 female participants. Regarding educational background, the distribution of residents was as follows: 145 with an associate degree, 113 with a bachelor’s degree, 70 with a master’s degree, and 42 with a doctoral degree. Regarding political affiliation, there were 103 non-affiliates, 89 members of the Communist Youth League, 154 Party Members (including Party activities and probationary Party members, and 24 other political Party members.

Residents value diversity of information source in DI, preference for complexity in NFC, freedom of information selection in IFP, trust for we-media in MT, institutional trust in PT (Figure 1). Specifically, the highest mean score was for “diversity of information source” in DI, which is consisted of DI1 (Mean = 4.049) and DI2 (Mean = 3.086). However, the other two indicators of DI, that is “diversity of opinion” and “diversity of content,” performed significantly lower on average than “diversity of information sources.” In the NFC construct, residents showed a strong inclination towards engaging with complex information, as evidenced by the highest mean score for “enjoyment of cognitive stimulation” (Mean = 4.334) and a similarly high score for “preference for complexity” (Mean = 4.035). For IFP, high mean scores for items such as “freedom of information selection” (Mean = 4.318) and “freedom of speech expression” (Mean = 4.226) indicate strong perceptions of the importance of information freedom. However, “freedom of speech protection” had a relatively lower mean score (Mean = 3.555), reflecting some concerns about state interference. Regarding MT, the mean scores were relatively low, with “trust for we-media” being the highest (Mean = 2.922), indicating general skepticism towards media reliability among residents. PT scores were generally high, with the highest mean scores for “institutional trust,” which consisted of PT1(Mean = 4.429) and PT2(Mean = 4.202), suggesting a high level of political trust among the residents.

Our PLS-SEM analysis demonstrated that DI significantly influences PT directly and indirectly through the mediating roles of IFP and MT. According to Table 3, DI exerts a positive effect on PT (β = 0.121, p < 0.05), indicating that exposure to more diverse information sources tends to strengthen citizens’ institutional confidence within the context of the proposed structural model. Furthermore, DI significantly enhances IFP (β = 0.351, p < 0.001) and NFC (β = 0.343, p < 0.001), suggesting that increased digital information diversity enhances perceptions of information freedom and the need for cognition. In turn, NFC shows a significant positive effect on IFP (β = 0.396, p < 0.001), MT (β = 0.139, p < 0.05), and PT (β = 0.292, p < 0.001), indicating that individuals with a higher need for cognition are more likely to perceive information as free, trust media, and exhibit higher political trust. However, although the statistical directionality aligns with the hypothesized causal chain, the cross-sectional design prevents us from fully verifying temporal precedence. Therefore, these results should be interpreted as model-based causal inferences rather than fully confirmed real-world causality. Additionally, IFP significantly influences MT (β = 0.180, p < 0.01), and MT significantly influences PT (β = 0.167, p < 0.01), highlighting the mediating roles of information freedom perception and media trust in the relationship between digital information and political trust. These results support hypotheses 1 to 8 proposed by our research framework.

Our PLS-SEM analysis further revealed the complex mediating pathways from DI to MT and PT, emphasizing the significant roles of IFP and MT in fostering PT. Other than phrased path analysis, we also investigated the interrelationship among individual’s DI, NC, IFP, MT, and PT. Through precise statistical analysis, we uncovered a series of complex mediating pathways illustrating the multi-step relationships from DI to the construction of trust in political and media systems. According to Table 4, the influence of DI on IFP through NFC (β = 0.136, p < 0.001) is notably strong, underscoring the role of NFC in enhancing perceptions of IFP. Furthermore, the indirect effect of DI on MT through IFP (β = 0.063, p < 0.01) indicates that DI contributes to MT via information freedom perceptions. The pathway from NFC to MT through IFP (β = 0.071, p < 0.05) highlights the importance of cognitive needs in shaping media trust via information freedom. Additionally, the combined indirect effects of DI on MT through both NFC and IFP (β = 0.024, p < 0.05) and the subsequent path from DI through NFC to PT (β = 0.048, p < 0.05) illustrate the multi-step relationships from diverse information to political trust. The mediating effects of information freedom and media trust are further supported by the significant pathways from DI through IFP and MT to PT (β = 0.01, p < 0.05) and from NFC through IFP and MT to PT (β = 0.012, p < 0.05). These findings emphasize the critical roles of both information freedom perceptions and media trust in fostering political trust. While these mediated effects provide important theoretical insight, they represent statistical mediation rather than definitive causal mechanisms and alternative pathways. Future longitudinal or experimental research will be necessary to confirm whether the same directional pattern persists when temporal ordering can be observed.

Through cluster analysis, we identified three distinct demographic groups characterized by their varying levels of media usage and political trust, which are “Informed and Trusting,” “Isolated but Trusting” and “Partially Informed and Skeptical” as shown in Table 5. Analysis of variance across the cluster categories revealed significant differences in media usage and political trust, with F-values ranging from 7.74 to 209.88, all with p-values less than 0.01. This indicates statistically significant variations among the groups on these dimensions.

First, “Informed and Trusting” (cluster 1) constitutes 45.68% of responses, which is characterized by having extensive educational or familial connections in Hubei, demonstrating high frequencies of media usage and levels of political trust. Notably, the mean ± standard deviation for indicators of media usage for digital public forums and the political trust series highlighted this group’s high mean and low variability in political trust, distinctly contrasting with the other two clusters. Notably, this cluster demonstrates a confluence of high media consumption and high political trust, challenging the conventional assumption that diversified information exposure necessarily heightens skepticism or critical detachment. Instead, it suggests that in contexts where media and governance narratives align, informational abundance may reinforce rather than erode institutional legitimacy.

Secondly, “Isolated but Trusting” (cluster 2) accounts for 32.43% of our responses, which have fewer connections in Hubei and lower media usage frequencies, yet exhibit high levels of political trust. Despite lower media engagement compared to “Informed and Trusting,” the mean values of political trust indicators such as China’s political system suitability and capabilities of officials, remain high, though they are lower than those observed in “Informed and Trusting.” This pattern reveals a form of affective or habitual trust sustained even under informational isolation, indicating that institutional trust in hybrid systems may not require sustained exposure to diverse information, but may instead rely on normative expectations or performance-based legitimacy.

Third, “Partially Informed and Skeptical” (cluster 3) makes up 21.89% of the responses, which has minimal connections in Hubei and moderate media usage, coupled with lower levels of political trust. This finding is supported by the mean ± standard deviation for media usage for social media and civil servant professionals, where the PT mean values are significantly lower than those of the other 2 clusters, indicating a lower level of trust in the political system. This configuration illustrates an emergent pattern of reflexive trust formation, wherein moderate media use, combined with disconnection from state-validated narratives, appears more conducive to evaluative skepticism than to either passive confidence or media-led trust.

Our multigroup path analysis shows significant cross-cluster differences in how DI, NFC, and MT affect IFP and PT. As shown in Table 6, the effect of DI on IFP is notably stronger among the “Partially Informed and Skeptical” group compared with the “Isolated but Trusting” group (Δ = −0.359, p = 0.005), suggesting that individuals in this cluster perceive greater informational autonomy when exposed to diverse sources. In contrast, the impact of NFC on IFP is significantly higher in the “Isolated but Trusting” cluster than in both “Informed and Trusting” and “Partially Informed and Skeptical” (Δ = 0.397 and −0.332, p < 0.01), indicating that those with limited media exposure but high cognitive motivation relies more on perceived freedom to evaluate information. The path from MT to PT shows only a marginal difference between “Informed and Trusting” and “Isolated but Trusting” (Δ = −0.449, p = 0.083), implying that media trust does not translate into political trust equally across groups, likely due to differing expectations of media credibility.

Taken together, the three clusters reveal distinct logics of trust formation that complicate linear exposure-based models of political trust. “Informed and Trusting” aligns with a mediated performance-legitimacy pattern, while “Isolated but Trusting” illustrates an affective or habitual trust that persists despite limited exposure. Meanwhile, “Partially Informed and Skeptical” embodies a cognitive reflexivity more commonly associated with fragmented or algorithmically filtered information environments.

This study examines a hybrid media environment characterized by both digital diversification and institutional information governance. By incorporating Diversity of Information (DI), Need for Cognition (NFC), Information Freedom Perception (IFP), Media Trust (MT), and Political Trust (PT) into a single framework, the study enriches existing models of trust formation that have traditionally emphasized either institutional performance or media effects alone. The integration of NFC as a mediating construct contributes to an expanded understanding of how individuals cognitively process diverse information sources and form political evaluations, extending the application of the Elaboration Likelihood Model (ELM) in a post-crisis context.

This study also advances research on the multidimensional nature of political trust by demonstrating that trust formation is not solely a direct output of media use but emerges through layered perceptual and cognitive pathways. The findings suggest that perceived information freedom serves as a pivotal intermediary between media structure and institutional trust, revealing that users’trust is shaped not only by the content they encounter but also by their subjective sense of autonomy within the information environment (van Dijck et al., 2018; Diamond and Renfrew, 1997). This provides empirical support for recent work that emphasizes the importance of perceived control and transparency in media-mediated trust processes.

The identification of three distinct clusters, “Informed and Trusting,” “Isolated but Trusting,” and “Partially Informed and Skeptical,” illustrates that different trust mechanisms coexist within the same media system. Rather than assuming a uniform relationship between media exposure and trust, our findings indicate that political trust may be sustained through alternative pathways, including performance-based legitimacy, credibility, and cognitive engagement in diverse information environments (Cacioppo and Petty, 1982; Petty and Cacioppo, 1984). This contributes to a more nuanced view of media and trust research, suggesting that the effects of information diversity operate conditionally rather than universally. A healthy cycle of information flow and public discourse co-evolves with media freedom, highlighting the importance of diverse and independent media channels in contemporary society (Sui, 2018). Post-pandemic policies often address immediate issues but have long-term implications. The formation of public opinion and political trust is both situational and latent (Lempert, 2003; Yoshikawa et al., 2020), meaning that while specific events can shape immediate reactions, the broader social environment continuously molds political cognition. Our research emphasizes the need to consider the enduring and latent effects of policies in studies of social issues.

Taken together, this study broadens the scope of political trust theory by showing that informational diversity, cognitive motivation, and perceived communicative autonomy function interdependently in shaping trust, particularly in post-pandemic contexts where information plays an outsized role in public legitimacy (Zhang and Sha, 2024; Hazelbarth, 1997). By situating these mechanisms within a hybrid media ecology, the study calls for a more context-sensitive framework for understanding political trust, one capable of accounting for both structural constraints and agency-based cognitive processes.

Rather than viewing political trust as a direct consequence of media consumption, the findings of this study reveal a more intricate psychological and communicative pathway from information environments to institutional evaluations. The validated model shows that Diversity of Information exerts both direct and indirect effects on political trust, with much of its influence transmitted through cognitive activation (NFC), perceived autonomy (IFP), and evaluative credibility (MT). This layered mechanism suggests that political trust is not only a function of what information is accessed, but how individuals make sense of their informational agency within mediated environments.

The first and second hypotheses are validated, which showed that diversity of information has significant and positive effects on political trust and information freedom perception. This coincides with findings of Himelboim and Limor (Himelboim and Limor, 2008) and Ananny (Ananny, 2014), and therefore further confirms the research hypothesis. When individuals are exposed to a diverse range of information sources and viewpoints, it can lead to a more nuanced understanding of political issues (Stroud, 2008). It is well known that Chinese citizens have been exposed to political information via a variety of channels based on different periods, which have all been undergone the control of central government (Lu, 2013). After the whole society entered into 21st century, the internet and mobile phones further expanded the channels available for political information, changing the political media landscape. Moreover, the changing media and new information technologies have also empowered the Chinese people with even more autonomy in choosing different information channels (Zhang and Fleming, 2005; Lu, 2013).

The third and fourth hypotheses are validated as well, which showed that information freedom perception has a positive impact on media trust and media trust has a positive effect on political trust, forming a chain mechanism. Although media trust and political trust have been studied, respectively, (Jones, 2004), here we used media trust as a mediator between information freedom perception and political trust to see whether it can be a “bridge” to affect political trust. Scholars have proved that media trust can significantly positively affect political trust in a general context, including political selection (Lu et al., 2019), political persuasion (Brosius et al., 2019), climate issues (Duan and Miller, 2021), etc. In our study, we further proved that information freedom perception has a positive influence on media trust in a relatively strict media environment in China, and media trust can further positively influence political trust, which has stepped further to connect the constructs to a broader context rather than just trust.

By connecting DI, NFC, IFP, MT, and PT within a single model, the study also extends the theoretical utility of the Elaboration Likelihood Model. The results suggest that in crisis and post-crisis settings, political trust may emerge from a dual-route process in which both central and peripheral cues operate simultaneously. The role of NFC further highlights the importance of individual-level variance in processing public information, indicating that cognitive engagement enhances the effects of informational diversity on perceived autonomy and, ultimately, institutional trust. Hypotheses 5 to 8 have been supported by our data analysis, which has resonated with previous studies (Xiao, 2013; Maksl et al., 2015; Lu et al., 2019; Vraga and Tully, 2021). The diversity of information has positive effects on people’s need for cognition, and people’s need for cognition has a positive impact on information freedom perception, media trust, and political trust. Some scholars might think of a need for cognition as part of media literacy (Craft et al., 2017; Maksl et al., 2015). According to the persuasion theory proposed by Hovland (Hovland et al., 1953), the effects of information diversity might be complex, and there will be single-cue and dual-cue messages that differently encourage individuals to form attitudes based on their cognitive engagement. Furthermore, the self-cleansing nature of the internet suggests that misinformation often disintegrates as contradicting facts emerge (Van Dijk, 1995), allowing for cognitive renewal through the alternating exposure to positive and negative information. This process challenges any prevailing narrative dominance, a phenomenon often referred to as the “reversal of news,” where an influx of diverse information disrupts a monopolized information field (Schuck and de Vreese, 2012). The need for cognition plays a crucial role here; without it, individuals might not engage deeply enough to activate the central route of information processing. Ultimately, the singularity of information sources fails to initiate comprehensive cognitive engagement, underscoring the importance of individual agency in the interaction with diverse information environments.

These findings underscore the need to reconceptualize the relationship between media systems and political trust beyond classical assumptions of democratization or liberalization. As post-pandemic societies transition from crisis communication to long-term legitimacy rebuilding, the informational preconditions of trust, such as perceived transparency, cognitive empowerment, and mediated credibility, take on renewed significance. In this way, the present study contributes to broader efforts to theorize how political trust is constructed, sustained, and reconfigured within changing communicative environments.

This study demonstrates that political trust in post-pandemic China is shaped by both the direct influence of information diversity and the indirect effects transmitted through information freedom perception and media trust. Because the three clusters identified in the analysis exhibit different combinations of DI, IFP, MT, and PT, policy interventions should not adopt a uniform communication strategy but instead strengthen trust through differentiated pathways. The results indicate that managing how information is perceived and processed is as important as managing what information is supplied.

Enhancing political trust through diverse and credible media is critical for achieving SDGs, particularly SDG 3 (Good Health and Well-being) and SDG 16 (Peace, Justice, and Strong Institutions). During COVID-19, the dissemination of accurate information significantly impacted public health responses and compliance with health guidelines (Han et al., 2023). Governments and media organizations should collaborate to ensure that information is accurate and accessible to all. Policy measures such as source traceability labels, fact-checking disclosures, and multi-actor information briefings can convert diversity into perceived freedom, thereby reinforcing media trust and, indirectly, political trust. This mechanism aligns with SDG 16 by improving institutional legitimacy not only during crises but in routine governance. By fostering a media environment where diverse viewpoints are shared and misinformation is actively countered, political trust can be strengthened, leading to better health outcomes and more robust public support for health initiatives (Tsfati and Ariely, 2014). This approach not only supports immediate health-related goals but also builds a foundation for long-term trust in public institutions, which is essential for sustainable development.

Promoting media literacy and encouraging cognitive engagement among citizens are essential strategies for improving political trust and achieving SDG 17 (Partnerships for the Goals). The significant effect of NFC on IFP, MT, and PT suggests that cognitive activation operates as a trust-building catalyst. Thus, media-literacy policy should not be limited to teaching skills for identifying misinformation, but should be reframed as a mechanism for enabling central information processing and reducing distrust. Media literacy programs can equip individuals with the skills to critically evaluate information, discern credible sources, and understand the mechanisms of media influence (Maksl et al., 2015). These programs should be integrated into educational curricula and community outreach initiatives to foster a well-informed and discerning public. Moreover, encouraging cognitive engagement through diverse information environments can enhance individuals’ need for cognition, which has been shown to positively influence perceptions of information freedom and trust in media and political institutions (Cacioppo and Petty, 1982; Schreurs and Vandenbosch, 2021). Collaborative efforts among governments, educational institutions, and media organizations can create a more informed citizenry that is better equipped to participate in democratic processes and support sustainable development initiatives.

Finally, the post-pandemic findings indicate that political trust is not a short-term response to crisis messaging, but a cumulative attitude shaped by information experience, perceived autonomy, and institutional performance over time. As the immediate urgency of COVID-19 recedes, communication policy must shift from crisis-based trust mobilization to long-term trust maintenance. Sustained transparency practices, participatory channels for feedback, and continued visibility of evidence-based decision-making will prevent trust decay and institutional volatility in future disruptions. This long-term view reframes trust not as a by-product of compliance, but as a governance asset that requires continuous communicative investment.