Printer Friendly

Are you an opinion giver, seeker, or both? Re-examining political opinion leadership in the new communication environment.

The introduction of the opinion leadership concept into the communication field (Katz & Lazarsfeld, 1955; Lazarsfeld, Lerelson, & Gaudet, 1944) marked an important turning point in the history of communication studies, moving the paradigm from strong to limited media effects models. In their seminal book Personal Influence, Katz and Lazarsfeld (1955) described opinion leaders as "both more generally exposed to the mass media, and more specifically exposed to the content most closely associated with their leadership" (p. 316). Many communication studies have examined opinion leadership and its relationship to persuasion, diffusion, and mass media use in the last half century (Powell & Shelby, 1985; Robinson, 1976; Shah & Scheufele, 2006; Troldahl & van Dam, 1965; Weimann, 1994).

This study revisits the concept of political opinion leadership and followership in the context of the current social and media environments. Many previous studies conceptualized opinion leadership and followership as competing concepts. However, the changes in the social and media environments have made the conceptualization of opinion leaders and followers much more complex and dynamic. In addition to the fact that one could play the role of opinion leader about one issue while being an opinion follower about another issue, one could consider oneself an opinion leader in one situation (depending on time, space, and people) while playing the role of a follower in another situation for the same issue. Even when individuals play the role of opinion follower, they may not just be passive followers but can actively seek out, compare, and evaluate multiple sources of opinions and their qualities. It is also possible for individuals to consider themselves neither opinion leaders nor followers for issues that do not have any relevance to their lives or for issues that are too difficult for them to comprehend. Considering these complex cases, we propose that opinion leadership and followership should be examined as complementary and interactional rather than dichotomous. In examining the relationship between opinion leadership and followership, we propose four groups: opinion givers/seekers, opinion givers, opinion seekers, and nondiscussants. The primary goal of the current study is to compare these groups' socioeconomic and demographic factors, media connectedness, curating skills (critical and active opinion selection and exchange), offline political participation, and political participation on social network services (SNSs). The study focuses on monomorphic opinion leadership and followership concerning politics and political affairs.

Opinion Leadership and Followership

The concepts of opinion leadership and followership were introduced in Katz and Lazarsfeld's (1955) research on mass media effects of political campaigns. Before this research, the mass media were understood as having strong power to influence individuals' perceptions, opinions, attitudes, and behaviors (DeFleur & Ball-Rokeach, 1990). When Katz and Lazarsfeld researched the effect of radio and print media on voting decisions in the 1940 presidential election in the United States, they discovered that the effect of the mass media was not as strong as it was perceived to be. Instead, they found personal influence of opinion leaders to be as strong as (or even stronger than) the mass media.

Katz and Lazarsfeld continued to conduct studies on the influence of opinion leaders on a range of topics and in various contexts, including small group settings. They established a connection between media effects and opinion leadership and proposed a two-step flow model of communication, which explicates that the effect of mass media on individuals is buffered by opinion leaders who follow the mass media closely and convey media messages to others in their community. The concept of opinion leadership and its relationship to media use continued to be an important topic for subsequent studies, including, most notably, diffusion of innovation studies (Rogers, 1962). In such studies, opinion leaders are explained as playing a crucial role in triggering the diffusion process.

In many previous studies, opinion followers have been defined as a counterpart to opinion leaders. Opinion followers are, by definition, those influenced by opinion leaders. They are less likely to follow the mass media as closely as opinion leaders (Katz & Lazarsfeld, 1955). In the two-step flow model of communication, opinion followers are at the final stage of information flow--mass media messages flow from media institutions to opinion followers via opinion leaders.

Opinion followers are more likely to regard opinions of those whom they perceive to be opinion leaders (based on socials status, expertise, etc.) as more credible than mass media messages (Katz & Lazarsfeld, 1955; A. L. Lee, 2010; Shoham & Ruvio, 2008). Recently, A. L. Lee (2010) found that although patients often seek drug information from the mass media and small print media, they are more likely to follow the opinions of pharmacists when they have to decide which drugs to take. Shoham and Ruvio (2008) found that whereas opinion leaders for computer hardware and software are more likely to read relevant magazines and visit computer stores, opinion followers are significantly less likely to do so. In addition, opinion followers tend to be less knowledgeable about new computer products than opinion leaders.

The Relationship Between Opinion Leadership and Followership

How are opinion leadership and followership related? The findings in previous studies provide mixed answers to this question. Several studies have reported negative correlations between the two factors (Lassar, Manolis, & Lassar, 2005; Shoham & Ruvio, 2008). For example, based on a survey in Israel, Shoman and Ruvio (2008) found that opinion leading and opinion seeking had a fairly high negative correlation (r = -0.56). On the other hand, several other studies have reported positive correlations between the two (Bertrandias & Goldsmith, 2006; Flynn, Goldsmith, & Eastman, 1996). A positive correlation between opinion leadership and followership challenges the commonly held notion that opinion leadership and followership are at opposite ends of a scale.

A complementary relationship between opinion leadership and followership has been previously proposed. Merton (1949) and Katz and Lazarsfeld (1955), for example, have noted that opinion leaders in one specific area may be opinion followers in other areas. In their 1965 study, Troldahl and van Dam found that opinion leaders were also likely to seek others' opinions. The authors found that two-thirds of those who gave opinions to others on public affairs topics also sought others' opinions, making them both opinion givers and seekers (Troldahl & van Dam, 1965). Robinson (1976) also found that 68% of opinion givers also received opinions but that only 26% of non-opinion givers received opinions. Robinson (1976) proposed a three-category typology of interpersonal influence: opinion givers (32%), opinion receivers (those who receive opinions from others but do not give them to others, 17%), and nondiscussants (those who neither give nor receive opinions, 51%). Weimann (1991) proposed four groups of strong to weak influenceability by employing the strength of personality measure.

Complex and contradictory findings about the relationship between opinion leadership and followership are further reinforced by critical changes in the social and communication environments in the last decade. First, how individuals form and maintain their social networks has changed. Compared to the mid-20th century when the concept of opinion leadership was first introduced by Katz and Lazarsfeld (1955), people's membership and belongingness to social communities have weakened (Putnam, 2001). People's social networks are less defined by clear membership in certain communities and more by fluid connections with various groups (Wellman, 2001). Many early studies examined opinion leaders and followers in geographically well-defined and relatively small community contexts (Katz & Lazarsfeld, 1955; Rogers & Cartano, 1962). In the current social environment where an individual can connect to multiple communities at the same time, the understanding of opinion leadership and followership deserves reconsideration.

Second, the diversification of communication media is likely to challenge the hitherto defined concepts of opinion leadership and followership. How individuals seek and find information today has changed significantly when compared to a time when people relied mainly on mass media. Today, the Internet and mobile technologies enable individuals to be more selective in seeking opinions from various sources, including new and old media and other people. Such a media environment brings about two main changes in opinion flow. First, the distinction between news and opinions is blurred. Although early studies of opinion leadership did not make a clear distinction between news and opinions (Lazarsfeld et al., 1944), opinion leaders were largely conceptualized as those who closely followed news and others' opinions, and followers were conceptualized as those who were more likely to seek opinions of leaders when making decisions (Robinson, 1976). In the current media environment where people tend to collect information not only from mass media but also from other sources on the Internet, it has become more difficult to clearly distinguish what counts as "factual news" and "opinions" (Rieder, 2010). In particular, news and opinions often flow together on the same media platform (Marwick & boyd, 2011). News and opinions produced and shared by mainstream media, independent media, and people become mingled together on social media. This trend has been intensifying as more and more people get news, information, and opinions on mobile media.

Another change that the current media environment brings about is the increased interactivity and channel diversity in news flow. A clear distinction between opinion leaders and followers can be made only when there is a linear and one-way flow of news in mass media. Until recently, most news was produced by professionals in the media industry and disseminated to the audience. Opinion leaders were the ones who closely followed the news from mass media and disseminated news and opinions to the followers in their communities. In the current media environment, however, the flow of news has become nonlinear, interactive, and recursive. Following news and opinions requires more than passively receiving information, and leading opinions requires extensive following of news and opinions from various other channels. Several recent studies have argued that the role of opinion leaders will be weakened in the current media environment because most people, including those who might have been classified as opinion followers, now have direct access to diverse information sources (Bennett & Manheim, 2006). Other studies have noted that opinion leaders in the online environment are those who monitor various information sources and select news and opinions to share with others (Turcotte, York, Irving, Scholl, & Pingree, 2015). Active and selective following of news and opinions is becoming a necessary condition to be an effective opinion leader (Schafer & Taddicken, 2015).

Re-Examination of Opinion Leadership and Followership

Considering changes in the current social and communication environment and conflicting findings in past studies about the relationship between opinion leadership and followership, we propose to re-examine opinion leadership and followership in this study. More specifically, we propose that viable opinion leaders are also likely to be "effective followers" (Kelley, 1988). Effective followership indicates the ability to pay attention to other people's opinions and incorporate them into one's own decision-making and opinion-formation processes (Kelley, 1988; W. Lee, Cha, & Yang, 2011). To examine interactions between opinion leadership and followership, we propose four groups: opinion givers/seekers, opinion givers, opinion seekers, and nondiscussants. We use the terms givers and seekers rather than leaders and followers in our typology to take a more neutral stance. Givers and seekers are more concrete terminologies in describing opinion-related motivation and behaviors for a particular issue. Also, the notion of opinion seeking connotes a more active agency than that of opinion following. The four categories of opinion giving and seeking we suggest in this study are based on Robinson's (1976) categories (opinion givers, opinion receivers only, and nondiscussants) and Weimann's (1991) four categories of influentials.

The main goal of this categorization is to compare the four groups and examine whether those who are high in both opinion giving and seeking are distinguished from others in terms of the criteria that have traditionally been associated with opinion leadership and of a newly proposed criterion, curating skill (the ability to critically select and actively share contents).

RQ1: Are individuals categorized into four distinct groups: opinion givers/seekers, opinion givers, opinion seekers, and nondiscussants?

Political Opinion Giving/Seeking and Relevant Factors

We frame our research to compare the four groups of opinion giving and seeking in terms of the five criteria based on previous opinion leadership studies and changes in the media environment: socioeconomic status, media connectedness, curating skills, offline political participation, and political participation on social network services.

Socioeconomic and Demographic Factors

An existing body of literature points to positive relationships between socioeconomic status and opinion leadership. In particular, opinion leadership is associated with higher income and higher education (Katz & Lazarsfeld, 1955; Rogers & Cartano, 1962). Furthermore, previous studies indicate that opinion leaders are more likely to be older and male (Shah & Scheufele, 2006; Weimann, 1994). Based on the literature, we propose research question 2:

RQ2: Do members of the four groups of political opinion giving and seeking differ in income, education, age, and gender?

Media Connectedness

The concept of opinion leadership has been associated with higher and broader connections to mass media (Katz & Lazarsfeld, 1955; Rogers & Cartano, 1962). Recent studies have found that opinion leaders are more likely to follow political news in mass media, particularly newspapers and magazines, more closely than others (Kim, 2006; Nisbet, 2006; Shah & Scheufele, 2006). On the other hand early studies conceptualized opinion followers as those who are less connected to mass media (Rogers & Cartano, 1962; Shoham & Ruvio, 2008). Based on the literature, we propose research question 3:

RQ3: Do members of the four groups of political opinion giving and seeking differ in political media connectedness?

RQ3a: Do members of the four groups of political opinion giving and seeking differ in overall connectedness to different types of media for political information?

RQ3b: Do members of the four groups of political opinion giving and seeking differ in connectedness to specific media for political information?

Curating Skills

In addition to general media connectedness, one of the media literacy skills that have emerged and been deemed important in the last decade is the ability to critically select content from widely available sources and share the selected content with people in social networks (Jenkins, 2007). Recently, both academic and nonacademic works have used the term curators to refer to those who have the ability to collect and share content on interactive media platforms (Hogan & Quan-Haase, 2010; Pew Research Center, 2014; Samph, 2012). Samph (2012) states that in the sea of information available to individuals these days, more and more people rely on those who select useful information and share it with others. Hogan and Quan-Haase (2010) note that curatorship in digital platforms emphasizes a selection of meaningful content from abundant data as compared to the traditional role of curators to select scarce and precious objects. A report published by the Pew Research Center (2014) differentiates creators and curators: Creators are those who upload content created by themselves, whereas curators are those who upload content created by others.

Two main criteria define curating skills: critical seeking and active sharing. Critical curating skills include monitoring content from the various channels available and critically selecting and evaluating the content suitable for making decisions on a specific topic or issue (Jenkins, 2007). Curators tend to go beyond collecting content for themselves to actively sharing it with others, just as museum curators collect and display content to the public. The networked nature of the online environment has made it much easier for people who have skills and motivations to select and share content with others (Carah, 2014; Quan-Haase & Martin, 2013). Our fourth research question thus proposes to examine whether the four groups of opinion giving and seeking differ in the level of curating skills.

RQ4: Do members of the four groups of political opinion giving and seeking differ in the level of curating skills?

Offline Political Participation

Past studies have inquired whether opinion leaders are also more likely to participate actively in political and civic affairs than followers. Early studies of opinion leadership found that opinion leaders tend to participate in formal and informal organizations more than followers (Katz & Lazarsfeld, 1955; Rogers & Cartano, 1962). Robinson (1976) found that opinion givers were more likely than opinion receivers and nondiscussants to attend meetings or rallies, belong to an organization, and give money to political parties. Troldahl and van Dam (1965) reported that opinion givers and askers were more likely to be members of social clubs or organizations and to hold office in social organizations than inactives. Only a few recent studies have investigated the direct relationship between opinion leadership and civic and political participation (Rhee, Kim, & Kim, 2007a; Scheufele & Shah, 2000; Shah & Scheufele, 2006; Weimann, 1994). Shah and Scheufele (2006) found in both a cross-sectional study and a panel study that opinion leadership had a direct effect on civic participation. Based on past studies, we propose the following research question:

RQ5: Do members of the four groups of political opinion giving and seeking differ in offline political participation?

Political Participation on Social Network Services

A few studies have previously examined the relationship between opinion leadership and participation in the online environment (Ma, Lee, & Goh, 2014; Rhee et al., 2007a). Rhee and colleagues (2007a) found that online opinion leaders have higher intentions to participate in political activities than others. Ma and colleagues (2014) found that among the factors that are likely to influence individuals' tendencies to share news on social media--such as number of friends, network homophily, strength of ties, opinion leadership, opinion followership, and perceived credibility of news--opinion leadership was found to be the strongest factor. Based on these recent studies, we propose to examine the relationship between opinion giving and seeking and political participation on social network services (SNSs).

RQ6: Do members of the four groups of political opinion giving and seeking differ in political participation on social network services (SNSs)?

Method

An online survey was conducted in Seoul, Korea in October 2014. A reputable Seoul-based survey research firm was hired to administer the survey in Korean. All survey procedures were thoroughly monitored by the researchers. Survey respondents were recruited from the online panel directory of the survey research firm, which contains about 1 million people. An email invitation was sent to 4,280 potential respondents who met our study criteria (between the ages of 19 and 59 and residing in Seoul). We used a cluster sampling procedure with three criteria: (a) gender, (b) age (20s, 30s, 40s, and 50s), and (c) 25 subdistricts, called ku, of Seoul. Among the 1,874 invited potential respondents who visited the online survey website, 625 completed the survey. Thus, the participation rate of this survey was 14.6% (625 out of 4,280). In the final sample, 27 respondents were excluded because (a) they did not meet our basic stratifying conditions such as region or age, or (b) their answers were suspicious (e.g., cases in which the log file showed that the survey was completed unusually quickly). Eventually, a total of 598 Seoul residents between the ages of 20 and 59 from 25 ku districts were included in the sample.

Measures

Political opinion giving was measured by six items (Appendix 1) that were modified from the measures of opinion leadership developed by Katz and Lazarsfeld (1955), Rogers and Cartano (1962), and Flynn et al. (1996). (1) For each item, respondents indicated their positions on a five-point Likert scale: 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, and 5 = strongly agree.

Political opinion seeking was measured with six items derived from Flynn et al.'s (1996) measure of opinion seeking. Flynn and colleagues (1996) generated an eight-item measure of opinion seeking and tested it in five studies. Six out of eight items of opinion seeking loaded well into a factor. Our study used Flynn et al.'s six items as a measure of political opinion seeking (Appendix 1), with positions indicated on the same five-point Likert scale as was used to measure political opinion giving.

A principal component analysis of 12 items was conducted to test the discriminant validity of opinion giving and seeking. Kaiser-Meyer-Olkin's measure of sampling adequacy was 0.89, above the recommended value of 0.6, and Bartlett's test of sphericity was significant ([chi square] (66) = 4570.46, p < .001). All communalities were above 0.3 (see Table 1). The extraction was based on eigenvalue and the varimax rotation method. Two factors were derived (shown in Table 1), and these were consistent with the proposed constructs of opinion giving (M = 2.61; SD = .81; range = 4; minimum = 1, maximum = 5; reliability alpha = .91) and seeking (M = 2.81; SD = .74; range = 3.67; minimum = 1, maximum = 4.67; reliability alpha = .89). One item (OG1: "I make decisions on political affairs without relying on other people's opinions") was eliminated because of low communality (0.310) and low factor loading (0.491).

Curating skills was measured by 10 items (Appendix 1). The items were developed based on critical and active dimensions in Kelley's (1988) effective follower concept. Kelley's measure was modified based on five focus group discussions (with five or six individuals in each session) that we conducted before the construction of the survey on people's media connections. The respondents gave their answers on a five-point Likert scale (M = 3.11, SD = 0.56, reliability alpha = .87).

Media connectedness was measured by the question, "How helpful are the following media for you to obtain news or information on sociopolitical affairs?" The measure of helpfulness was used instead of the frequency of use to reflect the degree of dependency on particular media (Ball-Rokeach, 1985). The respondents were given a five-point Likert scale ranging from 1 = not at all helpful to 5 = very helpful for nine forms of media: TV channels (M = 3.63, SD = 1.00), offline newspapers (M = 3.16, SD = 1.14), online news (M = 3.51, SD = 0.92), radio (M = 3.04, SD = 0.98), magazines (M = 2.74, SD = 0.98), books (M = 3.06, SD = 0.97), portal sites (M = 3.56, SD = 0.87), blogs (M = 2.90, SD = 0.97), and online video content (M = 3.10, SD = 0.91). Overall media connectedness was calculated by averaging the means of the helpfulness scores for media types (M = 3.19, SD = 0.61).

Offline political participation was measured by asking respondents to choose all activities in which they had participated in the previous calendar year from a list of eight activities (indicated as yes or no). The activities were "voted" (93.4%), "encouraged others to vote" (47.4%), "participated in political demonstrations or attended speeches" (5.8%), "signed a petition supporting a particular party or candidate" (7.5%), "participated in meetings supporting a political party or politician" (7.5%), "made a political donation" (4.6%), "volunteered for political organizations" (4.3%), and "expressed opinions to the government or a political party on political issues" (7.4%). Respondents' answers were summed up to construct a scope variable of offline political participation (M = 1.74, SD = 0.99).

SNS political participation was measured by asking how often respondents participate in seven types of political activities on social network services (SNSs). For each type of activity, seven answer categories were provided: (1) never have done it; (2) have an experience of doing it; (3) once or twice a year; (4) three to ten times a year; (5) once or twice a month; (6) several times a week; and (7) every day. The activities were obtaining information on political issues on SNSs (M = 3.31, SD = 1.98), visiting SNS pages of candidates or parties (M = 2.15, SD = 1.49), reading postings of candidates or parties (M = 2.26, SD = 1.49), participating in political groups on SNSs (M = 1.69, SD = 1.25), posting messages on political issues on SNSs (M = 1.81, SD = 1.38), writing comments on political issues on SNSs (M = 1.99, SD = 1.47), and sharing other people's postings on political issues (M = 1.97, SD = 1.45). The average score for the seven activities was calculated to derive the value of the SNS political participation variable (reliability alpha = .92).

Political interest was measured as one of the control variables by asking, "How much are you interested in politics in your everyday life?" Five answer categories were provided: (1) "not interested at all," (2) "not interested," (3) "neutral," (4) "somewhat interested," and (5) "very interested" (M = 3.16, SD = 0.96).

Income was measured by asking respondents to indicate their monthly household income (including bonuses). Seventeen income categories were provided that ranged from "less than 1 million Korean won" to "20 million won or more" (median income category: 4-4.49 million won). Education was measured by asking what level of final education the respondents had completed. Six categories were provided ("middle school graduate or lower," "high school graduate," "attending a university or technical school," "university or technical school graduate," "attending a graduate school," and "postgraduate degree or higher"). About 64% were college graduates. Respondents were asked to provide their ages (M = 39.39, SD = 10.62) and to identify their genders (male: 49.8%, female: 50.2%).

Results

Table 2 demonstrates bivariate correlations of variables included in the study. It should be noted that opinion giving and seeking have a significant positive correlation (r = .41, p < .01).

Four Groups of Opinion Giving and Seeking (RQ1)

To derive four groups of opinion giving and seeking (RQ1), K-means cluster analysis was conducted as it is one of the most efficient techniques for finding a user-specified number of clusters (Tan, Stenbach, & Kumar, 2006). Based on previous literature and research question 1, the number of clusters was set at 4. Maximum iteration was set at 10. The results are presented in Table 3. Out of 625 total respondents, 228 (36.5%) belonged to the opinion givers/seekers group, 148 (23.7%) to opinion givers, 142 (22.7%) to opinion seekers, and 107 (17.1%) to nondiscussants. While the cluster analysis derived four groups proposed in research question 1, it should be noted that the level of opinion seeking of the opinion givers group is higher than that of nondiscussants. Respondents in the opinion givers group engage in opinion seeking but much less than that of the opinion givers/seekers group.

One alternative grouping of individuals is based on a two-group option, which was supported by the traditional method of dividing individuals into opinion leaders and opinion followers. To confirm that the four-group option is better than the two-group clusters, we conducted an analysis with the number of cluster set at two. We found that one of the two groups was higher than the other for both opinion giving (M = 3.1 versus M = 1.8) and seeking (M = 3.1 versus M = 2.2), which did not clearly distinguish opinion givers from seekers. In addition, we conducted cluster analyses with the number of clusters set at three and five to compare with our proposed four-group option. The result indicated that the clusters that were generated from these alternative analyses were hard to interpret and did not derive significant differences between clusters. Therefore, we conclude that the four-group option as conceptualized in the current study is better than other alternative options. ANOVA tests indicated that mean differences of the four groups for both opinion giving and seeking were significant (Table 3). All pairwise mean differences were also significant for both opinion giving and seeking.

Socioeconomic and Demographic Factors (RQ2)

Four groups--opinion givers/seekers, opinion givers, opinion seekers, and nondiscussants--were first compared in terms of income, education, age, and gender. Analysis of variance (ANOVA) was employed for the comparison of income, education, and age. Income and educational differences among the four groups were not statistically significant. Age difference was significant (F[3,621] = 5.56, p < .001). Opinion givers (M = 42.37) were significantly older than opinion givers/seekers (M = 38.30) and opinion seekers (M = 38.06). Chi-square tests indicated that significant differences existed among the four groups for gender ([chi square]2[3] = 21.95, p < .001). Males were more likely to be opinion givers/seekers than females, and females were more likely to be opinion seekers than males (Table 4).

Media Connectedness (RQ3)

Media connectedness was measured by overall media connectedness and specific media connectedness. Analysis of covariance (ANCOVA) controlling for income, education, age, gender, and political interest reveals that there is a significant group difference in overall media connectedness (F[3,603] = 11.1, p < .001). Pairwise comparisons indicate that opinion givers/seekers (M = 4.19) have significantly higher overall media connectedness than opinion givers (M = 3.12) and nondiscussants (M = 2.82) (Table 5). There was no significant mean difference between opinion givers/seekers and opinion seekers (M = 3.78).

With regard to the connectedness to specific media, the main effects of the four groups (controlling for income, education, age, gender, and political interest) were significant for blogs (F[3,603] = 16.69, p < .001), magazines (F[3,603] = 8.09, p < .001), online video content (F[3,603] = 7.43, p < .001), portal sites (F[3,603] = 7.10, p <.001), radio (F[3,603] = 6.14, p < .001), TV (F[3, 603] = 4.80, p < .01), and books (F[3,603] = 4.67, p < .01). Main effects were not significant for paper-based newspapers and online news sites.

In pairwise comparisons, opinion givers/seekers were more connected to magazines and books than other groups (Table 5). For portal sites, blogs, and online video sites, opinion givers/seekers had similar connectedness as opinion seekers but were better connected than opinion givers and nondiscussants. No significant pair-wise differences were found for television, online news sites, and paper-based newspapers.

Curating Skills (RQ4)

The result of an ANCOVA test indicated that there was a significant mean difference in curating skills among the four groups (F[3,603] = 35.76, p < .001). Pairwise comparison tests showed that opinion givers/ seekers (M = 3.40) had significantly higher curating skills than the three other groups (Table 6).

Political Participation (RQ5, 6)

With regard to offline political participation, the main effect of four groups was not significant (F[3, 603] = 1.71, ns). However, pairwise comparison tests revealed that opinion givers/seekers (M = 1.92) and opinion givers (M = 1.88) were more likely to participate in offline political activities than opinion seekers (M = 1.61) and nondiscussants (M = 1.36). The four groups had a significant main effect on SNS political participation (F[3, 603] = 18.04, p < .001). Pairwise comparison tests showed that opinion givers/seekers (M = 2.81) were significantly more likely to participate in SNS political activities than all three other groups (opinion givers: M = 2.16, opinion seekers: M = 1.76, nondiscussants: M = 1.37) (Table 6).

Discussion

This study revisits concepts of opinion leadership and followership in the current social and media environments. Based on the survey data collected in Seoul, the study first found that opinion leadership and followership are conceptually distinct but are interrelated and complementary. We categorized individuals into four groups depending on their varying combinations of orientations toward opinion leadership and followership: opinion givers/seekers, opinion givers, opinion seekers, and nondiscussants. The study results indicate that opinion givers/seekers are distinct from other groups in several indicators of opinion leadership. A brief summary of the results is presented below.

Opinion givers/seekers are significantly younger than opinion givers, and a higher percentage of males than females are in the group. Opinion givers/seekers are connected to a broader range of media types compared to opinion givers and nondiscussants. They have higher connections to magazines and books than those in the other groups. Opinion givers/seekers have a significantly higher level curating skill and more experiences in SNS political participation than those in the other three groups, although their level of offline political participation is similar to that of opinion givers.

Opinion givers are the oldest group and have a higher percentage of males than do opinion seekers and nondiscussants. They have lower media connectedness for the overall media and for specific media types than do opinion givers/seekers. Their curating skills are significantly lower than opinion givers/seekers and are similar to opinion seekers. Their offline political participation is at the same level as opinion givers/seekers, but their SNS political participation is significantly lower than opinion givers/seekers.

Opinion seekers are the youngest group and have about twice the percentage of females as males. Their overall media connectedness is generally high, with the same level as opinion givers/seekers. In particular, they have high connectedness to portal sites, blogs, and online video sites. Opinion seekers have a moderate level of curating skills, lower than opinion givers/seekers but the same level as opinion givers. Their political participation both online and offline is less active than opinion givers/seekers and opinion givers.

Nondiscussants are slightly older than opinion givers/seekers but younger than opinion givers, and they have a slightly higher percentage of females than males. Nondiscussants have the lowest level of connections to the overall media and specific media types. They also have significantly lower curating skills than those in the three other groups. For both offline and SNS political participation, nondiscussants have significantly lower participation than opinion givers/seekers and opinion givers, but the same level as opinion seekers.

Implications of the Results

The positive correlation between opinion giving and seeking and the formation of four groups support the argument that opinion giving and seeking are not competing or mutually exclusive. The current study is one of the few recent attempts to distinguish individuals who are both opinion givers and seekers (opinion givers/seekers) from those who are only givers (opinion givers). In an urbanized and modernized social environment in which individuals are connected to multiple communities and are exposed daily to new information from multiple sources, it has become more crucial for opinion leaders to be able to actively and discerningly seek messages coming from various sources. That is, opinion seeking has become a more essential condition to being an opinion leader in the current communication environment. This is supported by the result that opinion givers/seekers are also the most connected to media. On the other hand, opinion givers were less connected to varying types of media, including both new and old media, than both opinion givers/seekers and opinion seekers.

The results of the current study demonstrate that opinion givers/seekers, rather than opinion givers, demonstrate characteristics of what past studies have described as opinion leaders--they actively follow the media and actively engage in interpersonal communication and civic activities. Our study shows that these opinion givers/seekers are not within the confines of the one-dimensional two-step flow of news and opinions. The possible changing nature of opinion leadership is also supported by the younger age of opinion givers/seekers when compared to opinion givers. The younger generation who grew up in the diversified media environment may have been trained to gather varied information and opinions to serve as opinion leaders. On the other hand, opinion leaders in the older generation may feel more comfortable using a limited number of media (e.g., the mainstream media) to consider themselves opinion leaders.

One of the intriguing results of the current study is the high percentages of opinion givers/seekers and opinion givers. More than one-third (36.5%) of the respondents belong to opinion givers/seekers, and about one-fourth (23.7%) belong to opinion givers. The combined percentage of the two groups is higher than past studies' percentages of opinion leaders, which ranged from 20% to 32% (e.g., Katz & Lazarsfeld, 1955; Robinson, 1976; Troldahl & van Dam, 1965). In addition to this study's methodological difference of distinguishing four groups of opinion givers and seekers, the large percentages of opinion givers/seekers and opinion givers may reflect the ways in which people interact in the current media environment. User-generated content has become prevalent in many interactive platforms, such as social media and blogs, and barriers to express one's opinion and to interact with others have become lower. In such a communication environment, more people may identify themselves as opinion givers. The results that the largest percentage of respondents are opinion givers/seekers and that they tend to be younger and more connected to new media platforms support this argument. The results call for more systematic measurements of opinion leadership to assess the validity of traditional opinion leadership measurements and to examine whether the meaning of opinion leadership is changing in the current communication environment.

With regard to individuals' connections to specific media types, none of the pairwise comparisons of the four opinion groups was significant for traditional mass media: television and newspapers. That is, when socioeconomic and demographic factors and political interest are held constant, opinion givers/seekers did not necessarily connect to mainstream mass media more than others. On the other hand, opinion givers/seekers did connect to specialized media (magazines, books) and online sites (portal sites, blogs, and video sites) more than other groups. These results indicate that opinion seeking in the current media environment goes beyond connecting to mass media. Opinion givers/seekers connect not only to mass media but also to other types of media to stay on top of news and information. As access to media becomes ubiquitous and people connect to television and newspaper content from multiple devices (not only TV sets but also PCs and portable media), opinion givers/seekers are more likely to be those who skillfully manage their connections to various types of media as a means of staying on top of what is happening in the world.

A new measure, curating skills, was introduced in the current study. Curating skills have become important criteria of media literacy in the current information and communication environment (Jenkins, 2007). Curating skills primarily are the ability to critically select important, useful, and reliable information (Kelley, 1988) and the ability to use interactive network channels to share information with others who would benefit from it. Recent studies have emphasized that curating skills have become crucial for opinion leaders, particularly those who lead opinions online (Rhee et al., 2007b; Schafer & Taddicken, 2015). The present study suggests that those who both give and seek opinions also have high curating skills, which are important components of literacy in the new media environment.

Our results regarding the relationship between political opinion giving and seeking types and political participation exhibited different patterns depending on whether respondents used online or offline media. For offline participation, both opinion givers/seekers and opinion givers were more likely to participate in political activities than opinion seekers and nondiscussants. In the case of online participation, on the other hand, opinion givers/seekers were more likely to participate in SNS political activities than the other three groups. While being consistent with the findings of the previous studies that suggest direct effects of opinion leadership on offline political participation (e.g., Robinson, 1976; Shah & Scheufele, 2006; Troldahl & van Dam, 1965), our results go further to suggest that on social network services, opinion givers/seekers are more likely to actively participate in political activities than opinion givers and opinion seekers. This result has important implications for political communication. It is feasible for those who give more than they seek opinions to engage in political activities offline, such as voting, participating in political demonstrations, and expressing opinions to the government or to a political party. On the other hand, for one to be politically active on SNSs, such as by posting political messages, participating in politics-related groups, and visiting pages of candidates or parties, being both an opinion giver and seeker likely promotes active participation. The present study does not differentiate opinion leadership in offline and online platforms, but future studies may build upon this study to further explore ways in which opinion givers/seekers and opinion givers exhibit different leadership patterns in offline and online networks and how these patterns are related to political participation.

Limitations and Future Implications

Several limitations of the study should be noted. First, the study was based on cross-sectional survey data. Although the relationships between variables in our research questions were based on theories and past research, any claims of causality should be taken with caution. Second, a cluster analysis was used to derive four opinion relations groups. Based on the current study, future studies can develop measurements unique to each of the four types rather than combining existing measurements of opinion leadership and followership. In addition, this study can be a basis on which to formulate concrete hypotheses to further test differences among the four groups. For example, future studies can explore whether mediating effects of curating skills and media connections exist between the four opinion leadership and followership types and political participation.

References

Ball-Rokeach, S. J. (1985). The origins of individual media-system dependency: A sociological framework. Communication Research, 12(4), 485-510. doi.org/10.1177/009365085012004003

Bennett, W. L., & Manheim, J. B. (2006). The one-step flow of communication. The ANNALS of the American Academy of Political and Social Science, 608(1), 213-232. doi.org/10.1177/0002716206292266

Bertrandias, L., & Goldsmith, R. E. (2006). Some psychological motivations for fashion opinion leadership and fashion opinion seeking. Journal of Fashion Marketing and Management: An International Journal, 10(1), 25-40. doi.org/10.1108/13612020610651105

Carah, N. (2014). Curators of databases: Circulating images, managing attention and making value on social media. Media International Australia, 150, 137-142.

Childers, T. L. (1986). Assessment of the psychometric properties of an opinion leadership scale. Journal of Marketing Research, 23(2), 184-188.

DeFleur, M. L., & Ball-Rokeach, S. J. (1990). Theories of mass communication (5th ed.). Boston, MA: Allyn & Bacon.

Flynn, L. R., Goldsmith, R. E., & Eastman, J. K. (1996). Opinion leaders and opinion seekers: Two new measurement scales. Journal of the Academy of Marketing Science, 24(2), 137-147. doi.org/10.1177/0092070396242004

Hogan, B., & Quan-Haase, A. (2010). Persistence and change in social media: A framework of social practice. Bulletin of Science, Technology and Society, 30(5), 309-315.

Jenkins, H. (2007). Media literacy: Who needs it? In T. Willoughby & E. Wood (Eds.), Children's learning in a digital world (pp. 15-39). Oxford, UK: Blackwell Publishing. doi.wiley.com/10.1002/9780470696682.ch1

Katz, E., & Lazarsfeld, P. F. (1955). Personal influence: The part played by people in the flow of mass communications. New York, NY: Free Press.

Kelley, R. (1988, November). In praise of followers. Harvard Business Review, 66, 142-148.

Kim, S.-E. (2006). Who is opinion leader? An exploring study on opinion leadership and media use. Korean Journal of Journalism & Communication Studies, 50(1), 91-114.

Lassar, W. M., Manolis, C., & Lassar, S. S. (2005). The relationship between consumer innovativeness, personal characteristics, and online banking adoption. International Journal of Bank Marketing, 23(2), 176-199. doi.org/10.1108/02652320510584403

Lazarsfeld, P. F., Lerelson, B., & Gaudet, H. (1944). The people's choice: How the voter makes up his mind in a presidential campaign. New York, NY: Columbia University Press.

Lee, A. L. (2010). Who are the opinion leaders? The physicians, pharmacists, patients, and direct-to-consumer prescription drug advertising. Journal of Health Communication, 15(6), 629-655. doi.org/10.1080/10810730.2010.499594

Lee, W., Cha, M., & Yang, H. (2011). Network properties of social media influentials: Focusing on the Korean Twitter community. Journal of Communication Research, 48(2), 44-79.

Ma, L., Lee, C. S., & Goh, D. H.-L. (2014). Understanding news sharing in social media: An explanation from the diffusion of innovations theory. Online Information Review, 38(5), 598-615. doi.org/10.1108/OIR-10-2013-0239

Marwick, A. E., & boyd, d. (2011). I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. New Media & Society, 13(1), 114-133. doi.org/10.1177/1461444810365313

Merton, R. K. (1949). Patterns of influence: A study of interpersonal influence and communications behavior in a local community. In P. F. Lazarsfeld & F. N. Stanton (Eds.), Communication research, 1948-9 (pp. 180-219). New York, NY: Harper and Brothers.

Nisbet, E. C. (2006). The engagement model of opinion leadership: Testing validity within a European context. International Journal of Public Opinion Research, 18(1), 3-30. doi.org/10.1093/ijpor/edh100

Pew Research Center. (2014). Social networking fact sheet. Retrieved from http://www.pewinternet.org/fact-sheets/social-networking-fact-sheet/

Powell, L., & Shelby, A. N. (1985). Para-political opinion leaders as audience: Communicative goals and media usage. Communication Research Reports, 2(1), 135-140.

Putnam, R. D. (2001). Bowling alone: The collapse and revival of American community. New York, NY: Simon & Schuster.

Quan-Haase, A., & Martin, K. (2013). Digital curation and the networked audience of urban events: Expanding La Fiesta de Santo Tomas from the physical to the virtual environment. International Communication Gazette, 75(5-6), 521-537. doi.org/10.1177/1748048513491910

Rhee, J. W., Kim, E.-M., & Kim, H. (2007a). Exploring online opinion leadership. Korean Journal of Journalism & Communication Studies, 51(3), 358-384.

Rhee, J. W., Kim, E.-M., & Kim, H. (2007b). Exploring online opinion leadership: A validity test of the concept in the digital age. In Conference Papers: International Communication Association. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&db=ufh&AN=26950169 &site=ehost-live

Rieder, R. (2010). A matter of interpretation. American Journalism Review, 32(4), 2.

Robinson, J. P. (1976). Interpersonal influence in election campaigns: Two step-flow hypotheses. Public Opinion Quarterly, 40(3), 304. doi.org/10.1086/268307

Rogers, E. M. (1962). Diffusion of innovations. New York, NY: Free Press.

Rogers, E. M., & Cartano, D. G. (1962). Methods of measuring opinion leadership. Public Opinion Quarterly, 26(3), 435. doi.org/10.1086/267118

Samph, T. (2012, April 26). The rise of the social media curator. Social Media Today. Retrieved from http://www.socialmediatoday.com/content/rise-social-media-curator

Schafer, M. S., & Taddicken, M. (2015). Mediatized opinion leaders: New patterns of opinion leadership in new media environments? International Journal of Communication, 9, 960-981.

Scheufele, D. A., & Shah, D. V. (2000). Personality strength and social capital: The role of dispositional and informational variables in the production of civic participation. Communication Research, 27(2), 107-131. doi.org/10.1177/009365000027002001

Shah, D. V., & Scheufele, D. A. (2006). Explicating opinion leadership: Nonpolitical dispositions, information consumption, and civic participation. Political Communication, 23(1), 1-22. doi.org/10.1080/10584600500476932

Shoham, A., & Ruvio, A. (2008). Opinion leaders and followers: A replication and extension. Psychology and Marketing, 25(3), 280-297. doi.org/10.1002/mar.20209

Tan, P.-N., Stenbach, M., & Kumar, V. (2006). Introduction to data mining. Boston, MA: Pearson Addison Wesley.

Troldahl, V. C., & van Dam, R. (1965). Face-to-face communication about major topics in the news. Public Opinion Quarterly, 29(4), 626. doi.org/10.1086/267365

Turcotte, J., York, C., Irving, J., Scholl, R. M., & Pingree, R. J. (2015). News recommendations from social media opinion leaders: Effects on media trust and information seeking. Journal of Computer-Mediated Communication, 20(5), 520-535. doi.org/10.1111/jcc4.12127

Weimann, G. (1991). The influentials: Back to the concept of opinion leaders? Public Opinion Quarterly, 55(2), 267. doi.org/10.1086/269257

Weimann, G. (1994). The influentials: people who influence people. Albany, NY: State University of New York Press.

Wellman, B. (2001). Physical place and cyber place: The rise of networked individualism. International Journal of Urban and Regional Research, 25(2), 227-252.
Appendix 1. Measurement Items for Political
Opinion Giving, Seeking, and Curating Skills.

                                            Mean (Standard
       Items                                  deviation)

OG     1. People around me listen to         2.88 (.889)
       my opinion on political
       affairs.

       2. People around me sometimes         2.64 (.915)
       make political decisions based
       on my opinion.

       3. I often persuade people            2.50 (1.000)
       around me to hold the same
       political opinion as me.

       4. I influence others'                2.58 (.961)
       thoughts on political affairs.

       5. People around me rely on my        2.44 (.897)
       advice on political affairs.

OS     1. I tend to listen to other          2.88 (.886)
       people's advice in making
       decisions on political
       affairs.

       2. I listen to other people's         3.11 (.908)
       opinions when making decisions
       on political affairs.

       3. I tend to ask people around        2.66 (.939)
       me before making decisions on
       political affairs.

       4. I prefer to hear other             3.08 (.943)
       people's opinions before
       making decisions on political
       affairs.

       5. I feel comfortable making          2.58 (.872)
       decisions on political affairs
       after hearing other people's
       opinion.

       6. I tend to make decisions on        2.58 (.872)
       political affairs after
       receiving other people's
       advice.

CS     1. When I want to refer to            3.18 (.722)
       other people's opinions on a
       particular issue, whom I refer
       to depends on the nature of
       the issue.

       2. When I want to make up my          3.34 (.791)
       mind on a particular issue, I
       try to listen to diverse
       opinions from different
       people.

       3. When I listen to someone's         3.31 (.826)
       opinion, I consider whether
       the person is an expert on a
       particular issue.

       4. When I listen to someone's         3.50 (.783)
       opinion, I consider whether
       the person is talking
       sincerely about the issue.

       5. I can appropriately select         3.19 (.783)
       people whom I want to refer to
       based on types of issues.

       6. I sometimes recommend to           2.80 (.860)
       others whose opinions to refer
       to for particular political
       issues.

       7. I can easily identify              2.94 (.811)
       people who will give me
       opinions on particular issues.

       8. If someone's opinion on a          2.95 (.848)
       political issue was helpful, I
       express support for the
       person.

       9. When I hear a helpful              3.02 (.867)
       opinion on a particular
       political issue, I share it
       with others.

       10. If I am not satisfied with        2.89 (.835)
       someone's opinion on a
       particular political issue, I
       explicitly express
       dissatisfaction.

Note: OG: opinion giving; OS: opinion seeking; CS: curating skills.


JOO-YOUNG JUNG

International Christian University, Japan

YONG-CHAN KIM

Yonsei University, Korea

Joo-Young Jung: jungjy@icu.ac.jp

Yong-Chan Kim: yongckim@yonsei.ac.kr Date submitted: 2016-01-05

(1) The six items were derived from the two items used in Katz and Lazarsfeld (1955), six items developed by Rogers and Cartano (1962), and six items included in Flynn et al.'s (1996) final scale. Items were selected to fit the topic (political affairs) and context (South Korea). Childers's (1986) scale was not used because it specified "friends and neighbors" in items, which did not align with the scope of the current study to measure opinion leadership in a wider communication environment.
Table 1. Principal Component Analysis of Opinion Giving and Seeking.

         Opinion giving   Opinion seeking   Communality

OG1           .491             -.262           .310
OG2           .816             .167            .693
OG3           .838             .220            .750
OG4           .784             .219            .662
OG5           .858             .246            .796
OG6           .830             .282            .768
OS1           .155             .797            .659
OS2           .200             .752            .605
OS3           .159             .803            .670
OS4           .130             .770            .609
OS5           .116             .806            .663
OS6           .100             .825            .691

Table 2. Bivariate Correlations of Variables.

                                      1         2         3

1. opinion giving
2. opinion seeking                    .41 **
3. income                             .11 **    .05
4. education                          .08 *     .05       .26 **
5. age                                .09 *     -.09 *    .18 **
6. political interest                 .55 **    .24 **    .07
7. overall media connectedness        .24 **    .23 **    .12 **
8. curating skills                    .56 **    .47 **    .15 **
9. offline political participation    .31 **    .12 **    .14 **
10. SNS political participation       .51 **    .30 **    .08

                                      4         5         6

1. opinion giving
2. opinion seeking
3. income
4. education
5. age                                .09 *
6. political interest                 .07       .14 **
7. overall media connectedness        .08 *     .18 **    .22 **
8. curating skills                    .07       .13 **    .51 **
9. offline political participation    .03       .12 **    .31 **
10. SNS political participation       .10 *     -.04      .40 **

                                      7         8         9

1. opinion giving
2. opinion seeking
3. income
4. education
5. age
6. political interest
7. overall media connectedness
8. curating skills                    .34 **
9. offline political participation    .19 **    .37 **
10. SNS political participation       .25 **    .40 **    .40 **

Note: * p < .05. ** p < .01.

Table 3. A Cluster Analysis of Opinion Giving and Seeking.

                                    Clusters

               Opinion
               givers/      Opinion      Opinion     Nondiscussants
               seekers       givers      seekers

Opinion          3.32         2.83         2.03           1.56
giving
(mean) (a)

Opinion          3.33         2.34         3.21           1.80
seeking
(mean) (b)
n (%)         228 (36.5)   148 (23.7)   142 (22.7)     107 (17.1)

(a) The mean difference of the four groups in opinion giving was
statistically significant: F = 510.27, p < .001. According to a
Tukey HSD test, all pairwise mean differences are significant.

(b) The mean difference of the four groups in opinion seeking was
statistically significant: F = 427.12, p < .001. According to a
Tukey HSD test, all pairwise mean differences are significant.

Table 4. Percentages for Gender in the Four Opinion Groups.

                     Opinion
                     givers/   Opinion   Opinion
                     seekers   givers    seekers   Nondiscussants

Gender      Male      41.8      27.0      15.8          15.4
            Female    31.2      20.4      29.6          18.8

Note: [chi square] [3] = 21.95, p < .001.

Table 5. Mean Comparisons of Overall and Specific Media
Connectedness Among the Four Groups of Opinion Givers and Seekers.

                     Opinion
                     givers/    Opinion    Opinion     Nondiscussants
                     seekers    givers     seekers

Overall media        4.19 (a)   3.12 (b)   3.78 (ab)   2.82 (b) (c)
connectedness
Magazines            3.04 (a)   2.71 (b)   2.64 (b)    2.34 (c)
Books                3.31 (a)   3.03 (b)   3.01 (b)    2.66 (c)
Portal sites         3.67 (a)   3.41 (b)   3.72 (a)    3.33 (b)
Blogs                3.19 (a)   2.75 (b)   3.00 (ab)   2.38 (c)
Online video sites   3.29 (a)   3.04 (b)   3.15 (ab)   2.75 (c)
Radio                3.19 (a)   2.98 (a)   3.07 (a)    2.77 (b)
Television           3.65       3.48       3.82        3.61
Online news sites    3.61       3.48       3.55        3.35
Paper-based          3.28       3.14       3.21        2.88
newspapers

Note: Letters (a), (b), and (c) indicate significant mean
differences in pairwise comparisons (p < .05). No pairwise
comparison was significant for television, online news sites, or
paper-based newspapers. Range: 1-5.

Table 6. Mean Comparisons of Curating Skills, Offline Political
Participation, and SNS Political Participation Among the Four Groups
of Opinion Giving and Seeking.

                    Opinion    Opinion    Opinion    Nondiscussants
                    givers/     givers    seekers
                    seekers

Curator skills      3.40 (a)   3.10 (b)   3.11 (b)      2.58 (c)

Offline political   1.92 (a)   1.88 (a)   1.61 (b)      1.36 (b)
participation

SNS political       2.81 (a)   2.16 (b)   1.76 (c)      1.37 (d)
participation

Note: Letters (a), (b), and (c) indicate significant mean
differences in pairwise comparisons (p <.05). Range: curator skills
1-4.6; offline political participation 1-7; SNS political
participation 1-7.
COPYRIGHT 2016 University of Southern California, Annenberg School for Communication & Journalism, Annenberg Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Jung, Joo-young; Kim, Yong-chan
Publication:International journal of communication (Online)
Article Type:Report
Geographic Code:9SOUT
Date:Aug 1, 2016
Words:9000
Previous Article:Reassessing "whose story wins": the trajectory of identity resilience in narrative contests.
Next Article:Je Suis Charlie? The framing of in-group transgression and the attribution of responsibility for the Charlie Hebdo attack.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters