|
This paper was presented at the Association for Education in Journalism and
Mass Communication in San Antonio, Texas August 2005.
If you have questions about this paper, please contact the author
directly. If you have questions about the archives, email
rakyat [ at ] eparker.org. For an explanation of the subject line,
send email to
[log in to unmask] with just the four words, "get help info aejmc," in the
body (drop the "").
(Jan 2006)
Thank you.
Elliott Parker
====================================================================
What makes deliberation possible?
Structural Equation Models of online and face-to-face deliberation process
by
Yun Jung Choi
Doctoral Student
Hyo Jung Kim
Master's Student
S. I. Newhouse School of Public Communications
Syracuse University
Contact:
Yun Jung Choi
S. I. Newhouse School of Public Communications
Syracuse University
215 University Place
Syracuse, NY 13244-2100
[log in to unmask]
*Student Paper
Manuscript submitted to the Communication Theory & Methodology Division
of the 2005 AEJMC, San Antonio, TX
What makes deliberation possible?
A test of online and face-to-face deliberation process
Abstract
This paper identifies factors that influence the deliberation process
by proposing a deliberation model. Six ideas, reciprocity, reasoned
discourse, freedom of expression, open-mindedness, empathy and public
interest were recognized as concepts that compose deliberation
construct, and deliberation was measured based on the six concepts.
The study found that those with high political interest are likely to
engage in political discussion and seek diverse political views
different from their own, and this lead them to exchange their
political opinions with those who have different political views from
them. A news concept, exposure to diverse political views was
introduced as a strong mediating variable that influence the
deliberation in the study. Internet media are found to have more
impact on the deliberation process than the traditional media.
As increasing number of people are discussing politic on the
internet, online deliberation model is also proposed. The online
deliberation model is similar to the face-to-face deliberation model,
but less variance in the online deliberation construct is explained
by the model. This indicates that there are more factors influencing
the online deliberation than the face-to-face deliberation.
Deliberation is a political process in which diverse ideas are
exchanged by participants of diverse backgrounds, which can
contribute to more informed and legitimate political decisions among
the public (Fishkin, 1995; Cohen, 1989). Deliberation is considered
as an integral process of democracy; however, whether the
deliberation process is taking place in the real world is
questionable because people like to talk with others who share
similar views with themselves and avoid discussing with opponents
(Jonas, Schulz-Hardt, Frey, & Thelen, 2001). This study provides a
model of deliberation by identifying factors that lead to the
deliberation process: how political interest, political knowledge,
media use, and frequency of interpersonal discussion influence the
public to engage in the deliberation process.
As an increasing number of people exchange their political viewpoints
with unknown others on the internet, the potential of online
deliberation is getting more attention. It is assumed by many that
online media and discussion forums allow more individuals and
minority groups to speak out, thus exposing the public to a more
diverse spectrum of views. Online discussions and forums are believed
to facilitate the online deliberation process. However, others are
concerned that online media and discussion forums create a fragmented
environment which encourages partisan cleavages and group
polarization (Jones, 2002).
The deliberation concept is linked with and provides the basis for
many political and social constructs, such as social capital, social
trust, political tolerance, and democracy (Mutz, 2002, Putnam, 2000).
However, the concept of deliberation is often used abstractly and has
not been thoroughly explicated. This paper identifies six dimensions
– reciprocity, reasoned discourse, freedom of expression, open
mindedness, empathy, and public interest – as the components of deliberation.
Deliberation
The concept of deliberation derives from deliberative democracy.
Waldman (2001) describes deliberative democracy as "a system in which
decisions are made not by coercion or bargaining among interests but
through a discursive process in which citizens collectively consider
and debate alternatives." Deliberation is a sub-concept of
deliberative democracy that is often refereed as an act of engaging
in political discussion and the quality of discussion that
contributes to deliberative democracy.
An early attempt to define deliberation has been made by Habermas
(1981, 1984). He acknowledges that in ideal situations people should
solve problems by inviting and encouraging arguments for all sides in
pubic sphere. Waldman (2001) describes deliberation as "reasoned
discussion among equals about public issues with the goal of
ascertaining the nest course of action so as to optimize the common
good." Rawls (1971/ 1999) defines deliberation as a process in which
the reasoned public with equal voice and positions freely
participates in deliberative discussion for a rationally motivated
consensus. Gutmann and Thompson (1996) contend that deliberation is
an ongoing process in which rational opinion is presented equally and
reciprocally. According to Bohman, deliberation is reaching reasoned
agreement among free and equal citizen (1996, p. 322).
Deliberation as a dependent variable
The attempt to operationalize and quantify the process of
deliberation has been made by a handful of scholars. In an
experimental study (Ryu, 2004), participants' deliberation about a
school issue was measured as a dependent variable with nine items
that construct four concepts of deliberation, willingness to express
opinion, attention to their discussion, understanding of their
discussion, and flaming. In a survey study (McLeod et al., 1999),
three questions about whether to attend a local forum where citizens
discuss local or community problems is called, the likeliness of
speaking up at the meeting, the likeliness of expressing an opinion
that is different from those of others at the meeting were asked to
gauge deliberation.
More thorough measurement of deliberation has been developed in
content analysis studies. In an analysis of discourse on abortion in
two German newspapers, Gehards (1997) identified representativeness
of the actors covered in the newspaper articles, the degree of
respect expressed toward other participants in the debates, degree of
justification of claims, and rationality of discourse quality as the
four core dimensions of deliberation. Recently, Steenbergen and his
colleagues (2003) developed a discourse quality index (DQI) that
serves as a quantitative measure of discourse of deliberation. Their
definition of DQI was based on the Habermas' discourse ethics. The
discourse quality index measured nine concepts – openness of
participation, justification of assertions and validity of claims
exchanged by participants, the concerns for common good, the idea
that participants should consider common good, mutual respect among
discussion participants, the constructive politics, the idea that
participants in a discourse should at least attempt to reach mutually
acceptable comprises solutions, and authenticity, the absence of
deception in expressing intentions – as indicators of deliberation.
Defining deliberation
Definitions of deliberation provided by political and communication
scholars indicate that deliberation is composed of two dimensions,
the actual act of engaging in political discussions to exchange
different views and the quality of the political discussion itself.
In this study both dimensions are considered as the components of
deliberation.
Based on the literature, six concepts – reciprocity, reasoned
discourse, freedom of expression, open mindedness, empathy, and
public interest – are identified as the essential components of
deliberation. Reciprocity refers to the actual behavior of engaging
in political discussion with opponents who have different views
whereas the rest of concepts represent the discourse quality. With
those six concepts, deliberation can be defined as "a process of
freely exchanging reasoned political views with those who have
different political views with open mindedness that can result in
understanding of different political views, which can contribute to
common good."
Reciprocity: Deliberation should accompany reciprocal
exchange of two opposing ideas. It involves two-way communication
processes where different political views are voiced and heard among
participants (Park, 2002). Participants should be wiling to express
their variety of political perspectives and views and further
persuade opponents who have different political views (Waldman,
2001). Therefore, disagreement is an essential part of deliberation.
Disagreement does not have to result in agreement in deliberation,
but the process of exchanging different perspectives and arguing for
their position should accompany the deliberation process.
Reasoned discourse: It is believed that much mass opinion is
disorganized (Converse, 1964) and without basis in belief or
knowledge (Herbst, 1993; Zaller, 1992). However, in deliberation,
reasoned ideas supported by some evidence or justification should be
exchanged (Waldman, 2001). Arguments that are not supported with
reasons and justifications are not readily understood by opponents
who have different political views, and thus no meaningful public
discourse can be made. When participants are fully aware of
justification of their political views as well as others, true
deliberation occurs (Park, 2000).
Cappella et al. (2002) proposes "argument repertoire" as one measure
of opinion quality. It provides a way to measure how well reasoned
each stated opinions are for one's own opinions. He thought that the
understanding of reasons for their own opinions and reasons why
others might disagree with my opinion provides one means of measuring
opinion quality.
Freedom of expression: Deliberation occurs when there are no
constraints on topics or point of views that can be discussed among
participants. The public should feel free to introduce new ideas and
perspectives to the discussion (Steenbergen et al., 2003). No one
should be prevented from exercising freedom of expression due to
internal or external coercions (Cohen, 1989).
Open mindedness: Participants in the deliberation discourse must have
respect for others and their opinions. Steenbergen et al. (2003)
identifies three dimensions of respect: respect toward the group,
respect for demands under discussion, and respect toward
counterarguments, arguments raised by opponents that contradict my
opinion. Respect for different perspective can elicit more complete
arguments with underlying reasons and justifications for such a view
(Stewart & Logan, 1998).
Empathy: A sense of empathy should be shared among participants in
deliberation. Deliberative participants develop a sense of empathy
after careful reflection of opponents' point of view (Rogers, 1980).
Although it is impossible to put oneself in another person's
position, empathy helps participants to pay more careful attention to
other's arguments and understand opposing views even though they do
not necessarily agree with their opponents' opinion.
Public interest: The discussion should involve public issues and the
consideration for the common good (Steenbergen et al., 2003). If the
discussion reflects self interest of a small group or the arguments
presented in the discussion harm the common good, the discussion does
not become deliberation.
Online deliberation
Deliberation is not limited to face-to-face situations. Deliberation
can occur on the internet through diverse channels such as online
forums, chat rooms, usernet groups, and listserv emails. People can
log onto the internet and join political discussion forums with
people who can not be encountered otherwise due to physical and
social constraints. This study will look at how online deliberation
differs from the traditional deliberation process that occurs in
face-to-face situations.
The potential of the internet as the deliberative public sphere has
been evaluated. Engaging in chat room dialogue exposes participants
to a variety of opinions, attitudes, and sources of information that
they may have never encountered otherwise (Weger & Askhus, 2003). The
internet provides a perfect place to find different views of a very
diverse group of people who are at the same time open to such
difference and disagreement needed for deliberation.
The internet also allows citizens to "break our public silence" and
become involved in public communication" (Fernback, 1997). The
internet provides a more liberated environment for those who are
afraid to speak up their voices because anonymity is relatively
guaranteed on the internet, as long as they do not have to log on
with their real names. The internet may liberate people from
psychological constraints and help them express opinions that are
different and sometimes the opposite of their discussion partners' opinions.
However, the contribution of the internet to deliberation is
questioned by some. The internet may provide a virtual public sphere
where diverse groups of people can join and exchange opposing ideas,
but the question is whether all these different people actually do
find each other on the internet, or whether they seek the
like-minded. Sunstein (2001) expresses worries about the polarizing
effects of the internet might hold. He fears that people will seek
the like-minded to talk on the internet. This is possible because the
internet provides very segmented online forums where only those
people with similar opinions can meet and discuss, unlike the real
world where people are sometimes forced to mingle with people with
different opinions.
The online deliberation process is very different from the
deliberation process that occurs with face-to-face interactions. The
diversity of the networked, liberated environment for free speech,
equality in participation provided by the internet may facilitate
deliberation, while segmented online forums may hinder deliberate
discussions on the internet.
Factors that influence deliberation
Factors that affect deliberation also have been identified.
Political Interest
Political interest is defined as the "degree to which politics
arouses a citizen's curiosity" (Van Deth 1990, p. 278), and political
information seeking and use of newspaper as a consequence of
generalized interest in political life (McLeod et al, 1996, 1999).
Political interest has been shown to be linked to news exposure,
political knowledge, and some dimensions of political participation.
Schuefele and Shah (2000) argued that political interest spurs
increased attention to news media and encourage other forms of
political information consumption. In their survey, respondents with
a higher level of political interest were more likely to expose
themselves to political contents of the news media.
Political interest is also known to be correlated with political
knowledge (Tilley et al.,2004). Many authors have tended to posit
political interest and media exposure as its primary exogenous causes
for political knowledge (Bartle, 2000). Tolbert et al. (2003) also
found that people with higher political interest and efficacy, media
consumption, and education have greater political knowledge from
their survey studies.
The relationship between political interest and exposure to diverse
views has been rarely studied. Rather, political interest has been
known to be positively related to political discussions. Schuefele &
Shah (2000) found that political interest is positively related to
civic activities and that it fosters engagement in collective
actions. We are particularly interested in political discussion among
various forms of civic activities. It can be assumed that political
interest has a positive relationship with the frequency of political
discussion.
In the sense that political interest spurs increased attention to
news media as well as and encourage participating in political
discussions, we can assume that political interest may lead to
exposure to diverse views.
Media Use
Media use, how much people read, watch and listen to news stories
through mass media, can increase political knowledge. Moy (2004)
found that attention to local news on TV and news in newspapers
enhanced perceptions of knowledge. Political knowledge is gained as
the consequence of the news consumption.
The role of media use has been given little attention in the
political participation literature (McLeod et al.,1999). Recently,
however, some researchers found that public affairs content in
newspaper and television is positively related to traditional forms
of participation (Chaffee & Kanihan, 1997; McLeod et al., 1996; Moy,
Scheufele, & Holbert, 1999). Particularly, McLeod et al. (1996) found
that consumption of both local newspaper and television public
affairs content to positively influence a citizen's willingness to
participate in a public forum. Based on these early studies, we
assume the positive relation between media use and the frequency of
political discussion.
In addition, Media use is also assumed to influence the exposure to
diverse views. Mutz and Martin (2001) assert that mere exposure to
mass media messages increase exposure to diverse views because media
reports tend to include a diversity of opinion to make stories look
more balanced. McLeod et al. (1999) also found the positive
correlation between media use and heterogeneity of network. Mutz and
Martin (2001) suggest that this role of media is rooted in the
relative difficulty of selectivity exposing oneself to those sources
of information. On the other hand, as for internet use other than
newspaper and television, such researchers as Chaffee (2001) concern
that the user-controlled nature of the internet might return campaign
communication to the selective-exposure pattern that obtained in the
era of highly partisan newspapers.
Exposure to diverse views
One's willingness to expose oneself to diverse point of views by
engaging in conversation with someone with different political views
or by consuming mass media whose opinion differ from mine is an
important gateway to deliberation. Deliberative discussion does not
happen unless one is exposed to different opinions. By the way, it
should be noted that this concept of exposure to diverse views is
distinguished from reciprocity, one of the six concepts which define
deliberation. Reciprocity refers to actual exchanging of ideas that
includes disagreement and persuasion among discussants, whereas
exposure to diverse views refers to paying attention to different
opinions in mass media and political discussion without expressing
one's own opinions.
The concept of exposure to diverse views is sometimes referred as
network heterogeneity (McLeod et al., 1999), which is defined as a
network of heterogeneous people in their social network in terms of
their demographics and political opinions. Individuals who compose a
network whose members come from diverse socio-political backgrounds
are more likely to encounter a diversity of viewpoints and opinions
in their political discussion (McLeod, Sotirovic, & Holbert, 1998).
This exposure to diverse viewpoints leads to deliberation. McKuen
(1990) addresses that through heterogeneous networks, comprised of
diverse opinion, public dialogue and deliberative democracy come alive.
In the classical two-step flow study, (Berelson et al., 1954,
Lazarsfeld et al., 1944) the term "cross-pressure" is used to
represent people's exposure to diverse opinions. "Cross-pressures"
are created by disagreement among participants, and these
cross-pressures lead to political withdrawal as a consequence of
psychic stresses produced by political uncertainty and mixed
political messages.
Mutz (2002) also presents a similar concept called "cross-cutting
exposure." She found that cross-cutting network increases awareness
of rationales for one's own viewpoints and also awareness of
rationales for oppositional viewpoints and these awareness increase
political tolerance. However, she also found that cross-cutting
exposure discourages political participation (Mutz, 2002). Exposure
to those with political views different from one's own creates
greater ambivalence about political options, and this makes it more
difficult to take part in political actions.
Two opposing views have been presented concerning the exposure to
diverse views. Encountering diverse viewpoints may lead to more
active political participation while this may also lead to confusion
among the public and thus withdrawal from the politics. Although two
opposing views are presented, it is difficult to deny that exposure
to diverse views provide foundation for the deliberative discussions
because exchanging different ideas does not take place unless
exposing oneself to other opinions precedes.
One important point to be made at this point is that exposure to
diverse views does not happen frequently. People try to avoid
political discussions that are likely to be controversial (Graber,
1984). Many express that politics and religion are topics they avoid
because they considered them potentially divisive. At the same time,
people are talking with same clusters of people and chance of meeting
someone who has completely different political view is decreasing as
the social network filters the kind of people we interact with in a
daily basis (Waldman, 2001; Huckfeldt, 2004).
Exposure to diverse views through mass media also educates and
prepares the public for deliberative discussion. However, whether
people actually look for media messages that contradict their
opinions is questionable. People tend to selectively expose
themselves to information and media outlets that confirm their
attitudes and opinions (Jonas, Schulz-Hardt, Frey, & Thelen, 2001).
Exposure to diverse views provide basis for the deliberation, but it
doesn't happen frequently.
Frequency of political discussion
The frequency of political discussion differs from deliberation in
that it just records the frequency of political discussion without
considering the quality of discourse. Carpini & Keeter (1996) suggest
that political discussion is related to political knowledge. It is
also suggested that the impact of hard news use on political
knowledge is moderated by a person's interpersonal discussion about
politics (Schuefele, 2002). The frequency of political discussion may
predict deliberation, because those who frequently engage in
political discussions are more likely to encounter opponents who have
different political views and exchange ideas with them. Thos who
avoid political discussion are not likely to be encounter discussants
who have different political views, and thus, deliberation can not occur.
Political Knowledge
In this study, political knowledge is conceptualized as factual or
current-event knowledge, which is knowledge about political issues or
politicians that are relevant in a contemporary political context.
Political knowledge is considered as a key form of citizenship and,
without this, citizens are unable to understand key issues (Perloff,
1998). With respect to participatory behavior, knowledge fulfills at
least three purposes (Wolfinger & Rosenstone, 1980): allows citizens
to make informed decisions, creates a sense of civic duty among
citizens, and increases familiarity with bureaucratic institutions
and political process (Scheufele, 2002). This political knowledge
will influence deliberation because without political knowledge,
people cannot make valid rational arguments, which is a key concept
in deliberation.
Demographics
Studies (Huckfeldt & Sprague, 1995; Verba et al., 1993) suggests that
man more than women, whites more than blacks or Hispanics, the
wealthy more than the poor, Republicans more than Democrats engage in
political discussion and exchange their views with others. Also those
with strong political ideologies are more likely to talk than those
with weak political ideologies.
Hypotheses
We propose that political interest, political knowledge, news
exposure, frequency of political discussion, and exposure to diverse
views influence deliberation directly and indirectly. The summary of
the hypotheses is presented in a proposed model of the deliberation
process (Figure 1).
[Figure 1 inserted about here]
H1: Political interest will have direct positive influence on
political knowledge, frequency of political discussion, and exposure
to diverse views.
H2: News exposure will mediate the relationship between political
interest and the three variables, political knowledge, frequency of
political discussion, and exposure to diverse views.
H3: Political knowledge, frequency of political discussion, and
exposure to diverse views will have direct positive influence on deliberation.
Finally, a research question about the online deliberation process is asked.
R1: How does the online deliberation process different from the
face-to-face deliberation?
Methods
In order to test the proposed theoretical model, an online survey was
conducted on the international journalism and mass communication
educators and their students. The survey was conducted as a part of
the World Inventory Media Use Survey Project conducted at a
northeastern university, where seven research teams
participated.[1] The online questionnaire was designed and
distributed using an online survey website called the Surveymonkey
(http://www.surveymonkey.com).
Sampling
Snow ball sampling method was used in order to reach broad levels of
participants in the world. The survey questions were sent to sampled
members affiliated with communication associations, IAMCR, WAPOR,
ICA, NCA, and AEJMC, and then the receivers were asked to forward the
messages to their students and colleagues. For IAMCR and WAPOR, the
entire list of members was included in the sample. For ICA and NCA,
every third email address in the alphabetized directory list was
included in the sample. For AEJMC, every second address in the
alphabetized directory list with an American suffix (.com, .net,
.org, .edu), and every address with a non-American suffix, such as
.il, .mx, .de, were included in the sampling. When email addresses
were not available, the next available address was used.
In order to increase the sample size for the international
respondents, websites for journalism and communication schools
located in countries other than the U.S. were searched and faculty
and students email addresses found from those websites were included
in the sample. As the results a total of 4610 survey invitation
emails were sent, and 807 people completed the survey.
Measurement
The analyses involved control exogenous variables and endogenous variables.
Exogenous variables
Control variables included age, gender, education, income, and race.
Age was measured by asking how old the participants were on their
last birthday. Education was accessed by asking the years of school
attendance. Income was measured by asking people to rate their house
hold income in comparison to other people in their country on a
nine-point scale from one (least income) to nine (most income). Fro
gender, male was coded 1 and female was coded 2. Whites were coded 1
and other ethnicities were coded 2.
Endogenous variables
News exposure
News exposure was measured separately for television news, newspaper,
and internet news. On average, how many days a week people spend
watching television, reading newspapers and browsing interne was
asked. On those days, how much time people send getting news from
those sources were measured in hours and minutes with open-ended
questions. The total days of week multiplied by the minutes of news
consumption per day yielded the news exposure score for each medium.
Political interest
Three questions were indexed to measure political interest
(Cronbach's Alpha = .93). How much respondents are interested in
politics; how worthy of time they think pursuing political issues
are, and how closely they follow news about political events were
measured on a five-point Likert-scale.
Political knowledge
Ten true or false questions on factual knowledge of current political
events and political figures were asked. Each of the ten items was
scored 1 for correct answers and 0 for incorrect answers. The
aggregation of the total correct scores became the political
knowledge score.
Frequency of political discussion
The frequency of political discussion in the face-to-face situations
and the online context was measured. For the face-to-face situations,
questions on how frequently people are engaged in political talks
with their parents, siblings, friends, teachers, spouse, and
co-workers or colleagues were asked on a five-point Likert-scale
raging from strongly agree to strongly disagree. The six questions
were indexed to yield a single score (Cronbach's Alpha = .75). The
frequency of political discussion on the internet was measured by
asking how frequently people share their political views with others
on the internet on a five-point Likert-scale.
Exposure to diverse views
Exposure to diverse political views was measured in two contexts –
the face-to-face situation and the online context. In the
face-to-face situations, the degree to which attention is paid to
diverse political views and diverse media with different political
views was measured to gauge exposure to diverse views.
For the online exposure to diverse view, the diversity of political
discussion forums respondents participate and the diversity of media
with different political views they consume were asked. Those two
items were indexed (Cronbach's Alpha = .91).
Deliberation
Six concepts – reciprocity, reasoned discourse, freedom of
expression, open mindedness, empathy, and public interest – that
compose the deliberation construct were measured with 14 Likert-scale
items. The same set of questions was asked twice for the face-to-face
situations and the online context.
Reciprocity was measured by four questions asking how frequently
respondents talk about politics with people who have different
political views and those who support different political views, how
frequently they try to convince others to change their political
views, and how frequently others try to convince them to change their
political views. The three items were indexed for the face-to-face
(Cronbach's Alpha = .77) and the online situations (Cronbach's Alpha = .61).
Reasoned discourse was measured with two open-ended questions.
Respondents were asked to write as many reasons they can list as to
why they feel favorable toward the political party they support or,
if being independent, the reasons why they are independent. The
second question asked respondents to write as many reasons as they
can think of for why others feel unfavorable toward the political
party they support, or why others might criticize about their being
independent. The number of reasons they provided and the quality of
their answers were coded based on Cappella et al. (2002). One point
was given for every reason the respondent wrote. However, when the
answers were irrelevant or did not make sense they were coded zero.
The combined score of the answers for the two questions produced the
reasoned discourse score.
Freedom of expression was assessed by asking about how freely
respondents can express their opinion even when others disagree with
them as well as how much opportunity they have to express their
opinions. The two questions were combined to produce a single score
for face-to-face (Cronbach's Alpha = .67) and the online models
(Cronbach's Alpha = .67).
Open-mindedness is measured by asking how often their opinions have
been trivialized by those who disagree with them, how often they
respect other people's opinion, and how often they ignore opinions
that are opposite to theirs. Due to a low reliability score, one
question on how often they respect other people's opinion was used
for both face-to-face and online measures.
Empathy was measured by asking how close their discussion partners
are to the respondents, how much they understand the perspectives of
those who disagree with them, and how much talking with people with
different political views helps them to build mutual respect with
their discussion partners. Due to a low reliability score, the
question on how close their discussion partners are to the
respondents was dropped, and two remaining questions were used to
produce indexes for both the face-to-face (Cronbach's Alpha = .57)
and the online (Cronbach's Alpha = .63) measures.
Finally, two questions about public interests were asked, but one
question on how much respondents are willing to withhold their
opinion if that can result in a better society was used in the final
measurement due to a low reliability score among the two items.
Analysis
First, missing values were treated with the method of full
information maximum likelihood estimates provided by the AMOS
program. Full information maximum likelihood estimates tends to be
less biased than other methods such as multiple imputation, listwise
deletion, and pairwise deletion (Arbuckle & Wothke, 1995). Second,
bivariate correlation matrices were created to examine zero-order
correlations for the face-to-face and online deliberation models
separately (Table 3 & 4). Third, structural equation modeling (SEM)
was carried out using the AMOS statistical program. Because
chi-square test is often unreliable with larger samples (Bentler,
1990), two additional model fit indices – the comparative fit index
(CFI) and the root mean square error of approximation (RMSEA) were
used to gauge the model fit. CFI values range from 0 to 1, which .95
indicating a good fit (Hu & Bentler, 1999). For the RMSEA, a value
below .06 indicates a good fit. Path significance was evaluated at
the .05 level.
Results
Means and standard deviations for the variables used in the study are
presented in Table 1. The average age for the sample is 39.89 years
old. The average education level is very high (Mean = 17.02 years,
SD= 5.59), which indicates that a lot of participants had above
Bachelor's degree. About the same proportion of male and female
respondents completed the survey (Table 2) and the majority of people
who participated in the survey are white (74.2%).
[Insert Table 1 about here]
[Insert Table 2 about here]
Two correlation matrixes were presented for face-to-face deliberation
model and the online deliberation model (Table 3 and Table 4). In
both matrixes, exposure to diverse view is correlated with many other
variables while political knowledge is not very well correlated with
other variables.
[Insert Table 3 about here]
[Insert Table 4 about here]
Two structural equation models were tested – one for the face-to-face
deliberation process (Figure 2, Table 5 and Table 6), and the other
for the online deliberation (Figure 3, Table 5 and Table 6). For the
online deliberation, three variables, exposure to diverse view,
frequency of political discussion, and deliberation construct, were
replaced with online exposure to diverse views, frequency of online
political discussion, and online deliberation respectively. The two
models did not produce satisfying goodness of fit values, so both
models had to be modified. As the result, the modified model of the
face-to-face deliberation has the following goodness of fit values:
chi-square = 188.16, d.f.= 116, p < .001; CFI = .96; and RMSEA = .02.
The modified online deliberation model has the following goodness of
fit values: chi-square = 167.43, d.f. = 112, p < .01; CFI = .967;
RMSEA = .02. Thus the two modified models has a very good model fit.
The chi-square is significant which doesn't suggest a goodness of
fit, but this could be related to the big sample size (Bentler, 1990).
[Insert Figure 2 about here]
[Insert Figure 3 about here]
[Insert Table 5 about here]
[Insert Table 6 about here]
The squared multiple correlations indicates that the face-to-face
model explains 50% of variance in deliberation, 11% in political
interest, 11% in internet news, 63% in exposure to diverse views, 24%
in frequency of political discussion, 10% in knowledge, 1% in
television news, and 0% in newspaper. The square multiple
correlations show that the online deliberation model explains 30% of
variance in online deliberation, 11% in political interest, 11% in
internet news exposure, 1% in television, 23% in exposure to diverse
views on the internet, 10% in online political discussion, 10% in
knowledge, and 0% in newspaper news exposure.
Deliberation measurement model
One goal of this study is to identify sub-dimensions that compose the
deliberation construct and see how those variables contribute. As
shown in Figure 2, a large variance in the deliberation construct is
explained by the six variables. For the face-to-face deliberation,
freedom to express ideas, reciprocity and the common good dimension
contribute to the deliberation construct better than four other
dimensions. As shown in Figure 3, freedom of expression and
reciprocity dimensions contribute to the online deliberation
construct as well. The coefficient for reasoned discourse in the
face-to-face deliberation measurement model is higher than the online
deliberation model, while coefficients for freedom of expression and
reciprocity are higher in the online deliberation.
Findings related to hypotheses
Three hypotheses about the face-to-face deliberation process has been
proposed. H1 predicted that the people with high political interest
has more factual political knowledge, participate in political
discussion more frequently and expose themselves to more diverse
political views. This hypothesis was supported. In the model,
political interest has direct paths to political knowledge, frequency
of political discussion, and exposure to diverse political views, and
all three paths are statistically significant.
The second hypothesis predicted that the relationships between
political interest and the three variables, political knowledge,
frequency of political discussion, and exposure to diverse views are
mediated by news exposure variables – newspaper, television news, and
internet news. This hypothesis is partly supported. Internet news is
found to mediate the relationship between political interest and the
frequency of political discussion. Those who are interested in
politics tend to get news from internet, and those who read a lot of
news from internet tend to involve more frequently in political discussions.
The third hypothesis proposed that political knowledge, frequency of
political discussion, and exposure to diverse views have positive
relationship with deliberation. This hypothesis is partially
supported. Frequency of political discussion, and exposure to diverse
political views have direct path to deliberation. As implied in the
correlation matrix, political knowledge which did not show many
correlations with other variables did not have influence on the
deliberation process. This indicates that frequency of political
discussion and exposure to diverse views mediate the relationship
between political interest and the deliberation.
The research question asks about the difference between face-to-face
and online deliberation process. As shown in Figure 3, a similar
model is proposed for the online deliberation process. Political
interest predicts political knowledge, the frequency of online
political discussion, and exposure of diverse views on internet. News
exposure also meditates this process similar to the face-to-face
deliberation model. Internet news exposure mediates the relationship
between political interest and the frequency of online political
discussion, and between political interests and the exposure to
diverse views on the internet. Those who are interested in politics
are more likely to read news from the internet, and those who read
news from the internet are more likely to engaged in online
discussion about political issues and view more diverse political
views on the internet. Unlike the face-to-face process, television
news exposure mediates the relationship between political interest
and the frequency of online discussion participation. However, a
negative coefficient is linked between television news exposure and
the online discussion. Those who are interested in politics are more
likely to watch television news, but those who watch television news
are less likely to participate in online political discussions.
Similar to the face-to-face situation, the frequency of online
discussion and the exposure to diverse views on the internet have
direct paths to the online deliberation construct, but knowledge does
not affect the online deliberation.
Discussions
This study tried to identify factors that influence deliberation by
proposing a deliberation model, and compare the face-to-face
deliberation process with the online deliberation model. First, six
variables that compose the deliberation construct were identified,
and how those six concepts are related to the deliberation process
have been examined. Overall the six concepts explained a lot of
variance in the deliberation construct for both face-to-face and
online models. For both models, freedom to express ideas,
reciprocity, and common good factors are identified as strong
contributors of deliberation. This is somewhat consistent with the
literature on the deliberation. Many scholars points out that the
reciprocity, exchanging ideas with discussants who have different
political views and trying to convince them to change their views or
vote, is the key factor that composes the deliberation. In order for
deliberation to occur, people should freely exchange ideas and try to
argue for their political views with those who do not agree with
their political views, and those exchanges should reflect the idea of
common good rather than group selfishness.
Political interest was identified as a strong predictor of
deliberation. Those who have a lot of interest in politics are more
likely to be engaged in political discussion and seek out diverse
political views that do not necessarily agree with their own views,
and these frequent discussion and exposure to diverse views lead them
to be engaged in deliberation more often.
The concept of exposure to diverse political view is introduced in
this study as a mediating variable between political interest and
deliberation. So far, concepts of heterogeneity of network (McLeod et
al., 1999), cross-cutting network (2002), and cross-pressure
(Berelson et al., 1954) were examined in relation with political
knowledge, political participation, and political discussion.
However, the concept of one's willingness to seek out diverse
political views that contradicts their own political views has not
yet been measured. This concept of exposure to diverse political
views may be developed into a valuable concept in the deliberation
process. In this study, those who are interested in politics are more
likely to seek opinions that contradict their own belief rather than
to surround themselves with similar ideas. Those who have a lot of
political interest are likely to talk with people who have different
political views and read newspapers that have different political ideologies.
Another finding of this study is that the news exposure through media
does not contribute much to the deliberation process. This
contradicts previous notion that newspaper reading and television
news viewing indirectly influence deliberation when mediated by
reflection (McLeod, Scheufele, & Moy, 1999), and increase political
participation by increasing knowledge. (Eveland & Scheufele, 2000).
Mutz (2002) also identified media news exposure as an important
factor that lead to exposure to diverse political views. One
interesting finding is that internet news reading is found to promote
political discussion and exposure to diverse political views, whereas
newspaper reading is neither influence by political interest nor
influence the political discussion. In the online deliberation model,
television news viewing even has a negative influence on the online
discussion, which means it has negative indirect impact on the online
deliberation. Internet media was more influential than the traditional media.
The study compared the face-to-face deliberation with the online
deliberation. The two models were not very different, but overall,
the face-to-face model explained more variance in the deliberation
(50%) than the online deliberation model did for the online
deliberation construct (30%). Also the path coefficients that depart
from political interest to deliberation is higher in the face-to-face
deliberation model than the online model. It can be assumed that more
factors that are not captured in the online model influence the
online deliberation process. In order to fully account for the online
deliberation process, more psychological variables, such as
self-efficacy and political tolerance should be included in future
studies. In the deliberation measurement model, reasoned discourse
dimension contributes more to the face-to-face deliberation model
than the online model, while freedom of expression and reciprocity
dimension contribute more to the online deliberation than the
face-to-face deliberation. This implies that people use more reason
to argue for their political views in the face-to-face situation than
on the internet, while on the internet, people are more freely
exchange ideas and more frequently encountering those who disagree
with their own political views.
Limitations of the study should be pointed out. The survey was
sent to international scholars with high education levels, who
relatively enjoy a high social status in the country they reside,
thus the education and income were not successful predictors.
Perhaps, due to the high educated sample, political knowledge was
found to have no influence on the deliberation process, which
contradicts previous findings. This lack of political knowledge
influence on the deliberation process may reflect the sampling
difference. Unlike other traditional political communication studies
that were conducted with the general U.S. population sample, this
study was conducted with international respondents. Another
possibility is that the knowledge score had a low variability. One
more thing to be mentioned is the cross-sectional nature of the data.
The SEM model was structured based on the cross-sectional data,
therefore the paths do not necessarily indicate causal relationships.
This study provided a way to measure deliberation with
questionnaires. Future studies should improve the deliberation
measurement based on this study's finding. For future studies,
deliberation process for specific issues should be studied in order
to take accounts the nature of discussion topics in the deliberation
process. Some may participate in deliberation for one topic while
they may not for another depends on the issues' salience and
relevance to them. When issues are relevant, people may exchange
their ideas with those who disagree with them. How issue's features
influence the deliberation process should be examined in future studies.
References
Arbuckle, J. L., & Wothke, W. (1995). Amos 4.0 user's guide. Chicago,
Il: Smallwaters corporation.
Bartle, J. (2000). Political awareness, opinion constraint and the
stability of ideological positions. Political Studies, 48, 467-484.
Berelson, B., Larzarsfeld, P., & McPhee, W. (1954). Voting. Chicago,
IL: University of Chicago Press.
Bentler, P. M. (1990). Comparative fit indexes in structural models.
Psychological Bulletin, 107, 238-246.
Bohman, J. (1996). Public deliberation: Pluralism, complexity, and
democracy. Cambridge, MA: MIT Press.
Cappella, J. N., Price, V., & Nir, L. (2002). Argument repertoire as
a reliable and valid measure of opinion quality: Electronic dialogue
during campaign 2000. Political Communication, 19(1), 73-93.
Carpini, D., Michael, X., & Keeter, S. (1996). What Americans know
about politics and why it matters. New Haven, CT: Yale University Press.
Chaffee, S. H. (2001). Attention to counter-attitudinal messages in a
state election campaign. Political Communication, vol. 18, no. 3, 247-272.
Cohen, J. (1989). Deliberation and democratic legitimacy. In A.
Hamlin and P. Pettit (Eds.), The good polity: Normative analysis of
the State (pp. 17-34). Oxford: Blackwell.
Converse, P. (1964). The nature of belief systems in mass publics. In
D. Apter (Eds.), Ideology and discontent, New York: Free Press.
Fishkin, J. (1991). Democracy and deliberation. New Haven, CT: Yale
University Press.
Fishkin, J. S. (1995). The voice of the people: Public opinion and
democracy. New Haven: Yale University Press.
Fernback, J., Jones, S. G. (Ed). (1997). The individual within the
collective: virtual ideology and the realization of collective
principles. In Virtual culture: identity and communication in cyber
society (pp. 36-54). Thousand Oaks, CA: Sage Publications.
Gastil, J. (1993). Democracy in small group: Participation, decision
making, and communication. Philadelphia: New Society.
Graber, D. (1984). Mass media and American politics. Washington,
D.C.: CQ Press.
Gutmann, A. & Thompson, D. (1996). Democracy and disagreement.
Cambridge, MA: The Belknap Press of Harvard University Press.
Habermas, J. (1984). The theory of communicative action, Vol. 1:
reason and the rationalization of society. Boston: Beacon Press.
Habermas, J. (1989). The structural transformation of the public
sphere. Cambridge, MA: MIT Press.
Habermas, J. (1995). Reconciliation through the public use of reason:
Remark on John Rawls's political liberalism. Journal of Philosophy,
92, 109-131.
Herbst, S. (1993). Numbered voices. Chicago, IL: University of Chicago Press.
Hu, L-Z, & Bentler, P. M. (1999). Cutoff criteria for fit indexes in
covariance structure analysis: Conventional criteria versus new
alternative. Structural Equation Modeling, 6, 1-55.
Huckfeldt, R., & Sprague, J. (1995). Citizens, politics, and social
communication: information and influence in an election campaign.
Cambridge, England: Cambridge University Press.
Huckfeldt, R., Mendez, J. M., & Osborn, T. (2004). Disagreement,
ambivalence, and engagement: The political consequences of
heterogeneous networks. Political psychology, 25(1), 65-95.
Jonas, E., Schulz-Hardt, S., Frey, D., & Thelen, N. (2001).
Confirmation bias in sequential information search after preliminary
decisions: An expansion of dissonance theoretical research on
"selective exposure to information". Journal of Personality and
Social Psychology, 80, 557-571.
Jones, D. A. (2002). The polarization effect of new media message.
International Journal of Public Opinion Research, 14(2), 158-174.
Kim, J., Wyatt, R. O., & Katz, E. (1999), News, talks, opinion,
participation: The part played by conversation in deliberative
democracy. Political Communication, 16, 1-33.
Lazarsfeld, P. Berelson, B., & Gaudet, H. (1944). The people's
choice. New York: Columbia University Press.
McKuen, M. (1990). Speaking of politics: Individual conversational
choice, public opinion, and the prospects for deliberative democracy.
In J. Ferejhon & J. Kuklinski (Eds.), Information and democratic
process, (pp. 59-99). Urbana: University of Illinois Press.
McLeod, J. M., Daily, K., Guo, Z., Eveland, W. P., Bayer, J., Yang,
S., & Wang, H. (1996). Community integration, local media use and
democratic processes. Communication Research, 23, 179-209.
McLeod, J. M., Sotirovic, M., & Holbert, R. L. (1998). Values as
sociotropic judgments influencing communication patterns.
Communication Research, 26(6), 623-654.
McLeod, J. M., Scheufele, D. A., Moy, P., Horowitz, E. M., Holbert,
R. L., Zhang, W., Zubric, S.. & Zubric, J, (1999). Understanding
deliberation: The effects of discussion networks on participation in
a public forum. Communication Research, 26, 743-774.
McLeod, J. M., Scheufele, D. A., Moy, P., Horowitz E. M., et al.
(1999). Understanding deliberation: The effects of discussion
networks on participation in a public forum. Communication Research,
vol. 26, no. 6, 743.
Mendelberg, T. & Oleske, J. (2000). Race and public deliberation.
Political Communication, 17, 169-191.
Moy, P., Scheufele, D. A., & Holbert, R. L. (1999). Television use
and social capital: Testing Putnam's time displacement hypothesis.
Mass Communication & Society, 2(1/2), 27-45.
Moy, P., McCluskey, M. R., McCoy, K. (2004). Political correlates of
local news media use. Journal of Communication, vol. 54, no. 3, 532-546.
Mutz, D. C., & Martin, P. S. (2001). Facilitating communication among lines of
difference: the role of the mass media. American Political Science
Review, 95(1), 97-114.
Mutz, D. C. (2002). The consequences of cross-cutting networks for
political participation. American Journal of Political Science, 46, 838-855.
Park, S. G. (2000). The significance of civility in deliberative
democracy. Paper presented at the conference of the Public Opinion
Research in the Digital Age (PORDA) project, Seoul, Korea.
Perloff, R. M. (1998). Political communication: Politics, press, and
the public in America. Mahwah, NJ: Lawrence Erlbaum Associates.
Putnam, R. D. (2000). Bowling alone. New York: Simon & Schuster.
Rawl, J. (1971/1999). Chapter VII. Goodness as rationality. In A
theory of justice. (pp. 347-396). Cambridge, MA: Havard University Press.
Rogers, C. (1980). A way of being. Boston, MA: Houghton Mifflin.
Ryu, S. (2004) When online meet offline: A new approach to the
operating mechanism of the ideal public sphere for deliberative
democracy. Paper presented at the annual meeting of Association for
Education in Journalism and Mass Communication, Toronto, Canada.
Scheufele, D. A. & Shah, D. V. (2000). Personality strength and
social capital: the role of dispositional and informational variables
in the production of civic participation. Communication Research,
vol. 27, no. 2, 107-131.
Scheufele, D. A. (2002). Examining differential gains from mass media
and their implications for participatory behavior. Communication
Research, vol. 29, no. 1, 46-65.
Sears, D. O., & Freedman, J. L. (1967). Selective exposure to
information: a critical review. Public Opinion Quarterly, 31, 194-213.
Steenbergen, M. R., Bachtiger, A., Sporndli, M., & Steiner, J.
(2003). Measuring political deliberation: A discourse quality index.
Stewart, J. & Logan, C. (1998). Together: Communicating
interpersonally. Boston, MA McGraw-Hill.
Sunstein, C. R. (2001). Republic.com. New Jersey: Princeton University Press.
Tilley, J., Sturgis, P., & Allum, N. (2004, May 7-8). Political
information and motivation: A case of reciprocal causality? Paper
presented at symposium on 'perceptions preferences and
rationalization: overcoming the problem of causal inference in the
study of political behaviour', Nuffield College, Oxford.
Van Deth, J. W. (1990). Interest in politics. In M.K. Jennings & J.
W. Van Deth (Eds.), Continuities in political action. A longitudinal
study of political orientations in three western democracies (pp.
275-312). Berlin: Walter de Gruyter.
Verba, S., Schlozman, K. L., Brady, H., & Nie, N. H. (1993). Race,
ethnicity and political resources: participation in the United
States. British Journal of Political Science, 23, 453-497.
Waldman, P. (2001). Deliberation in practice: Connecting theory to
the lives of citizens. In R. P. Hart and B. H. Sparrow (Eds.),
Politics, discourse, and American society: New agendas (pp. 151-171).
Lanham, MD: Rowman & Littlefield.
Weger, H. & Askhus, M. (2003). Arguing in Internet chat rooms:
Argumentative adaptations to chat room design and some consequences
for public deliberation at a distance.
Wolfinger, R. E., & Rosenstone, S. J. (1980). Who votes?. New Haven:
Yale University Press.
Zaller, J. (1992). The nature and origins of mass opinion. New York:
Cambridge University Press.
Table 1 Means and standard deviations
Variables
Mean
SD
N
Age (in years)
39.89
15.26
458
Education (in years)
17.02
5.59
441
Income*
4.36
1.62
455
Newspaper**
384.28
618.41
632
Television news**
401.15
667.24
640
Internet news**
455.24
792.77
645
Political interest ***
4.16
.89
564
Political knowledge ****
6.77
1.80
478
Frequency of political discussion (face-to-face)***
3.55
.74
431
Frequency of political discussion (online)***
2.08
1.11
560
Exposure to diverse views***
4.09
.92
565
Exposure to diverse views (online)***
3.07
1.13
558
Deliberation (Face-to-face)
Reciprocity***
3.46
.76
498
Reasoned discourse*****
7.03
3.90
394
Freedom of expression***
4.03
.79
505
Open mindedness ***
4.19
.71
501
Empathy***
3.92
.68
496
Public Interest***
3.94
.84
497
Deliberation (Online)
Reciprocity***
2.85
.97
482
Reasoned discourse*****
7.03
3.90
394
Freedom of expression***
3.29
1.29
492
Open mindedness***
3.87
.91
490
Empathy***
3.49
.85
478
Public Interest***
3.59
1.03
478
* Responses were coded 1= least income to 9= most income.
** Days a week X minutes a day spent with the media
*** Ranges from 1 to 5. 5 is the highest.
**** Number of correct answers, ranging from 0 to 10.
***** Number of reasoned provided.
Table 2 Percentages of gender and race
Variables
%
Gender
Male
49.7
Female
50.3
100.00%
(N = 461)
Race
African American
1.3
American Indian or Alaska Native
.2
Arabic
1.3
Black, but not African American
1.1
Central Asian
.4
East Asian
7.1
Hispanic
1.3
South Asian
4.8
White
74.2
Others
8.2
100.00%
(N = 462)
Deliberation Process
Table 3 Pearson correlations for variables in the face-to-face
deliberation model
Variables
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Age
- .30**
(455)
.44**
(436)
-.35**
(449)
-.10*
(456)
.32**
(453)
.05
(404)
.29**
(454)
.11*
(439)
-.14**
(442)
.06
(414)
.05
(410)
.07
(438)
.09
(436)
-.03
(437)
-.03
(434)
.01
(436)
.15**
(361)
Gender*
-
-.07
(438)
.14**
(451)
.02
(459)
-.20**
(455_
.03
(405)
-.23*
(456)
-.21
(442)
-.05
(445)
.07
(418)
-.01
(410)
-.07
(440)
-.06
(437)
-.03
(438)
-.09
(435)
.02
(437)
-.03
(361)
Education
-
-.15**
(436)
.09
(438)
.11*
(436)
.04
(389)
.09
(437)
.12**
(423)
-.08
(426)
.03
(397)
-.02
(392)
-.05
(424)
.10*
(420)
-.08
(422)
-.05
(420)
-.02
(422)
.09
(352)
Income
-
.02
(453)
-.22**
(451)
-.07
(402)
-.16**
(452)
-.14**
(437)
.002
(440)
.02
(412)
-.07
(404)
-.05
(438)
-.11*
(434)
.04
(435)
.05
(433)
.04
(435)
-.001
(361)
Race**
-
-.10*
(457)
-.08
(407)
-.03
(458)
-.03
(458)
.13**
(447)
.04
(418)
.10*
(411)
-.10*
(442)
-.10*
(439)
-.03
(440)
.004
(437)
-.03
(439)
-.16**
(363)
Political interest
-
.49**
(428)
.79**
(561)
.24**
(474)
.11**
(544)
.10
(515)
.05
(504)
.32**
(501)
.35**
(494)
.10*
(497)
.17**
(492)
.22**
(493)
.27**
(391)
Discussion (offline)
-
.46**
(430)
.13**
(417)
.002
(415)
.05
(394)
.04
(383)
.30**
(419)
.40**
(419)
.10*
(419)
.14*
(416)
.32**
(417)
.26**
(341)
Exposure to diverse views (offline)
-
.23**
(475)
.16**
(545)
.12**
(515)
.12*
(505)
.39**
(504)
.36**
(497)
.19**
(500)
.28**
(495)
.30**
(496)
.24**
(394)
Political knowledge
-
.04
(460)
-.01
(435)
.07
(428)
.14**
(457)
.21**
(435)
.09
(455)
.04
(453)
.10*
(454)
.11*
(375)
Internet news
-
.15**
(583)
.15**
(572)
.07
(490)
-.03
(484)
.06
(486)
.03
(481)
.11*
(482)
.01
(382)
Television news
-
.10*
(576)
.06
(460)
.01
(453)
.04
(456)
.10*
(452)
.10*
(453)
-.03
(359)
Newspaper
-
.06
(447)
.06
(441)
-.02
(444)
.03
(439)
.03
(440)
-.05
(352)
Freedom of expression
-
.25**
(494)
.27**
(497)
.26**
(492)
.28**
(493)
.13*
(389)
Reciprocity
-
.16**
(494)
.27**
(489)
.34**
(490)
.20**
(386)
Open mindedness
-
.47**
(493)
.41**
(494)
-.02
(389)
Empathy
-
.56**
(493)
.04
(387)
Public interest
-
.13**
(389)
Reasoned discourse
-
* Male was coded 1 and female was coded 2.
** Whites were coded 1 and others were coded 2.
Table 4 Pearson correlations for variables in the online deliberation model
Variables
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Age
-.30**
(454)
.44**
(436)
-.35**
(449)
-.09*
(456)
.32**
(453)
.00
(451)
.10*
(452)
.12*
(439)
-.14**
(442)
.06
(414)
.05
(410)
-.08
(427)
.03
(420)
-.04
(428)
-.03
(418)
-.02
(420)
.15**
(361)
Gender*
-
-.07
(438)
.14**
(451)
.02
(459)
-.20**
(455)
-.08
(453)
-.14**
(454)
-.21**
(442)
-.05
(445)
.07
(418)
-.01
(410)
-.10*
(429)
-.10
(421)
-.04
(429)
-.04
(419)
.02
(421)
-.03
(361)
Education
-
-.15**
(436)
.09
(438)
.12*
(436)
.03
(434)
.06
(435)
.16**
(423)
-.07
(426)
.03
(397)
-.02
(392)
.001
(415)
.10
(409)
.02
(414)
-.01
(408)
-.002
(409)
.09
(352)
Income
-
.02
(453)
-.22**
(451)
-.05
(449)
-.06
(450)
-.13**
(437)
.002
(440)
.02
(412)
-.07
(404)
-.001
(427)
-.05
(419)
-.04
(426)
-.01
(417)
-.01
(419)
-.01
(361)
Race**
-
-.10*
(457)
.11*
(455)
.07
(456)
-.03
(444)
.13**
(447)
.04
(418)
.10*
(411)
.04
(431)
.03
(423)
.06
(431)
.02
(421)
-.05
(423)
-.16*
(363)
Political interest
-
.21**
(557)
.40**
(555)
.24**
(474)
.11*
(544)
.10*
(515)
.05
(504)
.09*
(488)
.13*
(478)
.04
(486)
.10*
(474)
.17**
(474)
.27**
(319)
Discussion (offline)
-
.41**
(554)
.09*
(470)
.17**
(542)
-.01
(509)
.07
(500)
.44**
(489)
.46**
(479)
.20**
(487)
.20**
(475)
.29**
(475)
.15**
(390)
Exposure to diverse views (offline)
-
.13**
(471)
.23**
(539)
-.02
(508)
.13**
(498)
.20**
(487)
.29**
(477)
.24**
(485 )
.26**
(474)
.26**
(474)
.15**
(390)
Political knowledge
-
.04
(460)
-.09
(435)
.07
(428)
.12*
(448)
.17**
(438)
.08
(446)
.06
(439)
.06
(440)
.11*
(375)
Internet news
-
.15**
(583)
.15**
(572)
.09
(478)
.12**
(468)
.07
(475)
.08
(463)
.07
(463)
.01
(382)
Television news
-
.10*
(576)
-.05
(448)
-.06
(437)
-.08
(445)
-.02
(434)
-.11*
(437)
-.03
(359)
Newspaper
-
.04
(436)
.07
(426)
.05
(434)
.07
(422)
.05
(423)
-.05
(32)
Freedom of expression
-
.66**
(479)
.42**
(485)
.38**
(476)
.41**
(476)
.11*
(382)
Reciprocity
-
.34**
(480)
.37**
(471)
.39**
(471)
.22**
(376)
Open mindedness
-
.55**
(476)
.47**
(477)
-.01
(383)
Empathy
-
.69**
(474)
-.02
(375)
Public interest
-
.05
(377)
Reasoned discourse
-
* Male was coded 1 and female was coded 2.
** Whites were coded 1 and others were coded 2.
Deliberation Process
Table 5 Direct and indirect effects of exogenous variables
Source of Influence
Income
Age
Gender
(Male=1)
Education
Race
(white=1)
Political interest
Offline
Direct
-.13
.21
-.13
.00
-
Indirect
.00
.00
.00
.00
-
Online
Direct
-.14
.19
-.13
.00
.00
Indirect
.00
.00
.00
.00
.00
Frequency of discussion (offline)
Offline
Direct
.00
.13
.00
.00
-
Indirect
-.07
.11
-.06
.00
-
Frequency of discussion (online)
Online
Direct
.00
.00
.00
.00
.11
Indirect
-.03
-.01
-.06
.00
.00
Exposure to diverse views (offline)
Offline
Direct
.00
.00
-.07
.00
-
Indirect
-.10
.01
-.11
.00
-
Exposure to diverse views (online)
Online
Direct
.00
.00
.00
.00
.09
Indirect
-.06
.01
-.105
.00
.00
Political knowledge
Offline
Direct
.00
.00
-.16
.12
-
Indirect
-.03
.04
-.02
.00
-
Online
Direct
.00
.00
-.16
.12
.00
Indirect
-.03
.04
-.03
.00
.00
Deliberation (offline)
Offline
Direct
.00
.00
.00
.00
-
Indirect
-.07
.05
-.11
.00
-
Deliberation
(online)
Online
Direct
.00
.00
.00
.00
.00
Indirect
-.02
-.003
-.04
.00
.07
Television news
Offline
Direct
.00
.00
.00
.00
-
Indirect
-.01
.02
-.01
.00
-
Online
Direct
.00
.00
.00
.00
.00
Indirect
-.02
-.003
-.04
.00
.07
Internet news
Offline
Direct
.00
-.31
-.17
.00
-
Indirect
-.02
.04
-.02
.00
-
Online
Direct
.00
-.31
-.17
.00
.00
Indirect
-.02
.03
-.02
.00
.00
Table 6 Direct and indirect effects of endogenous variables
Source of Influence
Interest
Discussion (offline)
Discussion (online)
Exposure to diverse views (offline)
Exposure to diverse views (online)
Internet news
TV news
Political knowledge
Offline
Direct
.20
.00
-
.00
-
.00
-
Indirect
.00
. 00
-
.00
-
.00
-
Online
Direct
.20
-
.00
-
.00
.00
.00
Indirect
.00
-
.00
-
.00
.00
.00
Frequency of discussion (offline)
Offline
Direct
.52
.00
-
.00
-
.00
-
Indirect
.00
.00
-
.00
-
.00
-
Frequency of discussion (online)
Online
Direct
.20
-
.00
-
.00
.17
.00
Indirect
.03
-
.00
-
.00
.00
.00
Exposure to diverse views (offline)
Offline
Direct
.76
.00
-
.00
-
.09
-
Indirect
.02
.00
-
.00
-
.00
-
Exposure to diverse views (online)
Online
Direct
.40
-
.00
-
.00
.23
-.08
Indirect
.03
-
.00
-
.00
.00
.00
Internet news
Offline
Direct
.18
.00
-
.00
-
.00
-
Indirect
.00
.00
-
.00
-
.00
-
Online
Direct
.17
-
.00
-
.00
.00
.00
Indirect
.00
-
.00
-
.00
.00
.00
Television news
Offline
Direct
.20
.00
-
.00
-
.00
-
Indirect
.00
.00
-
.00
-
.00
-
Online
Direct
.10
-
.00
-
.00
.00
.00
Indirect
.00
-
.00
-
.00
.00
.00
Deliberation (offline)
Offline
Direct
.00
.41
-
.45
-
.00
-
Indirect
.56
.00
-
.00
-
.04
-
Deliberation (online)
Online
Direct
.00
-
.49
-
.12
.00
.00
Indirect
.17
-
.00
-
.00
.11
-.01
Figure 2. Modified Deliberation Model
[1] For more information, contact XXX at XXX University.
|
