The Influence of Political Awareness, Political Socialization, and Mass Media on Political Participation in Jambi Province

The purpose of this study was to determine the effect of political awareness, political socialization and mass media on the political participation of the people of Jambi Province in the regional head election. Political participation is a form of community involvement in exercising their political rights in general elections both at the national and regional levels. The embodiment of political rights will affect every decision-making related to the interests of the community. This study uses a quantitative approach that uses statistical techniques with PLS model analysis techniques. The study was conducted on the people of Jambi City as many as 100 people were selected based on area sampling techniques. The research findings explain that the mass media has no direct effect on political participation. Political participation is directly influenced by the variables of Political Awareness and Political Socialization. Therefore, it is suggested that the use of mass media in the post-conflict local election in Jambi Province needs to be improved, especially information related to the political rights of the community. The tendency that occurs in the community is more interested in obtaining information related to entertainment because people are still accustomed to listening culture rather than reading culture.


INTRODUCTION
Elections are a means of realizing democracy in Indonesia, both carried out at the central and regional levels in the election of people's representatives (legislative) and state or regional leaders (executive) (Antari, 2018). Through the general election accommodated the things that are desired and aspired to by a community. In the context of Indonesia, a period of general elections has taken place since independence in Indonesia until the current reform era. Through elections the process of maturation of democratic values runs gradually in the history of the Indonesian nation. The election according to the ideals of democracy emerged in the election of the people's representatives through 2004 and proceeded with a peaceful and successful atmosphere (Supriyanto et al., 2006).
Public participation in this election is an indicator of the implementation of the democratic process mandated by the Indonesian constitution. Political participation as an activity of a person or group to be active in political life (Boedihardjo and Meriam, 2010). Community participation in post-conflict local elections is voluntary and carried out consciously in realizing the political rights of citizens.
Regional head elections are part of general elections in Indonesia. Elections are held as an effort to transfer people's sovereignty to certain candidates to occupy political positions of power which are carried out by following the rules established by law (Junaidi, 2009). The purpose of the regional head election is to elect leaders who are at the regional level from the district / city to the provincial level (Prasetyoningsih, 2017).
Based on observations of several regions in Jambi province, it was found that many people were less active in their participation in the regional head elections. Based on the 2015 General Election of Regional Heads of Jambi Province as shown in Table 1, the participation of the Jambi Province community only reached 67.8 percent, far from the expected target of 77.5 percent (KPU, 2019).

Source: Indonesian General Election Commission (KPU, 2019)
Election of regional heads is a means to realize public participation in politics to realize democratic governance (Simamora, 2011). Thus, the low participation of the community in regional head elections is a benchmark for the level of democracy in an area (Antari, 2018). A study conducted by Alelaimat (2019) revealed that several factors can influence political participation, namely gender, age, family income and regional affiliation. Research conducted by Halim et al. (2018) identified several factors that influence people's political participation, namely social and economic conditions of the community, geographical conditions, parental affiliation, organizational experience, level of political knowledge and awareness, level of trust in the government and mass media socialization and formal discussions. People do not want to participate in politics due to technical factors related to the physical condition of voters, work, administration and do not trust political parties and candidates to be elected (Mustadin, 2013).
The results of research conducted by Casale and Resnick (2011), explain that the age factor in Africa is a strong predictor of voter behavior. The participation of young voters is significantly lower than older Africans. Erdogan and Kamiloglu (2014) in their research through a survey on Facebook also showed the same results, that less than 10 percent of Turkish youth were active in political parties and civil society activities. The reason is due to elements of culture. Activity on Facebook is largely determined by demographic factors and respondents' perceptions of politics.
The results of a study conducted by Kunz et al. Stated that reading a newspaper has the highest effect compared to other traditional news sources and the process of online political information communication has the strongest impact on internal efficacy. Where internal efficacy is known as a significant predictor in determining community political participation (Kunz et al., 2014). The findings of Ariadi et al. (2014) in their research states that political participation among youth will increase if the level of Facebook use increases, both in terms of duration and intensity as well as the number of users' friends.
This study seeks to examine the influence of political awareness, political socialization and mass media on the political participation of the people of Jambi Province in the regional head election. Increasing people's political participation is an interesting topic in realizing democratic governance in order to sustain sustainable development.

RESEARCH METHOD
This study uses quantitative methods to determine the effect of political awareness variables (X1) political socialization (X2) and mass media (X3) on the political participation (Y) of the Jambi Province community in the 2015 local elections (Pilkada) from data obtained through surveys of the selected object.
The object of research was selected using area sampling and purposive sampling techniques (Sugiono, 2013). The area side technique is used to determine the sampling areas of 11 districts and cities in Jambi Province. While the purposive sampling technique is used to determine the sampling area criteria from the district, sub-district and village / sub-district levels. To support the objectives of this study, the criteria for the object of the study were selected based on regions with very low political participation. The research object areas based on these criteria are Tanjung Timur Regency, Tanjung Jabung Barat Regency and Jambi City.
The number of samples is determined using the Solving formula (Nursalam, 2013), as follows: Where n is the number of samples, N is the population number and e is the error margin, which is 8 percent.
Data were analyzed using a variance based Structural Equation Modeling (SEM) equation model. PLS can be used as an alternative SEM model based on covariance which is more predictive of the resulting model. The PLS model is not based on assumptions such as normality of data and large samples based on variants (components). PLS is an alternative approach that shifts from covariance-based SEM to variants (Chin, 1998).
Covariance-based SEM tests causality / theory while PLS is more predictive of the model. The PLS model is not based on assumptions such as data normality and large samples. Latent variables in the PLS formula are aggregate linear from the indicators. Linear model specifications produce an error rate model that is formed from variables and dependent variables. The PLS model will produce estimates consisting of (a) estimated weight (producing latent variance values), (b) path estimation (determining the relationship between latent variables and indicators, (c) the average value and location of the parameters of each indicator and its latent variables.
Data analysis was performed using WarpPLS software version 4.0. The causal model contained in PLS has components in the form of measurement and structural models. The resulting structural model has two types of latent construct variables that are not observed, namely first and second order. From the measurement model there are observed indicators. The path coefficient estimation results identify the strength of the relationship between the independent variable and the dependent variable. The PLS model produced was assessed based on criteria (Ghozali, 2009).

RESULTS
Community participation in local elections is crucial to the realization of a democratic and quality government. Community participation in the Jambi regional head election in 2015 only reached 67.8% (KPU, 2019). This research was conducted to analyze the effect of exogenous variables on the level of political participation of the people of Jambi Province.

Description of research data
This research was conducted in areas where the level of community participation was low in the election of regional heads in Jambi Province, namely East Tanjung Jabung Regency, West Tanjung Jabung Regency and Jambi City. Of the 100 respondents surveyed obtained information about the level of political participation of the community and the factors that influence it, as explained in table 2. Source: Primary data processed, Jambi City 2019 The average score of public political participation in the study area in Jambi Province was 41.66. Based on a range of minimum and maximum scores, the average level represents a high level of political participation. However, the data division standard is quite large, so that the distribution of data does not only focus on the categories indicated on the average value but also focuses on the lower categories of political participation.
The average political awareness score also shows the same thing, where most of the people in the study area in Jambi Province have high political awareness but most of the others are still classified as low. The same thing is also illustrated by the average score of the variable political socialization and Mass media.

Evaluation of Outer Model or Measurement Model
To produce the best model as a predictor, an evaluation of the resulting model will be carried out in two stages, namely the constructs that form the first-order indicators, and the second-order which is formed by the first order. It is hoped that the two stages of evaluation will produce a high level of reliability and validity.

Evaluation of First Order Constructions
First order construct evaluation is done to find out the level of validity and reliability of the un dimensional structure indicator reflective. The model to be tested is a null model that has been designed previously, namely by evaluating the convergent validity and discriminant validity of each constructor indicator, as well as the value of alpha cronbach and compodite reliability of the indicator block (Ghozali, and Latan, 2012).

Convergent Validity
Convergent validity test results are used to determine the validity of the relationship of each indicator with its construct or latent variable. If the loading factor value is greater than 0.70, then the relationship is of high validity (Ghozali, and Latan, 2012).
The following is the value of the loading factor after indicators that have low validity are excluded from the model: In table 3, it appears that there is an increase in the loading value of each indicator after eliminating the indicator that has a low level of validity.

Discriminant Validity
Discriminant analysis is used to assess the validity of the indicators in the resulting model. According to Ghozali and Latan (2012), the level of indicator validity of a model is good if the correlation of items with the construct is greater than the correlation of the constructs. The results of the model discriminant validity analysis are shown in Table 4 below: Source: Primary data processed, Jambi City 2019 Irzal Anderson, Kuswanto, The Influence of Political Awareness, Political Socialization, and Mass Media on Political Participation in Jambi Province, JSSP, Vol.5, No. 2, December 31, 2021 The results of the discriminant validity analysis as shown in Table 4, show that the resulting model is in a good category, where all indicators with their constructs have a greater correlation than the existing constructs. This condition is also reinforced by the resulting average variance extracted (AVE) value greater than 0.5, as explained by Ghozali and Latan (2012). In more detail, the AVE value is shown in Table 5 below: Source: Primary data processed, Jambi City 2019

Reliability
The reliability level of the construct or latent variable is measured based on the value of composite reliability and Cronbach alpha. If the Cronbach alphanumeric value is greater than 0.7, then it is declared reliable (Ghozali, and Latan, 2012). The level of construct reliability that results from PLS data transfer is shown in Table 6 below: Source: Primary data processed, Jambi City 2019 Table 6 shows that all variables have a high level of reliability because the composite reliability and Cronbach alpha values are greater than 0.7.

Colinearity construct
A good model requires that there is no multicollinearity in formative constructs, namely by calculating the value of the variance inflation factor (VIF). Cenfetelli and Bassellier (2009);Petter et al., (2007) recommend a VIF threshold value of 3.3. If the VIF value is greater than 3.3, then the construct indicator must be eliminated from the model. By using the PLS program, the output value of VIF is obtained as shown in Table 7. Source: Primary data processed, Jambi City 2019 In table 7, it can be seen that all indicators of the construct do not experience multicollinearity where the VIF value does not exceed the value of 3.3 so that it can be used in the model.

Evaluation of Structural Models
There are several measures that can be used to evaluate the structural model, including the value of the coefficient of determination (R2), the value of the path coefficient, the Cohen effect size (f2), and predictive relevance (Q2).

Coefficient of Determination
The coefficient of determination shows the level of ability of the model as a predictor of the influence of exogenous variables on endogenous variables. The model is considered weak as a predictor if the value of R2 is less or equal to 0.19 (Chin, 1998).

Source: Primary data processed, Jambi City 2019
Based on Table 8, it shows that the R2 value for the Political Awareness variable is 0.401. The coefficient of determination explains that the ability of the model in predicting the influence of the variable Utilization of Mass Media and Political Socialization on political awareness is 40 percent. The value of R2 for the variable Political Participation is 0.417, explaining that the influence caused by the variable Utilization of Mass Media, Political Socialization and Political Awareness of Political Participation by 42 percent. Thus it can be concluded that the two models are considered moderate as a production tool.

Path coefficient (β)
The path coefficient value is considered significant if it is between -0.1 to 0.1 and significant if it is greater than 0.1 (Hass and Lehner, 2009). Based on the results of PLS output, as seen in Table 11 there is an insignificant path, namely the Mass Media Utilization Variable in stages Political participation because the coefficient value is less than 0.1, and is supported by its statistical t value (0.87) smaller than t table in alpha 5 percent (1.99), and a P value (0.38) greater than alpha (0.05).  Anderson, Kuswanto, The Influence of Political Awareness, Political Socialization, and Mass Media on Political Participation in Jambi Province, JSSP, Vol.5, No. 2, December 31, 2021

Source: Primary data processed, Jambi City 2019
More concisely, the results of the analysis of the estimated PLS model can be illustrated in the following path diagram:

Quality Index
Each part of the model requires validation of the measurement model, structural model, and the whole model which can be measured by the Goodness of Fit (GoF) index value (Ghozali, and Latan, 2012). GoF value in WarpPLS is more than equal to (> =) 0,1, more than equal to (> =) 0,25 more than equal to (> =) 0,36 indicates that the overall quality value of small, medium, and big. In this study, the GoF value generated from WarpPLS output was 0.665, thus overall the model in this study was considered fit because it had a GoF value> 0.36.
To determine the level of quality of the model measured using the index Goodness of Fit (GoF). If the value of GoF is equal to (> =) 0.1, indicates the quality of the small model. If more than equal to (> =) 0.25, then the quality of the middle model, and if more than equal to (> =) 0.36, indicates the overall quality of the large model (Ghozali and Latan, 2012). Based on the PLS output from this study, a GoF value of 0.67 was obtained. Thus the overall model has a great quality.

Hypothesis Testing
Hypothesis testing in this study uses the amount of the path coefficient and the P value and t count. A hypothesis is significant if the amount of the path coefficient which shows the level of significance if the value is greater than 0.1, the P value is smaller than 5 percent alpha and the t count is greater than the t table (Hass and Lehner, 2009). The results of the analysis using the PLS program obtained the P value and path coefficients as shown in Table 10 below:

Source: Primary data processed, Jambi City 2019
Hypothesis Testing 1 Hypothesis 1 states that political socialization has a direct effect on political participation. This hypothesis is significantly accepted at 1 percent alpha with the resulting path coefficient value of 0.278 greater than 0.1.

Hypothesis Testing 2
Hypothesis 2 states that the use of mass media has a direct effect on political awareness. This hypothesis is significantly accepted at alpha 1 percent with the resulting path coefficient value of 0.27 greater than 0.1.

Hypothesis Testing 3
Hypothesis 3 states that the use of mass media has a direct effect on political participation. This hypothesis is not accepted and is not significant at 5 percent alpha with the resulting path coefficient value of 0.07 less than 0.1.

Hypothesis Testing 4
Hypothesis 4 states that political socialization has a direct effect on political awareness. This hypothesis is significantly accepted at 1 percent alpha with the resulting path coefficient value of 0.430 greater than 0.1.

Hypothesis Testing 5
Hypothesis 5 states that political socialization has a direct effect on political participation. This hypothesis is significantly accepted at alpha 1 percent with the resulting path coefficient value of 0.382 greater than 0.1.

Hypothesis Testing 6
Hypothesis 6 states that the use of mass media has an indirect effect on political participation. This hypothesis is not accepted and is not significant at 5 percent alpha with the resulting path coefficient value of 0.075 which is smaller than 0.1.

Hypothesis Testing 7
Hypothesis 7 states that political socialization has an indirect effect on political participation. This hypothesis is significantly accepted at alpha 1 percent with the resulting path coefficient value of 0.119 greater than 0.1.

Regression Equations
The value of the path coefficient indicates how strong the influence of a variable on other variables. The higher the value of the path coefficient, the stronger the effect. Path coefficient values indicate the level of influence of exogenous variables on endogenous variables. The greater the coefficient value, the more influence exogenous variables have on endogenous variables. Based on the PLS output results obtained path coefficient values as shown in Table 12.

DISCUSSION
From the results of statistical analysis using the PLS model, it is known that each of the variables studied has a direct or indirect effect. The regional head election held in Jambi province has been followed by the Jambi community with a high level of participation. The high level of community participation is influenced by political awareness of the importance of exercising political rights in electing regional leaders. This is indicated by indicators related to political awareness that are definitions, goals, and political activities, applying political knowledge in daily life, enthusiastic about political developments in society, and following political activities in the communities in which they live their lives. The more aware of a person or society about politics, the higher the level of political participation will be. And vice versa, the lower the level of political awareness, the lower the level of political participation (Boedihardjo and Meriam, 2010). Political awareness is indeed very closely related to political participation in particular to elect leaders (regional heads) (Fatwa, 2016;Lee, 2017). Because the community is more aware that they are governed (by the leader), the community will increasingly demand that the right to be given the right to voice in order to influence policy making by the leader.
In a country that adheres to a democratic system this is a positive sign that reflects the community's concern about the political problems that occur in their country (Hanani et al., 2019). Conversely low participation is a bad sign, because it shows that the community is no longer paying attention to the problems of state and nationality.
In line with that, political awareness certainly does not just happen, but there is another fact, namely political socialization. The intense implementation of political socialization carried out by the state or political institutions certainly has an influence on people's participation in politics. As an agent of political socialization, family. The experience of participating in family decision making provides the abilities for the child to do political interaction, and makes it more likely to actively participate in the political system when he is an adult (Sahid, 2011).
The use of media is very important in the process of campaigning and modern political socialization (Mustofa, 2013). According to him, to win the competition in the electoral arena, the contestants from political parties compete by implementing sound political communication strategies. The mass media is not only an integral part of politics, but also has a central position in politics, able to arrange the reality of various events that occur to become a discourse that has the power to campaign for politics (Junaidi et al., 2018).
The important function of the mass media is not only in terms of education, sources of information, but also as entertainment. Trends that occur, people are more intense in choosing entertainment than news relating to politics (Makhshun & Khalilurrahman, 2018). So that the function of the mass media as a news disseminator to the public is not well realized. This is consistent with research findings that the mass media indirectly influences the political participation of the community in the regional head elections in Jambi province.

CONCLUSION
Based on the results of research and discussion that has been concluded, there are several indicators of each invalid variable used in the model, which is enthusiastic about political developments in society, following political activities in the community where they live, providing education to the community, political socialization in the family. Political socialization at work, direct political contact, political activities that are followed, participation in political activities, as apolitical.
All exogenous variables have a significant effect both directly and indirectly on political participation, except for the variable use of mass media. Thus it is suggested to the public to improve the system of using mass media as a means of information in the political field so that it will increase its participation in realizing democratic governance.