International Journal of Public Opinion Research Vol. 20 No. 2 ß The Author 2008. Published by Oxford University Press on behalf of The World Association for Public Opinion Research. All rights reserved. doi:10.1093/ijpor/edn019 Advance Access publication 3 May 2008 TESTING THE KNOWLEDGE GAP HYPOTHESIS IN SOUTH KOREA: TRADITIONAL NEWS MEDIA, THE INTERNET, AND POLITICAL LEARNING Sei-Hill Kim ABSTRACT Analyzing data from a survey of South Korean respondents, this study explores the role of news media in informing the audience about politics. In particular, it is examined whether different forms of news media function to increase the gap in political knowledge between socioeconomic classes. Consistent with findings in the United States, newspaper reading was positively related to political learning. Use of political web sites also indicated a small learning effect. Data supported the knowledge gap hypothesis. There was a considerable gap in political knowledge between highly educated and less-educated respondents. More importantly, the gap was even greater among heavy newspaper readers and among political Web users. These findings indicate that newspapers and the Internet may function to increase the gap between social classes. News media play an important role as an information source from which citizens learn about public affairs. This informational function is important in a number of respects, mostly because of the strong link between levels of political knowledge and participation in various civic activities (Rosenstone & Hansen, 1993; Verba, Schlozman, & Brady, 1995). Political learning also allows citizens to make careful evaluations of issues and candidates, promoting more informed political judgment (Kim, Scheufele, & Shanahan, 2005). An unintended effect in transmitting news, however, is an increase in the gap in political knowledge between social classes (Moore, 1987). Knowledge gap research (Tichenor, Donohue, & Olien, 1970) suggests that segments of the population with higher socioeconomic status (SES) acquire media-transmitted information at a faster rate than lower-status segments. The media, therefore, This article was first submitted to IJPOR January 8, 2007. The final version was received on August 30, 2007.
194 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH may function to increase the gap in various forms of knowledge including knowledge about politics (see Viswanath & Finnegan, 1996 for an overview). Using data from a telephone survey of respondents in South Korea, this study analyzes the role of news media in informing citizens about public affairs. In particular, the analysis includes not only such traditional media as newspapers and television, but also the Internet as a new source of political information. Does the Internet function simply as an online version of traditional news media (Margolis & Resnick, 2000)? Or does it offer an opportunity to learn beyond what is already available in other media (Tewksbury, 2003)? The effect of Internet use will be examined in detail, looking at whether the Internet may contribute to political learning above and beyond traditional news media. This study also explores the role of newspapers, television and the Internet in producing a gap in political knowledge between social classes. Researchers (Kwak, 1999; Eveland & Scheufele, 2000; Jerit, Barabas, & Bolsen, 2006) argue that news media may not only increase but also decrease the gap, depending upon the types of news media used (e.g., newspapers versus television news). In particular, the gap-narrowing or gap-widening effect of the Internet has been at the center of controversy among researchers. Does the Internet reduce the gap by making political information more available, accessible, and easier to follow particularly among uneducated lower classes (Anderson, Bikson, Law, & Mitchell, 1995; Gates, 1995; Dyson, 1997)? Or do people in the upper classes surf the Internet more often and use it more effectively, further widening the gap between the information rich and information poor (Negroponte, 1995; DiMaggio & Hargittai, 2001)? This study addresses this issue in detail, looking at whether each form of news media functions to widen or narrow the SES-based gap in political knowledge. While it is likely that the knowledge gap phenomenon can be found among citizens worldwide, the hypothesis has been tested mostly in the United States. South Korea offers an opportunity to examine whether the hypothesis can be applied to another country with a different political and cultural tradition. The country also represents one of the most wired places in the world, thus providing a great opportunity to test the effects of Internet use. As of 2004, for example, about 71 percent of South Korean households subscribed to highspeed Internet services (Borland & Kanellos, 2004). Taken together, this study offers an overall view of media effects on political learning in Korean politics. THE KNOWLEDGE GAP HYPOTHESIS According to the knowledge gap hypothesis (Tichenor et al., 1970), mass media function to expand, rather than narrow, the gap in knowledge between social classes: As the infusion of mass media information into a social system increases, segments of the population with higher socioeconomic status tend to acquire this information
TESTING THE KNOWLEDGE GAP HYPOTHESIS IN SOUTH KOREA 195 at a faster rate than lower status segments, so that the gap in knowledge between these segments tends to increase rather than decrease (pp. 159 160). EXPLANATIONS In their theorization, Tichenor and colleagues offer two separate explanations of how the media may produce an increasing gap. 1 First, there is a difference between social classes in their media use. In general, the upper classes use the media more often for information, gaining greater knowledge (Tichenor et al., 1970). People in the upper classes have a greater stake in what happens in the economy and politics, and thus are more motivated to become well-informed in the first place (Eveland & Scheufele, 2000). Most print media, where much of the available information about public affairs and science appears, are geared toward the tastes of the upper and middle classes because they are the key customers of advertisers (Donohue, Tichenor, & Olien, 1986). Over time, the difference in media use results in an increasing gap in knowledge. News media, therefore, may mediate the well-supported link between one s socioeconomic status and one s knowledge of politics (Scheufele, Shanahan, & Kim, 2002). More simply put, the upper classes are more politically knowledgeable largely because they use news media more often, thus acquiring more information about public affairs. Another explanation of how the media produce an increasing gap is rooted in the difference between classes in their ability to process information. Education provides cognitive skills which allow the upper classes to easily make sense of often complicated information, thus gaining greater knowledge from the media (Eveland & Scheufele, 2000). The highly educated also tend to have more prior knowledge, which helps them to be better prepared to understand the information (Tichenor et al., 1970). Learning from mass media, therefore, is moderated by one s social status education level in particular. When exposed to the same information in the media, persons with higher education gain knowledge more effectively, so that the initial gap between classes increases even further. NEWS MEDIA USE, POLITICAL LEARNING, AND GAPS IN POLITICAL KNOWLEDGE Studies have reported a significant association between news media use and political knowledge (see Chaffee & Frank, 1996 for an overview). Newspapers 1 In their original theorization, Tichenor et al. (1970) introduced five contributing factors (selective exposure, upper-middle-class oriented news reports, social contact, cognitive skills, prior knowledge) that may explain how the media may produce the knowledge gap. Some of them are in fact related to each other, and thus can be combined into two major explanations.
196 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH in particular are known as a major source of information about current issues (Stamm, Johnson, & Martin, 1997; Berkowitz & Pritchard, 1989), election candidates (Kim et al., 2005), and political parties (Chaffee, Zhao, & Leshner, 1994). With regard to television news, researchers argue that viewers are often passive and unmotivated, and thus may not learn as much as the self-selected and motivated newspaper readers (Robinson & Levy, 1986). Challenging this print superiority perspective, however, other researchers have in fact reported significant learning from television news (Garramone, 1983; Zhao & Chaffee, 1995; Eveland & Scheufele, 2000). As research on passive learning suggests (Krugman & Hartley, 1970), even unmotivated exposure to television news seems to produce meaningful learning (Graber, 1990). When it comes to Internet use, evidence is mixed as to whether its learning effects can be equivalent to the use of traditional news media. Certainly, Internet users can obtain essentially unlimited information about politics with relatively little effort, allowing themselves to investigate important issues in much greater depth (Bimber, 2001). As Nisbet and Scheufele (2004) point out, however, availability of information does not always lead to greater use or understanding. DiMaggio and Hargittai (2001) also note that the Internet is a supplementary medium through which traditional news organizations redistribute their information. Given that most users of online political information are heavy users of traditional news media (Althaus & Tewksbury, 2000), it is somewhat questionable how much learning may occur beyond what is learned from newspapers and television news. Nonetheless, a few studies have reported significant learning from the Internet, even after controlling for newspaper reading and television viewing (Norris, 2002; Norris & Sanders, 2003). Is political learning from news media moderated by one s education level? It is well supported in previous research that one s education level moderates how much one can learn from reading a newspaper. In general, highly educated readers tend to gain more knowledge from news articles, increasing the gap between socioeconomic classes (Tichenor et al., 1970; Gaziano, 1984; Kleinnijenhuis, 1991; Viswanath & Finnegan, 1996; Jerit et al., 2006). Where television news is concerned, researchers found that television might function as a knowledge leveler (Neuman, 1976). Because television presents information in less cognitively demanding ways, even those with weaker cognitive skills and less prior knowledge may gain significant information (Neuman, Just, & Crigler, 1992; Prior, 2005). Sophisticated and educated viewers, on the other hand, may find little to learn beyond what they already know because the hard news content of television is so limited and superficial (Eveland & Scheufele, 2000; Jerit et al., 2006). Consequently, television may function to decrease, rather than increase, the gap in knowledge between classes. Studies have provided support for the knowledge-leveling function,
TESTING THE KNOWLEDGE GAP HYPOTHESIS IN SOUTH KOREA 197 showing that the gap between high and low education groups is smaller among heavy viewers of television news than among light viewers (Kwak, 1999; Eveland & Scheufele, 2000; Holbrook, 2002). Although the Internet incorporates video, audio, graphic and text features in a mixed arrangement, the format of its journalistic and partisan information sources is in general closer to text-based newspapers rather than to visual media like television (Norris & Sanders, 2003). Effective online information-seeking requires such skills as purposeful searching, evaluation of source credibility, construction of interpretative frames and a certain level of literacy, which are also prerequisites for reading a newspaper (Bonfadelli, 2002). It is therefore reasonable to expect that people with higher education will gain greater knowledge from the Internet, resulting in a larger knowledge gap between classes. That is, the gap between high and low education groups would be larger among heavy users of political web sites than among light users. NEWS MEDIA USE IN SOUTH KOREA As of 2004, there are 59 national and 79 regional daily newspapers in South Korea (Korea Press Foundation, 2005). Three television networks and two news-only cable channels also offer a variety of news programs. According to recent estimates (Korea National Statistical Office, 2005), about 64 percent of South Koreans over 15 years of age read a newspaper and 77 percent watched television news in 2004. Over the last decade, the number of Internet users in South Korea has increased dramatically. Whereas less than 1 percent of the population used the Internet in 1996, the number has grown to about 70 percent in 2004 (Borland & Kanellos, 2004; Son, 2004). The Internet is widely accepted in Korean politics as well. Major news media (newspapers, magazines, and newscasts) all maintain online versions of their publications and broadcasting. As of 2004, there are also more than 13 online-only newspapers (Korea Press Foundation, 2005). In addition to online news media, most members of the Korean Parliament have their own web sites. This is also true for election candidates, nearly all government agencies, political parties, non-government organizations, and advocacy groups. Still, however, the use of the Internet for political information is relatively low. In 2004, for example, less than one out of five Koreans over 15 years of age read a newspaper online everyday (Korea National Statistical Office, 2005). Research in South Korea has provided support for the idea that news media play a significant role in informing the audience about public affairs (Kim, 1997; Lee, Kim, & Park, 1997). Studies also indicate that the highly educated upper classes in Korea are more likely to use news media than the lower-classes (Whang, 2001; Kim, 2003). Little effort, however, has been made in the country to test the role of news media in producing the gap in political knowledge between socioeconomic classes.
198 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH RESEARCH HYPOTHESES The first step in testing the knowledge gap would be to examine whether there is indeed a gap in knowledge between social classes. The first hypothesis holds that people with higher socioeconomic status typically operationalized by one s education level are more politically knowledgeable than their less-educated counterparts: H1: People with higher education will be more knowledgeable about politics than the less educated. The next two hypotheses examine the mediating role of news media use in producing the knowledge gap (socioeconomic status! news media use! political knowledge). It is hypothesized that people with higher levels of education will more likely use news media, which in turn allow them to gain greater knowledge, producing an increasing gap between social classes. Two conditions must be met. First, there should be a significant difference between classes in their news media use. Second, news media use should be related to increased political knowledge: H2: Highly educated citizens will use news media (newspapers, television news, political web sites) more often than the less-educated. H3: Heavy users of news media will be more knowledgeable about politics than light users. Finally, the next three hypotheses examine whether political learning from news media is moderated by one s education level. It is hypothesized that newspapers and the Internet will function to increase the knowledge gap, whereas television functions as a knowledge-leveler: H4: The knowledge gap between high and low education groups will be larger among heavy users of a newspaper than among light users. H5: The knowledge gap between high and low education groups will be smaller among heavy viewers of television news than among light viewers. H6: The knowledge gap between high and low education groups will be larger among heavy users of political web sites than among light users. METHODS Data for this study came from a telephone survey of the residents in Ansan, South Korea. The city of Ansan, located about 20 miles south-west of Seoul (the capital city of the country), has a population of 685,000. Using systematic sampling, households were randomly selected from the telephone directory. Within each household, interviewers selected one respondent aged 20 years or older for interview, using the last-birthday method. Interviews were conducted by trained graduate students majoring in communication. A total of 378
TESTING THE KNOWLEDGE GAP HYPOTHESIS IN SOUTH KOREA 199 respondents were interviewed between July 10 and August 2, 2003, with a cooperation rate (COOP1) of55 percent. 2 Even though the city of Ansan may well represent a typical midsize city in South Korea, the local nature of the sample and the small sample size make it hard to generalize the findings of the survey beyond their local or regional implications. The findings, therefore, should be interpreted with caution, as it is not the intention of this study to make assertions about the whole country. Traditionally, researchers have categorized political knowledge as either general or domain specific (Zaller, 1992; Delli Carpini & Keeter, 1994). General and chronic knowledge consists of civics-style information, while domain-specific knowledge represents facts about particular issues and policies (Jerit et al., 2006). In this study, a total of eight questions assessed respondents knowledge about politics (see Appendix for question wording). The first four tapped knowledge about current issues in politics. 3 These four items were then combined into a single measure representing issue knowledge (KR-20 ¼.69). The second four questions measured how knowledgeable respondents were about political institutions and processes. These four items were also combined into a single measure of one s civic knowledge (KR-20 ¼.72). Newspaper reading was measured with two items. On a six-point scale (0 ¼ never; 5 ¼ very often), respondents were asked how often they read articles about politics (M ¼ 2.12, SD ¼ 1.65) and economy (M ¼ 2.54, SD ¼ 1.86). These two measures were then combined into a single index ( ¼.85) representing the amount of newspaper reading. Television news viewing was also measured with two questions ( ¼.68) tapping how often (0 ¼ never; 5 ¼ very often) respondents viewed daily newscasts (M ¼ 3.75, SD ¼ 1.32) and investigative news programs (M ¼ 2.65, SD ¼ 1.33). Internet use was an additive index ( ¼.84) of three measures asking respondents how often (0 ¼ never; 5 ¼ very often) they used: online news sites (M ¼ 1.78, SD ¼ 1.41); the Internet to learn about the candidates during the last presidential election in 2002 (M ¼ 1.71, SD ¼ 1.36); Internet bulletin boards to exchange political views with other people (M ¼ 1.38, SD ¼.95). 4 Education was measured by asking respondents the highest degree completed (47.5 percent some college or more). Using the mean score, the sample was divided into high (some college or more, N ¼ 174) and low (high school diploma or less, N ¼ 199) education groups. For control purposes, the survey 2 Calculation of the cooperation rate followed the standard definitions put forth by the American Association for Public Opinion Research (2004). 3 These four issues were selected from an extensive review of news articles and newscasts and also from personal interviews with college professors in South Korea. Because knowledge about each issue was to be asked in a survey, it was necessary to select issues well known to ordinary Koreans. The issues selected were all controversial and prominent in the media during the last few months before the beginning of the opinion survey (July 10, 2003). 4 These low mean scores on Internet use measures are attributed in part to the large number of respondents who did not use the Internet at all (N ¼ 135 or 36 percent).
200 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH TABLE 1 Difference between high and low education groups in political knowledge and news media use Education Low (N ¼ 199) High (N ¼ 174) t Issue knowledge 2.04 (1.20) 2.53 (1.16) 3.96 Civic knowledge 2.15 (1.21) 2.52 (1.15) 3.08 Newspaper reading 3.91 (3.36) 5.55 (2.93) 5.07 Television news viewing 6.68 (2.21) 6.49 (2.21) 0.83 Internet use for political information 2.43 (3.11) 4.43 (3.12) 6.22 Note: Entries are mean scores with standard deviations in parentheses. Low education: High school diploma or less; High eucation: Some college or more. p <.001; p <.01; p <.05. also measured age (M ¼ 39 years, SD ¼ 12.26) and gender (60.6 percent female). This demographic information indicates that females and college graduates, compared with the national data, are somewhat over-represented in the sample. According to the 2005 Census (Korea National Statistical Office, 2005), females accounted for about 51.0 percent of the whole population. The same census shows that about four out of ten Koreans 20 or older (40.2 percent) completed some college or more. FINDINGS Testing the first hypothesis (H1) examines whether there is indeed a gap in political knowledge between highly educated and less-educated respondents. Table 1 shows that people with higher education are more knowledgeable than the less-educated about both current issues (Issue knowledge) and political processes (Civic knowledge). T-tests indicate that these differences are all statistically significant. Also examined is the question of whether the highly educated are more likely to tune into political news. Supporting H2, Table 1 shows that people with higher education read a newspaper more often than the less educated. The same table indicates that the highly educated use political web sites more often as well. There was no significant difference, however, in viewing television news. Political learning from news media (H3) and its interaction with education level (H4, H5, H6) were tested using a linear model, which included education, news media use and their interactions in predicting one s political knowledge (Table 2). In order to produce interaction terms, news media variables were dichotomized first, using mean scores, and then entered into the model. Respondents age and gender were also entered into the model as control variables.
TESTING THE KNOWLEDGE GAP HYPOTHESIS IN SOUTH KOREA 201 TABLE 2 Analysis of variance of political knowledge Issue knowledge Civic knowledge df Mean squares F df Mean squares F Age 1 0.01 0.00 1 0.57 0.43 Gender (female) 1 3.03 2.31 1 12.06 9.10 Education 1 7.56 5.77 1 1.56 1.17 Newspaper reading 1 17.66 13.48 1 12.77 9.63 Television news viewing 1 1.19 0.91 1 0.41 0.31 Internet use for political 1 5.11 3.90 1 2.17 1.64 information Newspaper 1 12.66 9.66 1 4.14 3.12 # reading Education Television news 1 0.31 0.23 1 0.52 0.40 viewing Education Internet use for political 1 5.65 4.31 1 1.81 1.36 information Education Within 363 1.31 363 1.33 Total N 372 372 p <.001; p <.01; p <.05; # p <.10. Supporting H3, newspaper reading was significantly associated with political learning (Table 2). There was a significant difference between heavy and light readers of newspapers (see also Figure 1), with heavy readers demonstrating greater issue and civic knowledge. Television news, however, did not show such a significant relationship to either current issue or civic knowledge. The Internet had a small but statistically significant relationship to issue knowledge. Its relationship to civic knowledge, however, was not statistically significant. H4, H5, and H6 examine if learning from news media (newspapers, television news, political web sites) is moderated by one s education level. Figure 1 graphs the interaction effects of news media use and education on two forms of political knowledge. As shown in the figure, newspaper reading indicated a considerable interaction with education. When exposed to the same information on newspapers, people with higher education tend to learn a lot more, widening the gap between classes. Table 2 shows that the interaction was statistically significant for issue knowledge, but only marginally significant for civic knowledge. Supporting H4, these findings indicate that the knowledge gap is larger among heavy readers of newspapers. Such a significant interaction effect was not found for television news viewing. As shown in Figure 1, people with higher education were not very different from the less educated in how much they learned from television news. The interaction was not statistically significant for either issue or civic knowledge (Table 2). H5, therefore, was not supported.
202 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH FIGURE 1 Interaction effects of education and news media use on political knowledge There was an interaction between Internet use and education in gaining current issue knowledge. As shown in Figure 1, the gap in issue knowledge between education groups was larger among heavy users of political web sites than among light users. Table 2 shows that this interaction was statistically significant. Supporting H6, this finding may indicate that political learning
TESTING THE KNOWLEDGE GAP HYPOTHESIS IN SOUTH KOREA 203 from the Internet is moderated by one s education level. That is, people with higher education tend to learn more from the Internet, increasing the knowledge gap between classes. When it came to civic knowledge, however, the Internet did not show such a significant interaction with education (Figure 1 and Table 2). DISCUSSION Analyzing data from a survey of South Korean respondents, this study explored the role of news media in informing the audience about politics. In particular, it was examined whether different forms of news media functioned to increase or decrease the gap in political knowledge between socioeconomic classes. Newspaper reading showed the most significant relationship with political learning, supporting the idea that newspapers function as the primary source of political information (Chaffee & Frank, 1996). Findings also support the idea that newspapers may increase the gap between social classes in regard to political knowledge. First, people with higher education read a newspaper more often than the less educated, thus learning more about public affairs. The initial gap between classes may therefore increase over time. Second, people s learning from a newspaper is moderated by their education level, so that the learning is greater among the highly educated. When exposed to the same information, the highly educated upper classes learn more effectively than the less educated, further increasing the gap between classes. These findings demonstrate that the well-known SES gap in political knowledge (Verba & Nie, 1972; Delli Carpini & Keeter, 1994) is produced, in large part, by the difference in newspaper reading between classes. That is, the upper classes are more knowledgeable largely because they read newspapers more often. Among those who do not read a newspaper often, as shown in Figure 1, there was in fact no difference between classes in either issue or civic knowledge. In other words, unless they read a newspaper often, the highly educated were not very different from the less educated in their knowledge of politics. This finding supports the idea that newspaper reading may mediate the link between one s social class and one s knowledge of public affairs. As researchers point out, the effect of socioeconomic status on political knowledge seems to depend on a host of communication variables, such as newspaper reading or use of news media in general (Verba et al., 1995; Nie, Junn, & Stehlik-Barry, 1996; Scheufele et al., 2002). Where television is concerned, news viewing was largely unrelated to political knowledge. There are a few explanations. First, the information on television is, in general, so limited and superficial that viewers may not be able to learn anything substantial (Eveland & Scheufele, 2000). According to Kim,
204 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH Han and Scheufele (2006), primetime daily news programs in South Korea report about 30 40 news items in their 1-hour newscasts, giving each item less than 2 min to cover. It is therefore quite unlikely that either the amount or the nature of political learning would be very substantial. Second, as Graber (1994) points out, the episodic nature of news coverage may also inhibit learning. Whereas episodic reports provide good pictures to show and interesting stories to tell, they are generally devoid of interpretive analyses and detailed information (Iyengar, 1991). Is broadcast news largely episodic in South Korea, too? Unfortunately, this question cannot be answered without an extensive content analysis, which goes beyond the scope of this study. Nonetheless, television news is by nature a story teller, focusing on narrative rather than substantive information (Wallack, Dorfman, Jernigan, & Themba, 1993). There is little reason to expect it is different in South Korea. No evidence was found to support either the gap-widening or gapreducing effect of television news viewing. First, highly educated respondents were not different from the less educated in the amount of news viewing. Second, there was no significant interaction between news viewing and one s education level. It is therefore hard to conclude that either the highly educated or the less educated learn more than the other from watching television news. Most importantly, political learning from television was hardly noticeable across the board. The Internet indicated a small relationship to political learning even after controlling for demographics, newspaper reading, and television news viewing. As Althaus and Tewksbury (2000) argue, the Internet may not replace traditional news media, but supplement them by providing more detailed information. That is, people may find an issue interesting or important from reading a newspaper or watching television, and then go to the Internet for further information. Use of the Internet for political information, therefore, may be a highly motivated and self-selected behavior. It is hard to argue that the Internet offers simply an online version of conventional news media. Surfing the web pages of advocacy organizations and candidates, or reading the articles on bulletin boards and political blogs, Internet users seem to learn beyond what they learn from other traditional media. Findings also indicate that Internet use may increase, rather than decrease, the gap in political knowledge between social classes. First, there was a significant difference in usage between classes, with the highly educated using political web sites considerably more often. Those in the lower classes may have limited access to a computer or an Internet service. 5 It is also likely that they simply have little interest in politics, and thus hardly use the Internet for 5 In 2002, about 83 percent of Korean households whose head was a college graduate had a computer at home. The number dropped to 67 percent when the head of the household had only a high school diploma (Korea National Statistical Office, 2005).
TESTING THE KNOWLEDGE GAP HYPOTHESIS IN SOUTH KOREA 205 political information (Bonfadelli, 2002). Second, highly educated Internet users tend to learn significantly more than their less-educated counterparts, further increasing the gap between classes. It seems that informational use of the Internet requires cognitive skills and literacy that may impede the less educated from learning a great deal. These findings support the pessimistic view that the Internet will reproduce existing patterns of political communication, further widening the gap between social classes (DiMaggio & Hargittai, 2001; Nisbet & Scheufele, 2004). As Norris (2001) points out, the Internet may function to reinforce inequalities of power and knowledge, producing deeper gaps between the information rich and poor, and between the activists and the disengaged. Overall, media effects were greater in informing citizens of current issues than in providing them with civic knowledge. Newspaper reading and Internet use, for example, were both more closely associated with issue knowledge. Significant interactions with education were also found only from issue knowledge. In fact, these findings are not very surprising given that most people learn civics-style facts in a textbook, whereas information about current issues is found mostly in mass media (Jerit et al., 2006). Before further discussing the findings, a few limitations of this study should be noted. First, despite the fast increase of Internet users in South Korea, there were still a considerable number of respondents (36 percent) who did not use the new technology at all. Use of the Internet for political information was particularly low, indicating that only a small number of people were using political web sites on a regular basis. Future research may likely find more significant or somewhat different results as the Internet becomes more available particularly for those in the lower classes. As Prior (2005) points out, however, it may not be the question of availability, but the lack of interest and cognitive skills that discourages citizens from using political web sites. Even as the Internet becomes more available, it is still questionable how willing or able the lower classes would be to use the new technology to acquire political information. News media variables in this study were all simple exposure measures, which may not be an optimal way to capture media effects. A more appropriate measure would be one that includes attention, in addition to exposure, to specific news content (Chaffee & Schleuder, 1986). The lack of attention measures is particularly a problem in this study, given that Internet use is by nature highly attention-driven behavior. It is somewhat inappropriate to draw a conclusion regarding differences between traditional media and the Internet unless attention to newspapers and television is measured and its effects are compared with the consequences of Internet use. Although a few studies (Zukin & Snyder, 1984) report that exposure alone can lead to political
206 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH learning, the news media variables in this study may be a crude and incomplete measurement of media use. Another shortcoming of this study is its cross-sectional data, which have limitations in making causal inferences. Uncontrolled individual variations in political knowledge and news media use may produce spurious relationships (Zhao & Chaffee, 1995). Also, given that the direction of causality is often unclear in cross-sectional research, findings in this study do not establish that news media use always precedes political learning. There is no way to rule out a reverse-causation explanation that knowledgeable citizens, rooted in other causes, use news media more often. With these and other shortcomings in mind, this study offers an overall analysis of media effects on political learning. Findings indicate that South Korean news media, particularly newspapers and political web sites, play an important role as a source of political information. Findings also provide support for the knowledge gap hypothesis. There is a significant gap in political knowledge between highly- and less-educated respondents. More importantly, the gap is even greater among heavy newspaper readers and among political Web users. These findings indicate that newspapers and the Internet may function to increase the gap between classes. Taken together, this study offers a compelling empirical demonstration that knowledge gaps may follow similar forms in South Korea as they do in the United States. This inequality of knowledge may lead to under-representation of the lower classes in many policy decisions, which in turn reproduces the same socioeconomic structure (Delli Carpini & Keeter, 1994). As McLeod and Perse (1994) note, knowledge and information translate into social power. Lack of knowledge may result in exclusion from social resources, thus leading to a lack of social power. News media, therefore, may function to enhance the information advantage of the upper classes by further increasing the knowledge gap that already exists between classes (Moore, 1987). APPENDIX: QUESTION WORDING OF POLITICAL KNOWLEDGE MEASURES ISSUE KNOWLEDGE Recently, a government project to build dikes on the west coast of the country has become controversial because of its potential environmental impacts. Can you tell me what the name of this project is? Is it the Saemangum or Shiwha project?
TESTING THE KNOWLEDGE GAP HYPOTHESIS IN SOUTH KOREA 207 The Supreme Prosecutors Office is investigating Mr Park Ji Won in connection with the allegation that he and former President Kim Dae Jung secretively forwarded hundreds of millions U.S. dollars to North Korea in return for the historic summit in 2000 with the North Korean leader Kim Jong-Il. Can you tell me who Mr Park was? Was he the head of Korean CIA or the presidential chief of staff? Which political party is currently in an escalating dispute over how to transform itself into a new coalition party? Is it the Grand National Party or the Democratic Party? The government proposed a plan to create a large computer data base of all the students in the nation (National Education Information System, NEIS), which has become very controversial because of its potential privacy problems. What is the official position of the National Teachers Union? Do they support or oppose the plan? CIVIC KNOWLEDGE Can you tell me how the head of each Province is selected? Is the head appointed by the President or elected by people? According to the Constitution, how long is the term of an elected congressman? Is it four years or five years? The system of checks and balances means the three tiers of government work independently of each other to balance decisions and to ensure that none operates beyond the reaches of its power. Two of these three tiers are the executive and the legislative branches. Can you tell me what the third is? Is it the military or the judiciary? According to the Constitution, who is Commander-in-Chief of the Korean military? Is it the Secretary of Defense or the President? REFERENCES Althaus, S. L., & Tewksbury, D. (2000). Patterns of internet and traditional news media use in a networked community. Political Communication, 17, 21 45. American Association for Public Opinion Research (2004). Standard definitions: Final dispositions of case codes and outcome rates for surveys (3rd edition). Lenexa, KS: AAPOR. Anderson, R. H., Bikson, T. K., Law, S. A., & Mitchell, B. M. (1995). Universal access to email feasibility and societal implications. Santa Monica, CA: Rand. Berkowitz, D., & Pritchard, D. (1989). Political knowledge and communication resources. Journalism Quarterly, 66, 697 701. Bimber, B. (2001). Information and political engagement in America: Search for the effects of information technology at the individual level. Political Research Quarterly, 54, 53 67.
208 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH Bonfadelli, H. (2002). The Internet and knowledge gaps: A theoretical and empirical investigation. European Journal of Communication, 17, 65 84. Borland, J., & Kanellos, M. (2004, July 28). South Korea leads the way. News.com. Retrieved December 8, 2005, from http://news.com.com/southþkoreaþleadsþ theþway/2009-1034_3-5261393.html?tag¼nl Chaffee, S. H., & Frank, S. (1996). How Americans get political information: Print versus broadcast news. The Annals of the American Academy of Political and Social Science, 546, 48 58. Chaffee, S. H., & Schleuder, J. (1986). Measurement and effects of attention to media news. Human Communication Research, 13, 76 107. Chaffee, S. H., Zhao, X., & Leshner, G. (1994). Political knowledge and the campaign media of 1992. Communication Research, 21, 305 324. Delli Carpini, M. X., & Keeter, S. (1994). The public s knowledge of politics. In J. D. Kennamer (Ed.), Public opinion, the press, and public policy (pp. 19 40). Westport, CT: Praeger. DiMaggio, P., & Hargittai, E. (2001). From the digital divide to digital inequality : Studying the internet use as penetration increases. Working paper, Centre for Arts, Cultural and Political Studies, Princeton University. Donohue, G. A., Tichenor, P. J., & Olien, C. N. (1986). Metro daily pullback and knowledge gaps: Within and between communities. Communication Research, 13, 453 471. Dyson, E. (1997). Release 2.0: Design for living in the digital age. New York: Broadway Books. Eveland, W. P. Jr, & Scheufele, D. A. (2000). Connecting news media use with gaps in knowledge and participation. Political Communication, 17, 215 237. Garramone, G. M. (1983). TV news and adolescent political socialization. In B. Bostrom (Ed.), Communication Yearbook 7 (pp. 651 669). Beverly Hills, CA: Sage. Gates, B. (1995). The road ahead. New York: Viking. Gaziano, C. (1984). Neighborhood newspapers, citizen groups and public affair knowledge gaps. Journalism Quarterly, 16, 556 566, 599. Graber, D. A. (1990). Seeing is remembering: How visuals contribute to learning from television news. Journal of Communication, 40(3), 134 155. Graber, D. A. (1994). Why voters fail information tests: Can the hurdles be overcome? Political Communication, 11, 331 346. Holbrook, T. M. (2002). Presidential campaigns and the knowledge gap. Political Communication, 19, 437 454. Iyengar, S. (1991). Is anyone responsible? How television frames political issues. Chicago: University of Chicago Press. Jerit, J., Barabas, J., & Bolsen, T. (2006). Citizens, knowledge, and the information environment. American Journal of Political Science, 50, 266 282. Kim, J.-K. (1997). An analysis on the uses and effects of election-related political news on television by the college student voter in the 1996 election for the member of the national assembly. Korean Journal of Journalism Study, 40, 5 48. Kim, H. (2003). An evaluation of the media and Internet campaigns. Paper presented to the 2002 Election Conference of the Korean Political Science Association, Seoul, Korea.
TESTING THE KNOWLEDGE GAP HYPOTHESIS IN SOUTH KOREA 209 Kim, S.-H., Han, M., & Scheufele, D. A. (May, 2006). Think about him this way: Priming function of television news and the public s evaluation of the president in South Korea. Paper presented to the 61th annual conference of the American Association for Public Opinion Research (AAPOR), Montreal, Canada. Kim, S.-H., Scheufele, D. A., & Shanahan, J. (2005). Who cares about issues? Issue voting and the role of news media during the 2000 U.S. presidential election. Journal of Communication, 55, 103 121. Kleinnijenhuis, J. (1991). Newspaper complexity and the knowledge gap. European Journal of Communication, 6, 499 522. Korea National Statistical Office (2005). Korea statistical yearbook 2005. Retrieved May 11, 2006, from http://www.nso.go.kr/e_book/yearbook05.htm Korea Press Foundation (2005). Readers are leaders. Retrieved November 11, 2005, from www.presskorea.or.kr/data/statistics01.asp Krugman, H. E., & Hartley, E. L. (1970). Passive learning from television. Public Opinion Quarterly, 34, 184 190. Kwak, N. (1999). Revisiting the knowledge gap hypothesis: Education, motivation, and media use. Communication Research, 26, 385 413. Lee, K., Kim, S., & Park, S. (1997). Economic news and the audience knowledge about the economy. Unpublished manuscript. Retrieved August 22, 2005, from www.ssmedianet.org/config/download.php?dir¼pds_data&filename¼ 961102_2.doc Margolis, M., & Resnick, D. (2000). Politics as usual: The cyberspace revolution. Thousand Oaks, CA: Sage. McLeod, D. M., & Perse, E. M. (1994). Direct and indirect effects of socioeconomic status on public affairs knowledge. Journalism Quarterly, 71, 433 442. Moore, D. W. (1987). Political campaigns and the knowledge-gap hypothesis. Public Opinion Quarterly, 51, 186 200. Negroponte, N. (1995). Being digital. New York: Knopf. Neuman, R. (1976). Patterns of recall among television news viewers. Public Opinion Quarterly, 40, 115 123. Neuman, R., Just, M. R., & Crigler, A. N. (1992). Common knowledge: News and the construction of political meaning. Chicago, IL: University of Chicago Press. Nie, N. H., Junn, J., & Stehlik-Barry, K. (1996). Education and democratic citizenship in America. Chicago, IL: University of Chicago Press. Nisbet, M. C., & Scheufele, D. A. (2004). Political talk as a catalyst for online citizenship. Journalism & Mass Communication Quarterly, 81, 877 896. Norris, P. (2001). A digital divide: Civic engagement, information poverty, and the internet in democratic societies. New York: Cambridge University Press. Norris, P. (2002). Tuned out voters? Media impact on campaign learning. Ethical Perspectives, 9, 200 221. Norris, P., & Sanders, D. (2003). Message or medium? Campaign learning during the 2001 British General Election. Political Communication, 20, 233 262. Prior, M. (2005). News vs. entertainment: How increasing media choice widens gaps in political knowledge and turnout. American Journal of Political Science, 49, 577 592.
210 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH Robinson, J. R., & Levy, M. (1986). The main source: Learning from television news. Beverly Hills, CA: Sage. Rosenstone, S. J., & Hansen, J. M. (1993). Mobilization, participation, and democracy in America. New York: Macmillan. Scheufele, D. A., Shanahan, J., & Kim, S.-H. (2002). Who cares about local issues? Media influences on local political involvement, issue awareness, and attitude strength. Journalism and Mass Communication Quarterly, 79, 427 444. Son, Y. (2004, September 7). Korea committed to bridging digital divide. The Korean Times. Retrieved December 8, 2005, from http://times.hankooki.com/lpage/ special/200409/kt2004090718035911440.htm Stamm, K., Johnson, M., & Martin, B. (1997). Differences among newspapers, television, and radio in their contribution to knowledge of the Contract with America. Journalism and Mass Communication Quarterly, 74, 687 702. Tewksbury, D. (2003). What do Americans really want to know? Tracking the behavior of news readers on the Internet. Journal of Communication, 53, 694 710. Tichenor, P. J., Donohue, G. A., & Olien, C. N. (1970). Mass media flow and differential growth in knowledge. Public Opinion Quarterly, 34, 159 170. Verba, S., & Nie, N. H. (1972). Participation in America: Political democracy and social equality. New York: Harper & Row. Verba, S., Schlozman, K. L., & Brady, H. E. (1995). Voice and equality: Civic voluntarism in American politics. Cambridge, MA: Harvard University Press. Viswanath, K., & Finnegan, J.R. Jr. (1996). The knowledge gap hypothesis: Twentyfive years later. In B. R. Burleson (Ed.), Communication yearbook 19 (pp. 187 227). Thousand Oaks, CA: Sage. Wallack, L., Dorfman, L., Jernigan, D., & Themba, M. (1993). Media advocacy and public health. Newbury Park, CA: Sage. Whang, Y.-S. (2001). A study of Internet use and political participation: Uses of political sites during the 2000 congressional election. Korean Journal of Journalism Study, 45, 421 456. Zaller, J. (1992). The nature and origins of mass opinion. Cambridge: Cambridge University Press. Zhao, X., & Chaffee, S. H. (1995). Campaign advertisements versus television as sources of political issue information. Public Opinion Quarterly, 59, 41 65. Zukin, C., & Snyder, R. (1984). Passive learning: When the media environment is the message. Public Opinion Quarterly, 48, 629 638. BIOGRAPHICAL NOTE Sei-Hill Kim (Ph.D., Cornell University, 2001) is an assistant professor in the Department of Communication & Journalism at Auburn University. His primary research interests are in political communication and media effects. Address correspondence to Sei-Hill Kim, Department of Communication, 220 Tichenor Hall, Auburn University, AL 36849, USA, E-mail: kimseih@auburn.edu