The role of social media in the political involvement of millennials

Purpose – This study aims to examine the relationship of information quality of social media, social media reputation,socialmedia politicalmarketing activities,trust andpolitical involvement of millennials. Methodology – The empirical analysis was conducted using a sample of 309 millennials. This study used online survey for the data collection. After passing reliability and validity tests, the data were analyzed with partial least squaresstructuralequationmodeling. Findings – The results show that information quality of social media has positive and signi ﬁ cant direct in ﬂ uence on reputation and trust. Information quality of social media also has a signi ﬁ cant indirect in ﬂ uence on trust through social media reputation. However, there is no signi ﬁ cant relationship between information quality and political involvement. Social media political marketing activities also have a direct and indirect signi ﬁ cant effect on political involvement through trust. Finally, trust also has a positive and signi ﬁ cant impact onpoliticalinvolvement. Practical implications – This research may contribute to the political marketing experts and politicians in increasing the quality and credibility of advertisements on social media, which will affecttrust andpolitical involvement of millennial generation. Moreover, politicians and political marketing experts who have an online-based community should optimize their marketing activities in social media to encourage positive behaviorandtrust from social mediausers. Value – This study has shown a more comprehensive model of the relationship between information quality of social media and political involvement. This study also reveals the signi ﬁ cant indirect effect of the trust on the relationship between information quality on social media, social media political marketing activities and political involvement.

El papel de las redes sociales en la implicaci on política de los millennials Resumen Prop osito -Este estudio examina la relaci on de la calidad de la informaci on, su reputaci on y las actividades de marketing político desarrolladas en las redes sociales, la confianza y la participaci on política de los millennials.
Conclusiones -Los resultados muestran que la calidad de la informaci on de las redes sociales tiene una influencia directa positiva y significativa en la reputaci on y la confianza. La calidad de la informaci on de las redes sociales también tiene una influencia indirecta significativa en la confianza a través de la reputaci on de las redes sociales. Sin embargo, no existe una relaci on significativa entre la calidad de la informaci on y la participaci on política. Las actividades de marketing político de las redes sociales también tienen un efecto significativo directo e indirecto en la participaci on política a través de la confianza. Por último, la confianza también tiene un impacto positivo y significativo en la participaci on política.
Implicaciones pr acticas -Esta investigaci on puede contribuir a que los expertos en marketing político y los políticos aumenten la calidad y la credibilidad de los anuncios en los medios sociales, lo que afectar a a la confianza y a la implicaci on política de la generaci on millennial. Adem as, los políticos y los expertos en marketing político que tienen una comunidad en línea deberían optimizar sus actividades de marketing en los medios sociales para fomentar un comportamiento positivo y la confianza de los usuarios de los medios sociales.
Originalidad -Este estudio muestra un modelo m as completo de la relaci on entre la calidad de la informaci on de los medios sociales y la implicaci on política. También revela el significativo efecto indirecto de la confianza en la relaci on entre la calidad de la informaci on en los medios sociales, las actividades de marketing político en los medios sociales y la implicaci on política.

Introduction
Internet and social media hold a significant impact on the social life and political participation of citizens (Koiranen et al., 2020). According to Safiullah et al. (2017), digital communication in the form of social media provides a space for politicians (candidates, government officials and members of political parties) to keep existing, where they can build and influence the public opinions. Social media has a positive impact on the success of campaign and on the political information dissemination (Newman, 2012).
Information quality presented on social media offers a collective knowledge development (Zhang et al., 2019). Social media reputation is also influenced by the presentation of information quality. In a marketing context, consumers also face high levels of information asymmetry and information overload of the contents, considering the vast volume of data in social media (Tang et al., 2012). Moreover, trust on social media information will affect the political involvement of its users. Trust can be used to assess information flow and credibility issues that often emerge in social media. According to Tang and Liu (2015), trust might improve the information quality and the credibility of social media by providing appropriate and accountable information.
Furthermore, Lock and Harris (1996) in their study found that the role of marketing strategy in the political marketing context is still limited. Moreover, attention toward social media as the information of political campaign has not been optimized in influencing political involvement (Kushin and Yamamoto, 2010). Several previous studies concluded that marketing strategy through social media has an important role in political campaign (Tudoroiu, 2014). Social media such as Facebook, Twitter and Instagram provide a space for the candidates to campaign their political information and also for voters to interact with the candidates (Williams, 2017). However, previous studies have not yet explained the relationship between trust and political involvement (S anchez-Villar et al., 2017). Therefore, to provide more information on the study area, this study aims to examine the relationship of social media, political marketing activities, trust and political involvement.
This study contributes in some areas. First, it provides a more comprehensive model of the relationship among the information quality of social media, reputation of social media, social media political marketing activities, trust and political involvement. Second, this study has analyzed the direct effect of social media marketing activities on the political involvement that has not been conducted in the previous studies yet (Gil de Zúñiga et al., 2014). Social interaction activities carried out using social media do not have a direct influence on community political involvement. Third, the previous study has only used political blog users as the population while this study expands it to social media users who potentially have different characteristics, because nowadays people have turned to the use of social media instead of using blogs or other conventional media in obtaining political information. Social media is believed to be able to build communication between candidates and voters and provide significant benefits and influence political decisions (Kahne and Bowyer, 2018;Al-Hussein, 2020). This study found that the information quality of social media has a significant indirect impact on trust that has not been conducted in the previous studies yet (S anchez-Villar et al., 2017). Finally, this study provides new practical insight in terms of allowing political marketing experts and politicians to optimize the quality and credibility of political advertisements in the social media, which will affect the millennial generation involvement in politics (Abdennadher et al., 2019).
2. Theoretical framework and empirical model 2.1 Theory of reasoned action Social media users are frequently exposed to various contents and insights which are shared and created by people from all society levels (Goyanes et al., 2021). In some cases, people will follow the opinion of a leader who publishes original and unique contents through the social media and which will potentially influence further attitudes (Casal o et al., 2020). Creativity is an important aspect of social media (e.g.; Instagram) that influences interaction intentions (Casal o et al., 2021). This condition is closely related to the Theory of Reasoned Action (TRA). TRA shows that people consider the consequences of the alternative behavior before getting involved in certain action (Fisbein and Ajzen, 1975). As the nature of social media is interrelated, it allows each individual to be exposed to the behaviors of others; those exposure are expected to create normative perception regarding certain individual behaviors in social media (Kim et al., 2015). According to Indahingwati et al. (2019), digital technology has changed the paradigm and perception of millennials in behaving. Hence, in this study, TRA is considered to be the most relevant theory to help explain the effectiveness of using social media as a source of valid and reliable political information contents.

Millennials
The millennial characteristics and uniqueness might become the opportunities and challenges for various aspects, such as economy, politics, education and health. Millennial, also known as generation Y, is a demographic group that comes after generation X. Generation Y who were born between 1980 and 2000 are deemed to be the most diverse and can accept diversity easily (Young et al., 2013). Most of generation Y understands the various methods of communications, ranging from face-to-face interactions to using electronic media, such as sending emails and using social media easily (Young et al., 2013). As digital media users, millennials have high expectations and level of mobility toward the flow of digital information. They like to search for certain information and latest news by reading digital newspapers on the internet (Flavi an and Gurrea, 2006).

Trust
Citizens have a high level of political trust when they are satisfied with the political situation (Lu et al., 2019). According to Casal o et al. (2017), satisfaction has an effect on consumer interaction intentions which in turn has a positive effect on actual behavior. Political communication pattern through the use of communication and information technology in a democratic party event has been able to present new media for political marketers. The existence of internet is a power capable of developing critical citizenship, promoting democratization and increasing the quality of democracy (You and Wang, 2019). Millennials consider social media as a global news channel and platform for information exchange (Helal et al., 2018). This group believes that information presented in social media is reliable and credible. According to Dabula (2017), political marketing activities through social media influence the trust and loyalty of young South African voters. If internet users trust the data given to them by their friends or colleagues online, they will be more likely to get involved in online political activities, such as sending political email, signing electronic petitions, reading political blogs and joining political groups on Facebook (Himelboim et al., 2012). In democracy, online behavior could result in political capital and trust (You and Wang, 2019).

Social media political marketing activities
Reading information in a digital version will be more preferable if the information provides a brief overview of the day's news and the latest updates of a news that has not yet been published in the printed version, or if the reader only wants one or two certain pieces of information (Flavi an and Gurrea, 2007). Social media is an online-based application program that provides convenience in interacting and sharing information contents. Social media marketing activities are part of online marketing activities created to complete the Webbased traditional promotion strategy, such as online advertisement campaign and email (Barefoot and Szabo, 2010). Social media marketing activities are not only applied to the marketing of products and services but can also be applied to the marketing of political parties. In social media political marketing activities, there are things that need to be considered (e.g. online interactivity, a clear social media and trends). The use of social media in political marketing, such as Facebook, Youtube and Twitter, has developed quite fast in the past 10 years (Okan et al., 2014). At the moment, social media has become an identity as well as a trend among the millennials. This group is perceived as digital generation who is the main target of political campaign and is widely discussed in political discourses (Irawanto, 2019). Their habit of using social media such as Facebook, Youtube, Whatsapp, Telegram and Twitter will create opportunities for political marketers to promote their political candidate through social media.

Empirical model, hypothesis development and proposed model
The quality of information reflects the user's perception of website content that is relevant, adequate, accurate and up-to-date (Zhou, 2012). According to Bolton et al. (2013), to obtain information, millennials choose to use social media. They are actively involved in social media sites to contribute, post, search and consume information (Fulton and Kibby, 2017). According to Johnson and Kaye (1998), millennials consider that the information obtained from online sources to be credible. This study refers to the quality of information as the millennial generations' perception of the quality of information presented on social media. The quality of information and communication using different social media (e.g. Facebook, Twitter, YouTube) will affect the credibility and reputation of social media perceived by its users (Keshavarz, 2021). Based on the arguments above, this study proposes the following hypothesis: H1. Information quality of social media has a positive and significant impact on the reputation of social media.
The quality of information has several criteria, including accurate, timely, relevant, economic and easy to understand and obtain. Information quality is a variable in an online interaction that has an impact on trust (Filieri et al., 2015). According to Kong et al. (2020), information quality is an important element in fostering trust for millennials. Trust in the information which is presented in social media is determined by the level of reputation of the social media. Social media has become a daily source of information on politics and latest events for millennials (Evans, 2019). They believe that social media is a source of information that has great credibility and reputation for obtaining political information. The information sources presented on a social media will be trusted once they contain accurate, relevant, sustainable and consistent information. According to Mardjo (2019), the information quality of social media has a positive impact on trust. A good quality information (e.g. quality of content, up to date and variety of information) will foster the confidence of the millennial generation in using social media to obtain political information. Thus, this study proposes the following hypothesis: H2. Information quality of social media has a significant impact on the trust.
Social media reputation might lead to consumer's or user's trust in the honesty and popularity of a social media (Zha et al., 2018). In this study, reputation refers to what extent the millennial generation as social media users perceive the credibility of social media as the source of political information. Reputation of social media that is perceived by its users will have an impact on trust (S anchez-Villar et al., 2017). Social media is an interactive tool of information communication technology that provides new possibilities and qualities in communication (Kijek et al., 2020). Millennials have high expectation and mobility on information needs. Social media is a choice for millennials to obtain various information and they will perceive social media as an information source with high credibility. High intensity of millennial generation in using social media may result in the emergence of the trust on social media. The longer they spend their time in using social media, the higher possibility they have to see and interact through the social media (Duffett and Wakeham, 2016). Currently, social media has a good reputation and has become a trend in presenting various news (Cacciatore et al., 2018) and has a positive impact on public trust as a medium for sharing their beliefs and opinions about politics (Shami and Ashfaq, 2018;Mardjo, 2019). Hence, this study proposes the following hypothesis: H3. Reputation of social media has a positive and significant impact on trust.
Information relevancy for each person will be different based on the usefulness of the information to the recipient. Political information contents which are obtained by individuals through social media will affect their active behaviors toward political campaign. According to Mohamad et al. (2018), information quality on social media, such as Facebook, has a significant impact on political participation. Mansoor (2021) has highlighted the significant role of quality of information in social media on government trust. McKnight et al. (2017) also assert that information quality is a significant antecedent of customer trust in business-to-business business. Moreover, Donati et al. (2020) have found that employees' trust leads to their organization involvement. Sharma and Klein (2020) also argue that trust is a significant influential factor of customer involvement. Therefore, trust on the quality of certain information which is presented on social media will have an impact on the user intensity in using social media. This means that trust is perceived as an important part in establishing the intention or involvement of users in political activities. Information quality is one of the variables in online interactions, having the most relevant impact in increasing the voters' trust (Filieri et al., 2015). In addition, Evans (2019) claims that social media has often been used among millennials as a media to obtain information and increase their insights on political knowledge. According to Abdullah et al. (2021), the quality of political information presented on social media platforms has a positive impact on youth political participation. Thus, the following hypothesis is proposed: H4. Information quality of social media has a positive and significant impact on political involvement.
Social media has been used by political marketers to introduce their political candidates to the public. According to Bode (2016), social media, as a forum where people share their thought, may become a medium for transmitting their political information. Moreover, in the political marketing context, social media is not only a medium to promote a candidate and political party to prospective voters; but it might also become a medium to maintain the relationship between the candidate and voters and a media to build public trust. According to Helal et al. (2018), recently, the millennial generation depend on social media in obtaining the trending global issues. They believe that social media is a tool that can be used to discover and exchange political information. According to Kim and Ko (2010), social media marketing activities have a significant effect on trust. Marketing activities which are conducted by experts in political marketing/advertising and politicians can optimize the quality and credibility of political advertisement through social media online. Thus, the following hypothesis is proposed: H5. Social media political marketing activities have a positive and significant impact on trust.
Social media has been adopted in various contexts, such as social, financial, business and political life as well as educational sector (Khan and Siddiqui, 2013). According to Khan et al. (2019), currently social media is widely regarded as a promising platform for carrying out promotion activities. Information presented in social media turns out to be able to trigger the intention of its users to make decisions related to the information, either in the form of product and service advertisement or political advertisement. The previous study conducted by Wells and Dudash (2007) has shown that recognition of social media has a positive relationship with political involvement situation because social media offers new insight for its users to obtain political information. Besides, political marketing on social media is an alternative for political candidates to build their relation with voters and to strengthen the public trust. They presume that social media would be able to influence people to participate in political activities. Moreover, White and Anderson (2014) stated that political communication continues to develop along with the advancement of technology; accordingly, political involvement through social networking sites has become a factor that cannot be denied in urging millennial generation political participation. In addition, according to Ciftci (2021), marketing activities in the form of political advertising through social media accounts contributed positively to voter involvement in election campaigns. Therefore, this study proposes the following hypothesis: H6. Social media political marketing activities have a positive and significant impact on political involvement.
Political involvement is reflected in the intention of people to get involved in a political activity. This involvement depends on information that they have obtained from several reliable sources. From a political perspective, social media has a significant impact on political marketing sector (Dabula, 2017). Political marketing activities through social media are considered capable of promoting political candidates. According to Dabula (2017), nowadays, more political leaders, political parties and politicians are aware of using social media to inform, communicate and engage with the voters. According to Ton and Kim (2016), the survey on 169 millennial generation respondents shows that the use of social media has certainly affected the millennial voters' decision in the United States presidential elections in 2016. Therefore, millennials believe that reliable information sources from credible social media platforms are able to increase their political knowledge. Social media provides its users with the ability to develop online connections or communities where people with the same interests can interact, get involved and share opinions and knowledge (Steenkamp and Hyde-Clarke, 2014). Currently, social media has been trusted as a credible platform in increasing political participation (Knoll et al., 2020). Therefore, the voters' trust in the information sources will lead to their involvement in politics. Thus, the following hypothesis is presented: H7. Trust has a positive and significant impact on political involvement.
Based on the description of the research hypothesis, the research model and proposed hypotheses can be illustrated accordingly (Figure 1).

Population and sample
The population of this research was all the millennial generation in Indonesia. As the overall population of this study was unknown, this study applied non-probability sampling for the data collection. The non-probability sampling is taken if the number of respondents is very large and not countable (Latan et al., 2020). The respondents were identified using a snowball sampling technique by the help of social media.

Measurement items and scales
The core section of quantitative research that frequently has an impact on the results of the research is the measurement item and scale. An adequate measurement item has to be able to capture the construct concept being measured. This research adopted the same measurement item as the previous empirical studies. Using the measurement item that has generally been deemed practical is better than developing a new measurement item because of the complexity of developing the scales (Latan et al., 2020). Information quality, reputation of social media, social media political marketing activities, trust and political involvement were measured by using the Likert's scale of five points ranging from 1 (strongly disagree) to 5 (strongly agree).

Data collection procedures
This study used several steps in the data collection procedure. First, in the questionnaire development, this study applied back translation procedure from English to Bahasa Indonesia and back to English. This aims to ensure the clarity of the questionnaire contents (Sekaran and Bougie, 2016). Second, after receiving the Table 2. final version of the questionnaire, we conducted a pre-test by sending it to 52 respondents for initial data analysis. This procedure minimized the potential bias that would affect the validity of the results of our studies. This stage included the calculation of possible measurement error in survey methods, such as method bias, response bias and social desirability bias, to improve the quality of the survey (Latan et al., 2020) and also to ensure that the questionnaires were understood by the respondents (Fowler, 2013;Latan et al., 2020). Third, we conducted the main study by distributing the questionnaires through social media and e-mail. Such distribution was followed up with text notification to ensure that the questionnaires sent were received by the respondents. This method is considered to be one of the best methods to reach a wide range of respondents with low cost and within a short time frame (Dillman et al., 2014;Latan et al., 2020). To increase the response rate, at the end of each month during the period of the research, we sent emails to the respondents as the reminders. To maintain the confidentiality of the respondents' personal data, we ensured them that their names and identity would not be disclosed in this research. The data collection was carried out during the period of April 2019 to January 2020.

Result
The hypothesis testing (see Figure 1) was done by using a structural equation modeling (SEM) technique through the partial least squared (PLS-SEM). This study used SmartPLS 3.2.9 as the software package. PLS-SEM was used in this research because it is more suitable for theory development and involves complex models. Moreover, this technique is effective in estimating the causal relationship in theoretical models based on empirical data (Hair et al., 2017). The criteria for evaluating the structural model (outer model) by using SEM-PLS are as follows: reliability testing that can be seen from the value of composite reliability and the Cronbach's alpha; convergent validity that can be seen from the loading factor and the average variance extracted (AVE) value; and discriminant validity that can be seen from the square root value of AVE and the correlation between the latent constructs.
The next step is to evaluate outer model through reliability test that aims to prove the accuracy, consistency and exactness of the instruments in measuring the construct. The reliability test was conducted by looking at the value of the composite reliability and Cronbach's alpha (>0.70) (Hair et al., 2017). Hence, it can be concluded that the data are reliable (Table 3). Convergent validity is related to the principles in which measures (manifest variables) from a certain construct should be highly correlated. Convergent validity testing was done by examining the loading factor value compared with the rule of thumb (>0.6) and then by looking at the AVE value compared with the rule of thumb (>0.50) (Hair et al., 2017). Based on the results of the convergent validity test, the loading factor on each construct has values larger than the rule of thumb (>0.60). The value of AVE for each construct is larger than the rule of thumb (>0.50) ( Table 3).
We adopted two criteria in evaluating the discriminant validity. First, we applied the Fornell and Larcker's (1981) criteria, where the square root from AVE variable must be higher than its correlation with other variables. Second, we evaluated the heterotrait-monotrait (HTMT) ratio from correlation. According to Henseler et al. (2015), HTMT is more sensitive in decreasing discriminant validity compared to other criteria. To show the discriminant validity, the HTMT between two constructs must be less than 0.90. These two criteria support the discriminant validity of all our variables (Table 3).

Structural model
The criteria for structural model assessment (inner model) by using SEM-PLS are as follows: R-square for dependent construct; and bootstrapping procedure (t-values >1.96 and significant level = 5%) to find out at the significant value.
The following is the result of the structural model evaluation (inner model) through the bootstrapping procedure for testing the hypothesis proposed in this research as presented in Table 4. Structural model is evaluated through the R 2 value and Q 2 value for the dependent latent construct. According to Hair et al. (2017), the rule of thumb for R 2 values is as follows: 0.75 for the strong category; 0.50 for the moderate category; and 0.25 for the weak category. The rule of thumb value for Q 2 > 0 indicates that the model has predictive relevance and the rule of thumb for Q 2 < 0 indicates that the model lacks predictive relevance. The R 2 value for each construct is obtained from the analysis results. The value of reputation of social media construct is 0.144. This means that the variability of social media can be explained by the information quality of social media variable in a model as much as 14.4%, which is, therefore, categorized as a weak model. Meanwhile, the Q 2 value for reputation of social media is 0.024 > 0, which means that the model has a predictive relevance. The value of trust construct is 0.283 > 0. This means that the variability of trust can be explained by the information quality of social media, reputation of social media and social media political marketing activities variables with the percentage of 28.3%, which, therefore, falls into the weak category. The Q 2 value of 0.149 is greater than 0, which means that the model has a predictive relevance. In addition, the political involvement construct of 0.211 means that the variability of political involvement can be explained by the information quality of social media, social media political marketing activities and reputation of social media through trust variables and the trust variable in the model as much as 21.1%, which, therefore, falls into the weak category. The Q 2 value of political involvement is 0.103 > 0, which means that the model has predictive relevance.  The evaluation of significant value was conducted by examining the path coefficient value from the testing results with PLS by calculating bootstrapping (Table 4). The results of path coefficient analysis show that (H1) information quality of social media has a positive and significant impact on the reputation of social media (b = 0.209; p < 0.05). Furthermore, (H2) the information quality of social media has a direct impact on trust (b = 0.323; p < 0.05). Moreover, H3 states that reputation of social media has a positive and significant impact on trust (b = 0.144; p < 0.05). Furthermore, the results show that (H4) information quality of social media has a direct positive and insignificant relationship on political involvement (b = 0.009; p > 0.05). Moreover, (H5) social media political marketing activities have a positive and significant impact on trust (b = 0.372; p < 0.05). Furthermore, the results show that (H6) social media political marketing activities have a direct and significant impact on political involvement (b = 0.268; p < 0.05). Furthermore, the results show that (H7) trust construct has a positive and significant effect on political involvement (b = 0.284; p < 0.05).
Surprisingly, this study found that social media information quality also has a significant indirect effect on trust trough social media reputation. Social media information quality also has indirectly and significantly influenced political involvement through trust. Finally, this study found that social media political marketing activities have a significant indirect effect on political involvement through trust.

Discussion
This study analyzes the effect of social media (e.g. information quality of social media and reputation of social media) and social media political marketing activities on political involvement by considering trust. This section contains the theoretical contributions and practical implications of our research results, presents the main limitations and gives recommendations for further research.
In general, we found empirical evidence for millennial generation political involvement model. Specifically, our main contributions are presented as follows. First, we found that social media has a positive effect on trust. This fact is in line with the TRA which shows that people would consider the consequences of alternative behaviors before getting involved in certain action (Fisbein and Ajzen, 1975). Our findings imply that social media is perceived well as a valid, reliable and reputable information source in providing political information. Several previous studies affirm our findings. For example, Johnson and Kaye (1998) stated that the majority of millennials believe that information obtained from online sources is credible. Filieri et al. (2015) stated that information quality has the most relevant impact on fostering trust in online interactions. However, we found evidence that information quality of social media has no impact on political involvement. This finding implies that various information, dynamic contents and latest information presented on social media without a boost of trust would have no impact on political involvement. Therefore, political communication pattern through the use of technology and information quality becomes the important element that must be considered by political marketers. This pattern is a challenge to optimize the strategy as an alternative method in overcoming the traditional political communication pattern (S anchez-Villar et al., 2017), considering the ever-increasing importance of interactive communication tool in Web 2.0 era ). This result is not able to support the previous findings (Mohamad et al., 2018).
Second, we found evidence that social media political marketing activities give a positive effect on trust and political involvement. Our findings imply that political marketing activities through social media are believed to be able to create intense communication between politicians and citizens. Communication through virtual media is deemed to be an effective form of Web-based political involvement (Valenzuela et al., 2012), by providing credible information sources (Johnson and Kaye, 1998). It should be noted that the b value coefficient on the relationship between social media political marketing activities and trust is the highest value among all the tested relationships; and this fact shows that social media political marketing activities construct is proven to be valid in assessing trust and political involvement. Social media users believe that political marketing activities in social media are able to increase their desire to get involved in democratic parties. The result of this finding is able to support the previous results (Kim and Ko, 2010).
Furthermore, the last finding of this research is that we are able to prove that trust on social media is able to give a positive effect on political involvement, which has not been explained well previously (S anchez-Villar et al., 2017). Our findings imply that trust plays a role in filtering and evaluating information flow and credibility issues that are often found in social media (Tang and Liu, 2015). This finding supports the previous findings (Lin and Ching Yuh, 2010), where trust has positive and significant impact on customer purchase intention. Moreover, this finding incorporates the role of trust as an indirect variable. The research results show that trust is not yet proven to be having a good role as indirect variable on the link between information quality of social media and political involvement. However, trust is proven to be valid as indirect variable between social media political marketing activities and political involvement. Surprisingly, we also found three significant indirect effects in our proposed research model. There is an indirect effect on the relationship between the information quality of social media and the reputation of social media. Likewise, the relationship between the information quality of social media on political involvement through trust, as well as the relationship between social media political marketing activities and political involvement through trust.
6. Conclusion 6.1 Theoretical contribution and managerial implication In summary, this study has some contributions, including theoretical development and managerial implication. First, this study found a more comprehensive model of the relationship of information quality of social media, reputation, trust, social media political marketing activities and political involvement of millennial generation in Indonesia. Theoretically, this study has contributed to the knowledge development on political marketing field by addressing a more comprehensive model of the relationship between information quality of social media and political involvement, especially in the millennial generation context. The information quality of social media has been found as a key factor in affecting trust, which is neglected in the previous studies and especially in the millennial voters' context. Second, this study found a significant indirect effect of the relationship between information qualities of social media on political involvement through social media reputation, political marketing activities and trust. These results provide a better understanding of the specific indirect influence relationship between the quality of information (e.g. variety of information, dynamic content, information content and up-to-date information) and encouraging political involvement. Interestingly, this study reveals that social media reputation, political marketing activities and trust have an essential role in building the perception of the millennial generation to be involved in democratic parties.
For the managerial implication, this study revealed that it may become guidance for political marketing practitioners (e.g. politicians and political party leaders) and political marketing consultants in treating the millennial generation and make them involved in the political agenda, such as general election. The decision-makers, e.g. in political party leaders, may create programs through social media to increase the millennial generation trust, which will affect their involvement in political agenda. The politicians should gain the millennial generation trust by using social media as their campaign media if they want to be elected. Therefore, political candidates have to concern with the information quality of social media and reputation of social media platforms, which will affect the millennial voters' trust and their willingness to involve in political agenda. The political candidates are also able to correct themselves, explain potential misunderstandings and discrepancies between promises and expectations that they offer through social media platforms. Moreover, the political candidates have an opportunity to create programs suitable for the millennial voters and share them through social media platforms. Therefore, it may increase the millennial voters' political involvement.
6.2 Limitation and future study Moreover, there are some limitations of our research. First, our research only focused on millennials; this means that the respondents of this research were homogeneous. Next, researchers need to consider the use of larger and more geographically diverse samples to provide more comprehensive picture of the political involvement of millennial generation. Second, this study only addressed millennial generation; however, the voters are not only millennial. Hence, the future research should include generation X or generation Y to get a better generalization. Third, our research only focused on one country; therefore, the result cannot be generalized to other countries. Further researchers are advised to expand the study to other developing countries, such as Southeast Asian countries (e.g. Thailand and the Philippines). Fourth, the R 2 and Q 2 values are still in the weak category. Hence, an opportunity to study other influential factors, such as socio-cultural factor, is open.