The Internet, selective learning, and the rise of issue specialists
First Monday

The Internet, selective learning, and the rise of issue specialists by S. Mo Jang and Yong Jin Park

Using national survey data (N = 1,208) in the U.S., the present study found that individuals relying upon the Internet translated their interest in the health care issue into issue–specific knowledge. However, those who depended on network TV, newspapers, and radio failed to display a high level of issue–specific knowledge, even when they were interested in the issue. The findings suggest that the Internet plays an important role in fostering issue specialists rather than generalists.


Who are knowledgeable citizens in an issue?
Traditional media and by–product learning
The Internet and selective learning
Discussion and conclusion




Theorists welcomed television as a “knowledge leveler” [1] that reduces the inequality in political knowledge (Eveland and Scheufele, 2000). 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 efficiently and seek entertainment single–mindedly. As a result, 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 and Iyengar, 2008; Park, 2011; Park, et al., 2012).

Despite the scholarly concerns about the decline of information generalists in the new information environment, relatively little attention has been paid to the advantage of information specialists, who are knowledgeable only within a particular domain of their interest. If people can seek their path of interest directly via the search function of the Internet, they will engage in more effective ways of information processing with their increased levels of motivation to learn and attention (Bandura, 1982). They are more likely to learn about what is going on in the world as well especially when the topics are personally interesting to them (Prior, 2007). Thus, in the emerging media environment, the most powerful driving force of knowledge acquisition in a certain domain would be individuals’ interest in the issue rather than conventional resources such as education or personal connections with experts.

Our study addresses this issue by examining whether the Internet, as compared to traditional media, facilitates the selective learning, driven by personal issue interests. Using a national survey about a health care reform bill in the U.S., we examine whether informed citizens in the health care domain consist of those generally educated or those specifically interested in the health issue. Then, the study investigates the role of different mediums in cultivating information specialists.



Who are knowledgeable citizens in an issue?

Habermas (1984–1987) posits that the functioning of a healthy democracy requires an informed citizenry whose attitudes and participation are based on a broad set of relevant and accurate information. 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 (Zaller, 1992). Second, overall levels of political knowledge in the U.S are frustratingly low (Lupia and McCubbins, 1998; Neuman 1986). Third, knowledge is unevenly distributed across the population and is associated with socioeconomic factors (Delli Carpini and Keeter, 1996).

However, relatively little has been examined concerning the distributed patterns of knowledge at individual levels. Who are informed citizens in each issue domain? Are their knowledge levels fluctuating or stable across domains? Reponses to these questions have varied 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 second model emphasizes that public opinion is stratified based on conventional resources such as education. Although the average citizen may not be knowledgeable in general, democracy functions owing to a small number of elites who are attentive, active, and well informed (Neuman, 1986). In the next few paragraphs we will examine each of these two arguments — the information specialist thesis and the information generalist thesis.

The information specialist thesis: Issue interest matters

Theoretically, the concept of issue public is a useful framework for developing hypotheses about why citizens are more likely to be information specialists rather than information generalists. The premise of issue public explains how American citizens engage in politics, although most of them show low level of political knowledge (Converse, 1964). Practically, most people have few resources and little motivation to pay attention to all of the nation’s social and political issues. Thus, citizens should be expected to concentrate on only a few issue domains and be selective in gathering and acquiring information within a domain.

Another important premise of issue public is that individuals need not be well educated to form attitudes regarding the issues they perceive as interesting or important. Prior research has indicated that perceived self–interest motivates individuals to obtain domain–specific knowledge and engage in policy evaluations (Berent and Krosnick, 1995). One explanation of this interest driven information–seeking and information–evaluating behavior is that the knowledge construct of a certain issue becomes more accessible when people are interested in the issue (Iyengar, 1990).

The information generalist thesis: Education matters

The information 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 owing to a small number of sophisticated, educated, and attentive elites (Price and Zaller, 1993; Zaller, 1992).

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 (cf., Grabe, et al., 2009; Krosnick, 1990). For example, more educated individuals are more familiar with political issues and more knowledgeable about political events (Neuman, 1986). 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 about other issues as well (Delli Carpini and 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 remarked [2], “The model is not wrong, but it is incomplete.” In addition, Delli Carpini and Keeter (2003) 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.’ [3] Adopting this perspective, the present study hypothesizes that although both the specialist thesis and the generalist thesis are theoretically reasonable, the specialist thesis will grow more convincing than ever before in this Internet era. Subsequently, the following section of the paper will provide a more detailed theoretical discussion of the effects of the new media environment on the growth of the specialists.

H1: Personal issue interest, compared to education, will be a stronger predictor of issue–specific knowledge.



Traditional media and by–product learning

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 news programs 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, and in–depth information, most citizens do not benefit from these media. For more than five decades, television has been the major source of political information.

Such traditional media environment offers ample opportunities for by–product learning (Downs, 1957; Lee, 2009; Zukin and Snyder, 1984). The rational theory posits that people collect information not only through active seeking behaviors but also through accidental exposure to information (Downs, 1957). Most important is that by–product learning enables individuals to minimize the information cost such as time and cognitive energy.

The features of by–product learning are well incorporated into the process of learning from traditional news media. Even viewers who do not have much interest in public affairs are likely to encounter news information on television regardless of their intentions.

Researchers have gathered empirical evidence of incidental learning in multiple contexts. Blumler and McQuail (1968) found that viewers were able to identify policies more accurately than indifferent non–viewers. Neuman, et al. (1992) showed that television was more effective for teaching people about low–salience issues, indicating incidental learning with low involvement.

Many researchers paid attention to the fact that incidental exposure leads to a decreased knowledge gap between more and less educated citizens (Kwak, 1999; Neuman, 1976). The gap narrows because less educated people are accidentally or occasionally exposed to TV news programs that are easily digestible, regardless of whether or not these viewers were particularly motivated to follow the news (Neuman, et al., 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 and Iyengar, 2008).

Another notable characteristic of traditional media outlets is homogeneous media content. The media content provided by centralized broadcast is ideologically moderate, non–controversial, and popular (Gerbner, et al., 1982). To the extent that the media do not cover various spectrums of areas, it can be that individuals’ personal tastes are ignored. Even if people have special interests in a particular domain, they might have difficulties obtaining relevant information through one–way publishing media. Taken together, in the traditional media environment characterized by by–product learning and homogenized information, the public is more accurately described as information generalists rather than as selectively informed specialists (Delli Carpini and Keeter, 2003).



The Internet and selective learning

Contrary to traditional news media, the Internet and related media technologies allow for audiences’ selective learning. The technological functions, such as menu options or Google search, enable individuals to seek information directly without having to wait for the mass media to provide information they want. These technological affordances fit well with specialists’ tendency to acquire information in only a few domains of their concerns (Sunstein, 2001). For example, in the traditional media environment, people are not able to develop their personal tastes because the mass media usually do not supply specialized information that might not appeal to other general viewers. Thus, if the mass media do not help people specialize in a particular topic, they either give up becoming specialists or need to make additional efforts. In contrast, in the new information environment, individuals can obtain issue–specific knowledge easily as long as they are interested 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 and Schleuder, 1986). 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 (de Waal and Schoenbach, 2008). A similar finding is also reported in political contexts. For example, after watching television, viewers were better able to recall the candidates’ statements about policy issues when they perceived those issues to be personally important (Holbrook, et al., 2005). More interestingly, they 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. There is a recent finding that selectivity in the use of the Web produces higher issue–specific knowledge, attitude extremity, and policy voting (e.g., Kim, 2009). Although such findings shed light on the relationship between the Web selectivity and issue–specific knowledge, few studies did directly compare the role of different types of media in fostering specialists. Therefore, this study will further test which types of media make greater contribution to the growth of information specialists.

RQ1: Does the new media environment facilitate the growth of information specialists?

H2: The relationship between personal issue interest and issue–specific knowledge (the information specialist thesis) will become stronger among those who rely on new media, than those who rely on traditional media.




Participants and demographic characteristics

The study used the secondary data from the Kaiser Family Foundation’s health tracking survey regarding health care reform. Telephone interviews were conducted with 1,208 U.S. adults between 9–14 April 2010, a few weeks after the health care reform bill was passed by the U.S. Congress and signed by President Obama in March.

A combination of landline (N = 800) and cell phone (N = 408) random digit dial (RDD) samples were used to represent all adults in the U.S. who have access to a phone. The cell samples were offered US$5 in exchange for their cell phone minutes spent during the interview. The response rate for the landline and cell sample was 22.4 percent and 21.6 percent respectively. Subjects reported their age (M = 51.5, SD = 18.0), sex (51.3 percent male), race (76.2 percent white), and household income (Median category=between US$50,000 and US$75,000).


Issue–specific knowledge

The study created an issue–specific knowledge index using nine dichotomous yes–no knowledge items about the health care reform bill that had been signed in March 2010. Using a split–half sample method, different sets of knowledge items were given to each half of the total sample. The issue–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 half) [4]. Two split–half samples were combined for further analysis (M = 5.8, SD = 2.0) [5].

Main source of information

Respondents were asked what is their main source of news and information about the health care reform bill (1 = cable TV channels, 2 = network channels, 3 = newspaper, 4 = the Web 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 were the most important source (38.9 percent cable TV channels, 16.4 percent network TV channels), less than 10 percent of the respondents relied mostly on the Web and blogs (7.7 percent).

Personal issue interest

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, party identification, and the number of media that people use. For this, we followed previous studies that examined the relationships between these control variables and political knowledge (Shen and Eveland, 2010). Education was measured on a seven–point scale, ranging from 1 = none or grade 1–8 to 7 = post graduate or professional schooling (M = 4.8, SD = 1.6). 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 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).




This study first assessed whether issue–specific knowledge is predicted by general education level or personal issue interest. Table 1 presents OLS multiple regression models predicting issue–specific knowledge. Model 1 consists of control variables including age, gender, income, party identification, and the number of media used. Model 2 combined education with model 1. Model 3 incorporated personal issue interest in addition to model 1. Finally, model 4 includes model 1 in conjunction with both education and personal issue interest.

Model 1 alone explains 11 percent of the variance in issue–specific knowledge. Gender and age are not significant predictors, but individuals with higher household income (β = .11, p < .01), Democrats (β = .15, p < .01), and those using diverse media (β = .26, p < .01) are more likely to have higher scores on the health care reform bill knowledge index.

To assess the information generalist thesis, model 2 included the education variable in addition to model 1. The education variable did not add a significant change to the variance initially explained by the model 1. R–square change = .00, F (1,930) = 2.75, p = .10. The coefficient for education was also not significant (β = .06, p = .10) at the conventional level. Thus, the information generalist thesis was not supported.


Table 1: OLS regressions predicting issue-specific knowledge.
Note: * p < .05; ** p < .01.
 Model 1Model 2
Model 3
(Model1+ Personal Interest)
Model 4
(Full Model)
Number of media.42**
Education .08
.06 .05
Personal issue interest .38**
R–square .12** .12**


In contrast, the results of model 3 suggest that personal issue interest is a significant predictor of issue–specific knowledge. Including personal issue interest in the model, the R–square increased significantly (R–square change = .02, p < .01). The coefficient of the personal issue interest variable was sizable as well (β = .16, p < .01). Finally, we included both education and personal issue interest in the model to see if the personal issue interest variable has explanatory power above and beyond the education variable. As expected, personal issue interest 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 information specialist thesis (H1), indicating that well–informed citizens in the health care domain are those who think that the issue matters to them personally, rather than those who are more educated in general.

The research question concerned the role of media environment in the growth of specialists. This cross–sectional study cannot directly compare the effects of the new media environment 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 us 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, we hypothesized that using the selective media (e.g., the Web and blogs) accelerates knowledge acquisition in the domain that people think is personally important, 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 personally important to them.


Predicting issue-specific knowledge with personal issue interest X main source of information
Figure 1: Predicting issue–specific knowledge with personal issue interest X main source of information (media type).
Note: This regression model includes control variables: gender, age, education, income, party identification, and the number of media sources used. The values on the graph represent unstandardized coefficients (b) of personal issue interest in the regression model.



Table 2: OLS regression predicting issue–specific knowledge X main source of information.
Note: * p < .05; ** p < .01.
The regression model includes control variables (gender, age, income, education, party identification, and the number of media sources used).
Main source of informationMain source of informationR–squareN
Standardized β  
Cable TV.09*
Network TV.10
Web sites.23*


The findings in Figure 1 and Table 2 provide support for the prediction (H2). As the information specialist thesis suggests, people in general tend to show higher issue–specific knowledge when they think the domain is of great interest 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 issue–specific knowledge because this type of media usually does not provide very detailed knowledge to viewers due partly to the limited time and space. In contrast, the relationship between personal issue interest and issue–specific knowledge remained significant among those who use the Internet (β = .32, p < .05) and cable TV channels (β = .11, p < .05) as a main source of information.

This result indicates that the Internet is probably the most efficient tool for individuals to translate their issue–interest into issue–specific knowledge. Notably, cable TV users also show a significant relationship between personal issue interest and issue–specific knowledge. Although it might be due to the relatively larger sample size than other source users, it is also possible that cable TV users are able to develop their interests owing to hundreds of cable TV channels that provide viewers with specialized content. Overall, the findings suggest that the new media, known to be more selective and specialized, are more efficient tools to help people cultivate their interests and become information specialists in the domain.



Discussion and conclusion

Responding to recent changes in the information environment, many scholars 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” are able to filter out political information more easily (Sunstein, 2001). The second concern is that the Internet facilitates audience fragmentation. As citizens tend to visit the Internet sites that are frequented by like–minded people, they may not be exposed to cross–cutting views, causing their attitudes to become even more extreme (Iyengar and Hahn, 2009; Stroud, 2010).

Although these 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 information specialist thesis, supported by the study, indicates that even though such a gap may appear, it is more likely to do so between those who are and are not interested in a particular issue rather than between people who are more and less educated. Furthermore, given that individuals report varying levels of interests across issues, the knowledge gap is not uniformly processed across a wide range of issue domains; thus the concerns over the increasing political information inequality may not be as threatening as we think.

Second, increased specialization may not necessarily trigger audience fragmentation. Insofar as 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 become more comfortable with learning through the new media, they tend to become more politically efficacious (Delli Carpini and Keeter, 2003). Furthermore, others, especially supporters of a pluralistic model of democracy, might argue that generalists are not necessary for democracy to work, and that issue specialists would achieve the same effect (Converse, 1964). In favor of this perspective, numerous studies (Gershkoff, 2006; Krosnick and Telhami, 1995) have highlighted the role of issue specialists in light of their issue–based participation. These studies indicated that issue specialists tend to exert pressure on government and make voting decisions on the basis of their issue positions.

Another important finding of this study is that the type of media plays a moderating role in the relationship between personal issue interest and learning. Interestingly, while patrons of network TV news, newspaper and talk radio do not reflect their knowledge in proportion to their issue interest, users of the Internet and cable TV news display a higher level of knowledge according to their issue interest. Supporting this view, Holbrook, et al. (2005), found that the relationship between personal issue interest and knowledge acquisition persists only when selective exposure and selective elaboration are allowed. If we juxtapose the present study with Holbrook, et al.’s (2005) 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 interest, and media use are far less clear. To make a stronger causal inference, future work is needed that involves experimental or longitudinal design.

One interesting question in this line of future research will be whether and to what degree the Internet activity remains selective. In this regard, we do not completely rule out the possibilities that other characteristics of the new media facilitate the growth of specialists. For example, as far as the degree of selectivity is concerned, visiting the regularly will be a different activity from typing a word in a Google search box. In addition, recent research (Lee, 2009; de Waal and Schoenbach, 2008) 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. End of article


About the authors

S. Mo Jang is a Ph.D. candidate in communication studies at the University of Michigan at Ann Arbor. His current research focuses on media effects, media psychology, science communication, and political implications of new media. Jang received his B.A. in mass communication from Seoul National University. Prior to entering the doctoral program at Michigan, he worked as a television journalist in Seoul Broadcasting System in South Korea. He has taught courses at Michigan, including media effects, managing information environment, and quantitative research methods. His conference papers were (and will be) presented at AAPOR, AEJMC, ICA, and WAPOR. He won paper awards at Mass Communication and Society Division, AEJMC 2011 and Mass Communication Division, ICA 2012.
E–mail: jangpro [at] umich [dot] edu

Yong Jin Park (Ph.D., Communication, University of Michigan; M.A., Communication Management, USC Annenberg) is Assistant Professor in Radio TV, and Film at the School of Communications at Howard University. His research centers on social and policy/political implications of new communication technologies, spanning the field of communication policy, media institutions, and new media users. His works (will) appear in Communication Research, Telecommunications Policy, Communication Teacher, Journal of Broadcasting & Electronic Media, Government Information Quarterly, Journal of Communication, Journal of Information Policy, Javnost — The Public, Sociology Compass, The Information Society, and Info: Journal of Policy, Regulation, and Telecommunications. In addition, he contributed to the East Asian section of the Privacy & Human Rights Report: 2005 by Electronic Information Privacy Center (EPIC) in Washington D.C. At Michigan, he taught courses on media institution, new media technology and policy, and research methods and worked as a research assistant to various projects related to new media policy, media consolidation, digital revolution, and political behavior/emotion (experimental studies), including a NSF–funded project. He won the Top Paper Award at Communication Technology Division, AEJMC 2005 and presented numerous articles in such conferences as TPRC, ICA, NCA, and IAMCR. At Howard, he is an active member of the Howard Media Group, a social science research group that examines the minority ownership and access disparities in broadcasting and new media industry.
E–mail: yongjin [dot] park [at] howard [dot] edu



The authors would like to thank Dr. Russell Neuman, Dr. Michael Traugott, Dr. Josh Pasek, Dr. Seungahn Nah, and Dr. Hoon Lee for their invaluable suggestions and comments for this manuscript.



1. Neuman, 1976, p. 122.

2. Neuman, 1986, p. 39.

3. Delli Carpini and Keeter, 2003, p. 145.

4. Relatively low Cronbach’s α does not necessarily indicate the limit of the knowledge index. Rather, this reflects the dichotomous nature of single items. Furthermore, 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 and Keeter, 2003). In other words, if knowledge items were highly correlated with one another, and reach a high level of reliability, it is hard to establish their discriminant validity. Faced with the trade–off, we chose discriminant validity rather than reliability since the main purpose of constructing the knowledge index in the current study was assessing factors influencing different levels of domain–specific knowledge.

5. For a simpler presentation, the issue–specific knowledge index is collapsed into low, medium, and high categories.



Albert Bandura, 1982. “Self–efficacy mechanism in human agency,” American Psychologist, volume 37, number 2, pp. 122–147.

Lawrence Bennett and Shanto Iyengar, 2008. “A new era of minimal effects? The changing foundations of political communication,” Journal of Communication, volume 58, number 4, pp. 707–731.

Matthew. K. Berent and Jon A. Krosnick, 1995. “The relation between political attitude importance and knowledge structure,” In: Milton Lodge and Kathleen M. McGraw (editors). Political judgment: Structure and process. Ann Arbor: University of Michigan Press, pp. 91–110.

Jay G. Blumler and Denis McQuail, 1968. Television in politics: Its uses and influence. London: Faber.

Steven H. Chaffee and Joan Schleuder, 1986. “Measurement and effects of attention to media news,” Human Communication Research, volume 13, number 1, pp. 76–107.

Philip E. Converse, 1964. “The nature of belief systems in mass publics,” In: David E. Apter (editor). Ideology and discontent. London: Free Press of Glencoe, pp. 206–261.

Ester de Waal and Klaus Schoenbach, 2008. “Presentation style and beyond: How print newspapers and online news expand awareness of public affairs issues,” Mass Communication & Society, volume 11, number 2, pp. 161–176.

Michael X. Delli Carpini and Scott Keeter, 2003. “The Internet and an informed citizenry,” In: Daivd M. Anderson and Michael Cornfield (editors). The civic Web: Online politics and democratic values. Lanham, Md.: Rowman and Littlefield, pp. 129–156.

Michael X. Delli Carpini and Scott Keeter, 1996. What Americans know about politics and why it matters. New Haven, Conn.: Yale University Press.

Anthony Downs, 1957. An economic theory of democracy. New York: Harper.

William P. Eveland and Daniel A. Scheufele, 2000. “Connecting news media use with gaps in knowledge and participation,” Political Communication, volume 17, number 3, pp. 215–237.

George Gerbner, Larry Gross, Michael Morgan and Nancy Signorielli, 1982. “Charting the mainstream: Television’s contributions to political orientation,” Journal of Communication, volume 32, number 2, pp. 100–127.

Amy R. Gershkoff, 2006. “How issue interest can rescue the American public,” doctoral dissertation in political science, Princeton University.

Maria E. Grabe, Rasha Kamhawi and Narine Yegiyan, 2009. “Informing citizens: How people with different levels of education process television, newspaper, and Web news,” Journal of Broadcasting & Electronic Media, volume 53, number 1, pp. 90–111.

Jürgen Habermas, 1984–1987. A theory of communicative action. Two volumes. Translated by Thomas McCarthy. Boston: Beacon Press.

Allyson L. Holbrook, Matthew K. Berent, Jon A. Krosnick, Penny S. Visser and David S. Boninger, 2005. “Attitude importance and the accumulation of attitude–relevant knowledge in memory,” Journal of Personality and Social Psychology, volume 88, number 5, pp. 749–769.

Shanto Iyengar, 1990. “Shortcuts to political knowledge: The role of selective attention and accessibility,” In: John A. Ferejohn and James H. Kuklinski (editors). Information and democratic processes. Urbana: University of Illinois Press, pp. 160–185.

Shanto Iyengar and Kyu S. Hahn, 2009. “Red media, blue media: Evidence of ideological selectivity in media use,” Journal of Communication, volume 59, number 1, pp. 19–39.

Thomas J. Johnson and Barbara K. Kaye, 2000. “Using is believing: The influence of reliance on the credibility of political information among politically interested Internet users,” Journalism & Mass Communication Quarterly, volume 77, number 4, pp. 865–879.

Young Mie Kim, 2009. “Issue publics in the new information environment: Selectivity, domain specificity, and extremity,” Communication Research, volume 36, number 2, pp. 254–284.

Jon A. Krosnick, 1990. “Government policy and citizen passion: A study of issue publics in contemporary America,” Political Behavior, volume 12, number 1, pp. 59–92.

Jon A. Krosnick and Shibley Telhami, 1995. “Public attitudes toward Israel: A study of the attentive and issue publics,” International Studies Quarterly, volume 39, number 4, pp. 535–554.

Nojin Kwak, 1999. “Revisiting the knowledge gap hypothesis: Education, motivation, and media use,” Communication Research, volume 26, number 4, pp. 385–413.

Jae Kook Lee, 2009. “Incidental exposure to news: Limiting fragmentation in the new media environment,” doctoral dissertation in journalism, University of Texas at Austin, at, accessed 12 April 2012.

Arthur Lupia and Mathew D. McCubbins, 1998. The democratic dilemma: Can citizens learn what they need to know? New York: Cambridge University Press.

W. Russell Neuman, 1986. The paradox of mass politics: Knowledge and opinion in the American electorate. Cambridge, Mass.: Harvard University Press.

W. Russell Neuman, 1976. “Patterns of recall among television news viewers,” Public Opinion Quarterly, volume 40, number 1, pp. 115–123.

W. Russell Neuman, Marion R. Just, and Ann N. Crigler, 1992. Common knowledge: News and the construction of political meaning. Chicago: University of Chicago Press.

Yong Jin Park, 2011. “Digital literacy and privacy behavior online,” Communication Research, published online 23 August, at, accessed 7 May 2012.

Yong Jin Park, Scott Campbell, and Nojin Kwak, 2012. “Affect, cognition and reward: Predictors of privacy protection online,” Computers in Human Behavior, volume 28, number 3, pp. 1,019–1,027.

Vincent Price and John Zaller, 1993. “Who gets the news? Alternative measures of news reception and their implications for research,” Public Opinion Quarterly, volume 57, number 2, pp. 133–164.

Markus Prior, 2007. Post–braodcast democracy: How media choice increases inequality in political involvement and polarizes elections. New York: Cambridge Universty Press.

Fei Shen and William P. Eveland, Jr., 2010. “Testing the intramedia interaction hypothesis: The contingent effects of news,” Journal of Communication, volume 60, number 2, pp. 364–387.

Natalie J. Stroud, 2010. “Polarization and partisan selective exposure,” Journal of Communication, volume 60, number 3, pp. 556–576.

Cass Sunstein, 2001. Princeton, N.J.: Princeton University Press.

John Zaller, 1992. The nature and origins of mass opinion. New York: Cambridge University Press.

Cliff Zukin and Robin Snyder, 1984. “Passive learning: When the media environment is the message,” Public Opinion Quarterly, volume 48, number 3, pp. 629–638.


Editorial history

Received 19 December 2011; accepted 3 April 2012.

Creative Commons License
This work is licensed under a Creative Commons Attribution–NoDerivs 3.0 Unported License.

The Internet, selective learning, and the rise of issue specialists
by S. Mo Jang and Yong Jin Park
First Monday, Volume 17, Number 5 - 7 May 2012

A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2020. ISSN 1396-0466.