Attribute agenda setting,priming and the media’s influence on how to think about a controversial issue

Abstract

Using a controversial issue in South Korea, a government plan to relocate the administrative capital, this study examines attribute agenda setting. Linking survey data to an analysis of news coverage, more specifically, the article explores how television news has influenced the way the public evaluates the plan and its various attributes, i.e. potential benefits and problems of the relocation. Findings support the idea that the media could increase public salience of certain attributes by placing them more prominently in news coverage. An important outcome of attribute agenda setting was its priming effect. Issue attributes emphasized in the media were functioning as important standards by which the audience evaluated the plan. These findings, taken together, enhance the intercultural validity of attribute agenda setting by demonstrating that the theory can be applied to South Korea, a country with a unique cultural tradition.

Keywords

attribute agenda setting media effects priming second-level agenda setting South Korea

News media may not exert a direct and instant influence on public opinion. Nonetheless, news coverage of politics has been shown to have a wide range of subtle, but still powerful, effects on what the public thinks about important issues (see Ansolabehere et al. [1993] for an overview). The agenda setting theory (McCombs and Shaw, 1972), for example, states that the media have the power to make certain issues more salient by placing them more prominently in news coverage or by devoting a large proportion of the news hole to the issues. As Cohen (1963: 13) suggested, the media ‘may not be successful in telling people what to think’, but there is much evidence suggesting that they are ‘stunningly successful in telling people what to think about’.

Priming is an impact that agenda setting can have on the way people evaluate public officials. According to Iyengar and Kinder (1987), priming refers to the process in which news media attend to certain issues and not others, and thereby shape ‘the standards by which governments, presidents, policies, and candidates for public office are judged’ (1987: 63). A prominent issue in the media becomes salient in people’s minds, and thus functions as an important dimension of evaluation when the public makes judgments about political figures. Although agenda setting is largely an effect limited to cognitive process, its outcomes may go beyond being cognitive (Kiousis, 2005). The media can influence the attitudes of the audience by telling them what is important to consider when making up their minds.

A recent theoretical development in agenda setting research captures well the link between agenda setting and the public’s attitudes toward political issues. Attribute agenda setting, or second-level agenda setting, deals with issue attributes, e.g. the reasons why people support or oppose a given issue. It is hypothesized that certain attributes made prominent by news coverage will also become salient in people’s minds (Golan and Wanta, 2001; McCombs, 1997). Also importantly, attribute agenda setting may produce a priming effect. As the media make certain attributes of an issue more prominent, according to attribute priming, the audience gives more weight to the same attributes when deciding whether or not to support the issue (Kim et al., 2002). The media, therefore, tell people not only which issue to think about but also how to think about it, an effect that may have important implications in shaping public opinion (McCombs and Estrada, 1997).¹

Using a controversial topic in South Korea, i.e. a proposed government plan to relocate the administrative capital, this study explores how television news can influence the way the public evaluates the issue. We content-analyze news coverage, examining which attributes of the issue were covered more prominently than others. Using an opinion survey, we then assess how important these attributes are when people evaluate the relocation plan. Findings from the content analysis are linked to the opinion survey in order to test potential attribute agenda setting and priming effects.

While attribute agenda setting has been tested in a number of countries outside the US, including Taiwan (King, 1997), Spain (Lopez-Escobar et al., 1998; McCombs et al., 1997), Japan (Takeshita and Mikami, 1997), the Netherlands (Meijer and Kleinnijenhuis, 2006), England (Dursun, 2005) and Israel (First, 1997; Sheafer, 2007), our study is one of the first attempts to explore attribute agenda setting and priming processes in South Korea. South Korea provides an opportunity to examine whether the theory can be applied to another country with a unique cultural tradition. It is our hope that the findings of this study will enhance the intercultural validity of the theory.

Attribute agenda setting, priming and dimensions of issue evaluation

Traditional agenda setting deals with the transfer of issue salience from the media to the public. Attribute agenda setting, an extension of agenda setting, is concerned with the salience of issue attributes. Widely conceived, these attributes include the properties, characteristics, or aspects of an object that can be employed to evaluate or think about the same object (McCombs and Evatt, 1995). It is hypothesized that the media’s emphasis on specific attributes for thinking about a given issue plays a powerful agenda setting function. Certain attributes emphasized in the media become salient in people’s minds, and thus function as the major attributes in public conceptions of an issue.

Priming research has consistently linked the salience of various issues to presidential evaluations (e.g. Iyengar and Kinder, 1987; Willnat and Zhu, 2001). As Price and Tewksbury (1997) rightly point out, there is no reason why other kinds of political evaluations might not also be subject to the priming effect. While traditional priming research has focused on the public’s evaluation of the president, attribute priming deals with the evaluation of social issues, either political or non-political. In particular, the theory addresses the media’s influence on the way the issues are judged. Certain attributes emphasized in news coverage become salient in people’s minds (attribute agenda setting), and thereby function as important dimensions of issue evaluation among the public (attribute priming).

In a sense that it deals with the question of which attributes or aspects of an issue are prominently placed, attribute agenda setting may sound similar to framing, the idea that the media, by highlighting certain aspects of an issue, can foster a particular ‘problem definition, causal interpretation, moral evaluation and/or treatment recommendation for the item described’ (Entman, 1993: 52). Some agenda setting researchers have considered framing as a natural extension of agenda setting, arguing that framing, as a term, should be replaced by attribute agenda setting (e.g. McCombs, 1997). Instead of calling for the convergence, more recently, other researchers (e.g. Kim et al., 2002; Scheufele and Tewksbury, 2007) have argued that attribute agenda setting and framing describe separate phenomena.

On a number of fronts, we maintain that agenda setting and priming are based on a set of theoretical premises that is different from framing. First, as Kim et al. (2002) point out, framing is still a fuzzy term; and there are multiple theoretical and operational definitions of framing in the literature. Some of them are somewhat similar to attribute agenda setting (e.g. Entman’s [1993] definition of framing as a process of selecting certain aspects of reality and making them more salient), while others appear to be quite different (e.g. Iyengar’s [1991] episodic and thematic frames; Gamson and Modigliani’s [1989] framing as an interpretive package or a story line).

Second, it is necessary to make a conceptual distinction between applicability-based models, such as framing, and accessibility-based models, such as agenda setting and priming (Price and Tewksbury, 1997). The way an issue is framed or described in the media provides the audience with important interpretive cues. If applicable to the pre-existing cognitive schemata of the audience, these cues will be used in subsequent processes of making sense of the issue (Scheufele, 2000). The framing effect is largely a function of how applicable the frame is to the existing schemata in receivers’ belief systems.

Both agenda setting and priming, on the other hand, rely on a memory-based model, which assumes that only the most salient issues in a person’s memory will become accessible, and thus likely to influence his or her decision making (Price and Tewksbury, 1997; Zaller and Feldman, 1992). Through agenda setting, the media influence ‘how much’ or ‘how recently’ a person is exposed to a particular issue, which in turn may determine how easily the issue will be retrieved from memory. By covering a certain issue more prominently or frequently, the media can increase the accessibility of the issue in people’s memory.

In this way, agenda setting is largely an argument limited to the question of how frequently an issue, or an issue attribute, is covered in the media (Scheufele, 2000). The root of the argument is quantitative, suggesting that greater frequency of coverage will make an issue or an attribute more likely to be used by the audience. Attribute salience is operationalized by counting how often a particular attribute is mentioned in the media (e.g. Golan and Wanta, 2001; King, 1997). Simply mentioning a particular issue attribute does not necessarily mean that the media framed the issue in terms of that attribute. A distinction should be made between the central theme or the frame of a story and the different aspects or attributes of the issue in the story. Operationally, the distinction is apparent in content analysis: identifying the frame defining the key theme of each news story vs keeping a tally of various attributes appearing in the story (McCombs, 2005).

Finally, and most importantly, the difference between framing and agenda setting matters mostly with respect to their assumed effects. In the framing model, the media have an effect not through the salience of an issue but through ‘the terminological or semantic differences in how an issue is described’ (Kim et al., 2002: 10). Different descriptions or framing of the same issue will be interpreted differently by the audience, producing different meanings, interpretations, or definitions. In the agenda setting model, the assumed effects are the salience attached to an issue; it is the salience of the issue itself, rather than the way the issue is described, that evokes audience responses in the subsequent judgments and decision making processes.

Research hypotheses

This study explores the processes of attribute agenda setting and priming during a national debate over a controversial government proposal in South Korea. We first analyze how the proposal and its potential benefits and problems (attributes) are covered in television news. An opinion survey then assesses which attributes of the issue are more salient than others in people’s minds. The first hypothesis (H1) examines if there is a significant correspondence between salient issue attributes in the media and the agenda of attributes among the public. We then link the salience of attributes to citizens’ overall evaluation of the proposal in order to test a potential priming effect (H2). We put forth the following hypotheses:

H1: Salient issue attributes in the media will be associated with the agenda of issue attributes among the audience (attribute agenda setting).

H2: Salient issue attributes in the media will be associated with important standards by which the audience evaluates the issue (attribute priming).

The issue of relocating the administrative capital

In January 2004, President Roh announced the New Capital Region Development Plan, which would in essence move the national capital from Seoul to another location. According to the plan, government and administrative functions would be relocated to a new city even though any sizable moving was not expected to happen until 2012 (BBC News, 2004). Right after the announcement, the plan was confronted with strong opposition, igniting nationwide controversy. In October 2004, the Constitutional Court ruled that the government’s plan was unconstitutional since the relocation was an important national matter requiring national referendum or revision of the constitution (Wikipedia, 2006). As a way to evade violating the court ruling, the Roh administration later came up with a revised and scaled-down plan, which would move almost all national government branches, except the Executive Branch, to a new administrative city instead of creating a new administrative capital (Wikipedia, 2006). In March 2005, lawmakers finally passed a new bill that allowed for relocating the country’s 12 economic and social ministries, while keeping the presidential office, the National Assembly, security and foreign affairs branches in Seoul (The Korea Times, 2005).

Most supporters and opponents based their opinions on five major attributes of the issue, all focusing on the potential consequences of the relocation. First, supporters claimed that relocating the core administrative functions would greatly enhance balanced development across the country. They argued that the plan would reduce Seoul’s overwhelming economic and political dominance over the rest of the country (Balanced development across the nation). Supporters also referred to many metropolitan problems, such as traffic congestion, pollution, overcrowding and urban sprawl. They believed that relocating the capital would be a good solution to these problems, significantly improving the quality of life in the Seoul area (Solution to metropolitan problems). Finally, it was mentioned that the relocation would greatly stimulate the local economy. Historically, the Chungnam province, where the new administrative city will be located, has always lagged behind other regions in terms of economic growth (Stimulating local economy).

Meanwhile, many opponents of the plan were concerned about the tremendous cost required to relocate the capital (Tremendous relocation cost). According to a government estimate (Kim T, 2004), the relocation would cost nearly 45 trillion Korean won (about US$45 billion). Opponents also pointed out that relocating the capital would necessarily lead to deregulation of the Seoul area, which would in turn cause many problems such as environmental degradation caused by urban renewal programs. It was argued that the relocation would phase out important policies and restrictions in the area necessary to maintain the quality of life (Deregulation of the Seoul area and resulting problems).

Methods

Content analysis

Television news data for this study came from primetime news programs of the three major television networks in South Korea: KBS News 9, MBC News Desk and SBS News 8. Each program is a one-hour newscast, which consists of about 30–40 news segments every night. Depending on its importance, a news item is covered in one or more segments. All news segments for our analysis were obtained from the Internet homepage of each program, which provided archives of transcripts and video clips made available for public use. These transcripts and clips, updated on a daily basis, cover the entire one-hour program aired each day.

In determining the time frame of analysis, we considered Stone and McCombs’s (1981) suggestion that at least two to six months might be necessary for multiple topics to move from the media to the public. To be safe, we first chose a period within the last 12 months before the beginning of our opinion survey (30 April 2005).² For economy of analysis, we then selected four specific months (June, July, October 2004 and March 2005), during which there were important new developments on the issue, producing a large amount of news coverage.³

To examine the salience of each issue attribute (Balanced development across the nation, Solution to metropolitan problems, Stimulating local economy, Deregulation of the Seoul area and resulting problems, Tremendous relocation cost), we manually scanned the entire news transcripts aired during the four months selected. A total of 456 news segments about the relocation plan were identified, and two coders analyzed them after having conducted a series of training and pilot-test sessions where they discussed and agreed upon what might and might not be considered a mention of each attribute. We calculated intercoder reliability by double-coding a sample of one month (March 2005). Intercoder reliability corrected for agreement by chance (Scott’s pi) ranged from .68 (deregulation of the Seoul area) to .81 (tremendous relocation cost) with an average reliability of .79.

In order to assess which attributes were more prominent than others, the coders counted how many total news segments mentioned each attribute. Our assumption was that the greater the number of news segments, the greater the prominence of the attribute in the media. Of the 456 segments found, 143 mentioned a reason (attribute) to support or oppose the capital relocation; and 112 referred to at least one of the five attributes we analyzed.⁴ In many cases, the same attribute was mentioned more than once in a single segment. No matter how many mentions were made, we counted them as one mention as far as they came from the same segment. There were also many segments that involved more than one attribute. Altogether, the five attributes were mentioned a total of 152 times in the 112 news segments.

Opinion survey

We carried out a telephone survey of residents in Seoul, South Korea. Using systematic sampling, households were randomly selected from telephone directories. Within each household, interviewers selected one respondent aged 20 years or older for interview, using the last-birthday method. Interviews were conducted by trained undergraduate students majoring in communication. A total of 527 respondents were interviewed between 30 April and 11 May 2005, with a cooperation rate (COOP1) of 41.2 percent.

Respondents’ support for the issue was measured with a single item asking, on a five-point scale (1 = strongly oppose; 5 = strongly support), how strongly they supported or opposed the capital relocation. In order to identify salient issue attributes, we then asked respondents, in an open-ended question, to mention a specific reason why the plan was a good or bad idea. Our assumption was that the most salient attribute must be the one most accessible in a person’s memory (Hodges and Wilson, 1993; Tversky and Kahneman, 1973).⁵ The salient attribute, therefore, should be brought to mind most easily and mentioned first. Two coders later categorized them into one of the five major issue attributes we selected. Again, intercoder reliability (Scott’s pi = .76) was calculated by double-coding 25 percent of the entire responses.

We also measured people’s opinions on each of the five attributes. On a five-point scale (1 = very unlikely; 5 = very likely), respondents were asked how likely it was for each attribute to happen as a consequence of capital relocation: Balanced development across the nation (M = 2.80, SD = 1.35), Solution to metropolitan problems (M = 2.88, SD = 1.26), Stimulating local economy (M = 3.02, SD = 1.21), Deregulation of the Seoul area and resulting problems (M = 3.56, SD = 1.18), Tremendous relocation cost (M = 4.40, SD = 0.89).

In order to assess the amount of news viewing, we used two questions tapping how often (1 = rarely; 5 = very often) respondents viewed (a) daily newscasts (M = 3.79, SD = 1.36) and (b) investigative news programs (M = 2.56, SD = 1.38). These two measures were then combined into a single item representing respondents’ exposure to television news (r = .49).

For control purposes, the survey measured a number of demographic and political orientation variables. Demographic variables included age (M = 45.65, SD = 14.90), gender (55.2 percent female), education (the highest degree completed, 52.9 percent some college or more) and income (the median monthly household income between 3 and 4 million Korean won). Respondents were also asked about their general interest in politics (1 = not at all interested; 5 = very interested, M = 2.72, SD = 1.34), ideology (1 = very conservative; 5 = very liberal, M = 2.84, SD = 1.00) and political party preferences (52.6 percent no party preference; 26.2 percent the Grand National Party; 10.4 percent the Uri Party; 10.7 percent others/‘don’t knows’).

Findings

Table 1 presents the frequency with which each attribute appeared in primetime news programs. In all three networks, Balanced development was mentioned most frequently, appearing in a total of 68 news segments (see ‘3 networks combined’ column (T) under ‘4 months total’]). Following next were Solution to metropolitan problems and Tremendous relocation cost, mentioned in 34 and 26 segments respectively. Stimulating the local economy and Deregulation of the Seoul area were mentioned relatively less often, each appearing in 12 segments only. In sum, balanced development was the most salient attribute in the media. Mentions of balanced development significantly outnumbered any other attributes, appearing at least twice to five times more often than others. Table 1 also reveals that the distribution of frequency was quite consistent across the four months examined, with balanced development being the most prominent attribute, followed by solution to metropolitan problems and tremendous relocation cost.

Table 1.

Number of news segments mentioning each attribute of capital relocation

	June 2004				July 2004				October 2004				March 2005				4 months total
Attributes	K	M	S	T	K	M	S	T	K	M	S	T	K	M	S	T	K	M	S	T
Balanced development	4	3	4	11	6	4	3	13	13	11	9	33	3	4	4	11	26	22	20	68
Solution to metropolitan problems	2	2	3	7	3	3	4	10	4	4	4	12	1	2	2	5	10	11	13	34
Stimulating the local economy	1	1	3	5	0	0	1	1	0	1	0	1	1	2	2	4	2	4	6	12
Deregulation of the Seoul area	0	1	1	2	1	0	0	1	1	1	2	4	2	1	2	5	4	3	5	12
Tremendous relocation cost	2	2	2	6	4	3	3	10	0	2	2	4	2	1	3	6	8	8	10	26

K = KBS News 9, M = MBC News Desk, S = SBS News 8, T = 3 networks combined.

Our first hypothesis (H1) deals with a correspondence between prominent attributes in the media and the agenda of attributes in people’s mind (attribute agenda setting). Table 2 shows the total number of people who mentioned each attribute, when asked in an open-ended question to give a reason why the plan was a good or bad idea. A considerable number of respondents (N = 86) did not answer the question. Many respondents (N = 59) simply rephrased their support or opposition without mentioning a specific reason (e.g. ‘This is not the best time to move the capital’; ‘There’s no reason to move’; ‘I simply like/hate the idea’; ‘It is good/bad for the country’). These invalid answers were grouped into a separate category (see Table 2). There were also a large number of respondents (N = 85) who mentioned a reason other than the five major attributes.

Table 2.

Reasons to support or oppose capital relocation (mentioned by respondents)

	Number of respondents mentioning each reason (%)
Reasons to support or oppose (issue attributes)	Whole sample (N = 527)	Light viewers (N = 263)	Heavy viewers (N = 264)
Balanced development across the nation	92 (17.5)	21 (8.0)	71 (26.9)
Solution to metropolitan problems	56 (10.6)	25 (9.5)	31 (11.7)
Stimulating the local economy	12 (2.3)	8 (3.0)	4 (1.5)
Deregulation of the Seoul area and resulting problems	13 (2.5)	6 (2.3)	7 (2.7)
Tremendous relocation cost	124 (23.5)	56 (21.3)	68 (25.8)
Other reasons	85 (16.1)	46 (17.5)	39 (14.8)
Invalid answers	145 (27.5)	101 (38.4)	44 (16.7)
Total	527 (100.0)	263 (100.0)	264 (100.0)

Note: Invalid answers include respondents who did not answer the question (N = 86) and those who simply rephrased their support or opposition without mentioning a specific reason (N = 59).

As shown in Table 2 (see ‘whole sample’ column), the most frequently mentioned reason was Tremendous relocation cost (N = 124 out of 527 total mentions, 23.5 percent), indicating that relocation cost was the most salient attribute in people’s minds. Following next were Balanced development across the nation (N = 92) and Solution to metropolitan problems (N = 56), accounting for about 17.5 percent and 10.6 percent of the all answers. A rank-order correlation reveals that there is a moderate correspondence (Spearman’s rho = .667) between the media (‘3 networks combined’ column (T) under ‘4 months total’ in Table 1) and the public (‘whole sample’ column in Table 2). This correlation, however, did not reach conventional levels of statistical significance (p = .110, one-tailed, N = 5).

More compelling evidence for attribute agenda setting emerged when we analyzed heavy and light viewers separately (see ‘light viewers’ and ‘heavy viewers’ columns in Table 2).⁶ A comparison of heavy and light viewers indicates that prominent attributes in the media are associated more closely with the attribute agenda among heavy viewers. For example, while only about 8 percent of light viewers (N = 21) mentioned Balanced development, i.e. the most prominent attribute in the media, a significantly larger number of heavy viewers (N = 71 or 26.9 percent) referred to the same attribute, possibly as a consequence of their greater exposure to the media. In fact, balanced development was the most frequently mentioned attribute among heavy viewers. The rank-order correlation between the media and heavy viewers was strong (Spearman’s rho = .872) and statistically significant (p = .027, one-tailed, N = 5). Light viewers, on the other hand, had a much smaller correlation (Spearman’s rho = .564), which was not statistically significant (p = .161, one-tailed, N = 5). Supporting H1, these findings indicate that news media, by placing greater emphasis on a certain attribute, can increase its salience among audience members, i.e. the transfer of attribute salience from the media to the public.

The open-ended format for the salience measure raises concern that our respondents might be unwilling to mention an attribute that ran contrary to their own views. Opponents of the relocation, for example, would not mention balanced development as an attribute not because it was not salient but because it was a reason to support, rather than oppose, the plan. A way to address this concern was to control for the effect of respondents’ own views (support for or opposition to the relocation), and then see the effect of news viewing on their mention of each attribute. We used a logistic regression model (see Table 3) with respondents’ mention of each attribute being the dichotomous dependent variable (1 = mentioned; 0 = did not mention). Support for the capital relocation and television news viewing entered the regression together after accounting for demographics, political interest, ideology and party preferences.

Table 3.

Logistic regressions predicting mention of each attribute (N = 527)

	Balanced development	Solution to metropolitan problems	Stimulating the local economy	Deregulation of the Seoul area	Tremendous relocation cost
Support for capital relocation	.860***	.622***	1.900**	–.084	–1.113***
TV news exposure	.256**	–.064	–.277	.032	.067
χ2 for the block	65.038***	21.443***	16.857***	.115	57.640***

Note: Entries are unstandardized logistic regression coefficients (b) after controlling for demographics (age, education, gender, income), political interest, ideology and political party preferences.

***p < .001,

**p < .01,

*p < .05.

Table 3 shows a positive coefficient (b = .256, p < .01) between news viewing and the mention of balanced development, even after controlling for the effect of one’s own view on the capital relocation. This finding supports the idea that viewing television news increased the likelihood that a person would mention balanced development, probably as a consequence of repeated exposure to the attribute, which had been highly prominent in the news. News viewing, however, was largely unrelated to mentioning other attributes probably because they were not as prominent in the media.

Drawing upon the notion of priming, our second hypothesis (H2) posits that issue attributes emphasized in news media will become significant dimensions of issue evaluation among the audience (attribute priming). In order to identify the significant dimensions, we used a regression model, where respondents’ support for capital relocation was regressed onto their opinions on five issue attributes (see Table 4). Demographics (age, education, income, gender) and political orientation variables (interest in politics, ideology, political party preferences) were also entered into the regression for control purposes. In this way, we assessed how important or influential each attribute was in predicting respondents’ support for or opposition to capital relocation.

Table 4.

Attributes predicting support for capital relocation

	Light viewers (N = 263)	Heavy viewers (N = 264)	Interactions (N = 527)
Balanced development across the nation	.162**	.353***	.275***
Solution to metropolitan problems	.302***	.223***	.325***
Stimulating the local economy	.051	.125*	.041
Deregulation of the Seoul area and resulting problems	–.082	–.045	–.063
Tremendous relocation cost	–.193***	–.184***	–.187***
TV news exposure			.040
Exposure × Balanced development across the nation			.075*
Exposure × Solution to metropolitan problems			.011
Exposure × Stimulating the local economy			.061
Exposure × Deregulation of the Seoul area and related problems			–.002
Exposure × Tremendous relocation cost			–.001
R² (%)	26.0***	39.0***	38.9***

Note: Entries are standardized regression coefficients (β) after controlling for demographics (age, education, gender, income), political interest, ideology and political party preferences. For the interaction block, only variables in previous blocks are controlled for in order to minimize multicollinearity.

***p < .001,

**p < .01,

*p < .05.

The regression model, when applied to light and heavy viewers separately, revealed somewhat different prediction powers of each attribute (see ‘light viewers’ and ‘heavy viewers’ columns in Table 4). Among light viewers, Solution to metropolitan problems (β = .302, p < .001) was the most influential predictor, followed by Tremendous relocation cost (β = –.193, p < .001) and Balanced development (β = .162, p < .01). A rank-order comparison revealed that there was not a statistically significant correlation (Spearman’s rho = .667, p = .110, one-tailed, N = 5) between the salient attributes in the media (‘3 networks combined’ column (T) under ‘4 months total’ in Table 1) and the influential attributes among light viewers (‘light viewers’ column in Table 4). In particular, balanced development – the most salient attribute in the media – was only the third influential attribute in shaping light viewers’ support for capital relocation.

The same attribute, however, was the most influential dimension (β = .353, p < .001) of issue evaluation among heavy viewers (see Table 4). Solution to metropolitan problems (β = .223, p < .001) and Tremendous relocation cost (β = –.184, p < .001), the second and the third prominent attributes in the media (see Table 1), were also the next two significant predictors of issue evaluation. The rank-order correlation between the media agenda and heavy viewers was .975 and statistically significant (p = .003, one-tailed, N = 5). Supporting H2, these findings indicate that prominent issue attributes in the media function as important standards by which the audience evaluates a given issue. When evaluating the relocation plan, heavy viewers gave the largest weight to balanced development, probably as a consequence of their repeated exposure to the attribute in news coverage.

To further examine the difference between heavy and light viewers, we tested potential interactions between the amount of news exposure and the importance of each issue attribute. We again used a regression model (see ‘interactions’ column in Table 4), which entered five interaction terms along with control variables, the measures of attribute opinions and the amount of news exposure. The regression analysis revealed a significant interaction effect from Balanced development (β = .075, p < .05). This significant interaction indicates that balanced development, i.e. the attribute prominently emphasized in the media, was far more influential among heavy viewers than it was among light viewers as an evaluative dimension of the capital relocation.

Discussion

Our findings provide compelling evidence that the media play a role in setting attribute agenda. By placing different degrees of emphasis, news media influenced the salience of certain attributes in the audience’s minds. Balanced development across the nation, an attribute hardly mentioned by light viewers, was considerably salient among heavy viewers when the media placed it prominently in their news coverage.

An important outcome of attribute agenda setting was its priming effect. Balanced development, the most prominent attribute in the media, was the strongest predictor of heavy viewers’ support for the relocation plan. Taken together, these findings suggest that mass media can successfully make certain aspects of an issue more or less salient, and thus prime which pieces of information will be used more prominently when citizens make up their minds on the issue. These interconnected processes of attribute agenda setting and priming can play a key role in shaping public opinion and in building public consensus on controversial issues.

Before further discussing our findings, we must point out a few shortcomings of this study. First, our use of an open-ended format had several shortcomings as a measure of attribute salience. Most importantly, a significant number of respondents gave invalid answers (no response, N = 86; simply rephrasing an opinion without talking about a reason, N = 59) when asked to mention a reason why the plan was a good or bad idea. Because the question was asked in an open-ended format, we had only limited control over the way the question was answered. The two-sided nature of the issue further complicated the open-ended measure. It is likely that many respondents were unwilling to mention a specific attribute that ran against their own views even when the attribute was highly salient. Our operationalization of attribute salience, therefore, had limitations in its validity as well as in its reliability.

What other measures are available then? Which measures are more appropriate for measuring attribute salience? Even though perceived importance has been adopted in many agenda setting studies, researchers (e.g. Nelson et al., 1997) demonstrated that there were distinctive differences between perceived importance of an issue and its actual salience. A latency test can be an appropriate measurement in the sense that the test directly measures the amount of time that it takes a respondent to answer a given question (e.g. Bassili, 1995). The rationale is that the longer it takes a person to answer a given question, the less salient or accessible the evoked information was in his or her memory. The tests of response latency, however, are easily plagued by random measurement error, and thus require an elaborate technical setup, including specially equipped telephone interviewing facilities. Developing a more valid and reliable method to measure attribute salience will greatly enhance attribute agenda setting research. The question of how to develop it, of course, is beyond the scope of our study, but will be an important topic for future research.

Our research design does not allow us to make a strong inference about causality. Certain topics, such as a natural disaster or a war, are by nature more drastic than others, and thus easily picked up by the public and also by the media at the same time (Cobb and Elder, 1972). When this is the case, the correspondence between the media’s and the public’s agendas is spurious; and the two variables simply go together without one influencing the other. In our study, however, a significant media–public correspondence was identified only from heavy viewers. If the nature of issue attributes, not a media effect, produces the correspondence, it cannot be explained why a prominent attribute in the media is highly salient among heavy viewers, but not as highly salient among light users of television news.

With these shortcomings in mind, our findings bolster accessibility-based theorizing and research showing that the media’s emphasis on particular issues produces agenda setting and priming effects. The media seem to influence attribute accessibility, the ease with which a specific issue attribute is retrieved from memory. Also importantly, we found that more salient attributes in the media were more likely to influence one’s evaluation of the issue. These attribute agenda setting and priming processes are probably a subtle but still powerful way news media can influence public opinion; and it will be a continuous interplay between social elites, journalists and other institutions that determines which aspects (or attributes) of a given issue will be highlighted in public discourse (Van Dijk, 1996). In sum, we conclude that the media, by emphasizing or de-emphasizing certain attributes of an issue, have the power to tell the audience ‘what is important to think about’ when making a judgment about the issue.

Our study offers a useful replication of a new and still-developing line of research in agenda setting. The findings reported here will expand the scope of the theory by demonstrating the transfer of attribute salience, as well as issue salience, from the media to the audience. The notion of attribute priming is particularly new; and our findings do support the idea that attribute agenda setting produces a priming effect. We certainly demonstrated that agenda setting and priming operate together.

Our study also enhances the intercultural validity of the theory. Findings provided a compelling empirical demonstration that attribute agenda setting and priming may follow similar forms in South Korea as they do in the US and other countries. Despite the cultural differences in these countries, the geographically expanded replications may be attributed, according to McCombs et al. (1997), to the similarities of the political and media systems. Since the democratization of the country in the early 1990s, South Korea has maintained a reasonably democratic political system with the media relatively unfettered by the government (Kim and Han, 2005).

Agenda setting studies in the US have suggested that newspapers are in general more likely than television news to dominate agenda setting (McCombs, 2004). Because of its limited and superficial hard news content, television has been also understood to play only a minor role in gaining information about political issues (Chaffee and Frank, 1996). In this regard, the agenda setting role of television news in South Korea is pronounced. Our findings indicate that news viewing is consistently associated with attribute agenda setting and priming effects. This is probably because South Korean news programs are operating in an environment somewhat different from their American counterparts. According to an audience survey in 2004, about 77 percent of the Korean population (15 years or older) ‘regularly’ watched news on television,⁷ while only about 60 percent of American adults (18 or older) in the same year answered that they had watched television news ‘yesterday’ (Pew Research Center, 2004). Although a direct comparison of these two statistics is hard to make, it may at least suggest that South Koreans watch television news considerably more than Americans do. It is also important to note that the primetime news programs in South Korea are all full one-hour programs with no commercial break, which allow for delivering a large amount of information in great detail, which may potentially produce substantial cognitive effects.

Footnotes

Notes

References

Althaus

Kim

(2006) Priming effects in complex information environments: Reassessing the impact of news discourse on presidential approval. The Journal of Politics 68: 960–970.

Ansolabehere

Behr

Iyengar

(1993) The Media Game: American Politics in the Television Age. New York: Macmillan.

Bassili

(1995) Response latency and the accessibility of voting intentions: What contributes to accessibility and how it affects vote choice. Personality and Social Psychology Bulletin 21: 686–695.

BBC News (2004) S Korea Chooses New Capital Site. Available at: news.bbc.co.uk_2_hi_asia-pacific_3554296.stm.

Chaffee

Frank

(1996) How Americans get political information: Print versus broadcast news. The Annals of the American Academy of Political and Social Science 546: 48–58.

Cobb

Elder

(1972) Participation in American Politics: The Dynamics of Agenda-Building. Baltimore, MD: The Johns Hopkins University Press.

Cohen

(1963) The Press and Foreign Policy. Princeton, NJ: Princeton University Press.

Dursun

(2005) News coverage of the enlargement of the European Union and public opinion: A case study of agenda-setting effects in the United Kingdom. Paper presented at the biennial international conference of the European Union Studies Association, Austin, TX.

Entman

(1993) Framing: Toward clarification of a fractured paradigm. Journal of Communication 41: 51–58.

10.

First

(1997) Television and the construction of social reality: An Israeli case study. In: McCombs

Shaw

Weaver

(eds) Communication and Democracy. Hillsdale, NJ: Lawrence Erlbaum, 41–50.

11.

Gamson

Modigliani

(1989) Media discourse and public opinion on nuclear power: A constructionist approach. American Journal of Sociology 95: 1–37.

12.

Golan

Wanta

(2001) Second-level agenda setting in the New Hampshire primary: A comparison of coverage in three newspapers and public perceptions of candidates. Journalism and Mass Communication Quarterly 78: 247–259.

13.

Herman

Chomsky

(1988) Manufacturing Consent: The Political Economy of the Mass Media. New York: Pantheon.

14.

Hodges

Wilson

(1993) Effects of analyzing reasons on attitude change: The moderating role of attitude accessibility. Social Cognition 11: 353–366.

15.

Iyengar

(1990) The accessibility bias in politics: Television news and public opinion. International Journal of Public Opinion Research 2: 1–15.

16.

Iyengar

(1991) Is Anyone Responsible? How Television Frames Political Issues. Chicago, IL: University of Chicago Press.

17.

Iyengar

Kinder

(1987) News that Matters: Television and American Opinion. Chicago, IL: University of Chicago Press.

18.

Kim

S-H

Han

(2005) Media use and participatory democracy in South Korea: News media vs. entertainment media. Mass Communication and Society 8: 133–153.

19.

Kim

S-H

Scheufele

Shanahan

(2002) Thank about it this way: Attribute agenda-setting function of the press and the public’s evaluation of a local issue. Journalism and Mass Communication Quarterly 79: 1–25.

20.

Kim

(2004) Capital relocation ads annoy Seoul City. The Korea Herald. Available at: www.koreaherald.co.kr/archives/result_contents.asp?id=200407310027&query=annoyseoul.

21.

King

(1997) The press, candidate images, and voter perceptions. In: McCombs

Shaw

Weaver

(eds) Communication and Democracy. Hillsdale, NJ: Lawrence Erlbaum, 29–40.

22.

Kiousis

(2005) Compelling arguments and attitude strength: Exploring the impact of second-level agenda setting on public opinion of presidential candidate images. Press/Politics 10: 3–27.

23.

Krosnick

Brannon

(1993) The impact of the Gulf War on the ingredients of presidential evaluations: Multidimensional effects of political involvement. American Political Science Review 87: 963–975.

24.

Lopez-Escobar

McCombs

Lennon

(1998) Two levels of agenda setting among advertising and news in the 1995 Spanish elections. Political Communication 15: 225–238.

25.

McCombs

(1997) New frontiers in agenda-setting: Agendas of attributes and frames. Paper presented at the annual convention of the Association for Education in Journalism and Mass Communication, Chicago, IL.

26.

McCombs

(2004) Setting the Agenda: The Mass Media and Public Opinion. Cambridge: Polity Press.

27.

McCombs

(2005) A look at agenda-setting: Past, present and future. Journalism Studies 6: 543–557.

28.

McCombs

Estrada

(1997) The news media and the pictures in our heads. In: Iyengar

Reeves

(eds) Do the Media Govern? London: Sage, 237–247.

29.

McCombs

Evatt

(1995) Issues and attributes: Exploring a new dimension in agenda setting. Communication Sociedad 8: 7–32.

30.

McCombs

Shaw

(1972) The agenda setting function of mass media. Public Opinion Quarterly 36: 176–187.

31.

McCombs

Llamas

Lopez-Escobar

Rey

(1997) Candidate images in Spanish elections: Second-level agenda-setting effects. Journalism and Mass Communication Quarterly 74: 703–717.

32.

Meijer

M-M

Kleinnijenhuis

(2006) Issue news and corporate reputation: Applying the theories of agenda setting and issue ownership in the field of business communication. Journal of Communication 56: 543–559.

33.

Nelson

Clawson

Oxley

(1997) Media framing of a civil liberties conflict and its effect on tolerance. American Political Science Review 91: 567–583.

34.

Pew Research Center (2004) News Audience Increasingly Polarized. Available at: people-press.org/reports/display.php3?PageID=833.

35.

Price

Tewksbury

(1997) News values and public opinion: A theoretical account of media priming and framing. In: Barett

Boster

(eds) Progress in Communication Sciences: Advances in Persuasion. Greenwich, CT: Ablex, 173–212.

36.

Scheufele

(2000) Agenda-setting, priming, and framing revisited: Another look at cognitive effects of political communication. Mass Communication and Society 3: 297–316.

37.

Scheufele

Tewksbury

(2007) Framing, agenda setting, and priming: The Evolution of three media effects models. Journal of Communication 57: 9–20.

38.

Sheafer

(2007) How to evaluate it: The role of story-evaluative tone in agenda setting and priming. Journal of Communication 57: 21–39.

39.

Stone

McCombs

(1981) Tracing the time lag in agenda-setting. Journalism Quarterly 58: 51–55.

40.

Takeshita

Mikami

(1997) How did the media influence the voters’ choice in the 1993 general election in Japan? Analysis of agenda setting. Keio Communication Review 17: 27–41.

41.

The Korea Times (2005) President Roh’s reasoning unconvincing. Available at: times.hankooki.com/plaza/plaza_domestic.php?cur_date=20050328&page=7.

42.

Tversky

Kahneman

(1973) Availability: A heuristic for judging frequency and probability. Cognitive Psychology 5: 207–222.

43.

Van Dijk

TA.

(1996) Power and the news media. In: Paletz

(ed.) Political Communication in Action: States, Institutions, Movement, Audiences. Cresskill, NJ: Hampton Press, 9–36.

44.

Wikipedia (2006) New capital. Available at: en.wikipedia.org/wiki/Gongju.

45.

Willnat

Zhu

J-H

(2001) Newspaper coverage of public opinion in Hong Kong: A time-series analysis of media priming. Political Communication 13: 231–246.

46.

Zaller

Feldman

(1992) A simple theory of the survey response: Answering questions versus revealing preferences. American Journal of Political Science 36: 579–616.