國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
29
3.2 Measurement
All measurement of variables are discussed as followed:
Offline political participation was measured by six items adapted from Andolina, Keeter, Zukin, & Jenkins (2003) and Willnat & Aw (2014). Respondents were asked whether during the past 12 months they (1) attended a public hearing or policy seminar; (2) contacted official or political actors via telephone or e-mail; (3) signed a hard copy petition on politics or public affairs; (4) attended a public rally or electoral campaign; (5) participated in any demonstrations, protests, or marches; (6) been involved in public interest groups, political action groups, and party committees.
Voting was not included though, since some of the respondents had not reached the voting age. Answers were recorded as 1 = never, 2= seldom, 3=sometimes, 4=often, and 5 = always so that the higher the number, the more citizens participated in politics.
Responses to each statement were added into a single index (Cronbachα=.859;
M=9.04; SD=3.87；range = 6 to 30).
Facebook use for political information was measured by two items, which were adapted from Gil de Zúñiga, Jung & Valenzuela (2012) and Wu (2013). The scale was the average of respondents' estimation, registered by a 5-point scale, with 1=never, 5=always: (1) I get informed about political issues and current events from Facebook Friends' posts of news, articles, or the videos.; (2) I search for information about political issues and current events from Facebook Friends, Fan pages, or clubs.
Answers were recorded as 1 = never, 2= seldom, 3=sometimes, 4=often, and 5 = always. Responses to each statement were added into a single index (r=.737; M=6.85;
SD=1.75; range = 2 to 10).
Facebook use for political expression was operationalized as an individual expressing an opinion on political issues or public affairs via Facebook, when the content relates to their political views, interests or participation. The concept was
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
30
measured by three items adapted from Kushin & Yamamoto (2010). The scale asked respondents to indicate their agreement level with the following statements: Facebook use helped me (1) to post personal experiences or thoughts related to politics or public affair; (2) to post or share photos or videos about politics; (3) to discuss politics with others by leaving the comments under the posts or sending the private messages.
Answers were recorded as 1 = never, 2= seldom, 3=sometimes, 4=often, and 5 = always. Responses to each statement were added into a single index (Cronbachα
=.861; M=9.21；SD=3.58; range = 3 to 15).
Connecting with public political actors, adapted from Tang &Lee (2013) ,was measured by asking the respondents whether they connected with public political actors through adding them as Facebook friends, following them, or joining their fan pages such as (1) legislatives; (2) social movement activists; (3) media commentators;
(4) government officials. On each item, the respondents could indicate none, one to three, four to six, seven to nine, ten or above. And then, the respondents were asked what proportion of them were taken as opinion leaders. The answers were summed up to form a 4–20 index (Cronbach α=.757; M=6.07; SD=0.98).
In addition, the researcher defined a number of variables that may affect Taiwanese youth’ political participation as control variables. The control variables comprise of demographic variables (gender, age, educational level, and income), daily hours spent on Facebook, media use (TV and newspaper use) and political antecedent variables (internal/ external/ collective efficacy and party identification), as these are variables that the literature has found to be related to offline political participation (i.e.
Verbal et al., 1995). Thus, these variables were introduced as controls to isolate potential confounding effects and present the most stringent model possible.
Political efficacy refers to the belief that one can make a difference on political matters. Some researchers have differentiated the concept into the internal, external,
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
31
and collective dimensions (i.e. Lee, 2006).
Internal efficacy refers to whether one believes in one’s own ability to understand and participate in politics. Respondents were asked to indicate their agreement level with the following statement: “(1) Although politics and the government seem so complicated, a person like me still can really understand what’s going on,” and "(2) I have enough knowledge to discuss political issues or join political affairs." The variable was measured on a five-point scale where 1 meant
“strongly disagree” and 5 meant “strongly agree.” (r=.763; M=6.98; SD=1.68, range=2 to 10)
Collective efficacy refers to the belief in the capability of the general public to act collectively to effect social change and is defined as perceptions of self-competence in regards to political and public affairs. “(1) Taiwanese collective action can change the society,” and "(2) Taiwanese collective action can affect the development of public affair." The variable was measured on a five-point scale where 1 meant “strongly disagree” and 5 meant “strongly agree.” (r=.813; M=7.6; SD=1.68; range=2 to 10) External efficacy refers to whether one sees the political system as responsive to public demands. Respondents were asked to indicate their agreement level with the following statement: Respondents were asked to indicate their agreement level with the following statement: “(1) Taiwan political system can effectively react to public opinion,” and "(2) Taiwan government can effectively react to public opinion." The variables were measured on a five-point scale where 1 meant “strongly disagree” and 5 meant “strongly agree.” (r=.818; M=4.55; SD=1.68; range=2 to 10)
Partisanship, adapted from Taiwan's Election and Democratization Study, refers to the long-term psychological attachment with a given party, in the form of shared attitudes and beliefs the tendency to support a particular group, cause, or viewpoint over another (Campbell et al., 1960). The respondents are asked whether
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
32
they support any particular political party including KMT, DPP, New Party, People's First Party, Taiwan Solidarity Union, and other parties, or they are non-partisanship.
Among these items, KMT, New Party, People's First Party was recoded as 1=pan-blue;
DPP and Taiwan Solidarity Union was recoded as 2=pan-green; If the answers are
“supporting none of the parties,” “refuse to answer,” “other parties”, respondents are recoded as 3=political independence.
Demographic variables such as age, gender, income, and educational level, were served as control variables in this study.
Time spent on Facebook refers to average hours spent on Facebook on a daily basis. It was measured by asking the respondents to report on the amount of time they spent on FB daily using a scale of 1-7, with 1 meaning “less than half an hour”, and 7 meaning “6 hours or above” (M=3.38; SD=1.75) .
News Media use combined two media attributes: the frequency and attention of TV and newspaper use. Frequency of media use was measured with a single item asking respondents to rate a 5-point scale (where 1=never and 5=always) how often they used the newspaper and TV to get information about current events, public issues, or politics. Media attention refers to how much attention he or she pays to news on TV and newspapers on a 5-point scale where 1 meant “very low” to 5 meant “very high.” Media frequency and attention were then summoned (M=11.83; SD=2.98;
range=4 to 20).
Political discussion was a measure with a single item asking respondents how often they talked about politics face to face with family or friends. The items are measured by a 5-point scale, where 1 meant “never”, 5 meant “always” (M=11.83;
SD=2.98; range=4 to 20).
‧
Table 3.1 Descriptive statistics of construct Construct Internal political activities on Facebook and the three forms of political efficacy were distinct from each other and whether the items measuring each of these variables were distinct from each other and that the items measuring each of these variables reflected the expected common underlying processes. It was expected that these analyses would identify two factors related to political activities on Facebook and three factors related to political efficacy.
The results were presented in Table 3.2. The researcher included all six items comprising two dependent measures, namely Facebook use for information and Facebook use for expression. As expected, the items of political activities on
‧
Table 3.2 Factor loading for the items of Political activities on Facebook
Items
Factors
Expression Information I normally get information about political issues
and current events from Facebook Friends' post of news, articles, or the videos.
.911
I search for information about political issues and current events from Facebook Friends, Fan pages, or clubs.
.808
I forward the information about politics or current affairs
.852
I post comments about personal experience of participating public affairs or political activities on Facebook, including articles, news, or videos.
.901
I post comments about politics or current events about articles, news, or videos
.898
I discuss politics or current events with my friends, family, or colleagues via Facebook personal messages.
.581
Eigenvalue explained variation
the total explanation of variance Cronbach’s α presented in Table 3.3. A confirmation factor analysis with principal components and varimax rotation revealed that the political efficacy items formed three distinct factors with eigenvalues over 1, namely internal efficacy, external efficacy, and collective
‧
respectively, with reliability tests of .763, .813, and .818.Table 3.3 Factor loading for the items of Political efficacy
Item Internal Although politics and the government seem so
complicated, a person like me still can really understand what’s going on
.887
I have enough knowledge to discuss politic issues or join political affairs
.893
Taiwanese collective action can change the society
. 895
Taiwanese collective action can affect the development of public affair
.919
Taiwan politic system can effectively react to public opinion
.918 Taiwan government can effectively react to
public opinion
.921
Eigenvalue 2.169 1.69 1.16
explained variation 36.15% 28.17% 64.31%
the total explanation of variance 83.65%
Cronbach’s α .763 .813 .818
Total Cronbach’s α .272
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
36