The effect of television media on beginner voters’ political participation

The political participation of beginner voters in the 2019 Indonesian General Elections in Belu Regency, East Nusa Tenggara Province (NTT) differentiate the current research from other studies. The television media were assumed to influence the political participation of beginner voters in the border region. Correlation test (r) was used to uncover the findings of this study. This study revealed a correlation between exposure to television mass media and political participation with 0.623 value, or strong correlation level category. Television media exposure also had a significant influence on political participation in the border district. Therefore, television broadcasting institutions which manage public frequencies play an important role in increasing the political participation of beginner voters at the border district by broadcasting political and election-related contents or programs.


INTRODUCTION
Mass media plays a vital role in conveying information about candidates, political parties, and the program of candidates and political parties themselves in general elections. Television media effectively deliver political statements of candidates to influence voters. Mass media reported two of Prabowo's controversial statements during the 2019 elections were "Indonesia will disappear in 2030' (Tempo.co, 2018) and "Indonesia will be extinct" (Tehusijarana, 2018). Jokowi's statement was no less controversial. He said, "be careful, there are 'sontoloyo' (foolish) politicians" and "there is 'genderuwo' politics" (Tempo.co, 2018). Citizen's political participation means citizens using their voting rights in the elections and in determining political policies, including influencing government policies. Jimenez organised citizen participation into nine dimensions, namely: 1) political trust; 2) participation in (political) associations; 3) participation in (social) associations; 4) campaign activity; 5) political knowledge; 6) political efficacy; 7) social trust; 8) political discussion; and 9) political understanding (Francisco & Jiménez, 2017). Dan Nimmo offered similar political participation concept: 1) Identification by political parties; 2) Registration to vote; 3) Voting in the general elections; 4) Taking part in the campaign (Nimmo, 2010). However, this study used Kennorton Hutasoit's concept of political participation which combines Jimenez's and Nimmo's political participation concepts (Hutasoit et al., 2020).
The first study of the media influence on politics was The People's Choice. In Southeast Asia, such as in the Philippines and Singapore, the media effectively influences the behaviour of audiences, including voters, in elections. In Manila, Philippines, mediation reinforces the implied but persistent racial hierarchy of another culture to which many local Filipinos are subscribed (Cabanes, 2014). Still in the Philippines, Rodrigo Duterte, who used the media to spread the issue of populism in a contemporary fashion, was influential in gaining votes and winning elections (Abinales, 2013). Jejomar Binay, who at that time served as the highest performing vice president with a tailcoat from Aquino III (Curato, 2017), lost to Duterte. In Singapore, the ruling People's Action Party (PAP) pledged to expand the political space and democratise it, but after winning the election brilliantly, it did not fulfil its promise (Abdullah, 2019). However, in cases involving religion where the state does not have a strong tendency (Abdullah, 2020), the PAP finally accommodates religious populism.
In the research titled "Frequency of Television Viewing And Association With Overweight And Obesity Among Women Of The Reproductive Age Group In Myanmar: Results From A Nationwide Cross-Sectional Survey," the categories of television viewing frequency was divided into the following three categories: (1) not watching television at all; (2) watching television less than once a week; (3) watching television briefly (Das Gupta et al., 2019). The frequency of watching television in one week is categorised into 1) very often (8 times); 2) Often (6-7 times); 3) Not often (4-5 times); 4) very rarely (2-3 times). In addition, the duration of watching television is categorised as follows 1) very long time (46-60 minutes); 2) average time (31-45 minutes); 3) Not long (16-30 minutes); 4) very short time (1-15 minutes) (Prabawati, 2013).
In the research of TV Viewership Quintiles, television viewers are distinguished in the Light (0 -1.6 hours/day), Light-Medium (1.6 -3.0), Medium (3.0 -4.6), Heavy Medium (4.6 -7.3), and Heavy (7.3 more) (Nielsen, 2012). Based on the National Survey by the Centre for Strategic and International Studies (CSIS) titled "Social, Economic and Political Orientation," it was revealed that millennials in Indonesia chose television as the primary source of information. As many as 79.3% watch television broadcasts every day, 8.5% 1-2 days every week, 5.8% 3-4 days every week, and 3.0% 5-6 days every week. Only 3.3% of claimed never to watch television (CSIS, 2017).
Based on the description above, this research question is: how much influence does television media exposure have on the political participation of beginner voters in Belu Regency?

METHODOLOGY
This study used a quantitative approach. The researcher intended to examine the effect of television media exposure on beginner voters' political participation in Belu Regency. Belu Regency, East Nusa Tenggara Province is a border area between Indonesia and Timor Leste. Researchers surveyed several villages that spread throughout the district in Belu Regency which directly bordered Timor Leste. The type of research in this study was explanative research. Research that uses survey methods in collecting data and testing the hypothesis of the influence of independent variables on the dependent variable is called explanatory research. The primary objective of explanatory research is to explain why events occur and to build, elaborate, extend, or test theory (Neuman, 2014).
In this study, 338 questionnaires were distributed to the respondents. Respondents in this study were beginner voters or teen voters. Beginner voters are first-time voters who use their voting rights in elections. In Article 198 paragraph (1) of Law Number 7 of 2017 concerning General Elections regulates that Indonesian citizens who are voting on the polling day are even 17 years old or older, have been married, or have ever been married have the right to vote. Someone aged 17 years included in the category of late teens with an age limit of 15 years to 19 years. This teenager has a more substantial interest in terms of career, spouse, and identity exploration. Teenagers, as an audience, are an age group that spends much of their time using media (Santrock, 2007).
Before using the questionnaire, the authors tested the validity and reliability of the questionnaire on 30 undergraduate students of Universitas Mercu Buana Jakarta on 1 April 2019. For television media exposure variables, there were 21 questions. For the political participation variable, there are seven questions. Based on the calculation of SPSS Statistics 21.0, the value of the product-moment correlation coefficient for all questions is more significant than 0.3, which means that all questionnaire questions are valid. The reliability test was done with Cronbach's Alpha with the results for the media exposure variable 0.95 and the political participation variable 0.677. Both are greater than 0.6, which means that the questionnaire is reliable. The questionnaires were distributed to beginner voters aged 15 to 16 years old who were married, and beginner voters aged 17 to 19 years old, who voted on the 2019 Election on 17 April 2019. The researcher used the questionnaire as a data collection tool when interviewing respondents. Researchers select respondents randomly. The selected respondents are beginner voters who have finished using their voting rights at the polling station. The polling station is located in the administrative area from Atambua City to Tasifeto Timor District, Belu Regency. Of all 388 distributed questionnaires, 282 of them were filled in completely and returned to the researchers. After collecting data, the authors processed the data using SPSS Statistics 21.0 to obtain the correlation and significance values needed to test the effect of independent variables on the dependent variable.

RESULTS AND DISCUSSION Results
Based on the 2017 population registration report, the Belu Regency had 213,596 population consisting of 106,782 men and 106,814 women, see Figure 1. The percentage of married beginner voters aged less than or equal to 16 years was 10.34%. According to (Pohan, 2017: 425), Indonesia is one of the countries with the highest percentage of child marriage in the world, ranked 37, or the secondhighest in the Association of Southeast Asian Nations (ASEAN) regions after Cambodia. The Education Level of the population in Belu District varied. As much as 32.69% of the population had not had or did not have a diploma which was fairly dominant. The second-highest ranking was people who had only graduated from elementary school, which was 30.44%. Based on data of BPS 2018, the percentage of Belu residents who could not read was 10.04%. In addition, in 2017, the number of poor in the Belu Regency is 33,950 people or around 15.92% of the poverty line of IDR 319,280 or SGD 30/capita/month.
This study took note of the characteristics of Belu Regency as an interstate border region. One of its characteristics was the number of Foreign and Domestic Passengers Visiting in 2017 by using the Crossborder Post (PLB) in Belu District, which consisted of 2,191 men and 2,872 women. Based on the voter participation Table 1, the level of voter participation in Belu District experienced fluctuations in the 2009 Presidential Election for about 70.98%. In the 2014 Presidential Election, it decreased to 59.45% while in 2019 Presidential Election increased to 71.89%. In this study, 338 questionnaires were distributed to the respondents. The questionnaires were distributed to beginner voters aged 15 to 16 years old who were married and beginner voters aged 17 to 19 years old. The questionnaires were distributed to beginner voters voted on the election day of the 2019 Election on 17 April 2019. The questionnaires distributed to beginner voters were randomly given to them directly after they had given their voting rights at polling stations/areas (TPS) that spread along the interstate roads in the administrative area from Atambua City to Tasifeto Timor Subdistricts. Of all 388 distributed questionnaires, 282 of them were filled in completely and returned to the researchers. There were 132 male respondents (46.80%) and 150 female respondents (53.20%). The percentage of beginner voters aged 15 years to 16 years who were married was 9.93%. The percentage of child marriage (15 and 16 years old) was close to Belu Regency records in 2017, which was 10.34%. The education level of respondents also varies. As much as 4.25% of the respondents had no education. Respondents who graduated from primary school was 6.38%, 20.21% graduated from secondary school, and 54.25% graduated from Post-secondary school. There were 14.89% of respondents who held a diploma degree.
Based on the results of the questionnaire, this researcher found that 269 or 95.39% of the total 282 respondents watched television media and 194 or 68.79% accessed new media. In contrast, 88 or 31.20% of respondents did not access new media. There were nine respondents or 3.19% who did not watch television and new media. Based on Table 2, the majority of respondents, 150 respondents or 53.28% of the total 282 respondents, watched television media eight times or more every week. The second highest frequency of television exposure was 6-7 times every week, with 61 respondents or 21.71% of the total of 282 respondents. In comparison, the number of respondents who had never watched television was minimal, only 2 respondents or 0.79% out of 282 respondents.  Table 3, the majority of respondents, 120 or 42.55% out of 282 respondents, watched television for 46 -60 minutes per watch. The second highest duration was 31 -45 minutes per watch with 72 respondents or 25.62% out of 282 respondents. Meanwhile, the number of respondents who had never watched television was minimal, only 2 respondents or 0.79% out of 282 respondents. Based on Table 4, the majority of respondents, which were 239 respondents or 84.75% out of 282 respondents, participated in discussions about politics, elections, presidential-vice presidential candidates, political parties, and legislative candidates. As many as 230 respondents or 81.56% out of 282 respondents knew and had a close relation or attachment to certain political parties. As many as 251 respondents or 89.00% out of 282 respondents checked names in the temporary voter list (DPS) or final voter list (DPT), either coming to the village in person or through online via KPU website. A total of 102 respondents or 36.17% out of 282 respondents attended the general meeting campaign or joined the campaign meeting, helped the success teams of presidential candidates, political parties, or specific candidates either directly by inviting others, or by installing campaign props or by using social media.

Test of Normal Distribution and R Test or Correlation Test
Before conducting a simple regression analysis, the Normality Test was first performed. The results of the normality test with variables X and Y are summarised as follows: Based on the testing rules, if the probability (Sig.) was> 0.05, Ho was accepted. It means the data is usually distributed. If the probability (Sig.) was <0.05, then Ho was rejected. It means the data is not normally distributed.
A correlation test was done based on the correlations table from the following data processing results: From the correlation test between television media exposure variables (X) and political participation (Y), a value of r = 0.623 was obtained. This value showed a strong positive relationship between exposure to television media (X) and voting participation (Y). The strong positive intention signifies a strong direct relationship between variables (X) and (Y). This meant that when the television mass media (X) rose, the level of voting participation (Y) also increased sharply.
A correlation value of X, which was 0.623, could be interpreted by the type of television media (X1), content (X2), frequency (X3), and duration (X4) of media exposure. These points could be seen in the table that researchers had calculated by using SPSS 25. Based on Table 8 above, it can be seen that the number ranged from X1 of 0.107, X2 of 0.550, X3 of 0.516, and X4 of 0.597. This shows that the type or television media station being watched had a relatively weak correlation with political participation. Meanwhile, the contents of television media, frequency and duration of television exposure had sufficient or moderate correlation.

Significance Test
The significance test showed how far the influence of independent variables is in explaining the variation of dependent variables. The results came from the comparison between tcount and ttable with the test rules. If tcount> ttable, Ho is rejected. Based on the table above, tcount for the exposure to television media (X) = 11.67 was higher than ttable with the formula: Ttable = t (α / 2) (n-2) = t (0.05 / 2) (282-2) = 7. The result: t1 count (11.67)> ttable (7), which means Ho rejected. The result showed that there was a significant influence between the exposure of television media (X1) and the level of political participation (Y).
The significance of the result can also be tested by using probability techniques with the rules of testing such as if Sig ≤ α, then Ho is rejected; if Sig> α, then Ho is accepted. From the Coefficient table (a), the value obtained for Sig was 0.00. For the α value, because it was applied to a two-tailed test, the α value was divided by 2, so the α value became: α = 0.05 / 2 = 0.025. Because of Sig. = 0.00 <α = 0.025, then Ho was rejected. It means that there was a significant influence between the exposure of television media (X) and the level of political participation (Y).
Based on the description above, the linear regression model related to the effect of exposure to television media (X) on political participation (Y) is described as follows:

Discussion
This study found that 214 or 75.88% of respondents frequently and very often received news and information on Presidential and Vice-Presidential Candidates and Presidential and Vice-Presidential Programs via television media. As many as 213 or 75.53% of respondents often and very often get news and information about political parties and their programs through television media. As many as 206 or 73.04% of respondents often and very often get news and information about Legislative Candidates and their programs via television. As many as 200 or 70.92% of respondents watch television media containing news and information on Presidential and Vice-Presidential Candidates and Presidential and Vice-Presidential Programs with a duration of 30 minutes to 60 minutes. As many as 166 or 58.86% of respondents watch television media containing news and information about political parties and their programs for 30 minutes to 60 minutes. As many as 162% or 57.44% of respondents watch news and information about Legislative Candidates and its programs for 30 to 60 minutes. In determining the choice in the 2019 election, 166 or 58.86% of respondents agreed and strongly agreed that information from television or new media (social media) helped them make a choice. Based on the partial correlation test, the correlation value of the variables of contents of television media or X2 with voting participation is 0.550, frequency of television media exposure or X3 with voting participation is 0.516, and duration of television media exposure or X4 with voting participation is 0.597. All three had excellent or moderate correlation categories. However, The type or television media station or X1 is 0.107, relatively weak correlation with voting participation. The findings above show that respondents who often watch television that contains the presidentialvice presidential candidates, political parties, and candidates for members of Parliament have a moderate correlation with their voting participation. Likewise, respondents who watched television media which contained presidential and vice-presidential candidates, political parties, candidates for members of Parliament with a duration of 30 to 60 minutes, had a moderate correlation with their participatory voting.
Based on the simultaneous correlation test between television media exposure variables (X) and political participation (Y), a value of r = 0.623, that showed a strong positive relationship between exposure to television media (X) and voting participation (Y). The strong positive intention signifies a strong direct relationship between variables (X) and (Y). This meant that when the television mass media (X) rose, the level of voting participation (Y) also increased sharply.
The findings of the simultaneous correlation test explained that respondents who often watched television media containing news and information about presidential candidates, political parties, and candidates for Parliament with 30 minutes to 60 minutes duration of viewing had a strong correlation with respondents' voting participation. The previous research that television media has a moderate effect on political participation (Sorensen, 2019), political advertising variables on television and voting variables have a moderate correlation (Darmawati, 2015), and consumption of political talk shows has a direct correlation with political participation (Zaheer, 2016).
Based on the test of significance of influence, t1 count (11.67)> t table (7), that there was a significant influence of television media exposure (X1) on the level of political participation (Y). This means the television media that deliver news or information about the presidential and vice-presidential candidates, political parties, and parliamentary candidates who are accompanied by their programs have a significant positive effect on voting participation. The theoretical implication of this study is that television media has a significant effect on beginner voters in Belu Regency. Nevertheless, the fact that the respondents in this study were young people who had just voted in the polls at the election is worth to note. Therefore, this study could not find how many teenagers who watched television but did not vote in the election. This means that this research cannot explain how many young people watch television media that contains politics but are not influenced to participate in politics or at least vote in elections.

CONCLUSION
Based on the description above, this study concludes the exposure of television media which contains presidential and vice presidential candidates, political parties, and parliamentary candidates with frequent to very frequent frequencies and the duration of 30 minutes to 60 minutes significantly influence the voting participation of beginner voters in Belu Regency.
Thus, to influence the participation of beginner voters' elections in Belu Regency, presidential candidates, political parties, and parliamentary candidates can create interesting campaign content through television media. Content must also be frequently displayed on television media because television media is still the key source of information about politics and election in Belu Regency. This statement is evident from the findings that 269 or 95.54% out of 282 respondents watched television to get information about politics and elections. Besides television, new media is also a source of information about politics and election for respondents. In this study, 194 or 68.79% of 282 respondents used new media as a source of political and election information. Therefore, even though television media is still the primary source of information, in the future, there will be a shift in media sources of information along with the development of internet technology.