• Author: Grace, Hua Shu-Hua (2020)
  
• Keywords: social interaction ； political engagement ； opinion networks ； social media ； civil communication
  
• Publication: Review of Social Science, 2020. 14(2), 1-43. DOI: 10.30401/RSS.202012_14(2).0001

Research question and purpose

literature gap

Research methods

Analysis methods

Research results and findings

Research question

1. Why do Taiwanese people actively participate in politics?
2. Can social interaction situations with divergent opinions promote citizens’ active participation in politics?
3. What is the role of social media?

Citizenship vs. Political Participation

Social structure, political cognitive attitudes and cultural values are the main constructs to explain political participation. （Norris, 2002）

Negative vs. positive citizens

Micro-citizenship view

• Political communication: The main point of current research is that conflicts or disagreements have different views on the impact of participating in politics.
• Different opinions: the counter-mobilization effect (GroßEr & Schram, 2006; Leighley, 1995; McClurg & Lazer, 2014; Mutz, 2002a, 2006; Ulbig & Funk, 1999),
• Heterogeneous opinions have different political participation results due to personal experience (Beck, Dalton, Greene, & Huckfeldt, 2002) 。

• Technology intermediary: after the rise of social media, new social structures or networks emerged (Durkheim, 1893; Pescosolido & Rubin, 2000; Latour, 1991)
• Political mediation(McLuhan, 1964; Van Loon, 2007 ; Lister & Dovey, 2008）
• Networking of social media and linking political participation （Castells, 2007, 2012; Holt, Shehata, Strömbäck, Ljungberg, & Nord, 2012; Van Dijck & Poell, 2013）*  

H1. The more positive the awareness of civic norms, the more inclined to engagement.

H2. The greater the scope of daily contact, the more inclined to engagement.

H3. The opinion network is heterogeneous and more likely to engagement

H4. Social media is intermediary

sample

2014 Taiwan Social Change Survey, citizenship (1) and (2) data, Specific questions for Taiwan (political communication) ISSP, International Social Survey Program Common questionnaire Executed by the Academia Sinica, A total of 1875 interviews were successfully completed.

Poisson regression model

# The dependent variable obeys the Poisson distribution, 
# which is a count model of positive integers.
# Usually refers to the number of occurrences, 
# from 0 to positive integer infinity, 
# such as the frequency of bills being rejected, 
# the number of articles published or the frequency of coups, 
# also known as "discrete variables" (Long, Ch8, 1997),
# It is mainly due to the internal process and 
# structure of data generation, 
# so the analysis method of connecting variables 
# is not suitable.

1. The petition (signature) is signed;
2. Because of political, ethical (moral) or environmental reasons, refuse to buy or buy certain products in particular;
3. Participate in demonstrations;
4. Participate in political gatherings or campaign activities;
5. Find politicians or public officials to express your views;
6. Donate money to a certain social or political event, or help them raise funds;
7. Express your views through the media;
8. Express your political thoughts through the Internet.

Frequency distribution of social media users

# There are 1,021 users and 854 non-users of social media,
# with a total observation value of 1,875. It does not mean 
# that different groups do not come into contact with each 
# other. Instead, use social media interaction, field, or 
# technology. This article continues by analyzing and 
# comparing the relationship between media use to 
# politics or life through a cross-tabulation, and examining 
# whether there are differences in the use of new media.

# "Daily Contact" Questions:
# How many people do you interact with from morning to night
#  in a typical day? (Including nodding, greeting, speaking,
#  making phone calls, writing letters, and through computer
#  networks, whether the other party knows or does not know it is
#  counted.)
# (1) 0-4 people (2) 5-9 people (3) 10-19 people
# (4) 20-49 people (5) 50-99 people (6) more than 100 people
### Classification
# (1) High degree of contact,
# (2) Medium degree of contact,
# (3) Low degree of contact

The dependent variable “active political participation” is a positive integer accumulation, so the Poisson regression model is used for analysis and the total marginal effect (AME) is used as the report.

“Social media usage” is divided into two groups and modeled separately.

“Daily Contact” and “Opinion Network” are divided into four different types of social interaction scenarios,

(Models 1 to 3) are for non-users of social media,

(Models 4 to 6) are social media users

People have the need to connect with others in society. As the social circle expands, they are more likely to come into contact with strange or heterogeneous others. However, occasional or inadvertent daily contact may have opportunities to change stereotypes