Running head:
NEW MEDIA AND SPECIALISTS
Are You a Specialist, or a Generalist?:
The Role of the Internet in the Formation of Issue Publics
1
NEW MEDIA AND SPECIALISTS
2
Abstract
The present study revisits the question as to whether U.S. citizens are information specialists or
information generalists.
Although the literature has presented mixed views, this study provides
evidence that the changing information environment facilitates the growth of specialists.
Using
national survey data (n=1208) regarding the health care reform, the study found that individuals
sought issue-specific knowledge driven by their perceived issue importance rather than by
general education, and that this trend was conditional upon individuals’ primary source of media.
Specifically, individuals who relied upon the Internet and cable TV news were capable of
translating their interest in the issue into issue-specific knowledge.
On the other hand, those
who depended on network TV news, newspapers, and radio failed to display a high level of
issue-specific knowledge, even when they perceived the issue to be important to their family.
NEW MEDIA AND SPECIALISTS
3
Are You a Specialist, or a Generalist?
A number of theorists have welcomed television as a “knowledge leveler” (Neuman,
1976, p. 122) that reduces the inequality in political knowledge (Eveland & Scheufele, 2000;
Kwak, 1999; Prior, 2007). They suggest that incidental and habitual exposure to daily evening
newscasts leads to a narrowing knowledge gap between the more and less educated citizens.
More specifically, the less educated inadvertently benefit from watching television as they
become generalists who are aware of a wide range of political and social issues in spite of their
relatively low interest in politics. However, as the information environment changes, many
have been concerned about whether new media can fully serve a function of fostering generalists
(Sunstein, 2001).
By virtue of decentralized media outlets and increased user controllability,
individuals, especially those who are uninterested in politics, can avoid news and seek
entertainment single-mindedly. Consequently, mass publics might fail to obtain the political
information necessary for competent citizenship in a democratic society.
Additionally, the
knowledge gaps between the educated and uneducated, news junkies and entertainment fans, and
the ‘‘haves’’ and ‘‘have-nots’’ widen (Bennett & Iyengar, 2008; Prior, 2007; Wei & Hindman,
2011).
Despite the scholarly concerns about the decline of information generalists in the new
information environment, little empirical attention has been paid to the rise of information
specialists, who are knowledgeable publics within a particular domain of their interest.
The
changing media environment tends to cultivate information specialists by allowing them to seek
information selectively and acquire domain-specific knowledge efficiently (Kim, 2009).
However, one main obstacle to the research on this topic is that the concept of specialist can be
fully captured only when the researcher breaks down political issues into separate issues and
NEW MEDIA AND SPECIALISTS
highlights each issue distinctively.
4
For example, since individuals who are specialists in a
health care domain are not necessarily specialists in environmental issues, an issue-by-issue
approach is necessary.
Researchers need issue-oriented measures to address the issue of
emerging information specialists in this changing media environment.
Unfortunately, most
studies on the effects of the Internet on political knowledge (Althaus & Tewsbury, 2002;
Dalrymple & Scheufele, 2007; Eveland & Dunwoody, 2000) have paid little attention to
systematic differences across issue domains (cf., Kim, 2009).
They simply viewed citizens as
an aggregated mass audience, lending little credence to the role of the new media in fostering
specialists.
On the other hand, the idea of issue publics provides a useful theoretical tool to address
the political implications of the rise of specialists in the new media environment (Kim, 2009).
The issue public hypothesis posits that citizenry consists of issue publics, or pluralistic small
groups of people who are passionately concerned about certain issues (Converse, 1964).
It is
highly plausible that individuals who seek specialized and in-depth information through the
Internet will become members of issue public on a certain issue. That is, the new media
environment facilitates the formation of issue publics by producing specialists in various issue
domains.
The present study proceeds in two steps.
The first aim of the study is to explore
whether the concept of information specialist is actually useful for describing the formation of
the public within a certain domain. Using a national survey about a health care reform bill in
the U.S., the study examines whether knowledgeable publics in this specific domain consist of
information specialists who are particularly interested in this domain, or information generalists
who are generally knowledgeable about every issue, including the health care issue.
I examine
NEW MEDIA AND SPECIALISTS
5
each of these two competing theoretical arguments in the context of health care reform.
the study investigates the role of new media in fostering specialists.
Second,
The study will examine
whether presumably more selective media (e.g., Internet) are actually more effective in
producing specialists as compared to traditional media (e.g., network television, radio,
newspaper).
Distribution of Public Knowledge
Normative democratic theory posits that the functioning of a health democracy requires
an informed citizenry whose attitudes and participation are based on a broad set of relevant and
accurate information (Habermas, 1984).
According to a voluminous literature on political
knowledge, at least three theses have been widely accepted. First, levels of political knowledge
are consequential to various democratic values, including participation, representation, and
abilities to form coherent and stable attitudes (Delli Carpini & Keeter, 1996; McDevitt & Chaffee,
2000; Zaller, 1992).
Second, overall levels of political knowledge in the U.S are frustratingly
low (Lupia & McCubbins, 1998; Neuman 1986).
Third, knowledge is unevenly distributed
across the population and is associated with socioeconomic factors (Delli Carpini & Keeter
1996).
However, relatively little consensus has been reached concerning the patterns of
knowledge at individual levels.
Who are informed citizens in each issue domain?
knowledge levels fluctuating or stable across domains?
Are their
Reponses to these questions have
varied widely but have generally stemmed from two theoretical models.
The first model posits
that individuals have varying interests and knowledge levels across domains and need not or
cannot be experts on every issue.
This model stresses the pluralism of public opinion.
The
next model emphasizes that public opinion is stratified. Although the average citizen may not
NEW MEDIA AND SPECIALISTS
6
be knowledgeable in general, democracy functions owing to a small number of elites who are
attentive, active, and are indeed well informed (Neuman, 1986).
In the next few paragraphs I
will examine each of these two arguments, that is, the specialist thesis and the generalist thesis.
RQ1: Are people who are well-informed in a specific domain specialists or generalists?
The Specialist Thesis
Theoretically, the concept of issue public is a useful framework for developing
hypothesis about why citizens are more likely to be specialists rather than generalists.
Converse (1964) invoked issue public to provide realistic explanation of how citizens can
respond to public policy in a fairly rational manner, despite their alarmingly low level of general
political knowledge.
For most people, once having managed their more pressing matters of
family, work, and recreation, they have few resources and little energy left to study every social
or political issue.
Since the cost for becoming well informed is substantial, individuals are
expected to focus their attention on a handful of issues at best.
Thus, the theory of issue publics
indicates that citizens are not generalists but specialists who are experts on a certain domain and
display a high level of domain-specific knowledge (Iyengar, 1986; Kim, 2009).
Two lines of empirical research have supported the view that the U.S. public is
composed of specialists rather than generalists. First, a number of studies (Krosnick, 1990;
Krosnick, & Telhami, 1995) reported that individuals are concerned only about a few issues and
attach varying degrees of attitude importance to each issue. Krosnick (1990) found no
significant correlations among various issue attitudes.
For example, individuals who are
information specialists in a foreign policy are not likely to be specialists in domestic social
welfare issues. The findings are also consistent with evidence that citizens hold the low level of
coherence across issue attitudes (Converse, 1964).
NEW MEDIA AND SPECIALISTS
7
While the first line of studies examines multiple domains and their relationships, the
second line of inquiries focuses on identifying informed citizens within a single domain.
The
latter type of inquiries attempts to investigate the specialist thesis by showing that individuals
seek domain-specific information not driven by general education level but by perceived issue
importance or personal self-interests (Berent & Krosnick, 1995; Kim, 2009).
Numerous studies
(e.g., Petty & Krosnick, 1995) have shown that people who perceive an issue as personally
important tend to form a more stable attitude toward the issue and be more cognitively and
behaviorally involved.
One explanation is that the knowledge construct of a certain issue is
more accessible when people perceive this issue to be important (Iyengar, 1990).
Another
explanation focuses on the selective information-seeking behaviors among specialists (Boninger,
Krosnick, Berent, & Fabrigar, 1995).
Specialists report a high level of domain-specific
knowledge because they are committed to collect information about the issue selectively.
Accordingly, Krosnick and his colleague (Krosnick, 1990; Krosnick & Telhami, 1995) argued
that personal issue importance is the best proxy for identifying specialists (cf., Price et al., 2006).
Since the present study does not deal with multiple domains simultaneously but focuses on the
single domain of the health care reform, the study will pursue the second type of inquiry.
The Generalist Thesis
The generalist thesis, perhaps the most widely supported proposition for explaining the
functioning of democracy, offers a rather different picture of the mass polity.
This approach
posits that despite the general paucity of political interest and knowledge among most American
citizens, democracy functions thanks to a small number of sophisticated, educated, and attentive
elites (Price & Zaller, 1993; Zaller, 1992).
Although people may be more informed about one
issue than the other, those who are well informed about one issue are likely to be well informed
NEW MEDIA AND SPECIALISTS
8
about other issues as well (Delli Carpini & Keeter, 1996).
In addition, this view indicates that
education is a significant source of information for political learning.
People who are more
educated are presumably equipped with sophisticated cognitive ability that enables them to
organize abstract ideas to understand complex political matters (Krosnick, 1990).
Ample studies (Delli Carpini & Keeter, 1996; Neuman, 1986) have also provided
empirical evidence supporting this view.
Educated individuals are more familiar with political
issues and more knowledgeable about political events (Neuman, 1986).
Delli Carpini and
Keeter (1996) were also skeptical about the existence of a multitude of distinct specialists.
They showed that knowledge about the United Nations is a good predictor of knowledge about
racial issues as well as international relations and concluded that if citizens are informed about a
certain topic, they are likely to be informed about other issues as well (Delli Carpini & Keeter,
1996).
On closer inspection, however, these scholars did not rule out the possibilities of the
specialist thesis.
After discussing the methodological difficulties of assessing the pluralistic
model, Neuman (1986, p.39) remarked, “The model is not wrong, but it is incomplete.”
In
addition, Delli Carpini and Keeter (2002) embraced the specialist thesis more explicitly in their
recent paper. While calling for more research on the effects of the Internet on the growth of
information specialists, they postulated, “(the Internet) will allow citizens to focus on the specific
levels of politics in substantive issues in which they are most interested” (Delli Carpini & Keeter,
2002, p. 145). Adopting this perspective, the present study hypothesizes that although both the
specialist thesis and the generalist thesis are theoretically reasonable, the specialist thesis will
have more explanatory power than the generalist thesis.
Subsequently, the following section of
the paper will provide a more detailed theoretical discussion of the effects of the new media
NEW MEDIA AND SPECIALISTS
9
environment on the growth of the specialists.
H1a (the generalist thesis):
Individuals who are more educated are more likely to be
well-informed in the issue (health care reform).
H1b (the specialist thesis):
Individuals who perceive the issue (health care reform) to
be personally important to their family are more likely to be well-informed in the issue.
H1c (the specialist thesis):
Perceived issue importance will be a stronger predictor of
domain-specific knowledge than education.
The Role of New Media in the Growth of Specialists
Traditional Media and Generalists
Before hundreds of cable channels penetrated American households, most people
watched television for several hours every night.
They relied primarily on the evening news
broadcasts by three network channels to catch up on what was happening in the world.
During
the heyday of network news, many Americans were exposed to the news partly because they
were followed by their favorite sitcoms or because all three channels aired the news at the same
time (Prior, 2007). Although some elite newspapers and magazines might provide selective,
detailed, in-depth information, most citizens do not benefit from these media (Delli Carpini,
2002). For more than five decades, television has been the major source of political
information.
That most citizens are generalists rather than specialists relates closely with the
traditional media environment described above.
environment deserve particular attention here.
Two characteristics of such information
First, the political information supplied by
traditional media outlets is sufficiently homogeneous and standardized (Neuman, 1991, Steiner,
1952). Media corporations aim to seek larger audience and maximize profits. To appeal to as
NEW MEDIA AND SPECIALISTS
10
many viewers as possible, and more importantly to disturb as few as possible, the media outlets
produce media content that is ideologically moderate, non-controversial, and popular (Gerbner,
Gross, Morgan, & Signorielli, 1982). Accordingly, individuals’ personal tastes are ignored.
Even if people have special interests in a particular domain, they might have difficulties in
obtaining domain-specific information through one-way publishing or broadcasting media.
Another feature of the traditional media environment is that public news exposure is
non-selective or incidental (Downs, 1957).
Political learning from watching television occurs
unintentionally and passively (Schudson, 1998; Zukin & Snyder, 1984).
Researchers suggest
that incidental and habitual exposure to the evening newscasts leads to a decreased knowledge
gap between more or less educated citizens (Jerit, Barabas, & Bolsen, 2006; Kwak, 1999;
Neuman, 1976). This narrowing gap occurs because less educated people are accidentally or
occasionally exposed to TV news programs that are easily digestible, regardless of whether or
not they were particularly motivated to follow the news (Neuman, Just, & Crigler 1992).
The
political information reaches not only those educated and attentive but also those with low levels
of political interest and knowledge, thus allowing the latter group to keep up even with their
more attentive counterparts (Bennett & Iyengar, 2008).
Taken together, in the traditional media
environment characterized by homogenized information and non-selective exposure, the public
is better described as generalists and non-generalists rather than as an assemblage of selectively
informed specialists (Delli Carpini & Keeter, 2002).
New Media and Specialists
Contrary to traditional new media, the Internet and related media technologies allow
for more diversity in media content and more selectivity in media use. First, the new media has
the potential for more diversity with decreased cost and lowered space constraints. Amateurs
NEW MEDIA AND SPECIALISTS
11
are capable of creating and distributing their ideas more freely, resulting in long-tail diversity
(Anderson, 2006).
learning.
Second, the new media induce audiences’ selective exposure and selective
Information is not given linearly, but is sought selectively through the technological
functions, such as menu options or Google search.
Here, a crucial juncture is reached where
these technological affordances fit with specialists’ tendency to obtain information in only a few
domains they are interested in (Kim, 2009).
For example, in the traditional media environment,
individuals have difficulties satisfying their personal tastes and cultivating them because the
mass media usually do not provide detailed, in-depth information about specialized topics.
Thus, if the mass media do not help people to specialize in a certain topic, they either give up
becoming specialists or need to make additional efforts.
However, since extra efforts often cost
monetary resources or require human networks, those with higher socioeconomic status are more
likely to be eligible for the benefits.
On the other hand, in the new information environment,
people can obtain domain-specific knowledge with ease and efficiency only if they have interests
in a particular topic.
A growing body of work lends more support to this view by highlighting differences
between selective or motivated learning and incidental or passive learning.
According to the
cognitive psychology literature, when individuals are allowed to seek their own path of interest,
their motivation to learn grows, subsequently leading to a heightened attention level (Bandura,
1982; Chaffee & Schleuder, 1986).
Waal and Schoenbach (2007) found that although
newspaper readership predicts higher awareness of societal issues as compared to non-readership,
the relationship disappears among those who have minimal interest in the first place.
finding is also reported in political contexts.
A similar
For example, Holbrook, Berent, Krosnick, Vissler,
and Boninger (2005) found that after watching television, viewers were better able to recall the
NEW MEDIA AND SPECIALISTS
12
candidates’ statements about policy issues when they had personally important attitudes towards
those issues. More interestingly, they further demonstrated that attitude importance increases
knowledge acquisition only when accompanied by selective exposure and selective elaboration.
Furthermore, Johnson and Kaye (2000) found that those who are politically interested rely more
on the Internet rather than television for news consumption.
More recently, Kim (2009) showed
that selectivity in the use of the web produces higher domain-specific knowledge, attitude
extremity, and policy voting.
Although this research sheds light on the relationship between
web selectivity and domain-specific knowledge, it did not directly compare the role of different
types of media in fostering specialists.
Therefore, this study will further test which types of
media facilitate greater growth of specialists.
RQ2: Does the new media environment facilitate the growth of specialists?
H2: The relationship between perceived issue importance and domain-specific
knowledge (the specialist thesis) will become stronger among those who rely on new
media, as compared to those who rely on traditional media.
--- Figure 1 about here --Method
Data come from the Kaiser Family Foundation’s health tracking survey regarding health
care reform.
Telephone interviews were conducted with a randomly selected national sample
between April 9 and 14, 2010.
Participants and Demographic Characteristics
1,208 US adults who are 18 or older participated in the survey through either landline or
cell phone.
Respondents contacted through a cell phone were offered $5 in exchange for their
cell phone minutes spent during the interview.
Subjects reported their age (M=51.5, SD=18),
NEW MEDIA AND SPECIALISTS
13
sex (51.3% male), race (76.2% white), and household income (Median category=between
$50,000 and $75,000).
Measures
Domain-specific knowledge.
The study created a domain-specific knowledge index using nine dichotomous yes-no
knowledge items about the health care reform bill that had been passed by Congress in March
2010. Using a split-half sample method, different sets of knowledge items were given to each
half of the total sample.
The domain-specific knowledge index was constructed by counting the
number of items answered correctly (0=all wrong, 9=all correct, Cronbach’s α=.626, and .567 for
each half1). Two split-half samples were combined for further analysis2 (M=5.8, SD=2).
Main source of information.
Respondents were asked what is their main source of news and information about the
health reform bill (1=cable TV channels or their websites3, 2=network channels or their websites,
3=newspaper or newspaper websites, 4=other websites and blogs, 5=conversation with friends
and family, 6=radio, 7=elected officials, 8=an employer, 9=community, 10=none of the above).
While the majority of the respondents reported that television channels or their websites were the
most important source (38.9% Cable TV channels, 16.4% network TV channels), less than 10%
of respondents relied mostly on general websites and blogs (7.7%).
Perceived issue importance.
Respondents provided their perceptions about how much the health care reform would
affect their family personally (1=nothing at all to 4=a lot).
Control variables.
The study includes six control variables: age, gender, income, education, party
NEW MEDIA AND SPECIALISTS
14
identification, and the number of media people use.
These control variables were chosen based
on previous studies that examined the relationships between these control variables and either
media selection or political knowledge (e.g., Shen & Eveland, 2010).
Notably, the number of
media people use is included in the analysis to extract the unique influence of their main media
and to control the influence of other media.
An index of the number of media sources used was
created by counting the number of media sources respondents used to get information about the
health care reform bill (M=2.7, SD=1.3).
Result
The study first assessed whether domain-specific knowledge is predicted by general
education level or personal issue importance.
predicting domain-specific knowledge.
Table 1 presents OLS multiple regression models
Model 1 consists of control variables including age,
gender, income, party identification, and the number of media used.
education with model 1.
1.
Model 2 combined
Model 3 incorporated personal issue importance in addition to model
Finally, model 4 includes model 1 in conjunction with both education and personal issue
importance.
Model 1 alone explains 11 percent of the variance in domain-specific knowledge.
Sex
and age are not significant predictors, but individuals with higher household income (β= .115,
p<.01), Democrats (β=.153, p<.01), and those using diverse media (β=.264, p<.01) are more
likely to have higher scores on the health care reform bill knowledge index.
To assess the generalist thesis, model 2 included the education variable in addition to
model 1. The education variable did not add significant change to the variance initially explained
by the model 1.
R-square change = .003, F(1,930)=2.745, p-value=.098.
education was also not significant (β=.06, p=.098) at the conventional level.
The coefficient for
Thus, the
NEW MEDIA AND SPECIALISTS
15
generalist thesis (H1a) was not supported.
--- Table 1 about here --In contrast, the results of model 3 suggest that perceived issue importance is a significant
predictor of domain-specific knowledge.
Including personal importance in the model, the R-
square increased significantly (R-square change =.023, p<.01). The coefficient of the perceived
issue importance variable was sizable as well (β=.16, p<.01). Finally, I included both education
and perceived issue importance in the model to see if the perceived issue importance variable has
explanatory power beyond the education variable.
As expected, perceived issue importance
remained significant (β=.16, p<.01), but education became even less meaningful in the model
(β=.03, p=.246).
Taken together, the data supported the specialist thesis (H1b, H1c) but not the
generalist thesis (H1a).
The study has shown that the specialist thesis better describes the distribution of publics
in the health care reform domain. The findings indicate that well-informed citizens in the
health care domain are those who think that the issue matters to them personally, not those who
are more educated in general.
Now, I turn to the next question by investigating the role of media environment in the
growth of specialists.
This cross-sectional study cannot directly compare the effects of the new
media environments on issue-specific political learning with those of traditional media
environment. However, the survey question asking “what is your main source of information
about the health care reform bill?” allowed me to compare the characteristics of people who rely
on the Internet with those who rely on television network news, cable news, newspaper, and
radio.
More specifically, I hypothesized that using the selective media (e.g., websites)
accelerates knowledge acquisition in the domain that people think is personally important,
NEW MEDIA AND SPECIALISTS
16
whereas using the non-selective media (e.g., network TV, or radio) is not so helpful for people,
even in the domain that people perceive to be important.
--- Figure 2 about here ----- Table 2 about here --Figure 2 and Table 2 provide support for my prediction (H2).
As the specialist thesis
suggests, people in general tend to have higher domain-specific knowledge when they think the
domain is of great importance to them.
However, this relationship disappears if people rely on
network TV news, newspapers, and radio to obtain information concerning the health care
reform bill.
Probably, network TV news, newspapers, and radio are not so ideal for people to
learn about domain-specific knowledge because this type of media usually does not provide very
detailed knowledge to viewers due to the limited time and space.
In contrast, the relationship
between perceived issue importance and domain-specific knowledge remained significant among
those who use websites and cable news channels as a main source of information. Notably, this
relationship became even stronger among the website users (β=.319, p<.05). These findings
suggest that website users and cable TV audiences rather than other media users (network TV,
newspaper, and radio) tend to cultivate their interests more efficiently. Thus, the media that
provide more specialized content are more likely to help people become specialists in the domain
of their interest.
Discussion
Responding to recent changes in the information environment, many are concerned that
these changes will make democracy more vulnerable. One such concern is that the knowledge
gap between the more and less educated may expand.
Since the Internet affords selective
exposure, the “haves” can seek political information even more efficiently while the “have-nots”
NEW MEDIA AND SPECIALISTS
17
are able to filter out political information more easily (Sunstein, 2001).
that the Internet facilitates audience fragmentation.
The second concern is
As citizens tend to visit websites that are
mostly frequented by like-minded people, they may fail to be exposed to cross-cutting views,
causing their attitudes to become even more extreme (Iyengar & Hahn, 2009; Stroud, 2010).
Although these views seem to be legitimate concerns, the findings of this study suggest
alternative perspectives.
First, concerns about the increasing knowledge gap are based on the
assumption that the knowledge gap widens between the more and less educated across a wide
range of issues.
However, the specialist thesis, supported by the study, indicates that even
though such a gap may arise, it is more likely to do so between those interested and those
uninterested in a particular issue rather than between more and less educated citizens.
Furthermore, given that individuals possess varying levels of interests in each issue, the
knowledge gap is not uniformly processed across a wide range of issues; thus the concerns over
the increasing political information inequality may not be as threatening as we think.
increased specialization may not necessarily trigger audience fragmentation.
Second,
To the extent that
the new media environment allows previously uninvolved citizens to cultivate an interest in
particular domains, the new media may well function as a gateway into other adjacent domains.
In addition, as people come to feel more comfortable with learning through the new media, they
tend to become more politically efficacious (Delli Carpini & Keeter, 2002).
Another important finding of this study is that the type of media plays a moderating role
in the relationship between personal issue importance and learning.
Interestingly, while patrons
of network TV news, newspaper and talk radio do not reflect their knowledge in proportion to
their issue importance, users of websites and cable TV news display higher level of knowledge
according to their issue importance.
Supporting this view, Holbrook et al., (2005) found that
NEW MEDIA AND SPECIALISTS
18
the relationship personal issue importance and knowledge acquisition persists only when
selective exposure and selective elaboration are allowed.
If I juxtapose the present study with
Holbrook et al.’s studies, the assumption is made that only two media, the Internet and cable TV,
allow selective exposure while the other media do not.
The findings from this study help answer some important questions about how the
changing media environment shapes the formation of the mass polity while paving the way for
future investigations.
However, these contributions must be qualified by several limitations.
First, the investigation into a single-issue domain, in this case health care reform, cannot be
generalized to other domains with confidence.
For instance, more polarized issues, such as
welfare policy and abortion, or nationally urgent issues, such as war or natural disaster, might
show entirely different pictures of the dynamics in the mass polity.
Second, as the survey data are cross-sectional in nature, relationships must be qualified
as correlational.
Although a number of predictors, such as demographics, are clearly exogenous,
the causal directions between knowledge, personal issue importance, and media use are far less
clear.
To make a stronger causal inference, future work is needed that involves experimental or
longitudinal design.
Finally, the study did not measure the degree of selectivity or content diversity directly.
The study simply relied on conventional assumption that new media, such as the Internet, will
provide more diverse media content and selective exposure as compared to more traditional
media, such as network television and radio.
This study cannot completely rule out the
possibilities that other characteristics of the media facilitate the growth of specialists.
One
interesting question in this line of future research will be whether the Internet activity is always
selective.
For example, as far as the degree of selectivity is concerned, visiting the
NEW MEDIA AND SPECIALISTS
19
newyorktimes.com regularly will be a different activity from typing a word in a Google search
box.
In addition, recent research (e.g., Lee, 2009; Waal & Schoenbach, 2007) suggests that the
Internet offers various opportunities for incidental as well as selective exposure.
For instance,
Facebook users are incidentally exposed to provocative news articles or YouTube clips that are
posted by one of their Facebook friends.
Thus, it will be interesting to see whether the
experience of social networking sites fosters specialists or generalists.
NEW MEDIA AND SPECIALISTS
20
Reference
Anderson, C. (2006). The Long Tail: Why the Future of Business Is Selling Less of More. New
York, Hyperion.
Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37,
122-147.
Bennett, L., & Iyengar, S. (2008). A new era of minimal effects? The changing foundations of
political communication. Journal of Communication, 58, 707–731
Berent, M. K., & Krosnick, J. A. (1995). The relation between political attitudes importance and
knowledge structure. In M. Lodge & K.McGraw (Eds.), Political judgment: Structure
and process (pp. 91-110). Ann Arbor: University of Michigan Press.
Boninger, D. S., Krosnick, J. A., Berent, M. K., & Fabrigar, L. R. (1995). The causes and
consequences of attitude importance. In R. E. Petty & J. A. Krosnick (Eds.), Attitude
strength: Antecedents and consequences (pp. 159-190). Mahwah, NJ: Lawrence Erlbaum.
Chaffee, S. H., & Schleuder, J. (1986). Measurement and effects of attention to media news.
Human Communication Research, 13, 76-107.
Converse, P. E. (1964). The nature of belief systems in mass publics. In D. A. Apter (Ed.),
Ideology and discontent (pp. 206-261). New York: Free Press.
Dalrymple, K., & Scheufele, D. (2007). Finally informing the electorate? How the
Internet got people thinking about presidential politics in 2004. Harvard International
Journal of Press/Politics, 12,96-111.
Delli Carpini, M. X., & Keeter, S. (1996). What Americans know about politics and why it
matters. New Haven, CT: Yale University Press.
Delli Carpini, M. X., & Keeter, S. (2002). The Internet and an informed citizenry. In D. Anderson,
NEW MEDIA AND SPECIALISTS
21
& M. Cornfield (Eds.), In The civic web: Online politics and democratic values, (pp.
129-156). Rowman and Littlefield.
Downs, A. (1957). An Economic Theory of Democracy. New York: Harper.
Gerbner,G., Gross, L., Morgan, M., & Signorielli, N. (1982). Charting the mainstream:
television’s contributions to political orientations. Journal of Communication, 32, 100127.
Gershkoff, A.(2006). How issue interest can rescue the American public. Doctoral Dissertation.
Princeton University.
Habermas, J. (1984). A theory of communicative action (T. McCarthy, Trans.). Boston:
Beacon.
Holbrook, A., Berent, M.K., Krosnick, J.A., Vissler, P.S., & Boninger, D. (2005).Attitude
importance and the accumulation of attitude-relevant knowledge in memory. Journal of
Personality and Social Psychology, 88, 749-769.
Iyengar, S. (1986). Analysis of ANES pilot study data. Paper presented at National Election
Study Pilot Study Conference, Ann Arbor, MI.
Iyengar, S. (1990). Shortcuts to political knowledge: The role of selective attention and
accessibility. In J. A. Ferejon, & J. H. Kuklinski (Eds.), Information and democratic
processes (pp. 160-185). Urbana: University of Illinois Press.
Iyengar, S., & Hahn, K. S. (2009). Red media, blue media: evidence of ideological selectivy in
media use. Journal of Communication, 59, 19-39.
Jerit, J., Barabas, J., & Bolsen, T. (2006). Citizens, Knowledge, and the Information
Environment. American Journal of Political Science, 50, 266-282.
Johnson, T. J., & Kaye, B. K. (2000). Using is believing: The influence of reliance on the
NEW MEDIA AND SPECIALISTS
22
credibility of political information among politically interested Internet users.
Journalism & Mass Communication Quarterly, 77, 865-879.
Kim, Y. M., (2009). Issue Publics in the New Information Environment : Selectivity, Domain
specificity, and extremity. Communication Research, 36, 254-284.
Krosnick, J. A. (1990). Government policy and citizen passion: A study of issue publics in
contemporary America. Political Behavior, 12, 59-92.
Krosnick, J. A., & Telhami, S. (1995). Public attitudes toward Israel: A study of the attentive and
issue publics. International Studies Quarterly, 39, 535-554.
Kwak, N. (1999). Revisiting the knowledge gap hypothesis. Communication Research, 26,
385-413.
Lee, J. K., (2009). Incidental exposure to news: limiting fragmentation in the new media
environment. Doctoral Dissertation. University of Texas at Austin. Retrieved
from http://repositories.lib.utexas.edu/
Lupia, A. & McCubbins, M. (1998). The democratic dilemma: Can citiznes learn what they need
to know? New York: Cambridge University Press.
McDevitt, M., & Chaffee, S. (2000). Closing gaps in political communication and knowledge:
Effects of a school intervention. Communication Research, 27, 259–292.
Neuman, W. R. (1976). Patters of recall among news viewers. Public Opinion Quarterly, 40,
115-123.
Neuman, W. R. (1986). Paradox of Mass Publics: Knowledge and opinion in the electorate.
Cambridge, MA: Harvard Press.
Neuman, W. R. (1991). The Future of the Mass Audience. New York: Cambridge University
Press.
NEW MEDIA AND SPECIALISTS
23
Neuman,W. R., Just, M.R., Crigler, A.N. (1992). Common Knowledge: News and the
Construction of Political Meaning. Chicago: University of Chicago Press.
Petty, R. E., & Krosnick, J. A. (1995). Attitude strength: Antecedents and consequences. Mahwah,
NJ: Lawrence Erlbaum.
Price, V., David, C., Goldthorpe, B., Roth, M. M., & Cappella, J. N. (2006). Locating the issue
public: The multi-dimensional nature of engagement with health care reform. Political
Behavior, 28, 33-63.
Price, V., & Zaller, J. (1993). Who gets the news? Alternative measures of news reception and
their implications for research. Public Opinion Quarterly, 57, 133-164.
Prior, M. (2007). Postbraodcasting Democracy. Cambridge Universty Press.
Schudson, M. (1998). The Good Citizen: A History of American Public Life. New York: Free
Press.
Shen, F., & Eveland, W. P. (2010). Testing the Intramedia Interaction Hypothesis: The
Contingent Effects of News. Journal of Communication, 60, 364-387.
Steiner, P. O. (1952). Program Patterns and Preferences and the Workability of Competition in
Radio Broadcasting. Quarterly Journal of Economics, 66, 194-223.
Stroud, N. J. (2010). Polarization and Partisan Selective Exposure. Journal of Communication,
60, 556-576.
Sunstein, C. (2001). Republic.com. Princeton, NJ: Princeton University Press.
Tewksbury, D., Weaver, A. J., & Maddex, B. D. (2001). Accidentally informed: Incidental
news exposure on the world wide web. Journalism & Mass Communication Quarterly,
78, 533-554.
Vissler,P.S., Holbrook,A., & Krosnick, J.A. (2008). Knowledge and Attitude. In W. Donsbach &
NEW MEDIA AND SPECIALISTS
24
M.W. Traugott (Eds.). Sage Handbook of Public Opinion Research. 127-140. Thousand
Oaks, CA: Sage
Waal, E.de, & Schoenbach, K. (2007). How print newspapers and online news expand awareness
of public affairs issues. Paper presented at the annual meeting of the International
Communication Association, San Francisco, CA.
Wei, L., & Hindman, D. B. (2011). Does the Digital Divide Matter More? Comparing the Effects
of New Media and Old Media Use on the Education-Based Knowledge Gap. Mass
Communication and Society, 14, 216-235.
Zaller, J. R. (1992). The nature and origins of mass opinion. New York: Cambridge University
Press.
Zaller, J., & Feldman, S. (1992). A simple theory of the survey response: Answering questions
means revealing preferences. American Journal of Political Science, 36, 579-618.
Zukin, C. & Snyder, R. (1984). Passive Learning:When the Media Environment Is the Message.
Public Opinion Quarterly, 48, 629–38.
NEW MEDIA AND SPECIALISTS
25
Footnotes
1.
Relatively low Cronbach’s α does not indicate the flaw of the knowledge index.
In order to
create a knowledge index that taps into multiple dimensions of knowledge (i.e., to increase
validity), reliability of the measure is inevitably compromised to some extent (Delli Carpini &
Keeter, 1992).
2. For simpler presentation, the domain-specific knowledge index is collapsed into low,
medium, and high categories.
3.
With this measure, CNN cable viewers and CNN.com users cannot be differentiated.
However, this does not compromise the validity of the measure because people regularly going
to CNN.com or Newyorktimes.com do not necessarily engage in selective news exposure.
In
fact, websites of news organizations provide visitors with a variety of opportunities for incidental
exposure to a wide range of issues (Tewksbury, Weaver, and Maddex, 2001) while blogs and
other specialized websites offer selective news exposure.
Similarly, Waal and Schoenbach
(2007) found that exposure to printed newspapers and to online newspapers did not make any
difference in terms of awareness of societal and political issues.
NEW MEDIA AND SPECIALISTS
26
Table 1.
OLS Regressions predicting domain-specific knowledge
Model 1
Model 2
(Model1+Educati
on)
Controls
Female
Age
Income
Model 3
(Model1+
Personal
importance)
Model 4
(Full Model)
b
ββ
(S.E.)
b
(S.E.)
β
b
(S.E.)
β
b
(S.E.)
β
-.17
(.13)
-.00
(.00)
.10**
(.03)
-.17
(.13)
-.00
(.00)
.07**
(.03)
-.04
-.14
(.13)
-.01
(.00)
.08**
(.03)
-.03
-.14
(.13)
-.01
(.00)
.06*
(.03)
-.03
.08#
(.05)
.06
.05
(.05)
.37**
(.080
.04
-.04
-.03
.12
Education
Perceived
importance
R-square
R-square change
from model1
Note: # p<.1. * p<.05.
-.03
.09
.38**
(.08)
.117**
.11**
.12**
.00#
** p<.01. n.s. = non-significant
-.05
.09
.16
.13**
.02**
-.05
.07
.16
.14**
.03**
NEW MEDIA AND SPECIALISTS
27
Table 2.
OLS Regression predicting domain-specific knowledge X main source of information
Main source
Perceived
RN
of
importance
square
information
b (S.E.)
Standardized β
Cable TV
.090*
.109
.142
416
(.044)
Network TV .096 n.s.
.131
.088
176
(.066)
Newspaper
.062 n.s.
.073
.123
150
(.085)
Website
.229*
.319
.282
82
(.097)
Radio
.105 n.s.
.158
.056
106
(.075)
The regression model includes control variables (sex, age, income, education, party identification,
and the number of media sources used).
NEW MEDIA AND SPECIALISTS
28
Type of Media
H2
Perceived Importance
H1b, H1c
Domain-Specific Knowledge
Education
H1a
Figure 1.
Hypotheses
NEW MEDIA AND SPECIALISTS
29
6.8
.319*
6.6
6.4
DomainSpecific
Knowledge
(DV)
n.s.
n.s.
6.2
n.s.
.109*
6
5.8
5.6
5.4
Cable (n=416)
Network
(n=176)
Webiste
(n=82)
Newspaper
(n=150)
Radio (n=106)
5.2
Low
High
Perceived Issue Importance (IV)
Figure 2.
Predicting domain-specific knowledge with perceived issue importance X main source of
information (media type).
Note: This regression model includes control variables: sex, age, education, income, party
identification, and the number of media sources used. The values on the graph represent
standardized coefficients (β) of perceived issue importance in the regression model.