WIWWIN!WWMHI)H!NlllilHWlUllUl

       

_‘I
(O
CDN

TH

'03

    

."- )
c e rs

C _‘>

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

New Media, New Politics: Political Learning Efﬁcacy and the
Examination of Uses of Social Network Sites for Political

Engagement

presented by

Rebecca A. Hayes

has been accepted towards fulﬁllment
of the requirements for the

Ph.D. degree in Media and Information Studies

 

 

Ly/fﬁ’i/aQﬁ IZ4M

{Majﬁrofefs’or’ 3 Signature

bezinb?Y 7075067

Date

MSU is an Afﬁrmative Action/Equal Opportunity Employer

nu-c---—-.-.-—-.-.---.-----n-u—o—»-.---—n-.-n--.---.---u-.-o—o-—--.--A. .o

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProi/Aoc&Pres/CIRCIDateDue.indd

 

NEW MEDIA, NEW POLITICS: POLITICAL LEARNING EFFICACY AND THE
EXAMINATION OF USES OF SOCIAL NETWORK SITES FOR
POLITICAL ENGAGEMENT
By

Rebecca A. Hayes

A DISSERTATION
Submitted to
Michigan State University
In partial fulﬁllment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY

Media and Information Studies

2009

ABSTRACT
NEW MEDIA, NEW POLITICS: POLITICAL LEARNING EFFICACY AND THE
EXAMINATION OF USES OF SOCIAL NETWORK SITES FOR
POLITICAL ENGAGEMENT
By
Rebecca A. Hayes

Online social network sites (SNSs) have become active political forums in the
United States and elsewhere. During the 2008 presidential election, the two major
candidates directed signiﬁcant resources toward the creation and maintenance of SNS
proﬁles and pages in an effort to reach out to young voters. These efforts garnered
substantial media and scholarly attention. One of the main demographic targets of these
sites, individuals aged 18-25, has historically been largely apathetic toward the political
process and has demonstrated a low level of engagement in politics. This may have
changed in the 2008 election with a winning campaign that was well-versed in online
social networking and engagement. Yet little is known about how these sites impact
young voters in terms of increased participation in the electoral process and the variables
that precede participation, such as political knowledge and efﬁcacy.

A person’s feeling of political efﬁcacy, the conﬁdence that one can both
effectively participate in and inﬂuence the political process, is a determinant of political
participation. This construct has been broken down into two components, internal and
external political efﬁcacy (IPE and EPE, respectively), and has been further extended into
the concept of political information efﬁcacy (PIE). The similarities between the
constructs of [PE and PIE, and their respective measurement scales, suggest that the

concepts may not be entirely different. There is a need for conceptual and operational

clariﬁcation and validation. A measurement instrument is needed to ﬁll the gap in the
literature regarding learning and informational antecedents to political efﬁcacy.
Preliminary research indicates that political candidate social network efforts are largely
tools of information provision. The proposed instrument will help answer the question of
whether SNSs can be valuable political engagement tools.

With these two issues in mind, the following multi-stage research was undertaken.
Using Bandura’s Social Cognitive Theory (198d; 1997) as a theoretical guide, the
concepts are examined and sorted, and a new scale, Political Learning Efﬁcacy, is
developed using both qualitative and quantitative methods. Next, the new scale is used in
a nationally representative survey of 18-25 year Olds (n=625) to examine the effects of
previous and continued use of political social networking on PLE, political knowledge
and participation with interesting and informative results.

Results show that use of social network sites for political purposes can have a
positive impact on political learning efﬁcacy and, thus, on knowledge and participation.
However, these results also demonstrate that political participation through social media
during the 2008 election was much lower than was anecdotally reported in the news
media. In addition, online sources of news and political information are not the preferred
informational source among this demographic. Contrary to conventional wisdom
regarding this group, traditional media such as newspapers and televised news are
reported as the preferred informational source.

This research provides a valuable new discipline-spanning measure to the political
efﬁcacy literature and helps to establish the value of candidate social networking in the

political communication arena.

Copyright by
Rebecca A. Hayes
2009

DEDICATION

For my parents, Joy and Randy, my husband, Steve, and my pseudo-kids, Buddy and
Phoebe. Thanks for putting up with me.

ACKNOWLEDGMENTS

Many people helped make my Ph.D. odyssey possible and even fun.

First, I have to thank Dr. Bonnie Reece. She is the reason I came back to Michigan State
for a Ph.D., and the opportunities she gave me made my experience at State truly
memorable. I admire her joie de vivre and appreciate the trail she helped blaze for all
women in academia.

To Nancy Ashley and Dr. Lucinda Davenport, thank you for always being there for every
doctoral student. You are the reason most of us are still sane.

I am also grateful to Dr. Rick Cole, for allowing me opportunities to explore endeavors
not usually allowed to graduate students and to take ideas and run with them. Thank you
to Tamara West for helping me run with them.

Dr. Robert LaRose taught me what it meant to be a graduate student, to appreciate theory,
and most importantly, helped me see what I was missing in the obvious. I remember
more from one class with him taken four years ago than from things I read this morning.
His valuable feedback on this work has made it immeasurably better.

While Dr. LaRose taught me what it meant to be a graduate student, Dr. Elizabeth Taylor
Quilliam taught me how to be a graduate student. She has been a great mentor and ﬁ'iend,
ﬁrst as a fellow graduate student and now as a faculty member.

I would like to sincerely express my appreciation to Dr. Rick Cole, Dr. Cliff Lampe, and
Dr. Robert LaRose. As a committee, their support, direction, and advice was invaluable.

Dr. Nora J. Rifon, my dissertation advisor, endlessly tolerated my somewhat strange
writing style and patiently helped me reﬁne my ideas and direct my energy into the right
conceptual nooks.

I am extremely lucky to have a live-in editor. My husband, Stephen Hensler, has been a
constant, patient, much-appreciated companion on this process. His assistance and un-
wavering support at all stages has been invaluable.

Finally, my parents, Joy and Randy Hayes. Their emotional, ﬁnancial, and statistical
support, literally at all hours of the day or night, made this possible.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. xi

LIST OF FIGURES ............................................................................................................. x

CHAPTER 1

INTRODUCTION ............................................................................................................... 1

CHAPTER 2

LITERATURE REVIEW .................................................................................................... 5
Historical Foundations of Political Communication ...................................................... 6
The Internet in Elections ................................................................................................. 8
The Social Network Phenomena .................................................................................. 10
Young People and Media .............................................................................................. 12
Young People and the Political Process ....................................................................... 14
Study Focus .................................................................................................................. 19

CHAPTER 3

THEORY ........................................................................................................................... 22
Political Efﬁcacy .......................................................................................................... 22
Political Information Efﬁcacy ...................................................................................... 24
Problems with Constructs and Measures ...................................................................... 25
Applying Social Cognitive Theory ............................................................................... 26
Political Learning Efficacy (Tentative Construct Deﬁnition) ...................................... 32

CHAPTER 4

METHOD ........................................................................................................ 34
Multi-Method Scale Development ............................................................................... 34
Qualitative Research to Inform Construct and Scale Development ............................. 34
Development, Reﬁnement and Validation of Construct and Measurement ................. 43
Scale Development and Validation Results .................................................................. 51
Survey Research ........................................................................................................... 65
Measures ....................................................................................................................... 66

CHAPTER 5

RESULTS .......................................................................................................................... 69
Political and Media Behavior ....................................................................................... 69

CHAPTER 6

DISCUSSION .................................................................................................................... 78
Limitations .................................................................................................................... 81
Implications for Campaigns .......................................................................................... 83
Conclusion .................................................................................................................... 84

vii

APPENDICES

Appendix A - Focus Group Guide ................................................................................ 86
Appendix B - Pilot Test Measures ................................................................................ 89
Appendix C - Full Survey ............................................................................................ 91
REFERENCES .................................................................................................................. 98

viii

LIST OF TABLES

Table 1: Factor Analysis of Proposed PLE Items ............................................................. 52
Table 2: Factor Analysis of Non-Media PLE Items ......................................................... 53
Table 3: Correlations and Scale Reliability for PLE Items ............................................... 54
Table 4: Correlations Between PLE Scale and Related Variables ................................... 55
Table 5: Factor Analysis of Proposed PLE Source Items ................................................. 56
Table 6: Correlations and Scale Reliability for Enactive Learning Items ........................ 57
Table 7: Correlations and Scale Reliability for Persuasion Items .................................... 57
Table 8: Correlations and Scale Reliability for Affective Items ....................................... 58
Table 9: Factor Analysis of Final PLE Source Items ....................................................... 59
Table 10: Factor Analysis of Efﬁcacy Scales ................................................................... 62
Table 11: Correlations and Scale Reliability of IPE and PIE as a Combined Scale ......... 63
Table 12: Political Uses and Actions on SNS ................................................................... 70
Table 13: SNSEngage Responses ..................................................................................... 71
Table 14: Preferred News Sources .................................................................................... 72
Table 15: Pearson Correlations of Expected Relationships .............................................. 76

ix

LIST OF FIGURES

Figure l. The triadic reciprocal causation model of Social Cognitive Theory .................. 26

Figure 2. Self-efﬁcacy’s effect on both behavior and environment within the model of

SCT .................................................................................................................................... 27
Figure 3. Sources of self-efﬁcacy ...................................................................................... 28
Figure 4. The implementation of sources of self-efﬁcacy and political learning efﬁcacy

into the triadic model of SCT. ........................................................................................... 32
Figure 5. Hypothesized relation of PLE, its sources, and its dependent variables ............ 33
Figure 6. A Social Cognitive Model of Political Learning ................................................ 60
Figure 7: Full PLE Model .................................................................................................. 75

CHAPTER 1
INTRODUCTION

Young people aged 18-25 have relatively low rates of voter participation,
typically about 20% less than Older age groups (CIRCLE, 2008). This group is known to
have a much lower conﬁdence in their political knowledge than the general population
(Kaid et al., 2007b); lack of information, knowledge, and attention from candidates is
often cited by this group as reasons for not voting (Kaid, McKinney, & Tedesco, 2007;
Hayes; 2008a). Their media choices vary signiﬁcantly from those of the general
population (Delli Carpini, 2000), and this has been noted for contributing to a lack of
exposure to traditionally informative political communication methods.

The media mix of this young population includes heavy use of social network
sites (SNSs), such as MySpace and F acebook, and Web sites where users have become
accustomed to a level of personal information and interactivity not afforded by traditional
political communication methods. As recently as 2006, political candidates began
posting proﬁles on these sites in order to reach out to younger voters and keep pace with
other candidates (Williams & Gulati, 2007). The 2008 presidential election supplied
some of the most complex and interactive candidate Web communication yet. With
varying degrees of success, the two leading candidates built their own social network
platforms within their main Web sites, in addition to maintaining a presence on the
popular public social network sites (Gizmondo, 2008). While candidates were and are
expending signiﬁcant resources to create and maintain these sites, limited research exists
on the impact of the sites on targeted groups and the tactic’s ability to increase

knowledge of or participation in the political process (U tz, 2009).

The 2008 presidential election saw record total turnout among young voters, and a
signiﬁcant increase in turnout from the previous presidential elections in 2000 and 2004.
While much of this is attributable to a dynamic candidate who appealed to young voters,
this candidate’s embrace of online recruiting tools indicates that these tools may be
effective in signiﬁcantly increasing the precursors to voting, such as political efficacy.

While the available research on political social network sites has only recently
begun to build and is there is promising research on the effects of the Internet overall on
political interest, involvement, and engagement. There is also some evidence that the
Internet, for a variety of reasons, could be more effective at promoting democratic
participation than traditional media (Jennings & Zeitner, 2003; Delli Carpini, 2004;
Lenhart, Madden, & Hitlin, 2005; Levine & Lopez, 2004; Lupia and Philpot, 2005;
Pasek, Kenski, Romer, & J amieson, 2006). These effects are expected to extend to
political SNS. Some of the ﬁrst studies in the area of SNSs show that they may have a
positive impact on a voter’s conﬁdence in his or her adequacy to engage in the political
process (Tedesco, 2007; Hayes et al., 2008), an important precursor to political
participation (Kim, 2003). SNSs also have the ability to reach those who are less
interested in politics, who tend to miss other online candidate communications such as
Web sites, since these sites have to be intentionally sought out rather than just stumbled
upon as one might incidentally ﬁnd the social network proﬁle of a candidate (Utz, 2009).

Political efﬁcacy, “the feeling that individual political action does have, or can
have, an impact on the political process” (Campbell, Gurin, & Miller, 1954 p. 187), is an
established precursor to political participation (Campbell et al., 1954; Milbrath & Goel,

1977). This larger construct of political efﬁcacy was sub-divided into the two concepts

of internal and external political efﬁcacy (Converse, 1972; Balch, 1974) due to repeated
ﬁndings of multidimensionality and concerns regarding the validity of the larger
construct (Morrell, 2003). Research has demonstrated the value and validity of each of
these separate concepts in isolation, with IPE in particular being noted as a robust
measure (Craig, Niemi, & Silver, 1990; Niemi, Craig, & Mattei, 1991; Morrell, 2003). A
new construct, political information efﬁcacy, has emerged, sharing remarkable
similarities with the construct of internal political efﬁcacy both in its construct deﬁnition
and most particularly in its measurement (Kaid et al. 2007a; Tedesco 2007), and may
muddying the conceptual waters. This new construct is not well-validated and three of
the four items used to measure it are shared with the measurement of internal political
efﬁcacy.

In addition to the need for conceptual and operational clariﬁcation, the earlier
measures of internal and external political efﬁcacy do not fully capture the range of
effects of online candidate communication through social networks. Other forms of
candidate communication, while providing information to voters, are generally
considered voter persuasion attempts. These new media sources provide political
information, promote learning, and encourage participation without the hard-sell of
traditional candidate communication methods, and current efﬁcacy measures do not
capture that. In order to address both the paucity of research in the area of effects of SNS
candidate communication and the lack of a relevant measurement instrument, a new
construct called Political Learning Efﬁcacy and an evaluative measure for it is developed
here using Social Cognitive Theory (Bandura 1986; 1997) and qualitative research as a

guide. This scale development process used both qualitative focus groups (n=36) and

quantitative survey research (n=358) to formulate the new construct and evaluate its
measurement. Survey research was then implemented with a large, nationally
representative survey of 18-25 year-olds (n=625) to measure whether previous and
continued use of political social network tools increase political knowledge and

participation through increases in political learning efﬁcacy.

CHAPTER 2
LITERATURE REVIEW

The mass media have been heralded as both an important source of political
knowledge (Sotirovic & McLeod, 2004; Eveland, Hayes, Shah, & Kwak, 2005) and a
potential cause of disengagement (Putnam, 2000), and studies of media effects on a
variety of political variables are an active and exciting ﬁeld of research. The most recent
media of interest is the Internet, and, by extension, social network sites such as MySpace
or Facebook. During the 2008 election year, nearly every candidate for ofﬁce at the
national, state, and local level had a Web site, and very likely had a social network proﬁle.
Web-based political communication is touted by many as the political tool of tomorrow,
but that doesn’t mean that the rich scholarly history based on more traditional methods of
political communication, such as the news media and televised advertising, need be
abandoned. This ﬁeld has evolved signiﬁcantly since its beginnings as a minor subﬁeld
of propaganda research (Rogers, 2004), and all future research has its foundations in the
past studies of Walter Lippman, Harold Lasswell, and Paul Lazersfeld, among others.

Originally concerned with the study of propaganda and later the effects of print
media and television on individuals’ vote choice (Rogers, 2004), political communication
has evolved in a parallel fashion in somewhat disparate ﬁelds, political science and
communication. In the past, the two areas rarely viewed the relationship between media
and politics in the same way; communication scholars primarily concerned themselves
with the processes of inﬂuence (the means), while political science scholars tended to
focus on outcomes of exposure (the ends) (Holbert, 2005). More recently, the ﬁelds have

started to come together in their use of similar outcome variables, such as knowledge,

efﬁcacy, and behavior in their studies of effects of various media, including new media

social network sites.

Historical Foundations of Political Communication

Considered by many to be the founding document of political communication,
Walter Lippman’s (1922) Public Opinion argued that controlling public opinion (through
propaganda) was key to controlling public behavior and laid the groundwork for the later
theory of media agenda setting. Lippman’s argument that media are crucial in creating a
free marketplace of ideas helped stimulate a normative concern in political
communication research (then called public opinion and propaganda research) about the
need for a free press and an informed citizenry in a strong democracy (Rogers, 2004).

While the origins of public opinion research began with Lippman, the study of
media effects began with Harold Lasswell. His 1927 work examined the effects of
German, French, American and British propaganda during World War I and it established
the ﬁve question model that many still follow in their media effects research: Who says
what to whom via which channels with what effects? (Rogers, 2004).

Finally, and perhaps most relevant to this work, the ﬁrst quantitative studies of
voting behavior and the media’s effects on it were conducted by Paul F. Lazarsfeld in the
19403. Lazarsfeld and his colleagues, utilizing a sample of 600 people in Erie County,
Ohio, conducted over 3600 personal interviews in the months leading up to the 1940
presidential election. Erie County was chosen as representative of the nation as a whole,
and the study was intended to measure the direct effects of the media on voting behavior.
Lazarsfeld found that only a very small number (54 out of the 600) changed their minds

over the course of the campaign, with information provided by the news media rarely

cited as the cause of a change. He concluded that the media had negligible direct effects
on voting behavior and proposed a theory of a two-step ﬂow of communication in which
interpersonal communication with opinion leaders plays an important factor in
determining media effects. This theory asserts that media information moves in two
different stages. First, individuals (opinion leaders) who pay close attention to the mass
media and its messages receive the information. Second, these opinion leaders pass on
their own interpretations of this information to those around them. The term ‘personal
inﬂuence’ was coined by Lazarsfeld to refer to the intervening process between the
media’s direct message and the audience’s reaction to that message. The two-step ﬂow
theory has improved understanding of how the mass media inﬂuence behavior and has
helped to reﬁne the ability to predict the inﬂuence of media messages on audience
behavior, and it helped explain how and why certain media campaigns may have failed
(Katz & Lazarsfeld, 1955; Katz, 1973).

From the time of Lazarsfeld’s work to through the early 19705, the media were
thought to have little direct effect on behavior or attitudes (Klapper, 1960; Johnston,
Hagen, & Jamieson, 2004; Rogers, 2004; Huber & Arceneaux, 2007). Propaganda
research was still popular, as was persuasion research related to televised news media
viewing, but neither demonstrated deﬁnitive media effects (Rogers, 2004). It was at this
time that McCombs and Shaw (1972) argued that a lack of research on the cognitive and
informational effects of political communication hindered the true understanding of
effects of mass media and that perhaps this was the cause of the lack of ﬁndings. Their
agenda-setting theory of the media, which postulated that the media doesn’t determine

how a person thought about an issue, but it did determine if they thought an issue was

important, became a dominant theory for years, and is noted for being the most frequently
cited work in political communication (Dearing and Rogers, 1996).

Even with McComb and Shaw’s argument related to problems in the research
base, there wasn’t a distinct shift in the research paradigm until the early 908, when
researchers began to focus primarily on the cognitive, affective and behavioral impacts of
political communication, advertising in particular (Johnson-Cartee & Copeland, 1997).
This shift in research paradigm uncovered many of the previously unrealized effects of
political communication. While deﬁnitive results related to persuasion were still elusive,
some candidate communication, notably traditional 30-second televised political
advertisements, were found to improve voter knowledge, political efﬁcacy, and interest in
campaigns (Brians & Wattenberg, 1996; Freedman, Franz & Goldstein, 2004; Kaid &
Postelnicu, 2007), and perhaps increase turnout (F inkel & Geer, 1998). Other researchers
demonstrated negative effects of political advertising, from decreased efﬁcacy (Austin &
Pinkleton, 1995), increased cynicism (Yoon, Pinkleton, & Ko, 2002), and reduced
turnout and disengagement (Ansolabehere, Iyengar & Valentino, 1 994; Ansolabehere &
Iyengar, 1995).

The Internet in Elections

The Internet has emerged as a signiﬁcant force in political campaigns in recent
years. Howard Dean in 2004 and Ron Paul in 2007 were both propelled to national
prominence primarily because of successful online efforts, and the potential of SNSs and
other online political tools were heralded throughout the news media during the latest
election cycle. Emerging as an entirely new media for political communication since

many of the studies on traditional campaign communication were conducted, so far,

research into the effects Of Internet political communication has shown promising results
in the areas of efﬁcacy, information, and involvement. Investigating the impact of
Internet political communication in the same manner and with many of the same
variables that traditional political communication has been will provide a valuable
comparison and an indicator of how powerful these new political communication tools
can be.

Survey research indicates the value that the Internet has to the political process,
with typical Web users found to be more likely to vote, more informed and more
politically efﬁcacious than the general public (Johnson & Kaye, 1998), and exposure to
political material online has been shown to increase political participation (Kenski &
Stroud, 2006). Web sites of political candidates have been drawing attention since the
1996 presidential election, when they were used primarily as “brochure-ware,” static
environments that duplicated other campaign literature (Tedesco, 2004). Recent online
efforts have become extremely sophisticated with many interactive elements such as polls
and videos, and this interactivity has been linked to increases in political efﬁcacy and
involvement (Hayes et al., 2008; Tedesco, 2007).

In addition to traditional Web sites, social network sites (SNSs) such as Facebook
and MySpace have, for the ﬁrst time in a presidential election, become active forums for
participation in a candidate’s campaign. During the Democratic and Republican
primaries of 2007 and 2008, all of the active candidates had active proﬁles on Facebook
and MySpace in addition to highly interactive Web sites, and the party nominees were
Obviously expending signiﬁcant resources maintaining their online presence. Both

candidates even established their own social networking sites (“McCainSpace” and

“MyBarackObama.com”). This recent surge in online presence begs the questions of
what the target audience thinks of this new way to reach out to young voters and what the
effect on them might be. Could SNS proﬁles of candidates provide the extra information
they need to increase their political conﬁdence and, as a result, vote?
The Social NetworkPhenomena

Deﬁning social network sites as “Web-based services that allow individuals to (1)
construct a public or semi-public proﬁle within a bounded system, (2) articulate a list of
other users with whom they share a connection, and (3) view and traverse their list of
connections and those made by others within the system,” boyd (sic) and Ellison (2007)
explicated the concept and summarized the small amount of research that has been done
in this rapidly emerging and dynamic aﬁeld. Launched in 2003 and 2004, MySpace and
Facebook rank consistently in the top ﬁve most popular Web sites in the United States
(Alexa, 2009). With 125 million and 250 million users respectively, MySpace and
Facebook have become seemingly essential habits and parts of the daily social lives of
America’s young people aged 18-25, who comprise approximately 40% Of the total
number of users for Facebook (cnn.com, 2009; F acebook, 2009; Steinﬁeld, Ellison, &
Lampe, 2008) and 18% of users for MySpace (TechCrunch, 2008). Nearly 80% of
Facebook users are under 34, and 66% of all online 18-25 year-olds have a proﬁle on one
of the two main sites (TechCrunch.com, 2007; Pew, 2009). Users of the sites may use
them to interact with established friends or to meet new people (Ellison, Steinﬁeld, &
Lampe, 2007) and many people use the sites, MySpace in particular, to discover new
music, trends, campaign for an issue, and build social groups. Users are able to discuss

whatever person, band or entity’s proﬁle they are visiting on the “wall” of that proﬁle, an

10

area that allows for real-time posting by “ﬁ'iends” of the proﬁle’s owner. In the past two
years in particular, these sites have been used to form groups around a particular topic,
idea, or person, further expanding the deﬁnition of SNS to allow groups or entities (such
as campaigns) to construct their own proﬁles and connect to others through them.

Wells and Dudash (2007) found that the top two sources of political information
for young people were discussions with family and friends and the Internet, and Pasek et
al. (2006) found that media use, especially the Internet, facilitates civic and political
engagement. The social network environment, which enables discussion and political
interaction with friends and family while online, could be the perfect intersection of these
ﬁndings by providing a setting for promoting engagement with a political campaign and
encouraging dialogue and information-seeking behavior while also, as Vitak et al. (2009)
suggest, making political participation among younger people a more normative behavior.
In addition, these authors argue that SNSs might increase political participation by
lowering the traditional barriers to political communication and participation through
instant connection and sharing with peers.

Political SNS activity received signiﬁcant media and scholarly attention early in
the 2008 election cycle. Preliminary focus group and survey research indicated that they
may not be as valuable as hoped, with many participants deeming the proﬁles “creepy”
(Hayes, 2009) or reacting negatively to the presence of candidates in what they consider
their personal space (Hayes et al., 2008). Other survey research demonstrated their
effectiveness in recruiting votes for a candidate (Williams & Gulati, 2007), and their
appropriateness as a political communication media (Vitak et al., 2009). Perhaps most

promising, Utz (2009), in a large survey and experiment involving Dutch young people,

11

found that SNS were effective at reaching previously uninvolved individuals through
unexpected or casual exposure. This indicates that SNS might be a more powerful tool
for reaching new voters than other online communications methods, which need to be
sought out with intent and are rarely just stumbled across (Utz, 2009).

While the potential of these sites is high, and the early research mostly promising,
what SNS users are actually using political SNS proﬁles and pages for is largely
unknown. In addition, what the impact is of exposure and continued use of candidate
social networking is only conjecture at this time. Like advertising, social network site
proﬁles provide a signiﬁcant amount of cognitive and affective information to users
(Hayes, 2008) suggesting that, as long as they provide adequate political information,
they may be able to afford the same type of informational beneﬁts to users that viewers of
political advertising receive.

RQI .' What political social network behaviors did young people undertake during and
after the 2008 election?

RQ2: Will young people report increased interest in the 2008 election because of social
network sites?

Young People and Media

The media consumption behaviors of young people differ signiﬁcantly from those of
people less than a generation ago; very few young people watch the evening news as their
parents once did, and even fewer read the newspaper. Eighty-seven percent of this
population is online, 58% every day, and when they do gather news for themselves, they
most likely view it on Web-based news sites. However, only 36% of young people say
they follow any sort of news (Pew Internet and American Life Project, 2005; Pasek et al.,

2006). Wells and Dudash (2007), in a large national focus group study, found that the

12

two primary sources of political information for young people are discussion with friends
and family (28.5%) and online sources (15%), with smaller numbers preferring cable
news and newspapers. Delli Carpini (2000), in his analysis of current data on young
people and participation, noted the value of Internet sources of political information to
young people, and further research has played out that proposition with growing evidence
that the Internet can be more successful in building young people’s knowledge of and
engagement in public affairs than many traditional media sources (Jennings & Zeitner,
2003; Lenhart, Madden, & Hitlin, 2005; Levine & Lopez, 2004; Pasek et al., 2006). In
focus groups, members of this demographic expressed a strong desire for more online
interactive functions to learn about campaigns to mirror the highly interactive Web sites
they choose to frequent in their non-political Internet activities. They also desire a
political information source that lacks the negativity so prevalent in other media (Hayes,
2009). SNSs and candidate Web sites are noted for their lack of negativity and high level
of interactivity (Hayes, 2008). This ﬁnding, combined with Wells and Dudash’s (2007)
ﬁndings that family and friends (who are linked together in a SNS environment) and the
Internet are the preferred political information sources for this demographic indicate that
SNSs might be highly valuable political tools. Thus, the possibility that SNSs have
become important sources of information and an avenue to increase engagement for
young voters is very real.

H1 : Young people will indicate a preference for online media over traditional media for
gathering political information

13

Young People and the Political Process

Engagement, Knowledge, and Political Eﬂicacy

Two schools of thought emerge in the political literature relating to political
participation and engagement. The ﬁrst conceptualizes participation as a conduit for
achieving inﬂuence, with Conway (2000) deﬁning political participation as actions that
citizens perform toward inﬂuencing different levels of government, and Best and Krueger
(2005) noting that individuals can take control of policy decisions that may impact them
through active participation in the process. The second school of thought, that political
participation and engagement is more related to speciﬁc activities and learning, such as
volunteering, donating, placing of signs and bumper stickers, etc. has been
conceptualized by Delli Carpini (2004) as being composed of 1) political behaviors such
as voting, communicating with representatives, and volunteering for a political cause or
candidate 2) holding a set of attitudes and beliefs about politics, 3) maintaining informed
positions about current issues, and ﬁnally 4) adherence to Democratic norms and values,
such as political efﬁcacy, civic duty and tolerance (Delli Carpini, 2004). Additionally, in
their deﬁnition of political participation, Kenski and Stroud (2006) include actively trying
to persuade others along with participation in political activities. This second school of
thought, with its emphasis on maintaining informed positions (through knowledge
accrual), persuasion, and social norms such as political efficacy, is particularly relevant to
this work.

Historically, young people aged 18-25 have relatively low rates of political
participation, with only 53% of those eligible to vote actually voting in 2008, a highly

contested presidential election year in which 64% of the US. population overall voted.

14

Among young people, 2008’s turnout reﬂected an increase of only 2% from 2004,
somewhat disappointing given the media and candidate attention this group received in
2008 (CIRCLE, 2008). However, looking back to the presidential election of 2000,
turnout among youth then was only 40% and in recent non-presidential election years
turnout has been closer to 22% (CIRCLE, 2005) compared to rates of 60 - 69% among
the general population (US. Census Bureau, 2004), thus 2008 was a actually a relatively
good year for young voter participation. However, many democratic countries see much
higher participation among young people, so there is still progress to be made. Increasing
participation among this demographic is of vital importance as there is a chance that if
people do not develop political habits by the time they are out of their twenties, they
never will (Delli Carpini, 2000).

There is evidence for strong links between political engagement and political
knowledge (Delli Carpini & Keeter, 1993). Political knowledge develops through the
political learning. Political learning is a participatory process in which “individuals
actively collect, store, modify, interpret, and incorporate new information with what they
already know about the world” (Sotirovic & McLeod, 2004, pp. 358). This information,
about institutions, processes, and the substance of national and state politics can come
from the media or other individuals and provides the base for the well-informed citizenry
that is necessary for any democracy (Delli Carpini & Keeter, 1991).

Scholars have found that over time, political knowledge in the United States has
been eroding somewhat. This ﬁnding is contrary to expectations based on higher
educational attainment by its citizens. Knowledge related to current events is lower than

it has been in the past and this erosion is partially attributed to the decline in political

15

interest observed from the 19405 and 50s to the 19803 (Delli Carpini & Keeter, 1991;
Delli Carpini, 2004) in addition to the decline in civics education in American schools.
Among young people, political knowledge has been reported as particularly low
(Sotirovic & McLeod, 2004), a troubling state of affairs for the future of our democracy,
as efﬁcacy and engagement, and thus, voting, depend on knowledge (Finkel, 1985; Delli
Carpini, 2004; Kaid et al., 2007).

Political knowledge is typically measured as an index; a series of questions
(between three and twelve, usually) relating to national and sometimes international
public affairs is administered. This approach is followed by the National Election Studies
(NES), and has been for years, providing a consistent measure over time Of political
knowledge in the country (though, as Delli Carpini and Keeter [1991] argue, since the
items can vary from year to year, it isn’t truly comparable). Questions usually include
who the vice president is, which party is more conservative, and sometimes who the
prime minister of the United Kingdom is, among others. Delli Carpini and Keeter (1993),
in their highly-detailed review and testing of a variety of knowledge questions, arrive at a
ﬁve-item index that includes party control of the American House of Representatives, the
percentage needed to override a presidential veto, the ideological location of the parties,
who has the power to declare a law constitutional or not, and ﬁnally, identiﬁcation of the
current vice president. Problems with this approach are noted by Sotirovic and McLeod
(2004), who argue for a measure of political knowledge that is not just regurgitation of
facts and requires some indication of processing of the information on the part of the

individual so that they may attach meaning and relevance to elements of information.

16

The mass media have been heavily studied as a contributor to political knowledge,
and, along with discussion with family and peers, is likely the source of most political
knowledge an individual has (Eveland et al., 2005). Every major mass medium ﬁom
newspapers, radio, television (news and entertainment), magazines, and now the Internet,
has been studied for effects on political knowledge. With the exception of entertainment
television, use of each type of medium shows a positive correlation with political
knowledge (Sotirovic & McLeod, 2004). Recent studies have indicated that the Internet
has a powerful potential in informing people, especially young people, who prefer it as an
information source over all others except discussion with friends and family (Wells &
Dudash, 2007). Delli Carpini (2000), in his analysis of existing data on young people and
participation, noted the (at the time) growing potential of the Internet to inform young
people, and ﬁnally, there is growing evidence that the Internet can be more successful in
building young people’s knowledge of and engagement in public affairs than many
traditional media sources (Jennings & Zeitner, 2003; Lenhart, Madden, & Hitlin, 2005;
Levine & Lopez, 2004; Pasek et al., 2006; Diddi & LaRose, 2006).

H2: Social network use for political purposes will be positively associated with political
knowledge

Reasons for Lack of Engagement

While there have been active efforts in recent years to involve younger voters in
political campaigns, such as Rock the Vote, which utilized public service announcements,
Web sites and debate-style programming to involve youth and is credited with producing
a much higher than usual turnout among young voters in 1992 (Tindell & Medhurst,
1998), it seems that traditional political communication methods, such as political

advertising, do not reach out and involve young people or address the issues important to

17

them. In a series of focus groups with Midwestern young people, many individuals noted
their lack of connection with most of the political spots they were seeing and many more
still expressed disdain for both the candidates and political process due to the negativity
so prevalent in current televised political advertising (Hayes, 2009). These traditional
ads are the primary media vehicle for more than 95% of campaign communication
spending, with $600 million spent on traditional TV and radio ads in 2004, and only $29
million spent online (Minnesota Public Radio 2004; Kaye 2006), though the amount
spent online increased signiﬁcantly during the run-up to the 2008 presidential election.
Many factors may contribute to lack of involvement among young voters, but with
televised campaign ads being the predominant form of political discourse most people are
exposed to, it is safe to say that they may be one of the factors.

Historically, there are many reasons young people may choose not to participate,
but in recent research they personally cite a lack of knowledge or information about a
given race or politics in general as a reason for not voting and usually have a lower
political efﬁcacy, the feeling that one can have an impact on the political process, than
the larger US. population (National Association of Secretaries of State, 1998; Kaid et al.,
2007). Political efficacy is a strong determinant of future engagement (Campbell et al.,
1954; Finkel, 1985; Niemi, Craig, & Mattei, 1991), and there is evidence that one’s
political efﬁcacy is formed at a young age (CIRCLE, 2005; Jennings & Niemi, 1981)
Young adults often lack the ability to participate; they are frequently deﬁcient in
knowledge about how the government works and do not know how to register to vote
(Delli Carpini, 2000). This lack of civic ability is tied to three factors: heavy media use

(television in particular) cutting traditional social and civic ties (Putnam, 2000), a serious

18

 

decline of civics education in schools, and a news media and government that does not
pay attention to the demographic (Delli Carpini, 2000). Clearly, the campaigns and
media environment of past years have not been providing, or they have been ignoring, his
demographic the tools and information they need to participate effectively.

This situation may have changed for the better in 2008, with estimates indicating
the highest total turnout ever for young people, even though participation still trailed that

of older Americans (CIRCLE, 2008). Obviously, part of this increased turnout was due F;

vi

k‘. _ . _~.

to a dynamic candidate who appealed to young people, but some of this candidate’s
appeal is attributable to his embrace of media popular with young people, such as
Facebook, MySpace, and his own social networking platform, for online organizing.
These media provided tools to increase involvement and stimulate “supporter
empowerment,” a self-identiﬁed goal of at least one candidate’s personal social
networking site. These sites’ apparent goals were to increase involvement in the process
through voter registration, information availability, interactive tools, and volunteer and
recruiting opportunities, and were not necessarily designed to persuade.

Study Focus

The main focus of this study is on whether political knowledge and efﬁcacy, and
thus engagement, can be increased through use of political social network tools.

A ﬁrst step in measuring any increase is identifying the precursors of knowledge
through development of a new measure of political efﬁcacy and learning called Political
Learning Eﬂicacy. This new measure will be valuable to the scholarly community as a
discipline-spanning measurement instrument and it will help in the identiﬁcation of truly

valuable political information sources. In the next section, contributing concepts and

19

theory will be presented toward development of this new concept; this will help us in our
goal of determining whether social network sites can be an impactful media in political
knowledge and engagement. It is expected that, as Internet use in general has a positive
impact on knowledge, participation and efficacy, that social network use for political
purposes will also have a positive impact.

Identifying any media that could have a positive impact on the electoral
engagement of young people is of vital importance to our democracy for two reasons:
First, scholars have noted the lack of control and efﬁcacy younger people feel, and the
concern is that a whole generation will grow up alienated from our civic institutions
(Delli Carpini, 2000). Second, in the past election, and in many elections, increasing
turnout of dedicated partisans is the goal of campaigns which are dealing with only a very
small number of undecided voters and very slim margins. Thus, the impetus is on
ensuring as many of your partisans get to the polls as possible, as persuading new
partisans is increasingly unlikely as the campaign wears on. Campaigns looking to
recruit new voters should look to the long-ignored demographic of young voters and the
media they frequent. The presidential election of 2008 demonstrated that young voters,
perhaps for the ﬁrst time since the early 19705, can be a powerful force in politics, and
for that to continue, knowledge on how to both foster their civic development and best
connect with them is needed. Considering the established research on the impacts of the
Internet on political variables and the high involvement younger individuals have with
SNS, the following hypotheses are posited:

H3: Use of social network sites for political purposes will be positively associated with
political participation

20

H4: Use of social network sites for political purposes will be positively associated with
political learning eﬂicacy

H5: Use of social network sites for political purposes will be positively associated with
internal political eﬂicacy

H6: Political learning eﬂicacy will be positively associated with political participation

21

CHAPTER 3
THEORY
Political Efﬁcacy

Political efficacy is one of the most inﬂuential and highly researched concepts in
political communication research (Craig et al., 1990; Niemi et al., 1991; Morrell, 2003).
Political efficacy was ﬁrst conceptualized as a uni-dimensional construct, “the feeling
that individual political action does have, or can have, an impact on the political process”
by Campbell et al. (1954, p. 187), and it has been part of the University of Michigan-
conducted American National Election Studies (N BS) in most years ever since. This
dataset, the oldest continuous series of survey data investigating electoral behavior and
attitudes in the United States, has allowed the steep decline in political efﬁcacy
experienced in the 19603 and 703 to be documented (Craig et al., 1990). Political
efficacy is strongly associated with political participation (Campbell et al, 1954; Milbrath
& Goel, 1977; F inkel, 1985; Niemi et al., 1991), and a decline parallel to that in political
efficacy has been observed in political participation over the same time frame.

While this has been a popular measure, the ﬁve items that originally made up the
scale have shown continual reliability and validity problems (Craig et al., 1990; Reef &
Knoke, 1999; Morrell, 2003). Later research revealed that many of the problems lay with
the scale actually measuring two separate dimensions, internal efﬁcacy (IPE), which
refers to a person’s beliefs about their competence to understand and to participate
effectively in politics, and external political efﬁcacy (EPE), which relates to beliefs on
the responsiveness of government to citizen demands (Converse, 1972; Balch, 1974;

Craig et al., 1990). Work in the late-19805 using the NES Pilot Study and the 1988 NES

22

Post-Election Study helped to validate a four-item scale of internal political efﬁcacy
(Craig et al., 1990; Niemi et al. 1991; Morrell, 2003). This scale was found to be robust
with strong reliability, but it has only been used intermittently by the NES (without
explanation), and has not been appropriately used by some researchers who do not follow
the validated wording of items or just make up their own measures without regard for the
validated scale (Morrell, 2003).
Internal Efﬁcacy Scale (Craig et al., 1990; Niemi et al. 1991):
Five Item, Likert-type Measures from Strongly Agree-Disagree Strongly
IPE] : I consider myself to be well qualiﬁed to participate in politics.
IPEZ: I feel that I have a pretty good understanding of the important political issues
facing our country.
IPE3: I feel that I could do as good a job in public oﬂice as most people.
IPE4: I think that I am better informed about politics and government than most people.
Two additional items were considered but rejected from NES for response set
reasons, though they are occasionally used by other researchers (Morrell, 2003), they are:
IPE5: I often don ’tﬁeel sure of myself when talking with other people about politics and
government.
IPE6: Other people seem to have an easier time understanding complicated issues than I
do. '
External political efficacy is not used as frequently in research as internal political
efﬁcacy, and its scale consists of only two items (Balch, 1974):
External Eﬂicaqv Scale
EPEI : People like me don 't have any say about what the government does.
EPE2: I don 't think public oﬂicials care much what people like me think
It is clear from the scale items used to measure these two constructs, and IPE in
particular, that they are measuring an overall, end-state, belief, not the source of that

belief (i.e. adequate knowledge, experience, peer observation, conﬁdence, self-efﬁcacy,

etc.).

23

 

A positive relationship has been found between internal political efficacy and
voting and campaigning (Pollock, 1983; Finkel, 1985), and political knowledge and
political communication usage signiﬁcantly increase internal political efficacy (Kim,
2004). Political participation seems to operate in a manner of reciprocal determination,
both creating and resulting from external political efﬁcacy (Finkel, 1985).

Political Information Efﬁcacy

Closely related to (and using three of four of the same scale measures of) internal
political efﬁcacy, political information efﬁcacy (PIE) measures a “voter’s conﬁdence in
his or her own political knowledge and its sufﬁciency to engage in the political process”
(Kaid et al., 2007, p. 1096). This construct was developed speciﬁcally to study feelings
of lack of information about campaigns and political issues, rather than an efﬁcacy
towards participating, among young voters, who frequently cite lack of knowledge as a
primary reason for not voting (Kaid, Tedesco, & McKinney, 2004; Tedesco, 2007).
Beyond a special issue of the American Behavioral Scientist that included four studies
utilizing PIE, an additional study by the same authors (Kaid & Postelnicu, 2005) and a
pilot study (Hayes, Zube, & Isaacson, 2008) it has not been extensively validated or
tested for reliability beyond the Cronbach’s alpha of +.87 reported by the original authors.
Utz (2009) attempted to use the scale in her study of Dutch youth and political social
networks, but had problems demonstrating more than a main effect of political
information efﬁcacywith her data and dropped analysis of the variable.

Political Information Eﬂicacy Scale (Tedesco, 2007)

(First three identical to internal political efficacy measure)

PIE] : I consider myself well-qualiﬁed to participate in politics

PIEZ: I think I am better informed about politics and government than most people

PIE3: I feel I have a pretty good understanding of the important political issues facing
our country

24

 

PIE4: If a friend asked me about the presidential election, I feel I would have enough
information to help my friend ﬁgure out who to vote for

Voters who demonstrate low political information efﬁcacy are less likely to vote
than those who feel more efficacious (Kaid et al., 2007a). It has also been shown that
viewing debates and paid political advertisements can increase an individual’s political
information efficacy (Kaid et al., 2007b) and that interactive Web-based sources of
political information such as chats, surveys, and games can signiﬁcantly increase political
information efficacy, but static, non-interactive, online political communication such as
brochure-ware type Web sites do not demonstrate the same results (Tedesco, 2007).

Problems with Constructs and Measures

 

While the concepts of internal and external political efﬁcacy are well-validated by
dozens of studies using the variables and the continued use in the NES surveys (Converse,
1972; Balch, 1974; Morrell, 2003), political information efficacy is not, and its utility and
validity as a separate concept is questionable due to the extreme closeness of its
measurement instrument to the measurement instrument of IPE. Political knowledge is
an precursor to increased levels of all three constructs, but all three constructs currently
neglect to measure the political learning one must undertake to become knowledgeable.
Thus, they do not yield a measure that can accurately evaluate the effects of interaction
with political communication tools, especially social network sites. Political uses of SNS
seem to combine all the behaviors known to increase these forms of efﬁcacy, such as
interactivity, political participation, and learning, so their impact is expected to be
signiﬁcant.

RQ3: Is there evidence for political information efficacy existing as a separate construct

or is it an extension of internal political efficacy?

25

 

With these conceptual issues in mind, we will look to Social Cognitive Theory
(Bandura, 1986; 1988; 1989; 1997; 2006) for guidance in both developing a new measure
of political efﬁcacy focused on learning and for building hypotheses related to effects of
political social network sites. Research used to build a measurement tool for the
improved construct and to determine the value of these concepts as separate constructs
will ﬁrst be described, and then survey research used to investigate the effects of previous
and continued political social network use by young people will be detailed.

Applying Social Cognitive Theory

According to Social Cognitive Theory (SCT), an individual is both the product
and the producer of his or her own environment. This theory examines learning in terms
of the interrelationship between behavioral, environmental, and personal factors (Bandura,
1986) and results in a triadic model where each factor inﬂuences the other factors bi-
directionally (Figure l). The environment in which an individual operates inﬂuences

cognitions, which in turn can effect the environment, and so on with each factor.

Personal (Cognitive,
Affective, Biological)

/

Behavior 4

V

Environment

Figure I. The triadic reciprocal causation model of Social Cognitive Theory

Bandura (1986) posits that SCT comprises a complex causal structure that
establishes the development of competency and the regulation of behavior through
knowledge accrual. SCT is noted for its emphasis on an individual learning vicariously

or through personal interactive experience. Given the observational (through status and

26

If‘f—ﬁ ‘—
3 .

news feed updates of ﬁiends) and interactive nature of social network sites, SCT is a
highly relevant theory for the study of the impacts of social networks. In addition, SNS
allows individuals to both produce and react to their online environments, ﬁtting nicely
into the model of SCT.
Self-Eﬂicacy

Along with expected outcomes, the very core of SCT is comprised by the notion
of self-efficacy, a cognitive component that Bandura (1986) describes as "people's
judgments of their capabilities (of which knowledge is only one) to organize and execute
courses of action required to attain designated types of performances" (p. 391). Self-
efﬁcacy has been shown to be a strong indicator of a person’s willingness to engage in a
certain behavior (Bandura, 1997), thus, as in Figure 2, a cognition (feelings of self-
efﬁcacy in this case) is inﬂuencing behavior. A resulting inability to perform said
behavior will further reduce self-efﬁcacy. A person’s environment, which could include
supportive, involved and/or persuasive friends who can serve as a model for the

individual’s behavior, also impacts self-efﬁcacy.

Self-Efﬁcacy

/ \

Behavior ‘ Environment

A
v

Figure 2. Self-efﬁcacy’s effect on both behavior and environment within the model of
SCT

Self-efﬁcacy beliefs derive from four principle sources: enactive mastery

experiences, vicarious (observational) experiences, persuasions and social inﬂuence, and

27

 

physiological and affective state (Bandura, 1997). Enactive mastery experiences, in
which the individual successfully completes a task, can build a strong sense of efﬁcacy.
Other efﬁcacy sources, such as comparison with peers, persuasion, and a positive
affective state, provide weaker levels of self-efﬁcacy. Additionally, a lower-level source
of self-efﬁcacy, such as observational learning, can lead to an action that would result in
a stronger sense of self-efficacy, such as enactive learning (LaRose & Eastin, 2004). It is
not a stretch to imagine a teen unwilling to attempt an intense skateboard trick until he
sees his peer successfully complete said trick and learning from their observations.

It is on these four self-efﬁcacy sources (Figure 3) that we will ﬁrst focus our
discussion of the theoretical development of the concept of political learning efﬁcacy.
Theoretically, PLE is a larger self-efﬁcacy concept contributed to by each of these four
sources, and the end goal of the construct development being a construct that predicts
both political knowledge and participation. The relation to political communication in a
SNS environment will also be noted, and hypotheses related to the theorized relationship

presented.

Self-Efﬁcacy

Enactive Learning

3

Vicarious Learning Affective State Persuasion

Figure 3. Sources of self-efﬁcacy

28

 

Vicarious (Observational) Learning

The vicarious learning function of SCT states that people learn by observing
others and model their behavior based on those expectations (LaRose & Eastin, 2004).
Kaid et al. (2007a) found that while some young people feel they lack political
information, many young people are politically informed and thus those individuals
could serve as models to others. Other young people actively seek out a “smart” friend to
ﬁnd out what happening politically and base their judgments on what that friend thinks
(Hayes, 2009). These model young people could have a positive effect on the political
conﬁdence of their friends through the vicarious learning mechanism.

Both leading social network sites, Facebook and MySpace, have a status update
function where activity is recorded and displayed to friends, and recorded activities can
include political actions such as becoming a supporter of a candidate or a cause, adding a
“bumper sticker” or button, donating to a candidate, and encouraging others to register to
vote. This activity could contribute to vicarious learning on the part of uninvolved
individuals, as in the example of the young people looking to their “smart” ﬁiend for
political guidance, and could encourage the observing individual to investigate the
candidate or issue further, resulting in enactive learning.

Hypotheses in this section are noted by HT, Theory Hypothesis as they are guided
speciﬁcally by Social Cognitive Theory and will be tested by the model of political
learning eﬂicacy.

TH] : Political vicarious learning will be positively associated with political learning

eﬁ‘icacy

29

 

Enactive Learning
While a failed attempt can undermine it, a successful enactive mastery experience
is the most important source of self-efficacy (Bandura, 1994). Individuals are likely to
avoid situations they feel they cannot master and view them as threats, so getting a person
to interact with a task, political participation, for example, in a non-threatening and
familiar environment may have a great impact on political learning and conﬁdence. l
Social network sites provide this familiar and non-threatening environment for

young people and the ability to discuss the task at hand with others in the same situation,

 

which may reduce the anxiety associated with a new task. Also, candidate social network
sites provide a base to begin political exploration and then provide links to other sources
of candidate information, such as the ofﬁcial candidate Web site and civic Web sites such
as Rock the Vote.
TH2: Political enactive learning will be positively associated with political learning
eﬂicacy
A ﬂective State

An individual’s belief in their ability to cope affects how much stress and
depression they experience in new or threatening situations, as well as their level of
motivation. Individuals who believe they will not be able to deal with threats experience
high anxiety arousal and dwell on their coping deﬁciencies, viewing the situational
environment as fraught with danger (Bandura, 1994).

Young people encountered mixed messages about registering to vote and voting
in the 2008 election; strong support and encouragement from a variety of sources to

register to vote, and a media and some nefarious political groups that tried to reinforce

30

the idea that the voting process is difﬁcult, time-consuming, and potentially threatening.
These mixed messages combined with the anxiety normally associated with performing a
complex behavior (registering and voting in this case) for what was likely the first time,
may have produced a troubled affective state in some individuals. Bandura (1994) noted
these individuals may “magnify the severity of possible threats and worry about things
that rarely happen” (p. 75). This worry could potentially reduce participation and turnout
unless it is addressed.

Candidate social networks have the opportunity to address this problem through
both information dissemination in a familiar and accessible environment and the ability to
link users to organizations that can provide accurate state-by-state information on poll
locations and assistance.

TH3: Political aﬁfective state will be positively associated with political learning eﬁicacy
Persuasion

Social persuasion is the ﬁnal way of strengthening a person’s belief that they can
succeed at a given task. Individuals who are persuaded verbally that they possess the
ability to master a task are more likely to participate in that task and are much less likely
to focus on personal deﬁciencies when something goes wrong (Bandura, 1994).
Persuasion boosts perceived self-efﬁcacy as well, as it may lead people to try harder to
succeed than they normally would, thus promoting the skill development that leads to
increased self-efﬁcacy (Bandura, 1994).

A frequent encounter during a political season is the activist who seeks to involve
other people. This may include organizations such as Rock the Vote, or it may be a friend

encouraging another ﬁ'iend to ﬁnd out more on an issue. As with vicarious learning,

31

 

contact with politically involved people may increase efficacy, but this time through
another mechanism, persuasion.

Social networks have included many opportunities for individuals to encourage
their ﬁiends to become involved in various things, including politics. You may invite
ﬁiends to join you in a group, become a supporter of an issue or candidate, or to get
together to discuss something. This ﬁmction of social networks may contribute to the

persuasion factor in political conﬁdence building. 1

 

T H4 .' Political persuasion attempts will be positively associated with political learning i ;
eﬁ‘icacy
Politicaﬂearning Efﬁcacy (Tentative Construct Deﬁnition)
The research related to internal political efﬁcacy, political information efﬁcacy
and self-efﬁcacy indicates that knowledge, leaming, or experience is key to efﬁcacy
gains. How is this accomplished, though? Through a combination of persuasion to seek
out information, modeling of politically informed individuals, eitherlow anxiety about an
individual’s ability to manage a new situation or subsequent information searches to cope

with this anxiety, and ﬁnally enactive experiences with the sources of information, just as

 

SCT predicts (Figure 4).
Personal
Political Learning Eﬂicacy
Behavior 4 t Environment
Enactive Political Learning Vicarious and Persuasive Factors

Figure 4. The implementation of sources of self-efﬁcacy and political learning efficacy
into the triadic model of SCT.

32

Political learning efﬁcacy is a person’s conﬁdence that they can seek out, obtain,
and process information relevant to a political decision. This construct deﬁnition differs
from internal political efﬁcacy (“a person’s beliefs about their competence to understand
and to participate effectively in politics”) and political information efﬁcacy (“3 voter’s
conﬁdence in his or her own political knowledge and its sufﬁciency to engage in the
political process”) in its focus on the probable causal mechanism of political efﬁcacy,
learning, and on the information accrual needed in order to make a decision and act on it. i”
This construct is likely an antecedent to internal political efﬁcacy as demonstrated by the
strong association between internal political efﬁcacy and knowledge (Kim, 2004), thus a
strong positive association between knowledge and PLE is expected. As knowledge and
the previous efﬁcacy constructs have been positively related to political participation, it is
also expected that increased political learning efficacy will lead to increased participation.
RQ4: Is there evidence for political learning efficacy existing as a distinct construct?
T H5 : Political learning eﬁ‘icacy will be positively associated with internal political
efficacy
T H6: Political learning eﬂicacy will be positively associated with political knowledge

TH 7: Political learning eﬁicacy will be positively associated with political participation

Vicarious

Learning

Political . . Internal , , , , ,

. Learning _, P011tlcal _> Political _> Political Particrpatron
Enactive <—> Eﬂicacy Knowledge Efﬁcacy

Learning

Aﬂ‘ective /

State

Persuasion

Figure 5. Hypothesized relation of PLE, its sources, and its dependent variables

33

CHAPTER 4
METHOD
Multi-Method Scale Development

The following research was completed in order to: 1. Develop a measurement tool
for the new construct of political learning efﬁcacy and for its sources, enactive and
vicarious learning, persuasion, and affective state, 2. Conﬁrm the theoretical propositions
and research questions set forth, and 3. Investigate the impacts of political social network
use on political learning efficacy and participation.

A multi-method study ﬁrst involving qualitative methods in the form of focus
groups (n=3 6) to further deﬁne the construct and inform the development of scale items
for both the construct and its sources (enactive and vicarious learning, persuasion and
affective state), followed by a pilot test (n=358) to evaluate the proposed scale items, was
undertaken. Finally, utilizing the new construct and its measure, a nationally
representative survey (n=625) of American 18-25 year-olds was completed to evaluate
the effects of prior and continued use of political social network sites. From this research
two highly valuable outcomes emerge: A new discipline-spanning construct and scale for
the measurement of political learning efﬁcacy and its sources, and an evaluation of the
impact of political social networks on young voters in terms Of efﬁcacy outcomes and
participation. Each stage in the research will be detailed separately here.

Mative Research to Inform Construct and Scale Development

A recognized weakness with current scale development practice is the lack of

attention given to the development of accurate construct deﬁnitions (Rossiter, 2002;

MacKenzie, 2003 Mowen & Voss, 2008), and that may be the root of the problem that is

34

seen with the overlapping constructs of internal political efﬁcacy and political
information efﬁcacy. MacKenzie (2003) argued that a good construct deﬁnition should
specify the construct’s conceptual theme in unequivocal terms so it is clearly
distinguished from other constructs. From the initial qualitative research completed here,
a precise construct deﬁnition of the new concept is developed. “Without adequate
construct deﬁnitions, it is impossible to adequately specify relationships among the l
constructs. Moreover, a precise deﬁnition provides clear guidance for the selection of i1
items and speciﬁes how the items relate to the construct” (Mowen & Voss 2008, p. 488). I
Padgett (1998) discusses a multi-method qualitative technique to inform
quantitative efforts in developing scales. Utilizing qualitative methods ﬁrst to explore
constructs, develop hypotheses, and then inform scale development can be of particular
value as it will elicit responses and data not available in the more common forms of scale
item development such as discussion with experts and theory-based item development
(which were also both completed for this study). Validity of constructs and scales in
quantitative research can be improved through the involvement of real-life situations and
observations through open-ended surveying, interviews, or focus groups with members of
the target population (Rowan & Wulff, 2007). By conducting qualitative research prior
to larger quantitative studies, fundamental insights from participants in speciﬁc
social/behavioral circumstances (e.g., young people making the decision to participate in
politics or not) can enhance the quality of the research (Rowan & Wulff, 2007). Used to
both improve the construct deﬁnition of PLE and to develop the scale items to measure it,
this initial qualitative research will help build a stronger and more valid construct and

measurement tool.

35

Focus Groups

Focus groups are a well-established research method for examining both political
communication and political constructs. While focus groups are an established research
method in social science, there are some signiﬁcant drawbacks, notably that of external
validity. Validity is addressed, and achieved to the best of the method’s ability, through
achievement of saturation of responses across groups, or simply by repetition of ﬁndings
across multiple groups (Krueger & Casey, 2000; Kern & Just, 1995). The minimum
number of focus groups given varies by source, but most state either two or three groups
as a minimum for valid results.

Four focus groups of one hour each and consisting of 5-11 participants (n=3 6)
were conducted with members of the undergraduate student population at Michigan State
University. The average age of the participants was 21.9 with a range from 20-24 years
old. Twenty participants were female (55.6%) and 16 male (44.4%), and they received
extra credit for participation. While this is obviously a convenience sample, a student
sample is well-justiﬁed to address the initial goals of this study, which speciﬁcally
concern how young people, 67% of whom are enrolled in colleges throughout the nation
(U .8. Census, 2006), gain political efﬁcacy through learning. A moderator and one
assistant taking notes were present for each group. Each session was recorded and later
transcribed by the assistant.

The type of focus group undertaken for this study is similar to a depth interview
where the individual opinions or statements are the unit of analysis. While the dynamics
of group discussion have an impact on the opinions expressed in any focus group,

questions were worded way to directly access the opinions of each individual

36

participating. The questions asked were grounded in Social Cognitive Theory,
speciﬁcally related to the four sources of self-efﬁcacy and how these might help an
individual learn during a political campaign. For example, a question related to vicarious
learning might ask if the participant had observed their friends on Facebook (through the
F acebook Feed function) interacting with a political campaign. An afﬁrmative answer to
that question would receive a follow-up question based on enactive learning, in which the
participant would be asked if they clicked on the link provided by the Feed and checked
out the candidate or cause. Other questions included general political warm-up questions
and inquiries into social networks and how they might have impacted participants’
political participation (Appendix A).
Data Analysis

Focus groups were recorded and transcribed, and transcriptions were the source of
ideas for both reﬁnement of the construct deﬁnition and for proposed scale items.
Focus Group Results

In a heartening ﬁnding, compared with results from a focus group study of similar
size and demographic proﬁle conducted in early 2007, politics and political participation
were viewed in a much better light in early 2009. In the 2007 data, nearly 85% of
participants reacted with negative words when asked to play word-association with the
word “politics” (Hayes, 2009); in this sample, most reacted with neutral (“parties,”
“election,” “voting,” etc.) or positive (“cheering,” “hope,” etc.) words. A very few gave
negative responses, such as “controversy” and “too many ads.” As noted, this is a
distinct shift from data taken only two years prior when negativity towards the political

process dominated. Participants also indicated that they felt more involved than they did

37

in the previous presidential election in 2004, this may have been because some of them
were not eligible to vote at the time, but others indicated that they had been turned off by
the negativity in 2004, or had only been told to vote by their parents or teachers, so that
they felt no real connection to either candidate.

Sources of Self-Eﬂicacy

Examining answers to questions related to the hypothesized sources of political
learning efﬁcacy yielded interesting results. In terms of persuasion, participants noted
that they had encountered multiple friends and acquaintances that both were encouraging
them not only to vote, but to ﬁnd out more about a candidate or cause through linking of
articles, Web pages, and social network proﬁles. Some participants noted that these
attempts sometimes got annoying, especially if the friend was particularly persistent.

Annoyance was also caused by persuasion attempts by the actual campaigns, with an
overwhehning number of emails or contacts coming after one signed up for updates
online.

“I was actually persuaded to join the oﬂicial Barack Obama thing (Web site contact list),
and I ended up quitting it quickly because I got so many messages saying ‘go out and tell
your ﬁ'iends ’, we ’re doing this, we’re doing that, and got annoying. ” — Male, 22

Few participants noted that on this particular campus it was unacceptable to be a
supporter of anyone but the most popular candidate, and when they expressed support for

the other candidate, it meant they were continually bombarded with negative information

about him and probing questions about why they would not support the popular candidate.

It was clear to many that discourse favoring the other candidate was not acceptable.

38

Experiences that this group had with enactive mastery experiences (learning by
trying things out) were varied. Some noted that their ﬁrst awareness of then-Senator
Obama was likely online, and many detailed how their ﬁrst exposure to him was not in a
campaign-generated or SNS communication, but in a fan-produced viral video on video
sharing site YouTube (“Crush on Obama,” the Obama Girl video) Every participant in
the two later focus groups had seen this video (the ﬁrst two groups did not bring it up),
and saw it early in the campaign. Some noted that viewing this video led to Web

searches for the candidate to see who he was (when this video was released, in early 2007,

 

Senator Obama had a much smaller name-recognition nationally than primary opponent
Senator Clinton).

While a few participants mentioned that the ﬁrst time they had interacted with
anything related to the campaign was on Facebook, it was not the norm. Multiple
participants did note that once they found an online source they liked and trusted, they
would rely on that site and visit it repeatedly and occasionally follow their links. In this
discussion, worries of trustworthiness and accurateness came up, with a few participants
noting the perceived bias of many news sites online, and how they sometimes do not
know which information to trust, especially when blogs enter the picture. A very few
remarked that they were turned off from Web-based news entirely by this perceived bias.
(When asked if they went online to search for information) “Yeah, well, honestly, not too
much because I didn ’t know which sites were going to be factual and what sites I could
actually trust” — Female, 21

While some were turned off by this perceived bias, more individuals noted the

value of a variety of opinions they could ﬁnd by bouncing around to different sites.

39

“I liked to do them all because everyone has their own opinions so it ’s interesting to ﬂip
ﬁ'om CNN to MSNBC to even the BBC or the smaller sites. It was cool because everyone
has their own way of broadcasting their own opinions. I wanted to be educated as I could
be on every side ” — Female, 21

“I just went to the most extreme sites ﬁ'om the mass media like F ox and then like CNN
and thought it would be nice to see both sides and also went to Ralph Nader ’3 site and
Bob Bar ’s just to see their opinions...I mean yeah, they were negative towards the
Republicans and Democrats... its useﬁrl to see it in the same way towards both parties” —
Male, 22

The vicarious learning function of social cognitive theory was in evidence
through both online and offline activities. Many participants remarked on the activity
they observed on F acebook and to a lesser extent, MySpace, during the later stages of the
campaign (the campus where this research was conducted is a “Facebook campus” with

most students choosing Facebook as their social media tool). News Feed updates (A
response to the question “What’s on your mind?” that appears on the top of proﬁles and
then appears on your ﬁiends’ feed) that included support for a particular candidate were
the most observed political activity, but participants also noted that they saw their ﬁiends
joining groups and that they had been the recipients of SNS “bumper stickers” and
campaign buttons online.

Offline, participants mentioned that they saw some of their friends volunteering
for the campaigns, and knew that those friends would be good sources of information
about that particular candidate while realizing they might not be a great source of

unbiased political information. Some participants lamented the fact that they were seeing

40

people who could not explain their reasons for voting for a particular candidate when
asked — it was just the cool thing to do for some people, and they had no information
behind their decisions.

Negative affective state impacting information seeking ability or participation was
not an issue with this group. None said they had been nervous about where or how to
vote, and that they had known how to check to see whether they were voting in the
correct precinct or not. Many of them either voted absentee or skipped class and went [1
home to their parents’ residence to vote. The most anxiousness actually seemed to be j
the previously mentioned problem related to ﬁnding accurate or trustworthy information
online, and a discussion of whether there would be a riot if Obama didn’t win. None
mentioned posting in discussion forums or having online conversations about the
campaign, but the reasons why not were not explored. In the future, it would be
interesting to examine whether any lack of posting is because of a fear of looking dumb
or facing criticism, an affective problem.

The most frequently mentioned information sources for this group were parents
and friends, online sources, and television news (both cable and broadcast), in line with
previous ﬁndings about media preferences for this group (Wells & Dudash, 2007). None
mentioned the newspaper, other than a brief mention of the campus paper. Very few
participants said they didn’t have enough information to make a decision, and nearly
every focus group member reported voting (a result that may be slightly inaccurate due to
social desirability), indicating that we had no individuals participating who experienced
the problems observed in earlier election cycles where lack of information was reported

as a reason for not voting (Kaid et al., 2007a). The one individual who admitted not

41

voting was concerned that neither candidate had really addressed the main issues he was
concerned with (though he did not state what they were). The strongest information
ﬁnding was the oft-mentioned angst about ﬁnding accurate or trustworthy information
online or on television, a signiﬁcant contributor to their affective state regarding
information surrounding the election. Quite a few students said most of their political
information came from their parents or ﬁiends and that they chose to ignore the news
media and its perceived biases (CNN, MSNBC and Fox News were all mentioned).
Discussion

From these focus groups, a few distinct trends emerged that relate to SCT. First,
young people perceive a problem with either accuracy or trustworthiness of information
presented by a variety of news sources and that impacts their information gathering
ability. In this same vein, some mentioned that once they had found information sources
they perceived as trustworthy, they visited that source repeatedly. These trusted (higher
status) sites are acting as role models for information. Second, persuasion by friends and
the campaigns can go too far, to the point where annoyance (and possible tuning-out)
kicks in. This persuasion could be both toward one particular candidate or, as SCT posits,
persuasion toward a person’s ability to complete a certain task, voting in this case. Third,
this group perceived the bombardment of information they were receiving in the 2008
election as a bit much and it contributed to a negative affective state regarding the
election, and some were happy when it was ﬁnally over. There was no perceived dearth
of information as has been reported by members of this demographic in previous
elections (Kaid et al., 2007a). And ﬁnally, fourth, many did think that SNS were an

interesting media to learn about the election from and used them to learn vicariously

42

about their friend’s political happenings. They had noticed their ﬁiends interacting with
candidates and causes online, and some had participated themselves in SNS groups
centered on election issues. This is a big change from focus group research from earlier
in the election cycle which found members of this demographic ambivalent or even
hostile to the presence of candidates on “their” turf. This illustrates the progression of
SNS political proﬁles from a perceived effort by candidates to try to look cool to just
another political communications tool (Hayes, 2009).

These focus group results will help form the basis, along with SCT, for scale
items to be proposed, tested, and validated in the next section. The results of the focus
group do not change the construct deﬁnition of political learning efﬁcacy as presented
earlier, but they do contribute valuable information toward the development of scale
items.

Development, Reﬁnemen_t_ and Vaﬂation of Construct and Measurement

Scale item development was guided by theory and focus group discussions. The
scale development procedure follows recommendations by DeVellis (2003) for clarity,
redundancy, uniform level of speciﬁcity, and length, and by Bandura (2006) for valid
self-efﬁcacy scales. The resulting 34 items are ll-point strongly-agree to strongly-
disagree scale questions and include 17 that measure the main construct of political
learning efﬁcacy and 17 that access the sources of self-efﬁcacy: vicarious and enactive
learning, persuasion, and affective state.

Bandura (2006) recommends as many response options as possible to make for a
more sensitive scale, with his strongest recommendation being for a 100—point scale.

This 100-point scale was not possible with the survey software available, so the 11-point

43

scale with a neutral middle point was settled upon. Bandura (2006) also recommends
response options different than Likert’s (1961) strongly agree — strongly disagree, and
suggests those that represent levels of certainty, that is, how conﬁdent the respondent is
that she can complete some task. As Bandura’s suggested wording is not consistent with
the political efﬁcacy items under comparison, which are typically measured with Likert-
type scales, wording for the new items included both conﬁdence and “I can” (as a
measure of capability and recommended by Bandura) in the text of the item rather than
the response. To ﬁrrther develop validity of items, they were discussed with a resident
expert on Social Cognitive Theory (R. LaRose, personal communication, March, 2009).
Main Construct and Proposed Scale Items to Measure

Political learning efﬁcacy (PLE) is a person’s conﬁdence that they can seek out,
obtain, and process information relevant to a political decision, and it is signiﬁcantly
contributed to by the four known sources of self-efﬁcacy, vicarious and enactive learning,
persuasion, and affective state. Differing from internal political efﬁcacy (“a person’s
beliefs about their competence to understand and to participate effectively in politics”)
and political information efﬁcacy (“a voter’s conﬁdence in his or her own political
knowledge and its sufficiency to engage in the political process”), PLE’s focus is on the
probable causal mechanism of political efficacy, learning, and on the information accrual
needed in order to make a decision and act on it. The following scale items for both the
main construct and its sources were constructed based on the construct deﬁnition, social
cognitive theory and a discussion with an expert in it, and the focus groups. Some media-

speciﬁc items were included to see if there was a difference in the tested group’s ability

44

to gather information between media and if different media contributed disproportionally
to PLE.
Proposed items to measure the overall construct Of PLE:

PLE] .' I am conﬁdent that I can learn about political issues if I want to

PLEZ: I am conﬁdent that I know where to go to locate good political information
PLE3: I am conﬁdent that I can seek out information related to a political decision in
general

PLE4: I am conﬁdent I can ﬁnd information related to a political decision online
PLE5: I am confident I can seek out information related to a political decision from my
ﬁiends

PLE6: I am conﬁdent I can seek out information related to a political decision from my
family

PLE 7: I am conﬁdent I can seek out information related to a political decision in the
newspaper

PLE8: I am conﬁdent I can seek out information related to a political decision on T. V.
PLE9: I am conﬁdent that I can understand most political information in general
PLEI 0: I am conﬁdent that I can understand most political information online

PLE] l : I am conﬁdent that I can understand most political information in the newspaper
PLE12: I am conﬁdent that I can understand most political information on T. V.
PLEI3: I am conﬁdent that I can ﬁnd trustworthy political information

PLE] 4: I am conﬁdent that I can ﬁnd accurate political information

PLEI 6: I am conﬁdent that I can ﬁnd unbiased political information

PLE] 7: I am conﬁdent that I could ask an intelligent question during an election

Sources of Self-Eﬂicacy: Vicarious and Enactive Learning, Persuasion and Affective
State

Vicarious learning is completed through the observation of peers and the
modeling of behavior based on the actions of peers (LaRose & Eastin, 2004). Political
behavior, including information seeking behavior, became much more visible to young
people in the 2008 election because of social network participation and the status updates
and news feeds that come with it.

Proposed items to measure vicarious learning:
Vic]: I have friends who know where to find good political information

Vic2: I have ﬁ'iends who know where to ﬁnd good political information online
Vic3: I have seen how my ﬁiends can ﬁnd political information online

45

A successful enactive mastery experience is the most important source of self-
efﬁcacy (Bandura, 1994), and experimental interactions with a task, political learning in
this instance, can encourage future interactions with it.

Proposed items to measure enactive learning:

EnaLI: If I don ’t ﬁnd the political information I am looking for in one source, I try to
ﬁnd another source

EnaL2: If a political information source does not have all the information I need, I will
visit their recommended sources of more information

EnaL3: If I like a source of information I am likely to come back to it repeatedly to see if
there are updates

EnaL4: I am conﬁdent I can learn to use a variety of political information sources

Social persuasion can strengthen a person’s belief that they can succeed at a given
task. Individuals who are persuaded verbally that they possess the ability to master a task
are more likely to participate in that task and are much less likely to focus on personal
deﬁciencies when something goes wrong (Bandura, 1994). Having friends who
encourage political knowledge building and participation could both serve as a model for
enactive learning and encourage and convince individuals that they are capable of
effective participation.

Proposed items to measure persuasion:
Perl : My friends encourage me to ﬁnd political information
Per2: My ﬁiends tell me that I should learn about the issues before I make a political
decision
Per3: I have been persuaded by ﬁiends or family to ﬁnd political information online
Per4: I am likely to read political information sent to me by a friend
Per5: I am likely to read political information sent to me by a family member
Per6: I am likely to check out apolitical candidate recommended by a ﬁiend
An individual’s belief in his or her ability to deal with stress effects their reactions

in new situations as well as their level of motivation to try new things. Individuals who

believe they will not be able to deal with new situations (perceived as threats) experience

46

 

high anxiety arousal and dwell on their coping deﬁciencies, viewing the situational
environment as fraught with danger (Bandura, 1994). These individuals are experiencing
a depressed affective state and are less likely to try new things (like participating
politically) as a result.
Proposed items to measure affective state:
Aﬂl : I am not apprehensive about ﬁnding the political information I need
Aﬂ2: Political information is somewhat intimidating to me (reverse coded)
A ﬂ3 : When I think about looking for political information, I get nervous (reverse coded)
Aﬂ4: When I am looking for political information, I get nervous (reverse coded)
AﬁS: I am conﬁdent that I could overcome my anxiety about getting the information I
needed to make a political decision
Pilot Test

Following the methodology of Craig et al. (1990) and Niemi et al. (1991)
established in the testing and validation of internal political efﬁcacy and political trust for
the National Election Studies Pilot Study, a pilot test with the developed scale items was
conducted to help determine the best items for the ﬁnal scales and to test the theory
hypotheses (HT s). Test construction following this method, starting with a theoretically
derived set of items and/or items derived from qualitative methods and then progressing
to a pilot test with members of the survey instrument’s target population to identify good
items, is a commonly recommended scale construction practice (DeVellis, 2003,; .Spector,
1992)

A convenience sample (n =358) of Midwestern college students using
Zoomerang’s online survey software was conducted in March 2009. While not ideal, a
large convenience sample is typically considered sufﬁcient for a pilot test as long as the

sample is representative of the population for which the scale is intended (DeVellis,

2003), and, as with the focus groups, a student sample is well-justiﬁed to address the

47

 

initial goals of this study, which speciﬁcally concern how young people, 67% of whom
are enrolled in colleges throughout the nation (U .S. Census, 2006), gain political efﬁcacy.
Nunnally (1978) suggests that a pilot test involving a sample greater than 300 is likely
sufﬁcient and other sources note that there should be between ﬁve and ten times the
participants as proposed scale items (Delli Carpini & Keeter, 1994). In either case, a
sample of size of 358 should be considered adequate for these pilot testing purposes.

In addition to the new PLE items detailed above, the pilot test included the }
established measures of internal political efﬁcacy, external political efﬁcacy, and political
information efﬁcacy in order to discriminate the new measure from these concepts and
answer the research question of whether political information efficacy exists as a separate
construct from internal political efﬁcacy. To allow for accurate comparison among the
scales, the ll-point response option with neutral middle point determined as best for the
new PLE measure and the available survey software, was used for all efﬁcacy scales.
Established measures of political knowledge (Delli Carpini & Keeter, 1994), opinion
leadership (Shah & Scheufele, 2006), political cynicism (Austin & Pinkleton, 1995), and
political trust (Miller, 1974), were also included as they are all concepts related to the
political efﬁcacy constructs. The inclusion of these measures will help determine validity
through examination of expected correlations and follows the recommendations of
Loevinger (1957), who implored those developing scales to “sample all possible contents
which might comprise the putative trait according to all known alternative theories of that
trait” (p. 659) (Appendix B). The inclusion of political behavior measures will also

contribute to the development of the model political learning efﬁcacy and its effects.

48

Data Analysis

The convenience sample (n=358) was disproportionately female (70.9%) and the
mean age of participants was 20.8. While disproportionately female, the sample was
ethnically representative of many Midwestern college campuses, comprised of 84.6%
Caucasian, 6.7% Asian, 5.0% African American, and 3.1% Hispanic participants.

Exploratory factor analysis and measures of internal consistency were used to
identify distinct measures, and correlations used to help validate the scales. Used to
explore data when the nature or number of factors that account for covariation between
variables is unknown, principle components analysis and exploratory factor analysis are
noted for their helpfulness in theory or measure building exercises (Stevens, 1996) and
they are considered an essential part of any scale development exercise involving a large
sample (DeVellis, 2003). This procedure allows us to see how the different variables
cluster into single factors, an indication that each item in the factor is measuring an
overall concept. Sufﬁcient factor loadings are somewhat subjective, but acceptable factor
loadings are typically noted as +0.65 for an item to be accepted as a part of that factor.
Factors with eigenvalues greater than 1.0 were included as the ﬁnal factor solution
(DeVellis, 2003). Finally, a common measure of internal consistency, Cronbach’s alpha,
is used to evaluate the quality and reliability of the scales, as it measures the extent to
which items share a commonalty of participant response, i.e., correlate highly with each
other, indicating a similar concept is being accessed by each item. Internal consistency of

each scale is evaluated with alphas of +0.70 considered acceptable (Nunnally, 1978)

49

Establishing Validity

This exploratory factor analysis assists with establishing discriminant validity of
the new measure. Clark and Watson (1995) note that a well-designed factor analysis can
play an integral role in enhancing the discriminant validity of a new scale, suggesting that
new and old measures of a construct or similar constructs be subjected. to a joint factor
analysis.

As discussed previously, IPE and PIE are very closely related measures, and it is
expected that they will load on the same factor. A high correlation between PLE, PIE
and IPE is also expected, as it is theorized that the new measure is an antecedent to IPE.
Similarly, based on previous research we would expect relationships between PLE and
measures of political knowledge (Delli Carpini & Keeter, 1994), political trust (Miller,
1974), political cynicism (Pinkleton, Um, & Austin, 2002), and opinion leadership (Shah
& Scheufele, 2006). These measures should load on different factors but the scales of
each should be correlated with IPE and PIE, either positively (knowledge, trust, and
leadership) or negatively (cynicism, which has been shown to have a direct negative
relationship to efﬁcacy (Austin & Pinkleton, 1995). The correlations establish construct
validity, the extent to which a measure performs as it should in relation to other measures
(DeVellis, 2003), as we anticipate the noted relationships.

It is expected that a number of the pilot-tested scale items will be discarded, and
that utilization of the exploratory procedures described will help reduce the proposed
measurement instrument of PLE to 4-7 items and the measurement of its sources

(vicarious and enactive learning, persuasion and affective state) to 2-4 items each.

50

Scale Development and Validafion Results

Political Learning Ejﬁcacy

Exploratory factor analysis (principle components analysis with varimax rotation)
revealed four distinct factors (eigenvalues > 1.0) from among the 16 items posed (Table
1). Factor I contained the ﬁrst four items, and included one item (PLE4) related to online
media, while all other items involving a speciﬁc media (PLE5-8 and PLE 10-12) held
together on two separate factors, indicating that media use may be determining where
these items load. All items related to the trustworthiness of media loaded on a ﬁnal
factor. All items performed well, loading strongly on one factor, with the exception of

PLE17 which performed poorly and loaded loosely across three factors.

51

a
Table 1: Factor Analysis of Proposed PLE Items

 

l

2

Factor

3

 

PLEl: I am conﬁdent that I can learn
about political issues ifl want to
PLE2: I am conﬁdent that 1 know
where to go to locate good political
information .

PLE3: I am conﬁdent that 1 can seek
out information related to a political
decision in general

PLE4: I am conﬁdent I can ﬁnd
information related to a political
decision online

PLES: 1 am conﬁdent I can seek out
information related to a political
decision from my friends

PLE6: I am conﬁdent I can seek out
information related to a political
decision from my family

PLE7: 1 am conﬁdent I can seek out
information related to a political
decision in the newspaper

PLE8: I am conﬁdent I can seek out
information related to a political
decision on T.V.

PLE9: I am conﬁdent that I can
understand most political
information in general

PLEIO: I am conﬁdent that 1 can
understand most political
information online

PLEI l: 1 am conﬁdent that 1 can
understand most political
information in the newspaper
PLE12: 1 am conﬁdent that I can
understand most political
information on T.V.

PLE13: 1 am conﬁdent that I can ﬁnd
trustworthy political information
PLE14: I am conﬁdent that I can ﬁnd
accurate political information
PLE16: I am conﬁdent that I can ﬁnd
unbiased political information
PLEl7: I am conﬁdent that I could
ask an intelligent question during an
election

Cumulative variance explained %

.815

.788

.817

.812

.583

42.34

.763

.710

.832

.715

.291

56.66

.609

.644

.789

.823

.100

62.24

.761

.814

.778

.281

69.73

 

‘3 . . . .
Varimax rotation With E rgenvalues > 1 specrﬁed, four factors extracted

explaining 69. 7% of the variance.

52

This breakup among factors of media-speciﬁc items precipitated a rethinking of
whether this scale should include media-speciﬁc items. Both theoretically and practically,
it made more sense to develop a scale that could be used in a variety of settings, not just
in media-effects research. Thus, a second factor analysis was completed, this time with
the media-speciﬁc items removed. In this case, two distinct factors emerge: one closely
related to the PLE construct deﬁnition as proposed, and one related to accuracy and non-

bias of the media (Table 2).

a
Table 2: Factor Analysis of Non-Media PLE Items

 

 

Component

1 2
PLEl: 1 am conﬁdent that 1 can learn about .823
political issues if 1 want to
PLE2: I am conﬁdent that I know where to go to .800
locate good political information
PLE3: I am conﬁdent that 1 can seek out .866
information related to a political decision in
general
PLE9: I am conﬁdent that I can understand most .64]
political information in general
PLE13: I am conﬁdent that 1 can ﬁnd trustworthy .434 .766
political information
PLE14: I am conﬁdent that 1 can ﬁnd accurate .371 .840
political information
PLEl6: 1 am conﬁdent that I can ﬁnd unbiased .866
political information
Cumulative variance explained % 56.74 73.94

 

a
Varimax rotation with Eigenvalues >1 specified two factors extracted
explaining 72.9% of the variance.

While the second factor is interesting and issues of accuracy and bias were
certainly important to the focus group participants who brought it up, it does not ﬁt well
enough to be included currently. Therefore, a scale composed of PLE 1, 2, 3, and 9

emerges.

53

When constructed, the new PLE scale shows strong reliability with a Cronbach’s
alpha of .840 and strong correlations between the items (Table 3). PLE9 is the weakest
item and does not seem to contribute as much to alpha, but its value as a theoretical
contributor to the construct deﬁnition of PLE (as the “understand” component) is high.

Thus, it will remain in the scale and its value will be reassessed in the next stage Of the

 

 

study.
a
Table 3: Correlations and Scale Reliability for PLE Items 1‘
PLEl PLEZ PLE3 PLE9 l

PLEI: I am conﬁdent that I can learn 1.000 .568 .612 .463 1
about political issues if 1 want to '
PLE12: I am conﬁdent that I know 1.000 .748 .505 '
where to go to locate good political
information
PLE3: 1 am conﬁdent that I can seek 1.000 .513

out information related to a political

decision in general

PLE9: 1 am conﬁdent that I can 1.000
understand most political

information in general

Item-total correlations .641 .730 .763 .566
Alpha if item deleted .808 .766 .754 .840
Alpha (for PLE scale) .340
Average correlation , 533

 

a All ﬁgures are Pearson ’s r (p >. 05) unless otherwise noted

As constructed, the PLE scale correlates in the expected direction and
signiﬁcantly with opinion leadership, political knowledge, and internal political efﬁcacy,
providing support to THs 5 & 6 There was no signiﬁcant relationship between PLE and
political trust or cynicism (as these two measures are considered opposites of each other
and have a strong negative correlation here, it is not unexpected that if one shows no

signiﬁcant correlation, the other also will not) (Table 4).

54

Table 4: Correlations Between PLE Scale and Related Variables

 

Political Opinion Political Political
PLE Cynicism Leadership Trust Knowledge IPE

 

O. O. O.

PLE Pearson Correlation l .042 .341 -.058 .367 .510
Political Pearson Correlation 1 .l 17‘ -.632“ .120‘ .121‘
Cynicism

Opinion Pearson Correlation l -. 144” .124. .333"
Leadership

Political Trust Pearson Correlation l -.O95 -.040
Political Pearson Correlation l .249"
Knowledge

IPE Pearson Correlation l

 

"‘ Correlation is signiﬁcant at the 0.01 level (2-tailed).

* Correlation is signiﬁcant at the 0.05 level (2-tailed).

Sources of Political Learning Efﬁcacy

Factor analysis for the four sources of PLE was a challenge to interpret, with one
source, persuasion, splitting across two factors (Table 5). Similar research relating to the
sources of self-efﬁcacy has also noted problems in identifying good measures for the
sources (R. LaRose, personal communication, March, 2009). Distinct measures emerge
and are discussed below, but further changes may have to be made in future

implementations.

55

a
Table 5: Factor Analysis of Proposed PLE Source Items

 

 

Component

1 2 3 4 5
ViCLl: l have friends who know where .816
to ﬁnd good political information
ViCL2: l have friends who know where .817
to ﬁnd good political information
online
ViCL3: I have seen how my friends can .686
ﬁnd political information online
EnaLl: If I don’t ﬁnd the political .792

information 1 am looking for in one

source, I try to ﬁnd another source

EnaLZ: If a political information source .757
does not have all the information I

need, I will visit their recommended

sources of more information

EnaL3: If I like a source of information .693
I am likely to come back to it

repeatedly to see if there are updates

EnaL4: 1 am conﬁdent I can learn to use .428 .548
a variety of political information

sources

Perl: My friends encourage me to ﬁnd .450 .673
political information

Per2: My friends tell me that 1 should .333 .662
learn about the issues before 1 make a

political decision

Per3: I have been persuaded by friends .259 .280 .630
or family to ﬁnd political information

Per4: 1 am likely to read political .838
information sent to me by a friend

Per5: I am likely to read political .811
information sent to me by a family

member

Per6: I am likely to check out a political .734
candidate recommended by a friend

Affl: I am not apprehensive about .498 .219

ﬁnding the political information I need

Aff2: Political information is somewhat .736

intimidating to me

Aff3: When I think about looking for .880

political information, I get nervous

Aff4: When I am looking for political .863

information, I get nervous

Aff5: I am conﬁdent that I could .429 .507
overcome my anxiety about getting the

information I needed to make a political

decision

Cumulative variance explained % 29.68 45.86 53.57 60.34 66.1

 

a
Varimax rotation with Eigenvalues >1 specified, ﬁve factors extracted,

explaining 66.1% of the variance. Ajf items were re-coded to account for
direction of wording.

56

As the vicarious learning dimension contains two media-speciﬁc items, this scale
will be reduced to a one-item measure as the decision to remove media-speciﬁc items
extends to the PLE sources as well. VicL3 is a valuable item if one removes the “online”
qualiﬁer, however, and it will be included in the later stage of the study to reassess ﬁt.

Enactive learning demonstrated three strong items out of four on the factor
analysis, with EnaL4 showing some multi-dimensionality. However, when one examines
the scale reliability (Table 6), EnaL4 shows itself to be a strong contributor to the alpha

of the overall scale and thus will remain and will be reassessed later.

a
Table 6: Correlations and Scale Reliability for Enactive Learning Items

EnaL l EnaL2 EnaL3 EnaL4
EnaL]: If I don’t ﬁnd the political 1.000 .575 .467 .552
information 1 am looking for in one
source, I try to ﬁnd another source
EnaL2: If a political information 1.000 .468 .400
source does not have all the
information 1 need, I will visit their
recommended sources of more
information
EnaL3: 1f 1 like a source of 1.000 .406
information I am likely to come back
to it repeatedly to see if there are

 

 

updates

EnaL4: 1 am conﬁdent I can learn to 1.000
use a variety of political information

sources

Item-total correlation .677 .597 .546 .553
Alpha if item deleted .689 .731 .756 .753
Alpha (for EnaL Scale) .786

Average correlation .473

 

a All ﬁgures are Pearson ’s r (p >. 05) unless otherwise noted

The persuasion dimension demonstrated a different problem, with the items
splitting into two factors, and one of those factors being relatively messy. On closer
inspection, it is Obvious that the wording of Perl -3 and Per 4-6 are causing them to

access entirely different concepts, the ﬁrst measuring whether friends are actively trying

57

to persuade, and the second being whether someone is likely to “cave” to a persuasion
attempt. Thus, the ﬁrst set of items more accurately captures what Bandura (1997)
intended when he discusses persuasion as a source of self-efficacy. These three items

correlate well and seem to form a relatively strong scale (Table 7).

a
Table 7: Correlations and Scale Reliability for Persuasion Items

 

 

Perl Per2 Per3
Perl: My friends encourage me to 1.000 .580 .550
ﬁnd political information
Per2: My friends tell me that I 1.000 .425
should learn about the issues before 1
make a political decision
Per3: l have been persuaded by 1.000
ﬁ'iends or family to ﬁnd political
information
Item-total correlation .699 , 566 .545
Alpha if item deleted .596 .708 .732
Alpha (for EnaL Scale) .763
Average correlation .518

 

a All ﬁgures are Pearson’s r (p >. 05) unless otherwise noted

Finally, enactive learning may be showing the effects of a response set problem —
the three items that were reverse coded (Aff2-4) held together. Nonetheless, these three
items form a strong scale (Table 8) and provide value in a cognitive shake-up to survey

participants in their reversed wording.

a
Table 8: Correlations and Scale Reliability for Affective Items

 

 

Aff2 Aft‘3 Aff4
Aff2: Political information is 1.000 .529 .474
somewhat intimidating to me
Aff‘3: When I think about looking for 1.000 .868
political information, I get nervous
Aﬂ’4: When I am looking for 1.000
political information, I get nervous
Item-total correlation , 5 13 _322 ,77 5
Alpha if item deleted .929 .641 .691
Alpha (for EnaL Scale) .832
Average correlation .624

 

a All ﬁgures are Pearson ’s r (p >. 05) unless otherwise noted

58

In order to fully vet the chosen sources of self-efﬁcacy items, one last factor
analysis was completed. As demonstrated in Table 9, each scale, with one exception,
loaded on its own factor. The lone vicarious item loaded with the persuasion items,
likely due to its emphasis on the behavior of friends. This reinforces the decision to add

VicL3 back to later surveys, omitting its media-speciﬁc wording.

a
Table 9: Factor Analysis of Final PLE Source Items

 

Component

Enactive Pers Affective
ViCLl: l have friends who know where .605
to ﬁnd good political information

EnaLl: Ifl don’t ﬁnd the political .817
information 1 am looking for in one
source. I try to ﬁnd another source

 

EnaL2: If a political information source .805
does not have all the information I need,

I will visit their recommended sources of

more information

EnaL3: Ifl like a source ofintbrmation l .742
am likely to come back to it repeatedly to
see if there are updates

EnaL4: I am conﬁdent I can learn to use .610 .402
a variety of political information sources

Perl: My friends encourage me to ﬁnd .864

political information

Per2: My friends tell me that I should .793
learn about the issues before I make a
political decision

Per3: I have been persuaded by friends or .738
family to ﬁnd political information
online

Aff2: Political information is somewhat .731
intimidating to me

Aff3: When I think about looking for .910
political information. 1 get nervous

Aff4: When I am looking for political .890
information, 1 get nervous

Cumulative variance explained % 37.5 55.51 66.18

 

a . . . .

Varimax rotation With E rgenvalues > 1 specrﬁed, three factors
extracted explaining 66.2% of the variance. Aﬁ' items were
recoded to account for direction of wording.

59

Bandura posits that these four dimensions are sources of self-efﬁcacy, and, as
such, should be related when regression is applied. Three of our four antecedents to
political learning efﬁcacy demonstrated substantial and signiﬁcant standardized [3
coefﬁcients, affective state ([3 = .258), vicarious learning ([3 = .156) and enactive learning

([3 = .440). Persuasion was more problematic, with an insigniﬁcant B of .005. The model

explained 41.9% ofthe variance in PLE. F(4, 353) = 63.616, p < .05, R2 = .419. PLE

was also able to predict 13.5% of the variance in political knowledge, with a standardized
B of .367 (Figure 6). This supports TH6. The replacement of the ﬁrst group of
persuasion items with the second group does not make for a better model. Theory
hypotheses I, 2, & 4 are supported, TH3 is not. The ﬁlll model of PLE as shown in
Figure 5 in the theory section will be tested with the results of the nationally
representative survey. In the pilot test survey, whether or not someone voted in 2008 was
chosen to measure participation. This was a poor decision, as over 77% reported voting,
a number too high to accurately test the relationships in the model, and signiﬁcantly
higher than the nationally-reported 53% tum-out among this demographic. In the full

survey another measure of participation will be used.

60

Vicarious Learning

B = .lS6*
Persuasion B = DOS

Political B = «367*

Learning Political
Eﬁ‘icacy —> Knowledge

4————>

Enactive Learning l3 = 440*
a——>

/

B = .258*
Affective State

Figure 6. A Social Cognitive Model of Political Learning
Pilot Test Research Questions Results

The ﬁrst theory research question of this study asked whether there was evidence
for political information efﬁcacy existing as its own construct separate from internal
political efﬁcacy. This research strongly indicates that PIE as measured is not a separate
concept from IPE (Table 10). The three items shared by both scales and the two
exclusive items all load on the same factor and have an excellent Cronbach’s alpha
of .890 as a combined scale (Table 11). Thus, it is concluded that while the current
construct deﬁnition of PIE touches on some worthy questions of a person’s conﬁdence in
their political knowledge, the current measurement instrument does not access anything
much beyond what the original IPE scale already accessed. IPE is a well-validated,
parsimonious construct and scale, and there can be little advantage in adding and
subtracting from it without a thorough investigation of proposed items and any changes

to the construct deﬁnition.

61

The second theory research question asked whether political learning efﬁcacy
stood alone as its own construct, and, as shown in Table 10, its items load strongly on
their own factor separate from IPE, PIE and items of external political efﬁcacy. PLE is a

separate construct.

a
Table 10: Factor Analysis of Efﬁcacy Scales

 

Component

IPE PLE EPE
IPElPIEl: I consider myself to be .792
well qualiﬁed to participate in
politics
IPE2PIE2: I feel that I have a pretty .787
good understanding of the important
political issues facing our country

IPE3PIE3: I think that I am better .847
informed about politics and

government than most people

IPE4: I feel that I could do as good a .810
job in public ofﬁce as most people

 

PIE4: If a friend asked me about the .777
presidential election, I feel I would

have enough information to help my

friend ﬁgure out who to vote for

PLE]: I am conﬁdent that I can learn .845
about political issues if I want to

PLE2: I am conﬁdent that I know .798
where to go to locate good political

information

PLE3: I am conﬁdent that I can seek .857
out information related to a political

decision in general

PLE9: I am conﬁdent that I can .605
understand most political

information in general

EPEI: People like me don't have any .876
say about what the government does

 

EPE2: I don't think public ofﬁcials .884
care much what people like me think

 

Cumulative variance explained % 49.94 71.62 80.05

 

a Varimax rotation with Eigenvalues >1 speciﬁed, three factors
extracted explaining 80.1% of the variance. EPE items were
recoded to account for direction of wording. Items shared by
the IPE and PIE measurement instruments are noted

62

a
Table 1]: Correlations and Scale Reliability of IPE and PIE as a Combined Scale
IPElPIEl IPE2P1132 IPE3PIE3 IPE4 PIE4

 

 

IPEIPIEI: I consider myself to be 1.000 .716 .625 .551 .580
well qualiﬁed to participate in

politics

IPE2PIE2: I feel that 1 have a pretty 1.000 .697 .533 .658

good understanding of the important

political issues facing our country

IPE3PIE3: 1 think that 1 am better 1.000 .613 .686
infomied about politics and

government than most people

IPE4: I feel that I could do as good a 1.000 .511
job in public ofﬁce as most people
PIE4: If a friend asked me about the 1.000

presidential election, I feel I would
have enough information to help my
friend ﬁgure out who to vote for

Item total correlations .728 .777 .785 .641 .715
Alpha if scale item deleted .860 .852 .847 .884 .863
Alpha (IPE - PIE combined scale) .890
Average correlation .617

 

a All ﬁgures are Pearson ’s r (p >. 05) unless otherwise noted

Discussion of Pilot Test

This exercise produced a seemingly valid scale for the measurement of political
learning efﬁcacy and also for its antecedents, with the exception of the problematic
measurement of persuasion. Factor analysis is able to discriminate between the new
measure of PLE and related measures of IPE, EPE, and PIE and correlations show a
strong relationship between three of ﬁve expected correlates. In addition, PLE was able
to predict political knowledge, which it was intended to do. The ability to predict
political knowledge was not as strong as anticipated, likely due to easy access to answers
to the political knowledge items online, no monitoring of whether participants stayed

with the survey page, and no time limits for completion.

63

The results of analysis of the PLE sources are very similar to what Bandura
(1997) posits, with enactive learning being the strongest determinant of self-efﬁcacy.

The problematic persuasion items may be accessing what came up repeatedly in the focus
groups: That individuals get annoyed and may tune out persuasion attempts when they
become too frequent. Participants may have experienced this in the 2008 election (four
months prior to the pilot test) and be under-reporting their responses as a result; the mean
of the scale responses for persuasion was signiﬁcantly smaller than that of either enactive
learning or affective state (5.92 vs. 8.62 and 8.20 respectively). It also may be
attributable to who the item is asking about; the enactive learning and affective state
items are asking about the participant themselves, and the persuasion items are asking
about the behavior of ﬁ'iends and family.

In addition to the development of the PLE scale and the measures of its
antecedents, the pilot test was also able to show that the measure of political information
efficacy either does not exist as its own construct or that the measurement instrument
used does not accurately access the construct. The ﬁndings reinforce earlier work noting
the robustness of the internal political efﬁcacy scale (Craig et al., 1991; Morrell, 2003)
and also of the underused external political efﬁcacy scale.

This pilot test has some distinct limitations, notably those of sample selection,
social desirability (77% of this sample reported voting, nationally only 53% of the sample
demographic actually did), and test length; survey fatigue on a 74-item survey could
easily have been an issue. Overall, this was an effective exercise however, as indicated
by the strong measure of PLE that emerged, and the unequivocal answers to both

research questions. The scale of PLE can be used together or separately from the

64

measures of its sources, and it should provide a valuable new addition to the political

communication literature.

Survey Research

In order to test the hypothesized relationships between young people, political
learning efﬁcacy, political knowledge, participation and political social network site use,
a survey was conducted in April 2009, six months after the 2008 election. This creates a
potential problem of fading memory of political actions taken during the election;
however, this time frame includes the signiﬁcant beneﬁt of a “cooling-off” period from
the media hype and social pressure associated with the election. This added beneﬁt may
mitigate some the social desirability and priming problems associated with many political
efficacy studies (including this study’s pilot test). In addition to hypotheses testing, this
survey also allowed for further testing and validation of the measurement and model
presented in the pilot test. An extremely common method in the social sciences, survey
research allows for the examination of relationships between variables.

Sample and Procedures

In early April 2009 an online panel of 5000 Americans aged 18-25 were invited to
access the survey through the online survey Web site Zoomerang. This panel is managed
by MarketTools, the parent company of Zoomerang, and participants receive points
towards merchandise for participating. The survey was automatically closed after 625
individuals completed it, which took approximately three days to achieve. Due to this
close-out procedure, an accurate completion rate is not calculable.

Of these 625 individuals, 64.5% were female, 34.1% male (1.4% gave no answer),

71.7% were Caucasian, 9.6% Hispanic, 8.5% African American, 6.2% Asian, and 3.6%

65

multi-racial. The average age was 22.9 years old. Almost 11 (10.9) percent reported not
being registered to vote, and 65.1% reported voting in the 2008 election (somewhat
higher than the 55% reported nationally for this age group). Twenty percent of the
sample identiﬁed as Independent, 22.1% Republican, 36.3% Democrat, and 17% didn’t
know their afﬁliation. More than 36 % of the sample had attended at least some college,
and 23% had only a high school diploma, 6.4% never graduated. Twenty percent had a
four-year degree, 8.2% a two-year degree. Use of both dominant social network sites
was represented (39.4% preferred Facebook, 27.7% preferred MySpace and 16.5% use
both) and only 16.5% of the sample reported not using either Facebook or MySpace (one
should note, however, that this sample only includes individuals with Internet access).
As other sources have reported, the average educational level for Facebook users was
signiﬁcantly higher than it was for MySpace users, t(417) = 9.883, p <.05, but there is no
signiﬁcant difference in age between the two dominant SNS.

Measures

Dependent variables

Knowledge was measured in the same manner as in the pilot test following Delli
Carpini & Keeter (1994). Five general political knowledge items form an index on which
one can score from 0 to 5 (Appendix C).

Political learning efﬁcacy and its sources are measured following the results of
the pilot test. A few items noted for further review during the pilot test were also
included to help further establish that the correct scale was chosen. VicL3, which
previously included media-speciﬁc wording, was included with the online mention

omitted.

66

Political participation was measured following a scale adapted and updated from
Rosenstone and Hansen’s (1993) common index of political participation, which includes
items about donating money, serving in public ofﬁce, displaying yard signs, etc. This
index was expanded to include items related to online participation, such as posting to a
political blog or watching a debate online, to update the scale to reﬂect current media
habits. One error was made in coding this item: no “none” option was given, and it was a
forced-response question, so individuals who did not participate politically at all were
required to report doing something they may not have done. Results on this variable may
be somewhat higher than they should be.

Independent Variables

Use of social network sites for political purposes was subdivided into three
different measures, two indices and one scale. Each index or scale accesses political
actions one might take on a SNS, but with increasing levels of commitment needed to
take the action. For example, the ﬁrst index, SNS political usage, asked if participants
had become a supporter of a candidate on an SNS, a very general action requiring of only
one click and taken by literally tens of millions of people during the 2008 election cycle.
On the other end of the spectrum, SNS political engagement asked about actions much
more indicative of true political engagement, such as participating in “real world”
political events learned about on an SNS or continued participation in political groups on
SNSs after the election was over. SNS political activity formed the middle ground, asking
about small political actions one might take on an SNS, from posting a wall message
regarding politics to taking a political quiz (Appendix C). This subdivision of use of SNS

for political purposes was both to look for differences between those taking actions that

67

require little commitment and involvement, and may be more indicative of a bandwagon
effect than anything else, and actions that take more thought, time and commitment to
undertake. This will also enable us to report on what young people were actually doing
on SNSs during the 2008 election.

SNS political usage (SNSPolUsage) was measured with four items describing the
most common political actions a person might take in a social network environment
during the recent election, from becoming a supporter of a candidate to posting a status
update that mentions something political. It was implemented as an index, with a binary
yes/no answer and each individual’s total score indicating the average number of actions
they took during the 2008 election.

SNS political activity (SNSPolAct) was implemented as an nine-item index added
to and adapted from Vitak et al. (2009). It included small political actions indicative of a
higher level of involvement with politics on SNSs, such as posting a status update
mentioning something political. Different than SNSPolUsage, SNSPolAct speciﬁcally
asked about activities undertaken in the past month and not during the election, to access
continued online involvement. Responses were provided in a list and participants could
check off what they had completed.

Political social network engagement (SNSEngage) went farther to ask about
speciﬁc and continued political actions taken in or because of an SNS environment.
Adapted and added to from Vitak et al. (2009), this seven-item scale asked, in ll-item
response (strongly disagree to strongly agree) format, whether individuals were more
interested in the election because of political SNS, or if they were continuing to stay

involved with the groups they joined during the election, among other things.

68

CHAPTER 5
RESULTS

Political agl Media Behavior

The ﬁrst research question of this study asked what political behaviors young
people were engaging while on social network sites. To answer this question, only the
self-reported users of social networks sites will be considered (n=522).

Low levels of usage of SNS for political purposes were reported. SNSPolUsage
(a = 0.818) had a mean of 1.70 (Max = 5.0, SD = 1.75), SNSPolAct (a = 0.529) M = 1.67,
Max = 9.0, SD = 1.51, and SNSEngage (a = 0.912, a scale rather than an index) displayed
a mean of 3.97 (Max = 11.0, SD = 2.47). So, the average person was engaging in less
than two forms of speciﬁc political action on SNS, and disagreed with most statements
regarding engagement on and political usefulness of SNS.

Of the uses made and actions taken on SNS taken by individuals, the most
frequent was becoming the supporter of a political cause (35.4%) or taking a quiz about

politics (33.5%). Other actions and uses are detailed in Table 12.

69

Table 12: Political Uses and Actions on SNS

 

Uses {SNSPolUsage} during2008 election Percent

 

 

 

Became supporter of cause 35.4
Visited SNS site of candidate 29.3
Became supporter of candidate 26.6
Joined a political group 25.5
Actions (SNSPolAct) in past month Percent
Took a quiz 33.5
Added/deleted political info from proﬁle 23.9
Posted political status update 23.6
Became a supporter of a candidate or cause 22.0
Discussed politics in a message 19.7
Added application or bumper sticker 13.8

' Posted political wall comment 12.5
Joined or leﬂ a political group 1 1.7
Posted a link about mlitics 7.3
n = 522

SNSEngage was found to be internally consistent (a = 0.912), though underlying
distribution of responses to SNSEngage was moderately positively skewed (.553). The
means on each item ranged from 3.46 to 4.51 (Max = 11.0), meaning that generally, this
sample was not engaging in the behaviors described and do not report ﬁnding political
SNS useful (Table 13). These results help in addressing RQZ, which asked whether
young people would indicate more interest in the 2008 election because of social network
sites. When asked speciﬁcally whether they were more interested in the election because
of information they encountered on SNS (SNSEngageS), the answer was a resounding no
(Table 13). Additionally, only 29.1% agreed with the statement that they use SNS to
share their political views. More than 84% of this sample use social network sites, yet
less than 40% of SNS users report any use of SNS for political purposes, even during a

high-proﬁle election.

70

Table 13: SNSEng_age Responses

 

 

 

Std.

Mean Dev.
SNSEngage]: I use social networking sites to share 4.01 2.830
political information with my friends
SNSEngageZ: I have participated in “real world” 3.46 2.883
political activities (like attending a rally or meeting)
because I heard about them on a social networking site
SNSEngage3: l was more likely to vote because of 3.66 2.880
political or candidate information I saw on a social
networking site
SNSEngage4: 1 still check out or participate in political 4.01 3.148
groups I joined during the 2008 election on social
networking sites
SNSEngageS: l was more interested in the 2008 election 3.62 2.859
because of information I received on social networking
sites
SNSEngage6: I use/have used social networking sites to 4.51 3.267
share my political views with others
SNSEngage7: I have found additional sources of 4.51 3.171

political information through social networking sites

 

n = 522, max = 11.0, strongly agree, min= 1.0, strongly disagree

Hypothesis 1 posited that young people will prefer online sources for their
political information. While 60.5% report using online news sites like CNN.com, only
36.3% report that online sources of any kind are their preferred source of information,
thus, this hypothesis is not supported. “Old” media, televised news and newspapers, are

still viable news sources for this demographic, and televised news is signiﬁcantly

preferred over other news sources, x2(6, N = 625) = 662.72, p > .05. Only a small

number report using political blogs (11.5%) and social network sites (17.5%) for ANY of

their political information. This runs counter to the media’s conventional wisdom about

this demographic.

71

 

Table 14: Preferred News

 

 

Sources

Percent
Televised news 43.5
Online news sites 29.4
Newspaper 9.8
Radio 6.4
Social Networking 5.3
Sites
Political Blogs 1.6
Other 4.0
n = 625

Political Learning Eﬂicacy and PLE Model

All four scales for the sources of self-efﬁcacy held up well in further testing.
Alphas are as follows: Enactive learning a = 0.876, vicarious learning a = 0.817 (VicLl
& 3) persuasion a = 0.851, and affective state a = 0.882. Under additional factor analysis,
the items loaded strongly on their speciﬁed factors, and vicarious learning exhibited none
of the problems it did in the pilot test when it loaded with persuasion. The PLE scale also
performed well, with an alpha of 0.870. The item in question in the pilot test, PLE9, held
well with the other items in this factor analysis (one factor loaded with an eigenvalue > 1,
total variance explained = 72.1%), so it will remain a part of the scale along with PLE1-3
for a four-item measure.

The data, with one exception, ﬁt the model of PLE well and in a similar fashion to
the pilot test data (Figure 2). Standardized B coefﬁcients are reported to account for
differences in measurement. The persuasion contributor to PLE performed poorly as it
did in the pilot test, and it should be reevaluated. While the persuasion items have face
validity, they may not be accessing truly what Bandura intended when he discussed

persuasion. It also may due to the recency of the 2008 election and the social pressure

72

this group reported encountering in the earlier focus groups; participants might not want
to admit that anyone else might have had any inﬂuence over them or they continue to
have some level of annoyance over the actions and pressure they experienced. Vitak et al.
(2009) found a similar aversion to persuasion in that attempting to persuade friends
politically via Facebook was not considered an appropriate behavior.

Of the sources of PLE, vicarious learning (.219**), persuasion (203") and
enactive learning (.183**) correlated the most strongly with political social network use,
and affective state correlated insigniﬁcantly. This corresponds to what was heard in the
focus groups, that participants remembered seeing their friends acting politically on SNS
and some ended up clicking on the links and following the links around the SNS,
sometimes supporting a candidate as a result.

There was no signiﬁcant difference between genders in level of PLE, and, while

there was a signiﬁcant relationship between educational level obtained and PLE, F (1 ,
623) = 14.985, p < .05, R2 = .023, B: .153, and educational level had only a small impact,

speciﬁcally and only between those who had less than a high school diploma and a four-
year college degree. Four-year degree holders had a signiﬁcantly higher PLE (M =
8.651) than non-high school graduates (M = 7.160), and had the highest PLE of all
educational levels, but none of the other direct comparisons were signiﬁcant, ANOVA
post-hoe with Scheffe, F(5, 619) = 4.205, p <.05.

The relationship between political learning efﬁcacy and political participation is
of signiﬁcant interest as increased participation is the goal of many engagement efforts
involving young people. The standardized B of 0.282 reported in Figure 2, which appears

somewhat low on ﬁrst glance, is actually reasonably good considering the distribution of

73

the data regarding political participation - participation was logarithmic, heavily skewed
toward the leﬁ. Nearly 49% of the sample reported only one or fewer units of
participation (an additive index, max = 11), and the mean participation was only 2.07.
Even this low number is likely slightly elevated, as this is where the instrument error, the
omission of the “none” option, occurred. From an analysis of the distribution of
responses, it appears that if anything was over reported, it was debate viewing (75.4% of
the sample and the ﬁrst option presented on the list).

The model helps answer TH 7 (unanswerable in the pilot test due to over-reporting
of voting), that political learning efﬁcacy will be positively associated with political

participation. A linear regression produced a signiﬁcant association with a somewhat
small R2, F(1, 624) = 53.9, p <.05, R2= .080, B = .282, VIF = 1.0. This small R2 may be

due to the distribution of the participation data, which was strongly positively skewed

(1.81) around low levels of participation. TH 7 is supported.

74

3on mam ==m K 3&3

 

com. n Na
83% mahomﬁwx
1.8.. u a
mi. N mg to. H ma who. n ma Allv
:ocaaﬁEmm EoE—om Al honor—mm Alll ammo—305— Alll Aouoﬁm 3.. 5m. H n MEEBmu maéoetm
12v. "n EoE—om IEN. H a 3250.“. 3mg. Mn M3535
EEBE 36.5ch _ _
1// owo. H a ll\\ coo H a towwczwamm
N
1%. u n
a... _ 8. n n

M3533 maﬁaeoﬁ

75

Social Network Sites

H2-H4 involved the relationship between use of social network sites for political
purposes and levels of participation and knowledge. As shown in Table 15, both
measures of political uses of social network sites are positively and signiﬁcantly

correlated with knowledge, participation, and political learning efﬁcacy.

Table 15: Pearson Correlations of Expected Relationships

 

 

 

PLE SNS SNS Knowledge Particip
Political Political ' ation
Usage Activity
PLE Pearson l .223M .227" .293“ .303"
Correlation
SNS Political Pearson l .503“ .225IMI .537“I
Usage ' Correlation
SNS Political Pearson 1 .200" .539“
Activity Correlation
Knowledge Pearson 1 .324“
Correlation
Participation Pearson g 1
Correlation
n=522 (SNS
users only)

More speciﬁcally, it was hypothesized in H2 that use of social network sites for
political purposes would be positively associated with political knowledge. This

hypothesis is supported for both measures of political uses of SNS. SNSPolUse, F (1 ,

520) = 27.664, p <.05, R2= .051, B = .225, VIF = 1.0; SNSPolAct, F(1, 520) = 19.674,p

<.05, R2= .036, B = .191 , VIF = 1.0. Standardized B is reported to account for

measurement differences. In an interesting ﬁnding, F acebook users (M = 4.301) had a
signiﬁcantly higher political knowledge than MySpace users (M = 3.31), t(4l7) = 8.546,
p< .05. This difference is likely due to the previously noted differences in educational

attainment between the two SNSs.

76

Hypothesis 3 put forward that usage of SNS for political purposes would be
positively associated with political participation. The strong correlations observed in
Table 15 between participation and both measures of usage are an indication that this
relationship works both ways, and this hypothesis is supported. Similar to the
knowledge ﬁnding, participants who preferred Facebook had a signiﬁcantly higher level
of participation (M = 2.756) than did MySpace users (M = 1.867), t(4l7) = 4.977, p< .05.

Hypothesis 4 suggested that the use of social network sites for political purposes

would be positively associated with political learning efﬁcacy. This was also supported,

SNSPolUse, F(1, 520) = 27.264, p <.05, R2: .050, B = .223, VIF = 1.0; SNSPolAct, F(1,

520) = 19.981, p <05, R2: .037, B = .192, VIF = 1.0. Again, Facebook users (M =

8.566) had a signiﬁcantly higher PLE than MySpace users (M = 7.961), t(4l7) = 3.019,

p< .05.

Finally, as internal political efﬁcacy is strong a strong predictor of political
participation in the PLE model, [PE is theorized as having an antecedent in PLE, and IPE
is such a prominent construct in the political communication ﬁeld, its relationship to

political social network usage was examined in H6. A signiﬁcant relationship was found

between both SNSPolUse, F(1, 520) = 72.726, p <.05, R2= .123, p = .350, and

SNSPolAct F(1, 520) = 55.826, p <.05, R2= .097, B = .311, VIF = 1.0. There was no

signiﬁcant difference between Facebook and MySpace users on this variable.

77

CHAPTER 6
DISCUSSION

This study attempted to develop and validate a new construct, make a theoretical
contribution to the social cognitive theory and political communication literature, and
shed light on a timely issue in political uses of social networks. This research allowed
the value of political information and candidate presence on social networks to be
measured and the new construct of political learning efﬁcacy to be developed and tested
in a large, representative sample. Existing literature in this area is presently sparse, and
most is conducted on small, student samples, so these data provides an excellent snapshot
of the media preferences, participation, and knowledge levels of American youth during
and aﬁer the 2008 presidential campaign.

The media choices being made by American 18-25 year olds (speciﬁcally, the
strong preference for televised news over other, online, choices) are not consistent with
the conventional wisdom regarding this demographic. Anecdotally, particularly in
traditional news media concerned with their own demise, the assumption has been that
this demographic strongly prefers online media and is eschewing “old” media. That
assumption does not play out in this research, a ﬁnding consistent with Wells and Dudash
(2007) who found, in a large national focus group study, that only 15% preferred online
sources of information. While this 15% was greater than the percentages preferring any
other information source except for family and friends in their sample, it is not a huge
number by any means. Additionally, during the 2004 campaign, Rainie et al. (2005)
found that 28% of 18-29 year-olds received most of the their campaign information

online, so the ﬁnding here that only 29% prefer online sources of information in 2008 is

78

only notable in that it may be indicative of a slowing of the adoption rate of online news
sources. In addition, these are self-reported preferred news sources and this self-report
may be slightly different ﬁ'om actual use.

While all relationships between political SNS use and political outcome variables
were signiﬁcant, and all SNS hypotheses supported, the impact of social network use for
political purposes is surprisingly low, contrary to high expectations put forward by both
the scholarly community and the news media. While it is not remarkable that someone
might not want to admit that a SNS had played an important role in a political decision,
one would think that, given the strong afﬁnity individuals seem to have for their chosen
SNS, if they had found valuable elements related to the election on their SNS they would
report it. This result is both similar and different to what the Pew Internet and American
Life Project found in their study of the 2008 election and Internet and new media use.
They found that across all age groups, 14% of Internet users used SNS for a political
purpose, which is in line with the ﬁndings of this study. Among users under 30, they
found higher usage; nearly half of their sample reporting using SNS for political purposes
(Smith & Rainie, 2008). This second number is higher than what was found in this study,
which may be partially attributable to the wider age range included in “younger” (18-25
in this study, 18-30 in the Pew study).

The strong positive correlation (Table 15) found between both forms of political
SNS usage and political participation likely indicates a bidirectional relationship; political
social network sites may be of great value to those who are already politically engaged or
those who are just ﬁred up on a particular issue. SNSs could serve as an outlet for their

views, providing a voice and a medium to individuals who previously have struggled

79

both to ﬁnd an audience and to feel that their views mattered to those in power, as
indicated by the historically low political efﬁcacy demonstrated by this demographic. In
fact, two of the strongest positive associations of this study were between political SNS
use and internal political efficacy (IPE), and IPE and political participation. In ﬁrture
research it would be worthwhile to investigate those linkages further as they may indicate
an impact of SNS use separate from persuasion or information accrual beneﬁts, such as
political empowerment and political social capital building.

While there does seem to be some informational beneﬁt to political social
network use, the impact on political learning efficacy and knowledge are small. Perhaps
most troubling is an examination of those who prefer social network sites for most of
their political information: these individuals have the lowest political knowledge of any
group, signiﬁcantly less than the next group, newspaper readers (who may be over-
reporting readership of the newspaper because it is a socially desirable action). While it
may seem obvious that someone who uses SNSs exclusively for his or her political
information might not be particularly politically engaged, and this may cause a lack of
knowledge; the data do not support this. There was no signiﬁcant difference in
participation between the SNS-preferring and those preferring other media for their
political information, only in knowledge.

Boulianne (2009) notes in her recent meta-analysis of studies examining the
impact of the Internet on engagement that political interest should be controlled for in
studies that consider engagement or participation. She found that studies that control for
political interest found few signiﬁcant effects of Internet use on engagement, but studies

that do not control for this variable are more likely to ﬁnd effects. Political interest was

80

not measured in this study, thus, this issue may be involved, as the effects found were
relatively small. In ﬁrture research regarding SNS use and engagement or PLE and
engagement, this variable should be included.

In a positive ﬁnding, it seems that any use of political social networking is related
to these small, but signiﬁcant, increases in political knowledge, PLE, and participation.
Both measures of social network site use for political purposes produced similar results,
so it seems that just completing some of the simple actions included under SNSPolUse,
such as supporting a candidate or cause, can either cause increases in information
gathering or may lead to other actions that do. In future research, however, the two
measures of use of SNS for political purposes should remain as separate scales, as they
do not hold together well under exploratory factor analysis and have stronger alphas
when treated as separate scales. From the low level of agreement with the
SNSPolEngage items, it does seem that only small numbers were truly politically
engaged in a social network setting.

Limitations

The limitations of this study lie ﬁrst in the problems of self-report and social
desirability. Voting and civic participation are actions that this group is strongly
encouraged to take part in through a variety of sources in media, on campuses, at home,
etc., but it has been shown that they do not participate as they should and this may be
personally embarrassing. Previous studies have encountered relatively serious problems
with self-report among this age group, but all possible steps were taken to minimize its
effects here. Though the problem of social desirability cannot be completely eliminated,

steps such as ensuring anonymity and value of honest responses, indicating that there is

81

no “right” answer, and urging participants to answer honestly, were taken (Oskamp &
Schultz, 2005). Additionally, it was fairly obvious from the questions that this was a
study looking for a connection between social networks and political outcomes;
participants could have been primed to answer in a certain way, and participants may
have been inclined to look up answers to the political knowledge items online.

Second, the sample for the main survey was purchased from a market research
ﬁrm. While there are problems with all samples, these individuals likely have taken
repeated surveys and some may know ways to quickly complete a survey without giving
much consideration to the questions. This may have some impact on the variance of the
scales used, as individuals could repeatedly be selecting one value as they head down the
items making up a scale. In addition, as this is an online survey research ﬁrm,
participants are likely to be Intemet-savvy with few individuals who are just learning
what resources are available on the Internet. As a result, levels of participation in online
politics may be elevated due the online experience level of participants, though
considering the extremely low online participation rates found, this seems unlikely.

Third, there was some error in some of the survey creation. As mentioned, the
“none” was omitted ﬁom the participation scale, resulting in a lack of a zero-point for
that variable. While this is a problem to be considered and resulted in what may be a
slightly elevated participation variable, research conducted by Vitak et al. (2009) during
the 2008 election found slightly higher political participation than this study did using the
same scale and the same demographic, though all the participants in that study were
current college students. Thus, the problem caused by this error may be minimal. In

addition to the omission of this response category, in a later analysis of the pilot test data

82

it was noticed that PLE17 (“I am conﬁdent that I could ask an intelligent question during
an election”) had not been included in the non-media speciﬁc exploratory factor analysis
(Table 2) that determined the PLE scale as administered in the ﬁnal survey. Thus, it was
left out of further data collection. When the factor analysis was redone after the survey
had been completed, it was found that PLE17 loaded strongly with the four variables
selected to make up PLE, and its addition bolstered the alpha for the scale. In future
research, it should certainly be included.

Implications for Campaigns

The ﬁndings indicate that while social network sites can provide some value to
users in the form of slightly increased political learning efﬁcacy and political
participation, they seem to be only a supplement to other sources of information. This
may indicate that they are just another political communications tool in the arsenal of
candidates, not quite the game-changing super weapon the media espoused. We have
seen, in the years between the 2004 and 2008 elections, resistance to political candidate
presence on SNS decrease and became accepted, even welcomed (Hayes, 2009; Utz,
2009; Vitak et al., 2009). We may be approaching a point where it has just become part
of standard political practice to include a social media campaign in your strategy.

What was not tested here was afﬁnity toward candidates who use social network
sites. It has recently been shown that SNS may have some value in creating a connection
between a candidate and the public (Utz, 2009). Perhaps the real value of political social
network site use is by and large social, not entirely informational, as was put forth by this

study. These sites were originally intended to provide a social connection between

83

individuals, and that may be reﬂected in the positive relationship between political trust
and social network use for political purposes reﬂected in this data.

For candidates, this may be the most valuable take-home message. Having a
viable presence on SNS is important, both for the need to look tech-sawy, but also to
make a connection and build trust with your constituents and/or potential voters. It is so
easy for the public to generalize all politicians as crooks, unresponsive, or, frankly, old
and out of touch. Putting a human face to your public presence through social networks
and generating an afﬁnity, as Utz (2009) found, may help to defray the negative
stereotypes associated with being a politician. Simply providing campaign information
that an individual could ﬁnd elsewhere online, is likely not going to help generate this
feeling of afﬁnity. Thus, the true social connectors that SNS do such a good job in
promoting, such as alumni status, taste in music or books, and family photos and
relationships, should be focused on by the candidate and his or her staff. In addition,
SNSs give individual users a voice, an easy way to express themselves politically. This
could be both a problem and an opportunity for candidates, as one needs to devote the
resources to manage a feedback wall on an SNS, and if you don’t have someone to
manage the feedback, controlling what shows up on that wall while still providing access
to supporters can be difﬁcult. Still, if one can manage the wall and even provide
responses to users, the chance to build afﬁnity and additional support is even greater.
Conclusion

This study provides a signiﬁcant contribution to an area where limited research
exists, and provides a new construct for use in political communication research. It gives

us a better picture of the effects of political social network site use, and has helped to

84

answer the question of whether young voters derive any value from them. It seems that
they do have value informationally, but their true value may be in a more social sense
than is fully examined here. The media hype over the potential of these sites should be
tempered and they should be investigated further.

As Mark Twain might say, the reports of the death of traditional media among
this demographic are greatly exaggerated. It is obvious that participants in this study still
respect traditional media enough to report preferring them as news sources, even if they
might not actually use them in them daily. As the traditional media evolves to include
and integrate online techniques (even social networking), this demographic may increase
their use of the content offered by traditional media outlets. Traditional media will fuse
and partner with new media and there will be less distinction between the two.

While some may deride the presence of candidates on social network sites as
pandering, this age group feels ignored and turned off by traditional communication
methods, and SNSs seems to be a way for young people to form connections with
candidates and issues. While their impact on political learning and participation may not
be great, given the high usage rate of SNS among this demographic, there is still potential
for SNSs to provide issue relevant information to young people in a manner that gives
them an active voice, a chance to participate in their own way, and a way to build afﬁnity
with those seeking or in power. Their use in politics should be examined ﬁrrther as they
become normal and accepted routes for communication between a candidate, potential

voters, and constituents.

85

APPENDIX A
FOCUS GROUP GUIDE
1. Introduction (5 minutes)

We would like the discussion to be informal, so there’s no need to wait for us to
call on you to respond. In fact, we encourage you to respond directly to the comments
other people make. If you don’t understand a question, please let us know. We are here
to ask questions, listen, and make sure everyone has a chance to share.

If we seem to be stuck on a topic, we may interrupt you and if you aren’t saying
much, we may call on you directly. If we do this, please don’t feel bad about it; it’s just
our way of making sure we obtain everyone’s perspective and opinion is included.

We do ask that we all keep each other’s identities, participation and remarks
private. We hope you’ll feel ﬂee to speak openly and honestly.

As discussed, we will be tape recording the discussion, because we don’t want to
miss any of your comments. No one outside of this room will have access to these tapes
and they will be deleted after our report is written.

Helping me is my assistant, Evan. He will be taking notes and be here to assist me if I
need any help.

11. Discussion Generation (50 minutes)

We are holding these groups to better understand how you gather political information
and make political decisions, and if you feel like you have enough information to make a
good decision. We are really interested in how you decided who to vote for in the 2008
election. I know this might be a little embarrassing to talk about, but even if you didn’t
vote, the reasons why are really, really important to us, so please speak freely, no one is
judging you.

Let’s get started!

Let’s go around and introduce ourselves, just give me your ﬁrst name and age, and then,
to make things fun, tell me the ﬁrst thing that comes to your mind when I say the word
“Politics”

That was all really interesting. Tell me, how involved you all feel in the American

political process?

Did that change with the last election? Did you feel the same way before the election?

86

What were candidates doing that helped you feel involved?

How did you all go about making a decision in the last election? You don’t have to tell us
your choice, just how you chose the best candidate for you?

-( if online not mentioned):

Did any of you use the Internet to ﬁnd out more about candidates or issues?

-(ifonline mentioned):

You mentioned (YouTube, web sites, Facebook, whatever)-how did this site help you?
Did this site or similar sites help any of the rest of you?

-(ifSNS not mentioned):

What about social network sites like Facebook and MySpace? Were you aware that the
candidates were using them?

Did you ever see a friend interacting with a candidate on their news feed?
Did a friend ever try to get you to support a candidate on one of these sites?
What do you think of the candidates being on these sites?

Ok, slightly different topic-did you feel like you had enough information to make a good
decision? Why or why not?

What could have helped you make a decision?
Did you ever discuss the election with friends?
Did that help you make your decision? Or did you try to persuade your friends?

Where you ever nervous or anxious about voting, especially with all the press about
incorrect registrations and fraud?

Ok, and please answer this honestly, because it really helps us. Did you vote in 2008?

Why or why not? Again, it really helps us to be honest. Not everyone votes and it is
really interesting to know why, even if it is just because you didn’t feel like it.

(depending on responses-gently pursue people who say they didn’t know enough/didn ’t
have enough info)

87

What could candidates do to help people like you?

Thanks, everyone; that was really helpful. Do any of you have anything to add that I
might have missed or you just thought of?

88

APPENDIX B
PILOT TEST MEASURES

Political Knowledge:

(Delli Carpini and Keeter, 1994)

Do you happen to know what job or political ofﬁce is now held by Joe Biden?

Whose responsibility is it to determine if a law is constitutional or not. . .is it the president,
the Congress or the Supreme Court?

How much of a majority is required for the US. Senate and House to override a
presidential veto?

Which party is generally more conservative?

Political Trust:

(Miller, 1974)

How much of the time do you think you can trust the government in Washington to do
what is right? (Just about always, Most of the time, Only some of the time)

Would you say the government is: (Pretty much run by a few big interests looking out for
themselves; Run for the beneﬁt of all of the people)

Do you think that people in government: (Waste a lot of the money we pay in taxes;
Waste some of it; Don’t waste very much of it)

Do you feel that: (Almost all of the people running the government are smart people who
usually know what they are doing; Or do you think quite a few of them don’t seem to
know what they are doing?)

Do you think that: (Quite a few of the people running the government are a little crooked;
Not very many of them are; Hardly any of them are crooked)

Political Cynicism:

(Pinkleton, Um, and Austin, 2002)

“Politicians lose touch with the people once elected”; “candidates for ofﬁce are only
interested in peoples' votes, not in their opinions”; “too many politicians only serve
themselves or special interests”; “it seems our government is run by a few big interests
who are just looking out for themselves”; and “politicians lie to the media and the

public.”

Opinion Leadership:

(Shah and Scheufele, 2006)

“I have more self conﬁdence than most of my friends”

“I like to be considered a leader,”

“I am the kind of person who knows what I want to accomplish in life and how to achieve
it,”

“I am inﬂuential in my neighborhood.” (modify for college kids)

Political Uses of Social Networks

(V itak, Smock, Zube, Carr., Ellison, & Lampe, 2009)
In the past week, which of the following have you done in Facebook?

89

(Check all that apply).

Posted a status update that mentions politics

Joined or left a group about politics

Become a "fan" of a political candidate or group

Posted a Facebook Note that has something to do with politics
Posted a wall comment about politics

Posted a photo that has something to do with politics
Discussed political information in a Facebook message
Discussed political information using Facebook's instant messaging system
Added or deleted an application that deals with politics

Posted a link about politics

Took a quiz that is about politics

Posted a photo of myself or others at a political event
RSVPed for a political event

Added or deleted political information from my proﬁle

90

APPENDIX C
FULL SURVEY
NOTE: Implemented through Zoomerangcom, appearance slightly different online

Thank you very much for agreeing to participate in this survey, it will take approximately
15 minutes to complete.

Your honesty is highly valued, so please answer truthfully.

(Political Learning Efficacy Scale Questions)

Now, please indicate your level of agreement or disagreement with the statements that
follow by recording a number from 0 to 10 using the scale below

0 1 2 3 4 5 6 7 8 9 10
Strongly Neutral Strongly
Disagree Agree
*1 am conﬁdent that I can learn about political issues if I want to

*I am conﬁdent that I know where to go to locate good political information

*I am conﬁdent that I can seek out information related to a political decision

*I am conﬁdent that I can understand most political information in general

I am conﬁdent that I can ﬁnd trustworthy political information

I am conﬁdent that I can ﬁnd accurate political information

(*PLE Scale)

(Sources of Political Learning Efﬁcacy)

(Vicarious Learning)

I have ﬁiends who know where to ﬁnd good political information

I have seen how my friends can ﬁnd political information

(Enactive Learning)

If I don’t ﬁnd the political information I am looking for in one source, I try to ﬁnd
another source

91

If a political information source does not have all the information I need, I will visit their
recommended sources of more information

If I like a source of information I am likely to come back to it repeatedly to see if there
are updates

I am conﬁdent I can learn to use a variety of political information sources

(Persuasion)

My friends encourage me to ﬁnd political information

My ﬁiends tell me that I should learn about the issues before I make a political decision
I have been persuaded by ﬁiends or family to ﬁnd political information online

(A ﬁective State)

Political information is somewhat intimidating to me

When I think about looking for political information, I get nervous

When I am looking for political information, I get nervous

Extra items (just curious)
My ﬁiends can help me make good political decisions

(IPE/EPE/PIE)

0 l 2 3 4 5 6 7 8 9 10
Strongly Neutral Strongly
Disagree Agree
(Internal Eﬂicacy)

I consider myself to be well qualiﬁed to participate in politics

I feel that I have a pretty good understanding of the important political issues facing our
country

I think that I am better informed about politics and government than most people
I feel that I could do as good a job in public office as most people
0 1 2 3 4 5 6 7 8 9 10

Strongly Neutral Strongly
Disagree Agree

92

(External Political Eﬁicacy)
People like me don't have any say about what the government does
I don't think public ofﬁcials care much what people like me think

0 1 2 3 4 5 6 7 8 9 10
Strongly Neutral Strongly

Disagree Agree
(Political Information Efficacy Scale (T edesco, 200 7))
I consider myself well-qualiﬁed to participate in politics

I feel I have a pretty good understanding of the important political issues facing our
country.”

I think I am better informed about politics and government than most people

If a friend asked me about the presidential election, I feel I would have enough
information to help my friend ﬁgure out who to vote for

(PLE Predictors/Outcomes)

Now, a slightly different format, please answer the following from the listed choices;
don’t be embarrassed if you don’t know!

(Political Knowledge)

Do you happen to know what job or political ofﬁce is now held by Joe Biden?
1. Secretary of State 2. Vice President 3. Senator 4. Governor 5. State
Representative

Whose responsibility is it to determine if a law is constitutional or not?
1. The President 2. Congress 3. The Supreme Court?

How much of a majority is required for the US. Senate and House to override a
presidential veto?
1. one half 2. three-quarters 3. two-thirds

Which party is generally more conservative?
1. Republicans 2. Democrats

93

Thank you so much for continuing this far, we are more than halfway done. Your
continued accurate and honest answers are a huge help to us.

(Opinion Leadership)

Please rate your agreement with the following statements on the 11-point scale, below
0 l 2 3 4 5 6 7 8 9 10
Strongly Neutral Strongly
Disagree Agree

I have more self-conﬁdence than most of my friends

I like to be considered a leader

I am the kind of person who knows what I want to accomplish in life and how to achieve
it

I am inﬂuential in my group of friends

(Political Participation)

We are interested in your actions during the 2008 election season. This is a list of some
things people do regarding government and politics, please let us know if you have done
any of the following in the past year (Check all that apply)

Watched a 2008 election debate on television

Watched a 2008 election debate (clips or entirety) on the Internet
Signed an online or paper petition

Attended a public meeting on town or school affairs

Wrote my congressional representative or senator

Served as an ofﬁcer of a club or organization

Wrote a letter to the newspaper regarding a political issue
Volunteered or worked for a political organization or candidate
Held or ran for political ofﬁce (including student government)
Posted comments on a political blog or political website
“Friended” a candidate online

(Social Networking Site Use)
Some ﬁnal questions...

Are you member of a social networking site like F acebook or MySpace?
Yes No

Which social networking site do you use for MOST of your online social networking?
Facebook MySpace Both Other

How many times do you visit your favored site per week?
0-2 times

94

3-5 times
6-10 times
More than 10 times

How much time do you spend on your favored site each time you visit?
1-30 minutes

30 minutes-1 hour

1-2 hours

More than 2 hours

Have you ever visited the social networking site proﬁle of a political candidate?
Yes No

Have you ever joined a political group on a social networking site?
Yes No

Have you ever become a friend or a supporter of a political candidate on a social
networking site?
Yes No

Have you ever become a supporter of a political causes on a social networking site?
Yes No

Have you ever noticed a friend becoming a supporter or fan of a political candidate or
cause on a social networking site?
Yes No

Please indicate your agreement with the following statements
0 1 2 3 4 5 6 7 8 9 10

Strongly Neutral Strongly
Disagree Agree

I use social networking sites to share political information with my friends

I have participated in “real world” political activities (like attending a rally or meeting)
because I heard about them on a social networking site

I was more likely to vote because of political or candidate information I saw on a social
networking site

I still check out or participate in political groups I joined during the 2008 election on
social networking sites ‘

I was more interested in the 2008 election because of information I received on social
networking sites

95

I use/have used social networking sites to share my political views with others

I have found additional sources of political information through social networking sites

In the past month, which of the following activities have you done on a social networking
site? (check all that apply)

Posted a status update that mentions something political
Joined or left a political group

Become a fan or supporter of a candidate or cause

Added an application or bumper sticker that mentions politics
Took a quiz about politics

Posted a wall comment about politics

Discussed politics in a message

Posted a link about politics

Added or deleted political information from my proﬁle

None

Which news/information sources do you use for information? (check all that you use)

Newspaper

Televised news

Online news sites (CNN, FoxNews, MSNBC, etc)
Political blogs (Politico, Wonkette, Drudge Report, etc)
Social networking sites (Facebook, MySpace, etc)
Radio (NPR, talk radio)

Other, please specify

Of these news and information sources, which do you use most often? (select one)
Newspaper

Televised news

Online news sites (CNN, FoxNews, MSNBC, etc)

Political blogs (Politico, Wonkette, Drudge Report, etc)

Social networking sites (Facebook, MySpace, etc)

Radio (NPR, talk radio)

Other, please specify

Finally, just a few personal questions

Were you eligible to vote in the 2008 Presidential Election?
Yes No '

People don’t vote for a variety of reasons. Were you able to vote in the 2008 Presidential
Election?
Yes No

96

What is your gender?
Male Female Prefer not to say

What is your age?
What is your ethnicity?

African American Asian Caucasian Native American Paciﬁc Islander Hispanic or
Latino Other, please specify

What is your educational level?

Some High School

High School

Some College

2-year Degree (Associates)

4-year Degree (Bachelors)

Graduate Degree

In general, do you consider yourself a

Republican Democrat Independent Green Don’t know

Thank you so much for participating!

97

REFERENCES

Ansolabehere, S., Iyengar, 8., Simon, A., & Valentino N. (1994). Does attack
advertising demobilize the Electorate? American Political Science Review, 88(4),
829-839.

Ansolabehere, S,, & Iyengar, S (1995), Going negative: How attack ads shrink and
polarize the electorate. New York, NY: The Free Press.

Atkin, C., & Heald, G. (1976). Effects of political advertising. Public Opinion Quarterly,
93(4), 901-910.

Austin, E. W. & Pinkleton, B. E. (1995). Positive and negative effects of political
disaffection on the less experienced voter. Journal of Broadcasting & Electronic
Media, 39(2), 215-236.

Alexa (2009). Top sites, United States. Retrieved from
http://www.alexa.com/topsites/countries/U S

Balch, G]. (1974). Multiple indicators in survey research: the concept “sense of political
efﬁcacy.” Political Methodology, 1, 1-43.

Bandura. A. (1977). Self-efﬁcacy: Toward a unifying theory of behavioral change.
Psychological Review, 84, 191-21 5.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice Hall.

Bandura, A. (1989). Human agency in social cognitive theory. American Psychologist, 44,
1 175-1 184.

Bandura, A. (1994). Self-efﬁcacy. In V. S. Ramachaudran (Ed.), Encyclopedia of human
behavior (V 01. 4, pp. 71-81). New York, NY: Academic Press. (Reprinted in H.
Friedman [Ed.], Encyclopedia of mental health. San Diego, CA: Academic Press,
1998)

Bandura, A. (1997). Self-eﬂicacy: The exercise of control. New York, NY: Freeman.

Bandura, A. (2006). Guide for constructing self-efficacy scales. In T. Urdan & F. Pajares
(Eds), Self-Eﬂicacy Beliefs of Adolescents. Greenwich, CT: Information Age
Publishers.

Best, S. J., & Krueger, B. S. (2005). Analyzing the representativeness of intemet
political participation. Political Behavior, 27(2), 183-216.

98

Boulianne, S. (2009). Does Internet use affect engagement? A meta-analysis of research.
Political Communication, 26(2), 193-21 1.

boyd, d. m., & Ellison, N. B. (2007). Social network sites: deﬁnition, history, and
scholarship. Journal of Computer-Mediated Communication, 13(1) Article 11.
Retrieved from http:/ljcmc.indiana.edu/vol 1 3/issue1/boyd.ellison.html

Burns, N., & Schlozman , K. L. (2001). The private roots of public action: Gender,
equality, and political participation. Cambridge, MA: Harvard University Press.

Campbell, A., Gurin, G., and Miller, W. E. (1954). The voter decides. Evanston, IL: Row
Peterson & Co.

Conway, M. M. (2000). Political participation in the United States. Washington DC: CQ
Press.

Center for Information & Research on Civic Learning (CIRCLE) (2005). The youth vote
2004. Retrieved from
http://www.civicyouth.org/PopUps/FactSheets/F S_Youth_Voting_72-04.pdf

Center for Information & Research on Civic Learning (CIRCLE) (2008). Turnout by
Education, Race and Gender and Other 2008 Youth Voting Statistics. retrieved from
http://www.civicyouth.org.

Clark, L. A., & Watson, D. (1995). Constructing validity: basic issues in objective scale
development. Psychological Assessment, 7(3), 309-319.

ComScore (2007). More than half of MySpace visitors are now age 35 or older, as the
site’s demographic composition continues to shift. Retrieved November 17, 2007,
from
http://www.comscore.com/press/release.asp?press=1019

Converse, PE. (1972). Change in the American electorate. In A. Campbell & P. E.
Converse (Eds), The human meaning of social change (pp. 263-337). New York,
NY: Sage.

Craig, C. C., Niemi, R. G., and Silver, G. E. (1990). Political efﬁcacy and trust: a report
on the NES pilot study items. Political Behavior, 12(3), 289-314.

Dearing, J. W., & Rogers, E. M. (1996). Agenda setting. Thousand Oaks, CA: Sage.

Delli Carpini, M. X. & Keeter, S. (1991). Stability and change in the US. public’s
knowledge of politics. Public Opinion Quarterly, 55(4), 583-612.

Delli Carpini, M. X. & Keeter, S. (1993). Measuring political knowledge: putting ﬁrst
things ﬁrst. American Journal of Political Science, 3 7(4), 1 179-1206.

99

Delli Carpini, M. X. (2000). Gen.com: youth, civic engagement, and the new information
environment. Political Communication, 1 7, 341-349

Delli Carpini, M. X. (2004). Mediating democratic engagement. In L.L Kaid (Ed),
Handbook of political communication research (395-434). New York, NY:
Lawrence Erlbaum Associates.

DeVellis, R. F. (2003). Scale development: Theory and applications. Thousand Oaks,
CA: Sage.

Diddi, A., & LaRose, R. (2006). Getting hooked on news: Uses and gratiﬁcations and the
formation of news habits among college students in an Internet environment. Journal
of Broadcasting & Electronic Media, 50, 193-210.

Drew, D., & Weaver, D. (2006). Voter learning in the 2004 presidential election: did the
media matter? Journalism and Mass Communication Quarterly, 83(1), 25-42.

Ellison, N. B., Steinﬁeld, C., & Lampe, C. (2007). The beneﬁts of Facebook “friends:”
social capital and college students use of online social network sites. Journal of
Computer-Mediated Communication, 12 (3), Article 1. Retrieved from

. http:// j cmc.indiana.edu/vol 1 2/ issue4/ellison.html

Eveland, W. P., Hayes, A. F., Shah, D. V., & Kwak, N. (2005). Understanding the
relationship between communication and political knowledge: a model comparison
approach using panel data. Political Communication, 22(3), 423-446.

Facebook (2009) Facebook Statistics, Press Room. Retrieved July 20, 2009 from
http://www.facebook.com/press/info.php?statistics

Finkel, S. (1985). Reciprocal effects of participation and political efﬁcacy: a panel
analysis. American Journal of Political Science. 29(4), 891-913.

Finkel, S. & Geer, J. (1998). A spot check: casting doubt on the demobilizing effect of
attack advertising. American Journal of Political Science, 42(2), 573-595.

Freedman, P., & Goldstein, K. (1999). Measuring media exposure and the effects of
negative campaign ads. American Journal of Political Science, 43(4), 1189-1208.

Freedman, P., Franz, M., & Goldstein, K. (2004). Campaign advertising and democratic
citizenship. American Journal of Political Science, 48(4), 723-741.

Hayes, R. A. (2008). Providing What they Want and Need on Their Own Turf:

Informational Content and Interactivity of Web and Social Networking Sites. Paper
presented at National Communication Association Annual Meeting, San Diego, CA.

100

Hayes, R. A., Zube, P. J., & Isaacson, T. E. (2008). Reaching Out on Their Own Turf:
Social Networking Sites and Campaign 2008. Paper presented at Web 2.0: New
Media and Politics, Royal Holloway University, Egham, United Kingdom.

Hayes, R. A. (2009). The Old and New in Political Communication: Are Young Voters
Receptive to Either? Paper presented at the American Academy of Advertising
Annual Meeting, Cincinnati, OH.

Huber, G. A., & Arceneaux, K. (2007). Identifying the persuasive effects of presidential
advertising. American Journal of Political Science, 51(4), 957-977.

Jennings, M. K., & Zeitner, V. (2003). Internet use and civic engagement: a longitudinal
analysis. Public Opinion Quarterly, 67, 311—334.

Johnson, T. J., & Kaye, B. K. (2000). Using is believing: the inﬂuence of reliance on the
credibility of online political information among politically interested Internet users.
Journalism and Mass Communication Quarterly, 77(4), 865-879.

Johnson-Cartee, K., & Copeland, G. A. (1997). Inside political campaigns: Theory and
practice. Westport, CT: Praeger.

Johnston, R., Hagen, M. G., & Jarnieson, K. H. (2004). The 2000 presidential election
and the foundations of party politics. Cambridge, UK: Cambridge University Press.

Kaid, L. L., Tedesco, J .C., & McKinney, M. S. (2004). Political Information Efﬁcacy and
Younger Voters. Paper presented at National Communications Association, Chicago,
2004.

Kaid, L. L., McKinney, M. S., & Tedesco, J. C. (2007a). Political information efﬁcacy
and young voters. American Behavioral Scientist, 50(9), 1093-1111.

Kaid, L. L., Postelnicu, M., Landreville, K., Yun, H.-J., & LaGrange, A. G. (2007b). The
effects of political advertising on young voters. American Behavioral Scientist,
50(9), 1137-1151.

Kaid, L. L., & Postelnicu, M. (2005). Political advertising and responses of young voters:
Comparison of traditional television and Internet messages. American Behavioral
Scientist, 49(2), 265-278.

Katz, E. (1973). The two-step ﬂow of communication: an up-to-date report of a
hypothesis. In B.M Enis and K.K. Cox (Eds), Marketing Classics (pp. 175-193).

Kenski, K., & Stroud, N. J. (2006). Connections between Internet use and political

efﬁcacy, knowledge, and participation. Journal of Broadcasting & Electronic Media,
50, 173-192.

101

Kern, M., & Just, M. (1995), The focus group method, political advertising, campaign
news and the construction of candidate images. Political Communication, 12, 127-
145.

Kim, E. (2003). Young people, media use, and voter turnout: an analysis of the 2000
national election study” presented at the annual conference of the International
Communication Association, San Diego, CA.

Kim, B. (2004). "Knowledge about Politics, Communication Behavior, and the Two
Dimensions of Political Efﬁcacy: An Analysis of the 2000 National Election Study"
Paper presented at the annual meeting of the International Communication
Association, New Orleans LA.

Klapper, J. T. (1960). The eﬁrects of mass communication. Glencoe, IL: Free Press.

Krueger, R. A., & Casey, M. A. (2000). Focus groups: A practical guide for applied
research. Thousand Oaks, CA: Sage.

LaRose, R., & Eastin, MS. (2004). A social cognitive theory of intemet uses and
gratiﬁcations: toward a new model of media attendance. Journal of Broadcasting &
Electronic Media, 48(3), 358-377.

Lenhart, A., Madden, M., & Hitlin, P. (2005). Teens and technology: Youth are leading
the transition to a fully wired and mobile nation. Retrieved from
http://www.pewintemet.org/pdfs/PIP_Teens_Tech_July2005web.pdf

Levine, P., & Lopez, M. H. (2004). Young people and political campaigning on the
Internet. Retrieved from
http://www.civicyouth.org/research/products/fact_sheets.htrn

Likert, R. (1961). The human organization: Its management and value. New York,.NY:
McGraw-Hill

Lippman, W. (1922). Public opinion. New York, NY: Free Press.

Loevinger, J. (1957) Objective tests as instruments of psychological theory.
Psychological Reports, 3, 63 5-694.

Lupia, A., & Philpot, T. S. (2005). Views from inside the Net: how Websites affect
young adults’ political interest. Journal of Politics, 67(4), 1122—1142.

MacKenzie, S. B. (2003). The dangers of poor construct conceptualization. Journal of the
Academy of Marketing Science, 31(3), 323—326.

102

Market Research (2007). College students online: A parallel life on social networks.
Retrieved from
http://www.marketresearch.com/product/print/default.asp?g=1&productid=1 53 7974

McCombs, M., &. Shaw, D. (1972). The agenda-setting function of mass media. Public
Opinion Quarterly, 36, 176-187.
Milbrath, L. & Goel, M. (1977). Political Participation (2nd ed.). Chicago, IL: Rand
McNally.

Miller, AH. (1974) Political Issues and Trust in Government: 1964-1970. American
Political Science Review. 68, 951-972.

Minnesota Public Radio (2004). Ad spending in presidential race triples that of 2000.
Retrieved from
http://news.minnesota.publicradio.org/features/2004/1 1/01_ap_adspending/

Mowen, J. C., & Voss, K. E. (2008). On building better construct measures: implications
of a general hierarchical model. Psychology and Marketing, 25(6), 485-505.

Morrell, M. E. (2003). Survey and experimental evidence for a reliable and valid measure
of Internet political efﬁcacy. Public Opinion Quarterly, 6 7(4), 589-602.

Niemi, R.G, Craig, S.C, & Mattei, F. (1991). Measuring internal political efficacy in the
1988 national election study. The American Political Science Review, 85(4), 1407.

Nunnally, J. C. (1978). Psychometric theory (2nd ed). New York, NY: McGraw-Hill.

Oskamp, S., & Schultz, P. W. (2005) Attitudes and opinions. Mahwah, NY: Lawrence
Erlbaum and Associates.

Padgett, D. K. (1998). Qualitative methods in social work research: Challenges and
rewards. Thousand Oaks, CA: Sage.

Pasek, J., Kenski, K., Romer, D., & Jamieson, K. H. (2006). America’s youth and
community engagement. Communication Research, 33(3), 115-135.

Pattie, C. & Johnston, R. (1998). Voter turnout at the British general election of 1992:
rational choice, social standing, or political efﬁcacy? European Journal of Political
Research, 33(2), 263-83.

Pollock, P. (1983). The participatory consequences of internal and external efﬁcacy.
Western Political Quarterly. 36(3), 400-409.

Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community.
New York, NY: Touchstone.

103

Rainie, L., Cornﬁeld, M., & Horrigan, J. (2005). The Internet and cammign 2004.
Retrieved ﬁ'om http://www.pewintemet.org/pdfs/PIP_2004_Campaign.pdf

Reef, M. J ., & Knoke, D. (1999). Political alienation and efﬁcacy. In P.R. Shaver, & J .P.
Robinson (eds. ) Measures of political attitudes, Volume 2 of Measures of Social
Psychological Attitudes (413-464). San Diego, CA: Academic Press.

Rogers, E. M. (2004). Theoretical diversity in political communication. In L.L Kaid (Ed)
Handbook of political communication research (3-16). New York, NY: Lawrence
Erlbaum Associates.

Rosenstone, S.J., & Hansen, M.J. (1993) Mobilization, participation, and democracy in
America. New York, NY: Macmillan Publishing Company.

Rossiter, J. R. (2002). The C-OAR-SE procedure for scale development in marketing.
International Journal of Research in Marketing, 19(4), 305—335.

Rowan, N., & Wulff, D. (2007). Using qualitative methods to inform scale development.
The Qualitative Report. 12(3), 450-466. Retrieved from
http://www.nova.edu/ssss/QR/QRl 2-3/rowan.pdf

Smith, A., & Rainie, L. (2008). The intemet and the 2008 election. Report of the Pew
Internet and American Life Project. Retrieved from
http://www.pewinternet.org/~/media//Files/Reports/2008/PIP_2008_election.pdf.pdf

Sotirovic, M., & McLeod, J. M. (2004). Knowledge as understanding: the information
processing approach to political learning. In L.L. Kaid (Ed), Handbook of Political
Communication Research (357-394). New York, NY: Lawrence Erlbaum Associates.

Steinﬁeld, C., Ellison, N., & Lampe, C. (2008, May). Net worth: Facebook use and
changes in social capital over time. Paper presented at the 58th Annual International
Communication Association conference, Montreal, Canada.

TechCrunch (2008). Facebook users up 89% over last year; demographic shift. Retrieved
from
http://www.techcrunch.com/2007/07/06/facebook-users-up-89-over-last-year-
demographic-

Tedesco, J.C. (2007). Examining Internet interactivity effects on young adult political
information efﬁcacy. American Behavioral Scientist, 5 0(9), 1183-1194.

Tindell, J. H., & Medhurst, M. J. (1998). Rhetorical reduplication in MTV’s Rock the
Vote campaign. Communication Studies, 49(1), 18-28.

United States Census Bureau. (2006). Statistical abstract of the United States. Retrieved
from http://www.census.gov/prod/www/statistical-abstract.htrnl

104

Utz, S. (2009). The (potential) beneﬁts of campaigning via social network sites. Journal
of Computer-Mediated Communication, 14(2), 221-243.

Vitak, J., Smock, A., Zube, P. J ., Carr. C. T., Ellison, N., & Lampe, C. (2009). “Poking”
people to participate: F acebook and political participation in the 2008 election. Paper
presented at the meeting of the International Communication Association, Chicago,
IL.

Wells, S. D. & Dudash, E. A. (2007). Wha’d’ya know? Examining young voters’
political information and efficacy in the 2004 election. American Behavioral
Scientist, 5 0(9), 1280-1289.

Williams, C. B., & Gulati, J. G. (2007). Social Networks in Political Campaigns:
Facebook and the 2006 Midterm Elections’, paper presented at Annual Meeting of
the American Political Science Association, Chicago.

Yoon, K., Pinkleton, B. E, & Ko, W. (2005). Effects of negative political advertising on
voting intention: an exploration of the roles of involvement and source credibility in

the development of voter cynicism. Journal of Marketing Communication, 11(2),
95-112.

105