Computers in Human Behavior 36 (2014) 114–121

Contents lists available at ScienceDirect

Computers in Human Behavior

journal homepage: www.elsevier.com/locate/comphumbeh

Revisiting civic voluntarism predictors of college students’ political

participation in the context of social media
Yeojin Kim a,⇑, Hyoungkoo Khang b,1
a
College of Communication and Information Sciences, University of Alabama, United States
b
Department of Advertising and Public Relations, College of Communication and Information Sciences, University of Alabama, United States

a r t i c l e i n f o a b s t r a c t

Article history: Utilizing and revamping the civic voluntarism model developed by Verba et al. (1995), this study
Available online 14 April 2014 examines civic voluntarism predictors and their influence on college students’ online and offline political
participation in the context of social media. Results showed that college students with more civic volun-
Keywords: tarism predictors (i.e., resources, psychological engagement, and recruitment) tend to participate more in
Civic voluntarism model political activities. The findings further revealed that individual subcomponents of each civic voluntarism
Political participation predictor have different effects on SNS and offline political participation. The findings also confirmed that
Social networking service
there is a mediating role of SNS political participation between civic voluntarism predictors and offline
College students
political participation.
Published by Elsevier Ltd.

1. Introduction However, despite this continuous attention to young voters, few

studies have paid particular attention to college students, who
From a historical perspective of democracy in the United States, comprise 60% of 18–29 year old eligible voters in the U.S. (CIRCLE
scholars have claimed that voluntary membership groups and staff, 2012; Niemi & Hanmer, 2010). According to a recent report
competitive elections propel civic engagement (Skocpol, Ganz, & on young voters by The Center for Information & Research on Civic
Munson, 2000). However, in response to the continuing decline Learning & Engagement, 71% of young voters in the 2012 presiden-
of young cohorts’ political participation in the U.S., scholars have tial election were college educated, and in the 2008 presidential
questioned the sustainability of democracy (Levine, 2007). Indeed, election, 70% had college experience (CIRCLE staff, 2012). Further-
youth constitute the least politically involved segment of the more, college students are typically on the verge of forming polit-
general population in the country (Kaid, McKinney, & Tedesco, ical identity, are geographically mobile, and their high percentage
2007), and this trend is not anticipated to reverse trend in the fore- of voter turnout is expected to make them an important segment
seeable future. During the 2012 presidential election, for example, of the future voting public in America (Niemi & Hanmer, 2010).
a Harvard Public Opinion Project poll measured youth voter turn- To date, however, scholars have substantially examined politi-
out at its lowest point since 18-year-olds were first permitted to cal participation and its antecedents, focusing mainly on socioeco-
vote in 1972 (Wirth, 2012). nomic status (Verba et al., 1995). Verba et al.’s civic voluntarism
Concerned with young voters’ lack of engagement in the elec- model has been recognized as a robust approach to accounting
toral process, scholars have acknowledged that this group’s voting for political participation (Whiteley, 2009) and investigating pre-
turnout and varied civic engagement (e.g., government and non- dictors of political participation. Indeed, the model can be applied
profit activities) are essential to maintaining and developing to examine, not only overall participation, but also a particular
democracy (Levine, 2007). This view seems to be corroborated by political activity. Considering the pervasiveness of social media in
research that suggests that young voters who are motivated to par- current society, however, the application of the civic voluntarism
ticipate in voting are more likely to get involved in an electoral model to political participation in the online environment raises
event or a political process as adults (Levine & Lopez, 2002). some concerns. First, the model was established based on the tra-
ditional model of political participation before the Internet was
⇑ Corresponding author. Address: P.O. Box 870172, Tuscaloosa, AL 35487-0172, widely available. Secondly, the model has overlooked media as
United States. Tel.: +1 2057653519. one of factors influencing political participation.
E-mail addresses: [email protected] (Y. Kim), [email protected] With the growing influence of the Internet, scholars have be-
(H. Khang). come increasingly interested in the effects of this new technology
1
Address: P.O. Box 870172, Tuscaloosa, AL 35487-0172, United States.

http://dx.doi.org/10.1016/j.chb.2014.03.044
0747-5632/Published by Elsevier Ltd.
 Y. Kim, H. Khang / Computers in Human Behavior 36 (2014) 114–121 115

on political participation and have tended to draw comparisons to such as newspapers, TV, and radio, that have been used as one-way
traditional predictors of offline participation (Boulianne, 2009; Cho transmitters of political information from governments or politi-
et al., 2009; Nie, 2001; Norris, 2002; Schlozman, Verba, & Brady, cians to citizens, the Internet offers an open space for two-way
2010; Shah, McLeod, & Lee, 2009). Scholarly inquiry into the effects communication between political entities and citizens as well as
of new media has been further invigorated as social networking among users. In addition to serving as a facilitator of political par-
services (SNS) have become significantly integrated into users’ dai- ticipation, the Internet has enabled general citizens to directly or
ly lives. In particular, scholars have paid attention to the potential indirectly participate in various political activities (de Zúñiga, Jung,
of these SNSs to motivate young people to engage in varied polit- & Valenzuela, 2012; Johnson & Kaye, 2003; Jung, Kim, & de Zúñiga,
ical activities because young cohorts, especially college students, 2011; Kaye & Johnson, 2002). As a more user-oriented and person-
are among the most frequent users of social media, and this group’s alized service, social media allows citizens to take part in politics in
social interactions and communications tend to be more firmly various ways, depending on their networks and connectivity, such
established through SNS (Ellison, Lampe, & Steinfield, 2009). as posting their political views, sending political messages via so-
Evidence demonstrated that between 67% and 75% of college-aged cial media, subscribing to political news, and signing up to volun-
youth in the U.S. use social media, while 90% of a sample of college teer for a campaign.
students from 126 U.S. universities and one Canadian university Despite the emergence of social media as an alternative form of
use social media (Junco, 2012; Lenhart, Purcell, Smith, & Zickuhr, the public sphere, few researchers have explored political partici-
2010; Smith & Caruso, 2010). pation through SNS, confining themselves instead to investigating
While research on social media has gained momentum, few the features of social media (Aral & Walker, 2012; Centola, 2010;
scholars have explored the antecedents of college students’ politi- Fowler & Christakis, 2010). Furthermore, findings on the associa-
cal participation in conjunction with the social media environ- tion between SNS usage and political participation have been
ment. In addition, few have examined the relationships of SNS inconsistent. Some scholars (Grant et al., 2010; Posetti, 2011;
usage with various determinants that might influence political par- Westling, 2007; Williams & Gulati, 2008) have suggested that
ticipation, especially during election cycles. In fact, past studies SNS demonstrates a significantly positive influence on political
tended to focus primarily on the effects of media usage on political participation by connecting people with government and politi-
participation, and seemed to overlook the pattern of determinants cians and eliciting their mobilization. Still others have suggested
that influence the relationship between media usage and political limited effects on voting turnout (Bryan, Walton, Rogers, &
participation in the social media context. Dwecka, 2011; Gerber, Green, & Larimer, 2008; Nickerson, 2008).
In this study, we intend to investigate how college students’ In addition, some scholars tend to focus more on examining the
SNS use influences political participation by examining anteced- direct relationship between SNS usage and specific aspects of polit-
ents of political participation. In particular, the applicability of ical activities, such as political information (Cantijoch, Jorba, & San
the civic voluntarism model (Verba et al., 1995) to the current so- Martin, 2008; Sweetser & Kaid, 2008), political attitude (Zhang,
cial media environment will be examined by focusing on college Johnson, Seltzer, & Bichard, 2010), or political participation (de
students, which is a segment of the population. By adapting the Zúñiga et al., 2012), while others (Bond et al., 2012) suggested an
civic voluntarism model to explain predictors and patterns of inconsistent relationship. Specifically, Bond and his colleagues
political participation, we specifically examine key antecedents of found that SNS messages directly affect online political self-expres-
political participation, such as SNS use, psychological engagement, sion and information seeking as well as offline voting behavior.
recruitment, and the relationships between these variables. The They also determined that SNS messages indirectly influence polit-
results of this study are therefore expected to expand our knowl- ical attitude and offline voting behavior via communication be-
edge of SNS usage and political participation, with a desired tween individuals with strong ties, such as close friends who
outcome of ultimately helping to build appropriate and effective engage in a face-to-face relationship. Despite inconsistent evidence
political campaign strategies to encourage young voters to engage regarding the effects of SNS use on political participation online
in the democratic process. and offline, the literature describes the potential of SNS as a
catalyst for political participation, while the effects of other tradi-
tional political variables, such as political interest, efficacy, or trust,
2. Literature review and theoretical background remain unknown.
Researchers have tended to focus mainly on specific platforms
2.1. SNS use and political participation of SNS, such as Facebook and Twitter. While Facebook requires a
reciprocal relationship between users to share their information
Since the 2008 U.S. presidential campaign, social media, includ- and opinions, Twitter allows interactions between users and non-
ing Twitter and Facebook has greatly expanded as a new form of users without special permission (Grant et al., 2010; Kwak, Lee,
political communication. SNS allows people to create and maintain Park, & Moon, 2010; Porter, 2009). As such, further study of the
networks of social relationships while also facilitating interactions usage of a distinct form of SNS and its association with political
between networks, thus allowing users to obtain more information participation is merited.
and engage in social interactions within a larger network of connec-
tions (Boyd & Ellison, 2007; Steinfield, DiMicco, Ellison, & Lampe, 2.2. Civic voluntarism predictors in the social media environment
2009). In the political context, SNS usage is expected to play a role
as an alternative communication channel with which to coordinate In the civic voluntarism model elaborated in previous studies
and mobilize political activity based on network-based interactions (Parry, Moyser, & Day, 1992; Verba, 1987; Verba & Nie, 1972),
between users. Verba et al. (1995) attempted to investigate ‘‘why people do not
Recognizing SNS as a common platform for motivating citizens become involved in political activities’’ and found that the primary
to connect with government and politicians and to engage in polit- reasons were ‘‘because they cannot,’’ ‘‘because they do not want
ical activities (Grant, Moon, & Grant, 2010; Posetti, 2011), scholars to,’’ and ‘‘because nobody asked’’ (p. 269). They also identified
have discussed its potential as a political campaign tool to mobilize three core components of political participation (‘‘resources,’’
citizens, facilitate political engagement, and eventually contribute ‘‘psychological engagement,’’ and ‘‘recruitment’’) and examined
to election victories (Tumasjan, Sprenger, Sandner, & Welpe, 2011; the similarities and differences of each component in relation to
Westling, 2007; Williams & Gulati, 2008). Unlike traditional media, various forms of political activity.
116 Y. Kim, H. Khang / Computers in Human Behavior 36 (2014) 114–121

2.2.1. Resources political information or knowledge based on cognitive engagement

Resources are comprised of time, money, and civic skills, all of encourages individuals to form political attitudes toward political
which are relevant for political activities. That is, people who have activity. In addition, the strength of one’s party identification is
more free time, money, and civic skills are more likely to partici- not only a powerful predictor of the direction of that individual’s
pate in political activities. Verba et al. (1995) regarded time and vote, but also promotes political engagement (Barkan, 2004; Verba
money as the resources that most directly affect political activity, et al., 1995; Zaller, 1992). Since individuals’ political participation
but made a distinction between these two components in terms relies on their own ‘‘voluntariness,’’ psychological engagement that
of politically relevant characteristics and unequal distribution. represents a psychological orientation to politics can explain why
For example, money, a more powerful predictor of political activ- people do or do not want to participate in political activity. People
ity, is more directly related to politically relevant conditions, who do not experience subjective psychological engagement with
including gender, race, and ethnicity, and is a more unevenly dis- politics might not use their resources to engage in politics. Verba
tributed resource among individuals than is time. Thus, Verba et al. (1995) argued that political information or knowledge differs
et al. (1995) suggested that more money made people more likely from other factors of psychological engagement in that it tends to
to support candidates. Verba et al. (1995) further defined civic be cognitive rather than psychological and objective rather than
skills as ‘‘the communications and organizational abilities that al- subjective. Considering the nature of political information, this as-
low citizens to use time and money effectively in political life’’ pect is excluded from the measurement in the current study and
(p. 304). That is, people who can communicate well or who are replaced by political trust, which was not included in the original
good at organizing and participating in meetings tend to partici- civic voluntarism model, but features more psychological and sub-
pate more in political activities that are less time-consuming and jective natures than political information. Verba et al. (1995) also
less costly, and they in turn get more involved in them. Hence, found that three of the components of psychological engagement
acquisition of civic skills is related to the nature of organization (political interest, political efficacy, and political information) are
as well as individual characteristics. Verba et al. (1995) suggested all significantly correlated with one another, while partisanship
a variety of measures of objective civic skills, including good lan- is only related to political interest. In addition, some scholars have
guage skills, the ability to communicate in English, experience suggested that trust in politicians or political institutions influ-
practicing communication, and organization skills in political and ences higher participation in politics (Barkan, 2004; Hetherington,
non-political organizations, workplaces, and churches. 1999; Hooghe & Marien, 2013; Valenzuela, Kim, & de Zúñiga, 2012;
In the social media environment, however, political participa- Zhang et al., 2010). Therefore, the following hypothesis is posited:
tion has moved from offline activities to the online sphere, and cit-
izens are increasingly participating in politics on social networking H2. Political interest, political efficacy, and political trust will be
services by expressing their opinions, eliciting public opinion, and predictors of psychological engagement of college students’ polit-
influencing decision-making offline. In this regard, political use of ical participation.
social media itself may reflect a new form of political participation,
and time spent using social media may be an important factor
influencing political participation. Furthermore, previous studies 2.2.3. Recruitment
have identified a positive relationship between social media usage The third and final component of the model deals with recruit-
and antecedents of political participation as well as political partic- ment by friends, relatives, or acquaintances in an individual’s inter-
ipation itself (Bond et al., 2012; Cantijoch et al., 2008; de Zúñiga personal network that leads to political activity. Although people
et al., 2012; Sweetser & Kaid, 2008; Vitak et al., 2011). In addition have resources and experience psychological engagement, they
to time spent on social media for political use, a user’s network size might not take part in political activity unless asked by others to
on social media can be considered a component of resources. For participate. In this sense, recruitment may play a role as a ‘‘trigger-
the purpose of this study, the term ‘‘political network size’’ on ing factor’’ (Verba et al., 1995, p. 273). Recruitment for political
SNS can be defined as ‘‘the number of people with whom the user participation occurs in different kinds of places that cultivate social
discusses politics’’ via SNS, which is based on Eveland and Hively networks, such as the workplace, church, and organizations.
(2009)’s conceptualization (p. 209). A higher number of political Recruitment seems to be very similar to civic skills in that those
networks may represent the civic skills to politically communicate who have a larger network size are likely to experience more occa-
in the social media context. For example, people who have a larger sions of being asked to participate in politics. As noted by Verba
network size on SNS may be more likely to exhibit higher commu- et al. (1995), however, civic skills do not necessarily coincide with
nication and organization skills on social media. In addition to recruitment because a direct request to participate in politics
household income, time spent on SNS for political purpose and net- differs from other types of political exposure. Further, recruitment
work size of SNS are, therefore, expected to demonstrate potential only refers to direct requests in the offline setting from the typical
predictors of resources that affect political participation. As such, perspective of the civic voluntarism model. However, it may also
H1 is tested. be applicable to direct requests on SNS in the social media context.
Given the importance of direct requests as a facilitator for
H1. Household income, time spent on SNS for political purpose, participation, we can assume that recruitment on SNS and offline
and network size of SNS will be predictors of resource for college is associated with political participation.
students’ political participation.
2.3. Civic voluntarism predictors, social media and political
participation
2.2.2. Psychological engagement
Verba et al. (1995) suggested psychological engagement with Among the three civic voluntarism predictors, Verba et al.
politics as a second component of the civic voluntarism model. (1995) emphasized individual resources and psychological engage-
Psychological engagement includes political interest, political effi- ment over recruitment network for most modes of political partic-
cacy, political information, and strength of party identification. In ipation. In particular, the traditional civic voluntarism model treats
particular, interest in politics tends to motivate people to actively resources as the most influential and recruitment as the least influ-
participate in political activities. A sense of political efficacy ential determinant of political participation. However, the compo-
provides the feeling that participation makes a difference, while nents of resources and recruitment have changed along with the
 Y. Kim, H. Khang / Computers in Human Behavior 36 (2014) 114–121 117

evolution of new media technology. Furthermore, it is anticipated 3.2. Measures

that the relative degree of importance among the three determi-
nants will be different in the social media context. As prior litera- 3.2.1. Resources
ture (Junco, 2012; Lenhart et al., 2010; McLeod, Scheufele, & Moy, Resources included participants’ household income levels, time
1999; Putnam, 1995; Putnam, 2000; Smith & Caruso, 2010) has spent on SNS for political use, and political network size on SNS.
indicated, in the context of social media, SNS and offline recruit- Household income was ranked from 1 (less than $20,000) to 6 (more
ment may be very influential for college students’ political partic- than $100,000). Time spent on SNS was measured as time spent
ipation. Thus, the three civic voluntarism predictors may affect SNS using SNS for political purposes. For this study, SNS included Face-
and offline political participation differently. Accordingly, the book and Twitter because these are the most frequently used SNS
following research questions were explored: platforms among college students. Specifically, time spent on SNS
was measured by asking participants, ‘‘How often do you use the
RQ1: Which civic voluntarism predictor demonstrates the following SNS to obtain political information?’’ Answers were
greatest effect on political participation? ranked on a 5-point scale from 1 (rarely) to 5 (all the time). The
RQ2: How do civic voluntarism predictors of participation affect two items of Facebook and Twitter were averaged to create a com-
SNS and offline political participation differently? posite index (a = .62). Modified from Eveland and Hively (2009)’s
definition, Network size on SNS was measured by asking two
In terms of SNS usage and political participation, to date, many open-ended questions: ‘‘How many people did you talk to via Face-
scholars (e.g. de Zúñiga et al., 2012; Valenzuela, Kim, & de Zúñiga, book about politics or public affairs during the past month?’’ and
2012; Valenzuela, Park, & Kee, 2009) have focused mainly on rela- ‘‘How many people did you talk to via Twitter about politics or
tionships between SNS usage and online participation, or between public affairs during the past month?’’ The two items were
SNS usage and offline participation. In light of the relationship be- averaged to create a composite index (a = .86).
tween online and offline political participation, there is a large
amount of evidence (Di Gennaro & Dutton, 2006; Hardy & 3.2.2. Psychological engagement
Scheufele, 2005; Raynes-Goldie & Walker, 2008; Vitak et al., Psychological engagement includes political interest, political
2011; Weinberg & Williams, 2006) that online participation is efficacy, and political trust. Specifically, political interest was mea-
now frequently performed and may further significantly affect sured by asking participants, ‘‘How interested are you in current
offline participation. Based on these findings, SNS political partici- U.S. politics?’’ with responses along a 5-point scale from 1 (not at
pation may also have a positive influence on offline participation. all) to 5 (very). Political efficacy was measured using four items
However, few studies have focused on the relationships between from Verba et al.’s (1995) 5-point political efficacy scale, with an-
online and offline political participation considering social media swers ranging from 1 (never) to 5 (all of the time). The four items
usage and all other possible predictors of political participation. were then averaged to create a composite index (a = .80). Political
As such, the following hypothesis was proposed: trust was measured by asking respondents to rank on a 5-point
scale from 1 (do not trust at all) to 5 (completely trust) the degree
H3. Political participation through SNS will be positively associ- to which they trust the U.S. Congress, politicians, political parties,
ated with political participation offline. the legal system, the police, and the U.S. government. The four
items were averaged to create a composite index (a = .87).

3. Research methods 3.2.3. Recruitment

Recruitment was measured by asking ‘‘Have you done X within
3.1. Sample and procedures the last 12 months?’’ for each activity. Respondents were asked to
choose among three responses (0 = never, 0.5 = yes, but not within
A web-based survey was conducted from October 23 to the last 12 months, and 1 = yes, within the last 12 months) for each
November 20, 2012, prior to and following the 2012 U.S. Presiden- of the following statements: ‘‘I have received direct personal
tial Election Day. The theoretical population of the study consisted requests to take some active role in a local, public, or political
of college-educated young people who are eligible to vote in U.S. issue;’’ ‘‘I have received direct personal requests to take part in
elections and are familiar with social media-based communication campaign work or make a campaign contribution;’’ ‘‘I have
such as Facebook and Twitter. By using a convenience sampling received direct personal requests to contact an elected official;’’
procedure, which is an acceptable and efficient method often used and ‘‘I have received direct personal requests to take part in a
for analytical surveys (Tewksbury & Althaus, 2000), a total of 348 protest, March, or demonstration.’’ The four items were averaged
undergraduates with different majors were recruited to take a to create a composite index (a = .78).
web-based survey, with links to the survey distributed via email
and college survey pools at large universities in the U.S. 3.2.4. Political participation
Participants were awarded course extra credit for participating in Political participation includes SNS and offline political
this study. After removing 29 incomplete responses, a sample of participation.
319 student participants was used for analysis in this study. SNS political participation included political participation on
Among a total of 319 respondents, 83 (26.2%) were male and Facebook and Twitter. Each political participation was measured
234 (73.8%) were female. Ages ranged from 18, the youngest age using six modified items from the Online Political Participation
of U.S. voters, (1 respondent or 0.3% of the sample) to 54 (1 respon- Scale from de Zúñiga et al. (2012). Political participation on Facebook
dent or 0.3%). The mean age of all respondents was 19.33 was measured on a 5-point scale from 1 (rarely) to 5 (all the time)
(SD = 2.99). The majority of participants identified themselves as in answer to the question, ‘‘How often do you use Facebook to X?,’’
freshmen (47.3%), followed by sophomores (25.6%), juniors including the following activities: ‘‘Post your political message on
(15.8%), seniors (11.0%), and not currently enrolled in college your Facebook,’’ ‘‘Post your response on others’ political view on
(0.3%). The racial distribution of the respondents was 86.8% others’ Facebook,’’ ‘‘Read others’ political opinion on others’
White-Non Hispanic, 7.3% African American or Black, 3.2% Asian Facebook walls,’’ ‘‘Subscribe to a political newsfeed/magazine,’’
or Asian American, 0.9% Latino or Hispanic, 0.9% Biracial (any ‘‘Sign up to volunteer for a campaign/issue,’’ and ‘‘Send a political
combination of the above), and 0.9% Other. opinion to others using Facebook message.’’ Political participation
118 Y. Kim, H. Khang / Computers in Human Behavior 36 (2014) 114–121

on Twitter was measured on a 5-point scale from 1 (rarely) to 5 (all Table 2

the time) in response to the question ‘‘How often do you use Zero-order Pearson correlation matrix for observed variables.

Twitter to X?,’’ including the following activities: ‘‘Post your (1) (2) (3) (4) (5) (6) (7) (8) (9)
political opinion on your Twitter feed,’’ ‘‘Reply with your political (1) 1
opinion to others’ tweets,’’ ‘‘Read others’ political opinion on (2) .023 1
others’ tweets,’’ ‘‘Retweet others’ political opinion,’’ ‘‘Subscribe to (3) .075 .139* 1
a political newsfeed/magazine,’’ and ‘‘Sign up to volunteer for a (4) .108 .164** .096 1
(5) .104 .138* .058 .220** 1
campaign/issue.’’ The twelve items of political participation on (6) .120* .026 .001 .163** .309** 1
Facebook and Twitter were averaged to create a composite index (7) .099 .122* .101 .181** .243** .120* 1
of political participation through SNS (a = .90). (8) .025 .326** .127* .341** .294** .149** .270** 1
Offline political participation was measured using six modified (9) .017 .192** .171** .362** .371** .195** .458** .521** 1
items from the Political Participation Scale (Valenzuela et al., Note: (1) Household income; (2) Time spent on SNS; (3) Network size of SNS; (4)
2009). Respondents were asked to respond to the question ‘‘Have Political interest; (5) Political efficacy; (6) Political trust; (7) Direct request; (8) SNS
you done the following activities within the last 12 months?’’ by political participation; (9) Offline political participation.
*
p < .05.
choosing one of three answers (0 = never, 0.5 = yes, but not within **
p < .01.
the last 12 months, and 1 = yes, within the last 12 months). The six
items were averaged to create a composite index (a = .82). p > .05). Household income was the third predictor of resource
The descriptive statistics for observed variables are found in (b = .11, p < .001). As such, H1 was partially supported.
Table 1. H2 predicted that political interest, political efficacy, and polit-
ical trust would be predictors of psychological engagement of col-
4. Results lege students’ political participation. The findings revealed that
political interest had the greatest impact on psychological engage-
To test the hypotheses and answer the research questions ment of college students’ political participation, but no statistical
simultaneously, we employed structural equation modeling significance was found (b = .95, p > .05). Political efficacy was the
(SEM) analysis based on the correlation analysis of observed vari- second strongest predictor of psychological engagement (b = .92,
ables as shown in Table 2. All the suggested independent and p < .001), and political trust was significant, but the third best pre-
mediating variables except for household income were signifi- dictor of psychological engagement to affect political participation
cantly related to SNS and offline political participation. Our (b = .47, p < .001). Thus, H2 was supported.
research hypotheses and questions include direct and indirect RQ1 assessed which civic voluntarism predictor demonstrates
relationships between observed and latent variables that the SEM the greatest effect on political participation, while RQ2 explored
analysis involves (Caplan, 2005; Holbert & Stephenson, 2002). In how civic voluntarism predictors of participation affect political
order to report the SEM analysis to test the relationship between participation differently through SNS and offline. Overall, Fig. 1
civic voluntarism predictors and SNS and offline political participa- indicated that all three civic voluntarism predictors, including re-
tion, LISREL 9.1 for Windows software package was used. source, psychological engagement, and recruitment, demonstrated
The initial SEM analysis was tested and revised by adding new statistically positive significant predictors of both offline and SNS
correlation paths, and the revised model reported full information political participation (p < .01). Additionally, the effect size of civic
maximum likelihood estimation with standardized estimated path voluntarism predictors was varied by the type of political partici-
coefficients. Out model had acceptable fit indices (v2 = 19.40, pation. Resource was a significant positive predictor of SNS politi-
df = 11, p > .05; RMSEA = .049; CFI = .983; NNFI = .946; cal participation (b = .14, p < .001), but insignificant positive
ECVI = .274). The results are presented in Fig. 1. predictor of offline participation (b = .01, p < .01). Psychological
H1 predicted that household income, time spent on SNS for engagement was a significant positive predictor of both SNS
political purpose, and network size of SNS will be predictors of re- (b = .28, p < .001) and offline political participation (b = .19,
source of college students’ political participation. As shown in p < .001). Recruitment was also a significant positive predictor of
Fig. 1, time spent on SNS for political use, as expected, exhibits both SNS (b = .20, p < .001), but was a significant predictor of offline
the greatest predictor of resource college students’ political partic- political participation (b = .32, p < .001). Among the three civic vol-
ipation (b = 1.71, p < .05). Network size of SNS was the second pre- untarism predictors, psychological engagement was the greatest
dictor of resource, but no statistical significance was found (b = .75, predictor of SNS political participation, while recruitment was
the greatest predictor of offline political participation. On the other
hand, resource was the least predictor of both SNS and offline polit-
Table 1
Descriptive statistics of observed variables with mean and standard deviation.
ical participation.
H3 predicted that SNS political participation would be posi-
Variable M SD tively associated with offline political participation. As expected,
Resource the findings revealed that SNS political participation (b = .37,
Household income 4.50 ($60,000 to $99,999) 1.63 p < .001) was a significant positive predictor of offline political par-
Time spent on SNS 2.38 (1–3 times a week) 1.22
ticipation. Therefore, H3 was supported. Significantly, this finding
Network size of SNS 8.63 (persons) 49.38
shows SNS political participation is an important mediator
between civic voluntarism predictors and offline political partici-
Psychological engagement
Political interest 3.40 1.05 pation. As shown in Fig. 1, civic voluntarism predictors have
Political efficacy 2.62 .70 indirect positive effects on offline political participation mediated
Political trust 2.83 .77 by SNS political participation.

Recruitment
Direct request 1.27 1.27 5. Discussion and conclusions
SNS political participation 3.36 4.92
In light of the emergence of SNS as an online form of the
Offline political participation 1.01 1.39
public sphere, this study revisited the determinants of political
 Y. Kim, H. Khang / Computers in Human Behavior 36 (2014) 114–121 119

Fig. 1. Standardized path model of civic voluntarism predictors and political participation. Note: p < .05, 
p < .01, 
p < .001.

participation identified by the civic voluntarism model (Verba The results of the study also demonstrated that psychological
et al., 1995). In particular, we examined the effects of resources, engagement, including political interest, political efficacy, and
psychological engagement, and recruitment on two SNS (Facebook political trust appears to be a consistent and strong predictor of
and Twitter) and offline political participation. Thus, the findings of both SNS and offline political participations. The findings of this
this study were expected to reveal similarities and differences in study provide political efficacy as an important determinant of
effects of those predictors on two different types of political psychological engagement. The correlation between political effi-
participation as well as different platforms of social networking cacy and participation has been considerably corroborated in past
services. Furthermore, this study suggests determinants of two studies (Finkel, 1985; Ginsberg, 1982; Kaid et al., 2007; Sweetser &
distinctive forms of political participation in considering college Kaid, 2008; Verba et al., 1995). However, in our findings, it has a
students’ participatory behaviors in the social media environment. limitation to guide further insight of the reciprocal effects of polit-
Accordingly, we anticipated being able to add meaningful ical efficacy and political participation. Future researchers may re-
knowledge to the civic voluntarism model in the social media con- solve this question by identifying the interaction effects of
text and suggest viable strategies to encourage people, especially predictors of psychological engagement and SNS and offline polit-
young voters, to engage in democratic procedures. ical participation.
Taken together, the results of the current study support previ- Recruitment has been considered a type of political mobiliza-
ous findings that people who demonstrate a higher degree of civic tion (Rosenstone & Hansen, 1993) and one of the three major
voluntarism predictors—resources, psychological engagement, and causes of political participation (Verba et al., 1995). As a triggering
recruitment—tend to participate in political activity more fre- predictor of political participation, recruitment was thought to be
quently. Furthermore, individual subcomponents of each civic vol- most directly related to political participation among the three ci-
untarism predictor have different effects on each type of political vic voluntarism components (Verba et al., 1995). The finding of this
participation. study is in line with findings of past studies in demonstrating the
In particular, SNS political participation was found to be a sig- striking explanatory power of recruitment in relation to offline
nificant mediator of offline political participation. Well before the political participation. Considering that recruitment efforts are ex-
emergence of social media, Putnam (2000) claimed that communi- pected to flourish on the various social networking services during
cation and social networking are crucial to the health of civic soci- upcoming political events, future researchers should investigate
ety in a democracy. Indeed, the interactive nature of SNS, along the effects of recruitment on different forms of political participa-
with its brevity, instantaneousness, and dissemination of salient tion in conjunction with different aspects of SNSs.
current issues, appears to complement the characteristics of con- Interestingly, our findings indicated that resource was a posi-
temporary voters, who are simultaneously fragmented and con- tive and significant predictor of SNS political participation, but
nected. In this respect, Twitter and Facebook have been somewhat insignificant predictor of offline political participation.
considered powerful social media platforms in political campaign- This finding is somewhat contradictory to the claim of the civic
ing and a silver lining for the democratic process due to their influ- voluntarism model (Verba et al., 1995), in which resource is con-
ence of fundraising, mobilizing, and communication endeavors sidered to be the most powerful component of three civic volunta-
during campaigns in recent elections (Grant et al., 2010; Posetti, rism predictors. Considering that we modified the factors of
2011; Williams & Gulati, 2008). However, the literature has failed Verba’s civic voluntarism model, embracing social media environ-
to consistently explain the direct relationship between SNS usage ment, these findings exhibit that amount of SNS usage and size of
and political participation (Bryan, Walton, Rogers, & Dwecka, social networks are likely to elicit online political participation
2011; Gerber et al., 2008; Nickerson, 2008). In this sense, our find- among young voters. The finding of the study, however, that there
ings of SNS political participation, including Facebook and Twitter was an insignificant relationship between SNS resource and offline
as a mediator of offline participation, will provide a meaningful political participation, leads us to further scholarly inquiry for clar-
way of understanding political participation in current society. ification of the association. In addition, our finding demonstrating
120 Y. Kim, H. Khang / Computers in Human Behavior 36 (2014) 114–121

the mediating role of SNS political participation between civic vol- CIRCLE staff (2012). Young voters in the 2012 Presidential election: The educational gap
remains. The Center for Information & Research on Civic Learning & Engagement
untarism predictors and offline participation will be helpful in
(CIRCLE). <http://www.civicyouth.org/wp-content/uploads/2012/11/2012-Exit-
understanding prior inconsistencies regarding relationships be- Poll-by-Ed-Attainment-Final.pdf>.
tween social media usage and political behaviors. de Zúñiga, H. G., Jung, N., & Valenzuela, C. (2012). Social media use for news and
Household income, which is a subcomponent of resources, was individuals’ social capital, civic engagement and political participation. Journal
of Computer-Mediated Communication, 17(3), 319–336.
a significant, but not strong predictor. This finding is also inconsis- Di Gennaro, C., & Dutton, W. (2006). The Internet and the public: Online and offline
tent with Verba’s civic voluntarism model that considers money as political participation in the United Kingdom. Parliamentary Affairs, 59(2),
the most influential factor of resource. In this respect, our finding 299–313.
Ellison, N. B., Lampe, C., & Steinfield, C. (2009). Social network sites and society:
provides a couple of intriguing implications. First, college student Current trends and future possibilities. Interactions, 16(1), 6–9.
may be less likely to be influenced by the money factor because Eveland, W. P., & Hively, M. H. (2009). Political discussion frequency, network size,
the costs of using social media platforms for political engagement and ‘‘heterogeneity’’ of discussion as predictors of political knowledge and
participation. Journal of Communication, 59(2), 205–224.
are much lower than the costs of using traditional media. Secondly, Finkel, S. E. (1985). Reciprocal effects of participation and political efficacy: A panel
the result of the study might be related to the characteristics of the analysis. American Journal of Political Science, 29(4), 891–913.
sample. Compared to other groups, college students are less likely Fowler, J. H., & Christakis, N. A. (2010). Cooperative behavior cascades in human
social networks. Proceedings of the National Academy of Sciences, 107(12),
to be influenced by resources such as money. Beyond gender, their 5334–5338.
socioeconomic status conditions are thought to be similar, and Gerber, A. S., Green, D. P., & Larimer, C. W. (2008). Social pressure and vote turnout:
their resources may be relatively evenly distributed among indi- Evidence from a large-scale field experiment. American Political Science Review,
102(1), 33–48.
viduals. In addition, parental income is not always directly linked
Ginsberg, B. (1982). The consequences of consent: Elections, citizen control, and
to a student’s resources (e.g., parents differ greatly in the amount popular acquiescence. Reading, MA: Addison-Wesley.
of support they provide their children in college), also making this Grant, W. J., Moon, B., & Grant, J. B. (2010). Digital dialogue? Australian politicians’
population less likely to be impacted by the variable. However, it is use of the social network tool Twitter. Australian Journal of Political Science,
45(4), 579–604.
out of topic for this study to examine the relationship between Hardy, B. W., & Scheufele, D. A. (2005). Examining differential gains from Internet
parental income and student’s money. As such, future studies are use: Comparing the moderating role of talk and online interactions. Journal of
encouraged to investigate political participation of college students Communication, 55(1), 71–84.
Hetherington, M. J. (1999). The effect of political trust on the Presidential vote,
considering their socioeconomic variables, including gender, age, 1968–96. American Political Science Review, 311–326.
parental income, and money. Holbert, R. L., & Stephenson, M. T. (2002). Structural equation modeling in the
A few limitations should be noted when interpreting the find- communication sciences, 1995–2000. Human Communication Research, 28(4),
531–551.
ings of this study. First, this study utilized convenience sampling, Hooghe, M., & Marien, S. (2013). A comparative analysis of the relation between
which may limit the external validity and reliability of the findings. political trust and forms of political participation in Europe. European Societies,
To lessen this problem to some degree, the current study obtained 15(1), 1–22.
Johnson, T., & Kaye, B. (2003). A boom or bust for democracy? How the Internet
samples from several large universities in the U.S. However, our influences political attitudes and behaviors. Harvard International Journal of
sample showed a high bias between male and female participants, Press/Politics, 8, 9–34.
although gender is not an important variable in this study. To over- Junco, R. (2012). The relationship between frequency of Facebook use, participation
in Facebook activities, and student engagement. Computers & Education, 58(1),
come this drawback, we identified correlation analysis of gender
162–171.
and other civic voluntarism predictors, but no statistical significant Jung, N., Kim, Y., & de Zúñiga, H. G. (2011). The mediating role of knowledge and
relationship was found, excluding political trust. Accordingly, fu- efficacy in the effects of communication on political participation. Mass
ture researchers are encouraged to use random samples in measur- Communication and Society, 14(4), 407–430.
Kaid, L. L., McKinney, M. S., & Tedesco, J. C. (2007). Introduction political information
ing the variables tested in this study with a balanced sample in efficacy and young voters. American Behavioral Scientist, 50(9), 1093–1111.
order to enhance the generalizability of the findings. Second, we Kaye, B., & Johnson, T. (2002). Online and in the know: Uses and gratification of the Web
revamped past studies’ measurement scales of online and offline for political information. Journal of Broadcasting & Electronic Media, 46(1), 54–71.
Kwak, H., Lee, C., Park, H., & Moon, S. (2010). What is Twitter, a social network or a
political participation in order to assess Facebook and Twitter newsmedia? In Paper presented at the 19th international world wide web
political participation. Therefore, future researchers should review conference, Raleigh, NC, USA. <http://delivery.acm.org/10.1145/1780000/
and replicate this measurement to consistently warrant its validity 1772751/p591-kwak.pdf?ip=130.160.173.172&id=1772751&acc=ACTIVE%20
SERVICE&key=C2716FEBFA981EF1B3784EA5F52A32A34C069BF3F1B56327&
and reliability. CFID=233584252&CFTOKEN=46473787&__acm__=1373913103_acc694c0501
cb04d639bdbdb77daf7a0>.
Lenhart, A., Purcell, K., Smith, A., & Zickuhr, K. (2010). Social media and young adults.
References Washington, DC: Pew Internet and American Life Project. <http://
pewinternet.org/Reports/2010/Social-Media-and-Young-Adults.aspx>.
Aral, S., & Walker, D. (2012). Identifying influential and susceptible members of Levine, P. (2007). The Future of Democracy: Developing the Next Generation of
social networks. Science, 337(6092), 337–341. American Citizens. Lebanon, NH: Tufts University Press.
Barkan, S. E. (2004). Explaining public support for the environmental movement: A Levine, P., & Lopez, M. H. (2002). Youth voter turnout has declined, by any measure.
civic voluntarism model. Social Science Quarterly, 85(4), 913–937. College Park, MD: The Center for Information & Research on Civic Learning &
Bond, R. M., Fariss, C. J., Jones, J. J., Kramer, A. D., Marlow, C., Settle, J. E., et al. (2012). Engagement (CIRCLE).
A 61-million-person experiment in social influence and political mobilization. McLeod, J. M., Scheufele, D. A., & Moy, P. (1999). Community, communication, and
Nature, 489(7415), 295–298. participation: The role of mass media and interpersonal discussion in local
Boulianne, S. (2009). Does Internet use affect engagement? A meta-analysis of political participation. Political Communication, 16(3), 315–336.
research. Political Communication, 26(2), 193–211. Nickerson, D. W. (2008). Is voting contagious? Evidence from two field experiments.
Boyd, D. M., & Ellison, N. B. (2007). Social network sites: Definition, history, and American Political Science Review, 102(1), 49–57.
scholarship. Journal of Computer-Mediated Communication, 13(1), 210–230. Nie, N. H. (2001). Sociability, interpersonal relations, and the Internet: Reconciling
Bryan, C. J., Walton, G. M., Rogers, T., & Dweck, C. S. (2011). Motivating voter turnout conflicting findings. American Behavioral Scientist, 45(3), 420–435.
by invoking the self. Proceedings of the National Academy of Sciences, 108(31), Niemi, R. G., & Hanmer, M. J. (2010). Voter turnout among college students: New data
12653–12656. and a rethinking of traditional theories. Social Science Quarterly, 91(2), 301–323.
Cantijoch, M., Jorba, L., & San Martin, J. (2008, August). Exposure to political Norris, P. (2002). Democratic phoenix: Reinventing political activism. Cambridge, UK:
information in new and old media: Which impact on political participation. In Cambridge University Press.
Paper presented at the APSA 2008 Annual Meeting, Boston, Massachusetts, USA. Parry, G., Moyser, G., & Day, N. (1992). Political participation and democracy in Britain.
Caplan, S. E. (2005). A social skill account of problematic Internet use. Journal of Cambridge, UK: Cambridge University Press.
Communication, 55(4), 721–736. Porter, J. (2009). Relationship symmetry in social networks: Why Facebook will go fully
Centola, D. (2010). The spread of behavior in an online social network experiment. asymmetric. <http://bokardo.com/archives/relationship-symmetry-in-social-
Science, 329(5996), 1194–1197. networks-why-facebook-will-go-fully-asymmetric/>.
Cho, J., Shah, D. V., McLeod, J. M., McLeod, D. M., Scholl, R. M., & Gotlieb, M. R. (2009). Posetti, J. (2011). The#spill effect: Twitter hashtag upends Australian political
Campaigns, reflection, and deliberation: Advancing an O-S-R-O-R model of journalism. Media shift. <http://www.pbs.org/mediashift/2010/03/the-spill-
communication effects. Communication Theory, 19(1), 66–88. effect-twitter-hashtag-upends-australian-political-journalism061>.
 Y. Kim, H. Khang / Computers in Human Behavior 36 (2014) 114–121 121

Putnam, R. D. (1995). Bowling alone: America’s declining social capital. Journal of Valenzuela, S., Kim, Y., & de Zúñiga, H. G. (2012). Social networks that matter:
Democracy, 6(1), 65–78. Exploring the role of political discussion for online political participation.
Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community. International Journal of Public Opinion Research, 24(2), 163–184.
New York, NY: Simon and Schuster. Valenzuela, S., Park, N., & Kee, K. F. (2009). Is there social capital in a social network
Raynes-Goldie, K., & Walker, L. (2008). Our space: Online civic engagement tools for site? Facebook use and college students’ life satisfaction, trust, and
youth. In W. L. Bennett (Ed.), Civic life online: Learning how digital media can participation. Journal of Computer-Mediated Communication, 14(4), 875–901.
engage youth (pp. 161–188). Cambridge, MA: MIT Press. Verba, S. (1987). Participation and political equality: A seven-nation comparison.
Rosenstone, S. J., & Hansen, J. M. (1993). Mobilization, participation and democracy in Chicago, IL: University of Chicago Press.
America. New York: Mac Millan. Verba, S., & Nie, N. H. (1972). Participation in America: Social equality and political
Schlozman, K. L., Verba, S., & Brady, H. E. (2010). Weapon of the strong? democracy. New York, NY: Harper & Row.
Participatory inequality and the Internet. Perspectives on Politics, 8(2), 487–509. Verba, S., Schlozman, K. L., & Brady, H. E. (1995). Voice and equality: Civic voluntarism
Shah, D. V., McLeod, J. M., & Lee, N. J. (2009). Communication competence as a in American politics. Cambridge, MA: Harvard University Press.
foundation for civic competence: Processes of socialization into citizenship. Vitak, J., Zube, P., Smock, A., Carr, C. T., Ellison, N., & Lampe, C. (2011). It’s
Political Communication, 26(1), 102–117. complicated: Facebook users’ political participation in the 2008 election. Cyber
Skocpol, T., Ganz, M., & Munson, Z. (2000). A nation of organizers: The institutional Psychology, Behavior, and Social Networking, 14(3), 107–114.
origins of civic voluntarism in the United States. American Political Science Weinberg, B. D., & Williams, C. B. (2006). The 2004 US Presidential campaign:
Review, 94(3), 527–546. Impact of hybrid offline and online meet up communities. Direct, Data and
Smith, S., & Caruso, J. (2010). ECAR study of undergraduate students and information Digital Marketing Practice, 8(1), 46–57.
technology (vol. 6). Boulder, Colorado: Center for Applied Research. <https:// Westling, M. (2007). Expanding the public sphere: The impact of Facebook on political
net.educause.edu/ir/library/pdf/ERS1006/RS/ERS1006W.pdf>. communication. <http://www.thenewvernacular.com/projects/facebook_and_
Steinfield, C., DiMicco, J. M., Ellison, N. B., & Lampe, C. (2009). Bowling online: Social political_communication.pdf>.
networking and social capital within the organization. In Proceedings of the Whiteley, P. (2009). Government effectiveness and political participation in Britain.
fourth international conference on communities and technologies (pp. 245–254). Representation, 45(3), 247–257.
ACM. Williams, C., & Gulati, G. (2008). What is a social network worth? Facebook and vote
Sweetser, K. D., & Kaid, L. L. (2008). Stealth soapboxes: Political information efficacy, share in the 2008 Presidential primaries. In: Paper presented at the 104th annual
cynicism and uses of celebrity weblogs among readers. New Media & Society, meeting of the American Political Science Association, Boston, Massachusetts, USA.
10(1), 67–91. Wirth, A. (2012). A democracy at risk. The Huffington Post. <http://
Tewksbury, D., & Althaus, S. L. (2000). Differences in knowledge acquisition among www.huffingtonpost.com/alex-wirth/millennial-voters_b_1974659.html>.
readers of the paper and online versions of a national newspaper. Journalism & Zaller, J. R. (1992). The nature and origins of mass opinion. Cambridge, UK: Cambridge
Mass Communication Quarterly, 77(3), 457–479. University Press.
Tumasjan, A., Sprenger, T. O., Sandner, P. G., & Welpe, I. M. (2011). Election forecasts Zhang, W., Johnson, T. J., Seltzer, T., & Bichard, S. L. (2010). The revolution will be
with Twitter: How 140 characters reflect the political landscape. Social Science networked the influence of social networking sites on political attitudes and
Computer Review, 29(4), 402–418. behavior. Social Science Computer Review, 28(1), 75–92.