Effects of social media users’ attitudes on their perceptions of the attributes of news agency content and their intentions to purchase digital subscriptions

Abstract

This study analyzes how social media users’ attitudes influence their perceptions regarding the attributes of news agency content and their intentions to purchase digital subscriptions. Their attitudes toward production activities influence their purchasing intentions and affect their use time of social media and news. Furthermore, their attitudes toward production activities influence their news perceptions and, subsequently, their perceptions of the attributes of news agency content. Because of these variables, their attitudes toward production activities affect their purchasing intentions. Social media users’ attitudes toward use activities influence their intentions to purchase digital subscriptions through their perceptions of the attributes of the content. The more often people use social media, the more they are likely to consume news. However, news use time does not influence purchasing intentions. Daily use of news agency content is controlled.

Keywords

Attitudes news agency content news perceptions produsers purchasing intentions social media users structural equation modeling

Attitudes of social media users

Social media websites attract more users than news websites do. According to Google (2012), Facebook ranked first on the list of the top 1000 websites in the world on 18 January 2012, while YouTube ranked second and Yahoo.com ranked third. Twitter ranked 15th. There were no news organizations in the top 20 websites. BBC ranked 48th and CNN.com ranked 60th. NYTimes.com (124th), Reuters.com (215th), WSJ.com (289th), and other major news websites fell even further behind. The popularity of social media is based on the ability of users to produce and distribute content.

User-generated content (UGC) is often produced via convenient digital platforms (Proulx et al., 2011). Wikipedia and OpenStreetMap are examples of platforms where both production and usage occur (Bruns and Schmidt, 2011), and the activities stem from open-content communities that are involved in the decision-making process (Wofford, 2012). The participants include creators, spectators, and inactives (Van Dijck, 2009). British news websites have promoted UGC initiatives with “Your stories” and “Have your say” segments in which journalists upload questions to which readers can post written replies (Hermida and Thurman, 2008; Thurman, 2008). Forums, chat rooms, and Wikis exemplify the context of users as producers, while Really Simple Syndication (RSS) feeds, grading, and polls capture the context of users as consumers (Jönsson and Örnebring, 2011).

People who are engaged in producing and consuming are categorized as “produsers”; they are concerned with collaboration and the continuous production, distribution, and consumption of improved content (Bruns, 2008). Produsers include numerous participants who use digital platforms to cooperate through unplanned encounters in the creation and development of diverse types of content (Bruns and Schmidt, 2011). A thorough investigation is needed of the complex relationship between social media users’ attitudes toward news content, their perceptions of news content, and their purchasing intentions. Although many news agencies aim to distribute large amounts of digital information through social media, few researchers have focused on social media users’ attitudes toward the content of news agencies. As the structure of a society becomes increasingly complex, news agencies “work full-time to access, select and filter, produce and edit news, which is then distributed via the media to network members” (Domingo et al., 2008: 329). The Chinese news agency Xinhua provides information through services similar to Facebook to attract new users (Kim, 2010). This attempt illustrates how news agencies are engaging in efforts to understand the attitudes of social media users. In addition, news agencies are seeking to broaden the scope of their customer base by attracting individuals from other news organizations and corporations.

As practical implications, this study’s findings call for news agencies to understand the attitudes of social media users toward their content. This is particularly relevant to news agencies, which are losing their position as the dominant provider of breaking stories because other news outlets instantly release stories on Twitter and other social media (Han, 2013). This situation calls for news agencies to understand the implications of social media for news production. News production is decentralized; thus, it involves collaboration with and participation by news users (Bruns, 2003). This study reveals the implications of social media users’ attitudes toward the news produced by news agencies. More importantly, this study’s findings will enable news agencies and their journalists to reconsider their roles as news producers and distributors.

Background information regarding Korea’s websites and the Yonhap News Agency

South Korea ranked first in the penetration of high-speed Internet (Cutler, 2010). Korea’s social media thrives on high transfer speed and a digital structure that is popular among users. On the list of the top 100 websites in Korea (Google, 2010), Naver (first), Daum (second), Nate (third), and cyworlds (fourth) are the leaders. Facebook ranked 14th. The news outlet Newsis ranked 37th, and the Yonhap News Agency (YNA) did not rank among the top 100 websites. Information users seem to prefer social media sites to the YNA web site.

Nonetheless, YNA is a dominant news provider. YNA has service contracts with 35 daily newspapers, 27 broadcasting stations, 30 local daily newspapers, 37 Internet-only newspapers, 11 portals, 23 specialized news websites, and 11 public organizations and associations (Yonhap News Agency, 2013). On 29 May 2003, Korea’s parliament passed a law promoting the services of news agencies to reinforce their freedom of expression, independence, and public responsibilities, and on 17 March 2004, the government introduced relevant regulations (Kim, 2009). Thus, YNA was able to engage in a service contract with the government and establish its status as the national news agency. The stakeholders of the news agency consist of the Korea News Agency Commission (30.77%), the Korean Broadcasting System (27.78%), the Munhwa Broadcasting Corporation (22.30%), and various newspapers (19.15%) (Chung, 2012).

YNA has three user categories. The first group includes national daily newspapers, broadcasting networks, and local daily newspapers. The news agency provides these news organizations with stories, photographs, and graphics. The second group includes corporations that receive customized financial information. Infomax, for example, provides financial news to banks and companies. The third group includes visitors to YNA’s homepage and subscribers to its news applications. While the first and second groups are traditional user groups, the third group includes produsers.

Prosumers versus produsers

Two terms have been circulating among researchers regarding social media users: prosumers and produsers. First, prosumers collectively produce and consume news content by adopting a free-writing approach and by using websites as sources (Kim et al., 2007). Thus, news is a process that allows for the active involvement of any interested person. The concept of prosumers originates from Alvin Toffler. Consumers contribute not only to revenues of producers but also to item design and market strategies through their feedback (Toffler, 1990). Consumers are involved in the design process and share information with producers. Thus, prosumers represent the integration of consumers and producers. Prosumers continue to give feedback to the producers; they remain engaged in the value chain—the linear progression from producers to distributors to consumers (Bruns, 2008). Prosumers prefer producing goods and services for their own consumption (Kotler, 1986). They are positioned in the middle of a production-to-consumption continuum, where production and consumption are roughly balanced (Ritzer et al., 2012).

In this sense, prosumers cannot control production because the integrated feedback loop from consumers to producers occurs under the parameters determined by corporate producers (Bruns, 2008). The relationship between producers and prosumers is one directional, and the concept of prosumers has a commercial connotation. The term ‘prosumers’ does not represent social media users who give feedback and produce content by collaborating with journalists or other users.

Produsers are not passive receivers or consumers (Bruns, 2003). With produsers, the line between producers and users is unclear; thus, users are transformed into produsers. News readers have become authors, and journalists have become readers (Robinson, 2007). Produsers relate to Web 3.0 by focusing on the creators; prosumers relate to Web 2.0 by focusing on documents; and consumers relate to Web 1.0, which emphasizes the computers (Grinnell, 2009). This categorization recognizes Web 3.0 as the Giant Global Graph, a system that incorporates the social connections of all networks into a single metanet.

The concept of produsers is characterized by the existence of a distributed but coordinated community; the absence of distinct categories of producers, distributors, and consumers; the endless string of users acting as users or producers; collaboration; and participation (Bruns, 2008). Produsers perform the role of gatewatching by writing stories and offering comments related to the news. Produsers focus on their emotions, empathize with community members, and demonstrate their capabilities of managing news technology while producing content (Sundet and Ytreberg, 2009). They are knowledgeable about new media technology, and they use their abilities and talents for producing and controlling their creative and informative works (Van Dijck, 2009). Thus, produsers distribute created content and share knowledge and culture with other users; in fact, their contributions may add value to the content.

Given the concept of produsers, this study relies on two theoretical frameworks: ambient journalism and expectancy-value theory. The following section contains a review of relevant scholarly works from these areas.

Social media users and ambient journalism

Social media allows people to redefine what news is by engaging in the activities of producing, distributing, and consuming news. This phenomenon is called “news as social experience” because people have access to multiple channels of news production and distribution (Hermida, 2010a: 2). Participatory culture has emerged in opposition to passive media spectatorship, as news producers and consumers are not separate; they are participants who interact with each other (Jenkins, 2006).

As complex social networks have emerged and communication technology has allowed for increased connection, society members can engage in the institutionalized communication that previously belonged only to news agencies and media; this is called “participatory journalism” (Domingo et al., 2008). For example, microblogging is a way of broadcasting with text messaging, which allows users to share information in less than 200 characters via multiple channels, such as websites, third-party applications, or mobile tools (DeVoe, 2009). As part of social media, microblogging is defined as “a new media technology that enables and extends our ability to communicate, sharing some similarities with broadcast” (Hermida, 2010b: 298). Social media, including microblogging, are weakening the gatekeeping role of journalists (Hermida, 2010b).

Social media provide an ambient information system that delivers content to spaces where users are positioned (Hermida, 2010b). There is a category of users who produce news content and one that includes users who distribute created news content to others. A third type includes users who combine production with distribution and engage the public sphere through discussion and participation (Jeon, 2012). These activities are considered a form of journalism called “ambient journalism,” which means “an awareness system that offers diverse means to collect, communicate, share, and display news and information, serving diverse purposes” (Hermida, 2010b: 301).

Under awareness systems, news content is delivered instantly to users. A person near an event can deliver breaking stories on Twitter (Sankaranarayanan et al., 2009). Retweeting is an example of a distribution activity. Retweeting news stories distributes content quickly and broadly to other users. Social media users can simultaneously combine production, distribution, and use activities. For instance, content created by the user community is likely to be available to all participants. This availability can lead to the sharing of content for communal purposes. Thus, users evaluate news stories distributed through social media by reading comments from other users, and they estimate the credibility of sources in the stories (Bruns, 2008). Evaluation and source estimation can contribute to improvement in the quality of stories. Facebook provides an interesting case exemplifying social media users’ activities. When Facebook tried to change its terms of service in February 2009, social media users resisted, leading Facebook to incorporate users’ demands in its new terms (Grinnell, 2009).

News users become involved in story selection by indicating the stories they enjoy reading and having a say in how journalists collect information and select news topics (Robinson, 2007). Users’ comments and online conversations contribute to a greater understanding of events and a broader view of the news (Jönsson and Örnebring, 2011). Thus, the attitudes that social media users form while engaging in the activities of news selection can influence their perceptions and evaluations of content from news agencies. People consider multiperspectival news to deal with a nation and to contain the perspectives of citizens rather than of leaders (Gans, 1979). News communicates the efficiency of governmental policies and represents public opinion while providing useful information about the government and related departments or agencies. Representative news is important because the media environment has changed (Gans, 2011).

Expectancy values, attributes of news agency content, and purchasing intention

How people perceive the probable outcomes of a behavior influences their intention to perform the behavior (Swanson, 1987). This relationship is based on the expectancy-value theory, which explains people’s behavioral intention as the perceived probability that an entity has a specific trait (Rayburn, 1996). People’s expectancy-value judgments lead them to form intentions to use certain media (Ajzen and Fishbein, 2005). Furthermore, the expectancies that people have tend to predict how satisfied they will be with television news programs (Palmgreen and Rayburn, 1985).

In this sense, this study reasonably expects social media users’ perceptions of news to influence how they evaluate news agency content, as these perceptions are linked to people’s expectations regarding news. For instance, when people expect a news story to be positive, they are likely to have a positive perception of the story (Cooper et al., 2001). More importantly, people’s expectations and attitudes toward media attributes influence their decision to consume the media (Leung and Wei, 1999). In other words, people’s expectancies affect their intentions to engage certain media (Ajzen and Fishbein, 2005). When people form favorable attitudes toward a free paper and the value of daily news is obvious, they read the free paper more frequently than before (Tsao and Sibley, 2004).

In general, people have a high level of resistance to the idea of purchasing digital subscriptions because they prefer free news portals (Kim and Kang, 2010). However, when the quality and relevance of news is considered, people have different attitudes toward subscriptions. They express their willingness to purchase digital news content for necessary and high-quality information or for coverage of unique interests (Lee and Lim, 2012). At the same time, individuals are willing to purchase content not provided by other websites. Paid content tends to be unique; examples include popular columns, crossword puzzles, and financial analyses (Herbert and Thurman, 2007). Therefore, it is reasonable to predict that social media users’ perceptions of news agency content can affect their intentions to purchase the content. In particular, if the YNA’s content is considered clear, accurate, and fair, social media users may intend to purchase the content in order to obtain necessary information or coverage based on certain values.

Proposed theoretical model

The review of relevant research underpins the notion that attitudes of social media users toward production, distribution, and use activities influence their perceptions of news. Furthermore, users’ perceptions of news can influence their perceptions of the attributes of news agency content, leading to an increase in their intention to purchase digital subscriptions.

In addition, this study considers the possibility that increased usage of social media may lead to increased engagement with news. This notion is based on the effects of time replacement and time enhancement. People who use the Internet and other information technologies frequently spend their time in leisure activities such as watching movies, volunteering, or attending sporting events (Robinson, 2011). The implication is that the more time a user spends with information technologies, the more time they have to engage in other social activities; in other words, usage time does not replace such activities. Internet users engage in social activities more frequently than do nonusers (Robinson and De Haan, 2006). Therefore, this study also investigates whether a similar time-enhancement effect is present for the usage times of social media and news. This study proposes the following theoretical model (Figure 1).

Figure 1.

Proposed theoretical model.

H1. Respondents’ attitudes toward production activities on social media will influence their perceptions of news.

H2. Respondents’ attitudes toward production activities on social media will have indirect effects on their intentions to purchase digital subscriptions to YNA through their perceptions of news and of the attributes of YNA content.

H3. Respondents’ attitudes toward production activities on social media will have indirect effects on their daily use time of news through their daily use time of social media.

H4. Respondents’ attitudes toward distribution activities on social media will influence their perceptions of news.

H5. Respondents’ attitudes toward distribution activities on social media will have indirect effects on their intentions to purchase digital subscriptions to YNA through their perceptions of news and of the attributes of YNA content.

H6. Respondents’ attitudes toward distribution activities on social media will have indirect effects on their daily use time of news through their daily use time of social media.

H7. Respondents’ attitudes toward use activities on social media will influence their perceptions of news.

H8. Respondents’ attitudes toward use activities on social media will have indirect effects on their intentions to purchase digital subscriptions to YNA through their perceptions of news and of the attributes of YNA content.

H9. Respondents’ attitudes toward use activities on social media will have indirect effects on their daily use time of news through their daily use time of social media.

Method

Online survey

We employed Macromillembrain (http://embrain.com/company/power.asp) to conduct a large-scale online survey of social media users from 18 July 2013 to 26 July 2013. The company, which specializes in surveys, randomly selected 17,785 people from its panel of 970,000 and e-mailed survey announcements to them. Of those selected, 4194 people participated in the online survey and 3054 completed it. This study eliminated 307 people who did not answer all of the questions, 201 who skipped certain series of answers, and 247 who remained after quota sampling. Finally, 2299 responses were submitted for analysis. We divided these 2299 responses into two samples of 1149 and 1150 to cross-validate the results. We conducted a pilot survey to evaluate the reliability and validity of the questionnaires, which led us to revise wording and assess the random presentation of the questionnaires.

Measures

Respondents’ attitudes toward their activities on social media

This latent variable was measured on a 7-point scale from 1 (not at all) to 7 (very much). By using previous research (Sundet and Ytreberg, 2009; Van Dijck, 2009), we developed 20 items.

Respondents’ perceptions of news

We measured this latent variable by referencing previous research (Gans, 1979, 2011) on a 5-point scale from 1 (not at all) to 5 (very much). We used three items.

Respondents’ perceptions of the attributes of the YNA’s content

This latent variable was measured on a 7-point scale from 1 (not at all) to 7 (very much). Based on previous research (Lee and Lim, 2012), we created six items.

Respondents’ intentions to purchase digital subscriptions to a news agency

Based on previous research (Hwang, 2001), we developed three items and measured them on a 5-point scale from 1 (not at all) to 5 (very much). Tables 1 and 2 present the wording of the items. We measured respondents’ daily use time of social media (M = 54.44, standard deviation [SD] = 58.86, range = 599) by asking, “how many minutes per day do you use social media?” Furthermore, we measured their daily use time of news (M = 43.67, SD = 36.67, range = 450) by asking, “how many minutes per day do you use news stories?”

Table 1.

Reliability and validity of attitudes toward production and use activities.

Latent variables	Factor loading
Attitudes toward production (M = 13.92, SD = 5.66, Cronbach’s α = .87, H = .91)
I have produced content for social media using my own ability. (Content production)	.64
My content on social media can change the world. (Changing world)	.79
I focus on my emotions while producing content on social media. (Focus on emotion)	.90
I feel like part of the community while producing content on social media. (Community member)	.87
Attitudes toward use activities (M = 14.30, SD = 3.06, Cronbach’s α = .75, H = .75)
Social media users’ access to information enriches the content of news. (Information access)	.75
Social media users are able to check the sources of a news story. (Source checking)	.65
Social media users are able to evaluate a news story through online search. (Story evaluation)	.71

SD: standard deviation.

All standardized factor loadings are significant at p < .001. Descriptions in parentheses refer to the manifest variables in Figures 2 and 3.

Table 2.

Reliability and validity of the perceptions of attributes of YNA content and purchasing intentions.

Latent variables	Factor loading
Perceptions of attributes of YNA content (M = 26.79, SD = 6.92, Cronbach’s α = .95, H = .95)
Yonhap News Agency allows us to easily understand new events or issues. (Understandable)	.82
Yonhap News Agency provides us with accurate information about events or issues. (Accurate)	.91
Yonhap News Agency offers factual coverage rather than arguments or opinions. (Factual)	.90
Yonhap News Agency covers the supporting and opposing positions of issues in a fair manner. (Fair)	.89
Yonhap News Agency provides diverse perspectives on current issues. (Diverse)	.85
Yonhap News Agency delivers in-depth coverage of events or current issues. (In-depth)	.83
Intentions to purchase paid YNA content (M = 8.01, SD = 2.98, Cronbach’s α = .86, H = .86)
I will use the content if its quality is better than its price. (Information quality)	.81
I will pay for using the content if the content is not available through portals. (Uniqueness)	.79
I will use the content if its content is necessary to me. (Necessity)	.85
News perceptions (M = 10.63, SD = 2.73, Cronbach’s α = .79, H = .82)
News discusses what citizens want. (Citizen)	.63
News represents public opinion. (Representative)	.82
News covers stories that satisfy the needs of every segment of the public. (Gratification)	.80

YNA: Yonhap News Agency; SD: standard deviation.

All standardized factor loadings are significant at p < .001. Descriptions in parentheses refer to the manifest variables in Figures 2 and 3.

Structural equation modeling analysis of the theoretical model

We checked multivariate normality and outliers by using IBM AMOS 22, which verified normality and the absence of outliers. Following previous research (Mueller and Hancock, 2008), we used latent variable path analysis (LVPA). We assess the model fit based on the ratio of chi-square fit statistics over degrees of freedom (DF), expressed as CMIN/DF. When the sample size is large, the chi-square fit statistics are significant, indicating a rejection of the proposed model (Wheaton et al., 1977). Researchers use the relative chi-square (χ²/df) to minimize sensitivity to large samples (Hooper et al., 2008). When the sample size exceeds 900 and the values of χ²/df are less than 5, the model fit is reasonable (Al-Gahtani and King, 1999). We used the model fit indices by following the guidelines cited in previous research (Mueller and Hancock, 2008): Normed Fit Index (NFI) ≥ .90, Non-Normed Fit Index (NNFI) ≥ .95, and Comparative Fit Index (CFI) ≥ .95, Goodness of Fit Index (GFI) ≥ .90 and Standardized Root Mean Square Residual (SRMR) ≤ .08, and Adjusted Goodness of Fit Index (AGFI) ≥ .90 and Root Mean Square Error of Approximation (RMSEA) ≤ .06. The Expected Cross-Validation Index (ECVI) compares models across samples. The model containing the smallest ECVI value has the greatest chance for replication (Byrne, 2010).

We used a two-phase process consisting of a measurement model and a structural model. We identified the acceptable measurement model by using confirmatory factor analysis (CFA). Standardized residual covariances and modification indices were used for determining the fit of the CFA model. The absolute values of standardized residual covariances should be less than 2. When modification indices contain large values in Par Change, the model needs to be modified (Arbuckle, 2012). Through these procedures, we eliminated 12 items containing values of standardized residual covariances larger than 2. The CFA model fit was good (χ² = 1101.833, p = .000, df = 260, CMIN/DF = 4.238, GFI = 0.928, AGFI = 0.910, CFI = 0.948, NFI = 0.933, RMSEA = 0.053, ECVI = 1.073).

For the structural model, we compared our proposed model with the CFA model. We found that the proposed model (χ² = 1389.705, p = .000, df = 313, CMIN/DF = 4.440, GFI = 0.916, AGFI = 0.898, CFI = 0.935, NFI = 0.918, RMSEA = 0.055, ECVI = 1.324) was less acceptable than the CFA model. The differences in the chi-square fit statistics were significant (287.872, df = 53, p < .001), and the ECVI value was greater than that of the CFA model.

If improvement is possible, the causal variables could be related to respondents’ intentions to purchase digital subscriptions. We found four items having low factor loadings (.435, .525, .540, .583); therefore, we eliminated them. This modification improved the model fit (χ² = 753.283, p = .000, df = 218, CMIN/DF = 3.455, GFI = 0.945, AGFI = 0.930, CFI = 0.964, NFI = 0.950, RMSEA = 0.046, ECVI = 0.757). Modification also yielded significant differences in the chi-square fit statistics (636.422, df = 95, p < .001). The variable of distribution (M = 9.86, SD = 2.21) had low reliability (Cronbach’s α = .58) and low validity (H = 0.58) because these values were below the criterion (0.70). We eliminated the variable and analyzed the model fit, which yielded significant improvement (χ² = 622.341, p = .000, df = 181, CMIN/DF = 3.438, GFI = 0.950, AGFI = 0.936, CFI = 0.968, NFI = 0.955, RMSEA = 0.046, ECVI = 0.629). The model containing the variable of distribution and the model without the variable produced significant differences in the chi-square fit statistics (130.942, df = 37, p < .001). Thus, we used the model without the variable of distribution.

Results

Of the 2299 respondents, 52.5% were male and 47.5% were female. The age distribution of respondents was 19.4% for ages 19–29 years, 26.1% for ages 30–39 years, 28.6% for ages 40–49 years, and 26.0% for ages 50–59 years. These characteristics were compared with those of the total population of Korea. According to Statistics Korea (2012), the total population of Korea was 50,004,000, 50.1% of whom were male and 49.9% of whom were female. There were 29,158,471 people between the ages of 20 and 59 years: ages 20–29 years (22.6%), ages 30–39 years (26.7%), ages 40–49 years (28.1%), and ages 50–59 years (22.5%). Thus, the characteristics of the sample used in this study were similar to those of the general population.

We used Cronbach’s α to measure the reliability of each factor, and the H coefficient provided by previous research (Mueller and Hancock, 2008) to estimate the construct validity of each factor. Values above .70 indicated good validity. Tables 1 and 2 show good levels of reliability as well as the validity of each factor. Table 3 shows standard deviations and correlations between latent variables.

Table 3.

Standard deviations and correlations between latent variables.

	1	2	3	4	5
1. Production	1
2. Use activities	.42**	1
3. YNA attributes	.28**	.38**	1
4. Intentions to purchase YNA content	.23**	.17**	.33**	1
5. News perception	.18**	.25**	.30**	.06*	1
SD	5.92	5.24	7.59	3.54	2.80

YNA: Yonhap News Agency; SD: standard deviation.

p < .01; *p < .05.

The produsers who had experience with producing or distributing video clips, audio clips, photographs, and graphics on social media constituted 13.6% of the 2299 respondents, with 14.7% in the first sample (1149) and 12.4% in the second sample (1150). These were the respondents whose answers were above 3 on a 5-point scale from 1 (not at all) to 5 (very much).

The respondents’ experience with YNA was a control variable measured by the time they spent using YNA content on a daily basis (M = 30.15 minutes, SD = 33.23). The results in Figures 2 and 3 contain standardized regression weights based on the data. In Figure 2, the daily use time of YNA content influenced respondents’ perceptions of attributes of the news agency content (β = .13, p < .001), intentions to purchase digital subscriptions to the news agency (β = .09, p = .009), and news use time (β = .44, p < .001). These patterns are consistent in Figure 3. In Figure 2, respondents’ attitudes toward production activities explained 2.1% of the variance in news perceptions and 27% of the variance in perceptions of the attributes of YNA content. The attitudes toward production and use activities explained 4.7% of the variance in social media use time, while accounting for 13.8% of the variance in respondents’ intentions to purchase digital subscriptions to YNA. The attitudes toward production and use activities explained 28.7% of the variance in news use time.

Figure 2.

Causal relationship of attitudes toward production and use activities with YNA content.

Figure 3.

Cross-validation of causal relationship of attitudes toward production and use activities with YNA content.

Figure 3 shows the cross-validation of the results from Figure 2 using the second sample (N = 1150). The attitudes toward production activities explained 3.8% of the variance in news perceptions, and the attitudes toward production and use activities explained 26.7% of the variance in respondents’ perceptions of the attributes of YNA content. Respondents’ attitudes toward production activities accounted for 4.1% of the variance in social media use time, while their attitudes toward production and use activities explained 20.6% of the variance in intentions to purchase digital subscriptions. Their attitudes toward production activities explained 18.6% of the variance in news use time.

The testing of the hypotheses was based on comparison of the results in Figures 2 and 3. As the first hypothesis predicted, respondents’ attitudes toward production activities on social media influenced their news perceptions (β = .14, p < .001), and this effect was consistent in Figure 3. Thus, the first hypothesis was supported.

As the second hypothesis stated, respondents’ attitudes toward production activities on social media had indirect effects on their perceptions of the attributes of YNA content through news perceptions (β = .03, p = .010). Furthermore, respondents’ news perceptions influenced their perceptions of the attributes of YNA content (β = .22, p < .001), while having indirect effects on their intentions to purchase digital subscriptions to YNA through their perceptions of the attributes (β = .06, p = .010). Similar effects were found in Figure 3. The second hypothesis was supported.

As the third hypothesis predicted, respondents’ attitudes toward production activities on social media influenced their social media use time (β = .15, p < .001) and had indirect effects on their news use time through social media use time (β = .04, p = .010). The same directional effects are evident in Figure 3. Accordingly, the third hypothesis was supported. Interestingly, respondents’ attitudes toward production activities also influenced their purchasing intentions (β = .16, p < .001), which is also evident in Figure 3. This study eliminated the variable of attitudes toward distribution activities according to the modification of the proposed model, and thus, the fourth, fifth, and sixth hypotheses about respondents’ attitudes toward distribution activities were not confirmed.

The seventh hypothesis predicted that respondents’ attitudes toward use activities on social media would influence their perceptions of news. This was not supported in Figures 2 and 3. The eighth hypothesis expected that respondents’ attitudes toward use activities on social media would have indirect effects on their intentions to purchase digital subscriptions to YNA through their perceptions of news and of the attributes of YNA content. Respondents’ attitudes toward use activities influenced their perception of the attributes of YNA content (β = .42, p < .001) and had an indirect effect on their intention to purchase digital subscriptions through their perceptions of the attributes of YNA content (β = .11, p = .010). However, there were no indirect effects through news perception. Similar patterns are evident in Figure 3. Thus, the eighth hypothesis was not fully supported.

As the ninth hypothesis predicted, respondents’ attitudes toward use activities on social media influenced their use time of social media (β = .10, p = .017) and had an indirect effect on their news use time through social media use time (β = .03, p = .03). However, as shown in Figure 3, respondents’ attitudes toward their use activities did not influence their social media use time (β = .03, p = .541). Thus, the ninth hypothesis was rejected.

The magnitudes and directions of path coefficients from the first and second samples were similar, except for the two path coefficients. In Figure 3, respondents’ attitudes toward their use activities did not influence their social media use time (β = .03, p = .541), unlike in Figure 2. In Figure 3, respondents’ news use time affected their intentions to purchase digital subscriptions (β = −.11, p < .001). However, this was not the case in Figure 2 (β = −.03, p = .369). Nonetheless, the values of the two coefficients were negative, which means that the more often people used news, the less they were willing to purchase digital subscriptions. This inconsistency could be related to differences in social media use time and purchasing intentions between the two samples. For the first sample, the magnitude of social media use time was 54.44 minutes per day, which was higher than that for the second sample (51.74 minutes). In the first sample, the magnitude of purchasing intentions was 8.01, which was higher than that of the second sample (7.80). These differences could contribute to changes in path coefficients, though the differences were not significant.

However, the model fit in Figure 3 was satisfactory (χ² = 784.356, p = .000, df = 197, CMIN/DF = 3.982, GFI = 0.940, AGFI = 0.923, CFI = 0.959, NFI = 0.945, RMSEA = 0.051), similar to the model fit in Figure 2 (χ² = 781.620, p = .000, df = 197, CMIN/DF = 3.968, GFI = 0.941, AGFI = 0.924, CFI = 0.959, NFI = 0.946, RMSEA = 0.051). Figure 4 portrays the revised theoretical model.

Figure 4.

Revised theoretical model.

Discussion

This study found that social media users’ attitudes toward production activities influence their perceptions of news, as well as their intentions to purchase digital subscriptions to a news agency. Furthermore, social media users’ attitudes toward use activities influence their perceptions of the attributes of news agency content. Their attitudes toward production activities affect their social media use time and have indirect effects on their perceptions of the attributes of and their intention to purchase digital subscriptions to a news agency. Their attitudes toward use activities have indirect effects on their purchase intentions through their perceptions of these attributes. The more often users engage in social media, the more time they spend consuming news.

This study is one of the first attempts to examine the plausible link between social media users’ attitudes and their perceptions of news agency content. This study highlights the importance of understanding the attitudes that social media users form toward such content. Therefore, the attitudes of social media users are important variables. In South Korea, KakaoTalk, a mobile messaging service, had secured more than 100 million members by July 2013, following its advent in March 2010. Furthermore, mobile messaging applications, such as NateOn, KakaoTalk, Skype, Daum MyPeople, and Naver’s Line, have dominated the computer messaging market (Chung, 2013). Users normally engage in producing content on Twitter, Facebook, or YouTube, and they demonstrate diverse use activities such as accessing information, checking sources, and evaluating stories. Their attitudes toward the activities of both production and use can positively influence their decision to purchase worthy content. The more favorable are users’ attitudes toward production activities, the more they think of news as citizen-centered, representative, and able to satisfy people’s needs. They are willing to purchase digital subscriptions to news agencies, and they spend more time using social media. The more favorable are users’ attitudes toward use activities, the more they view the attributes of the content published by a news agency positively. These relationships indicate that news users are not those who consume information passively but rather a new group of people called produsers. They challenge the monopolistic role of gatekeeping traditionally held by the news media; instead, they perform the task of gatewatching by producing news and posting comments about stories (Bruns, 2003). For instance, produsers post stories, replies, photographs, audio and video files, or graphics on Facebook, YouTube, and KakaoTalk, and share them with other users. Therefore, it is important that YNA produce news content consistent with what these produsers want. In this study, users consider the content of the news agency to be understandable, accurate, factual, fair, diverse, and in-depth.

As suggested by previous research, gatewatching as a form of content production, source checking, and story evaluation can facilitate multiperspectival news, characterized by citizen-centered, representative perspectives. Users are willing to pay for multiperspectival news as long as it contains high-quality information that is unique and necessary. These attitudes imply that digital subscriptions are a possible alternative for the survival of Korean online news organizations. Users’ perceptions of the news and of the attributes of YNA content are key predictors for the success of digital subscriptions.

Social media users’ news perceptions influence their perceptions of the attributes of YNA. This relationship is explained by expectancy-value theory. When people judge a news story to be positive in terms of expectancy values, this judgment affects their attitudes toward television news (Cooper et al., 2001). In this study, social media users consider that news should reflect citizens’ voices, represent public opinion, and satisfy people’s needs, which represent their judgments of expectancy values. Accordingly, these judgments influence their attitudes toward news agency content. These attitudes further influence social media users’ intentions to purchase digital subscriptions. This finding is consistent with previous research showing that people’s expectancy values and attitudes toward media attributes induce them to consume media (Leung and Wei, 1999). Likewise, people’s expectancy values against media influence their intention to use media (Ajzen and Fishbein, 2005).

The value of news content is determined by the preferences of news users. Thus, a redefinition of news is occurring inside the minds of social media users. The news agency that does not satisfy the perceptions of its users may be criticized for imitating the boldness of Reuters, a leading news agency. Reuters made the mistake of focusing too much on its status as a leading news agency (Bartram, 2003). Journalists need to pay attention to the role of news agencies as a news platform for social media users, given that our society is complex.

Because time spent on social media activities leads to greater news consumption, the effects of time enhancement are more visible than the effects of time replacement. People who use the Internet frequently are engaged in diverse social activities, such as watching movies or volunteering (Robinson, 2011). Therefore, Internet use time appears to enhance one’s social interactions. Similar relationships are found for use time of social media and use time of news. More time spent by users in sharing content and posting messages via social media leads to increased news use time. Users’ experience with news agency content affects their perceptions of the content, intention to purchase digital subscriptions, and news use time.

In summary, the findings of this study indicate that the attitudes of social media users toward production and use activities influence their perceptions of the attributes of YNA and their intentions to purchase digital subscriptions. Users’ perceptions of news influence their perceptions of the attributes of the news agency, and their social media use time affects the news use time. Therefore, news agencies should consider social media users’ attitudes toward production and use activities in their approach to making news. In particular, the finding that the social media users’ attitudes influence their intention to purchase digital subscriptions serves as an important guideline for developing a viable business model for a news agency.

Footnotes

Funding

This work was funded by the Korea News Agency Commission of 2013 and was partially funded by the Sogang University Research Grant of 2011 (grant number 201110068.01).

Author biography

Jeongsub Lim (PhD, University of Missouri) is an associate professor at the School of Communication of Sogang University. His research includes data journalism or computational journalism, and framing building and its effects.

References

Ajzen

Fishbein

(2005) The influence of attitudes on behavior. In: Albarracín

Johnson

Zanna

(eds) The Handbook of Attitudes. Mahwah, NJ: Lawrence Erlbaum Associates, pp. 173–221.

Al-Gahtani

King

(1999) Attitude, satisfaction and usage: factors contributing to each in the acceptance of information technology. Behaviour & Information Technology 18(4): 277–297.

Arbuckle

(2012) IBM SPSS Amos 21 User’s Guide. Amos Development Corporation. Available at: http://public.dhe.ibm.com/software/analytics/spss/documentation/amos/21.0/en/Manuals/IBM_SPSS_Amos_Users_Guide.pdf

Bartram

(2003) News agency wars: the battle between Reuters and Bloomberg. Journalism Studies 4(3): 387–399.

Bruns

(2003) Gatewatching, not gatekeeping: collaborative online news. Media International Australia Incorporating Culture and Policy: Quarterly Journal of Media Research and Resources 107: 31–44.

Bruns

(2008) Blogs, Wikpedia, Second Life, and beyond. New York: Peter Lang Publishing.

Bruns

Schmidt

(2011) Produsage: a closer look at continuing developments. New Review of Hypermedia and Multimedia 17(1): 3–7.

Byrne

(2010) Structural Equation Modeling with AMOS: Basic Concepts, Applications, and Programming. 2nd ed. New York: Routledge.

Chung

(2013) (Weekly SNS trends) KakaoTalk secures 10 million members. Available at: http://www.bloter.net/archives/157860

10.

Chung

(2012) How to deal with the ownership structure of publicly held news media. Available at: http://www.journalist.or.kr/news/articleView.html?idxno=29946

11.

Cooper

Burgoon

Roter

(2001) An expectancy-value analysis of viewer interest in television prevention news stories. Health Communication 13(3): 227–240.

12.

Cutler

(2010) Still shy of 2 million users, facebook quadruples in South Korea. Available at: http://www.insidefacebook.com/2010/11/04/facebook-southkorea/?utm_source=feedburner&;utm_medium=feed&utm_campaign=Feed%3A+InsideFacebook+%28Inside+Facebook%29&utm_content=Google+Reader

13.

DeVoe

(2009) Bursts of information: microblogging. Reference Librarian 50(2): 212–214.

14.

Domingo

Quandt

Heinonen

. (2008) Participatory journalism practices in the media and beyond. Journalism Practice 2(3): 326–342.

15.

Gans

(1979) Deciding What’s News: A Study of CBS Evening News, NBC Nightly News, Newsweek and Time. New York: Vintage Books.

16.

Gans

(2011) Multiperspectival news revisited: journalism and representative democracy. Journalism 12(1): 3–13.

17.

Google (2010) The 100 most-visited sites: Korea. Available at: https://support.google.com/adplanner/answer/3310841?rd=1

18.

Google (2012) The 1000 most visited sites on the web 1. Available at: http://topgooglesites.blogspot.kr/2012/01/1000-most-visited-sites-on-web-1.html

19.

Grinnell

(2009) From consumer to prosumer to produser: who keeps shifting my paradigm? (We do!). Public Culture 21(3): 577–598.

20.

Han

Y-H

(2013) The era of breaking news online is the crisis for the Yonhap News Agency. Available at: http://www.mediaus.co.kr/news/articleView.html?idxno=32805

21.

Herbert

Thurman

(2007) Paid content strategies for news websites: an empirical study of British newspapers’ online business models. Journalism Practice 1(2): 208–226.

22.

Hermida

(2010a) From TV to Twitter: how ambient news became ambient journalism. M/C Journal 13(2): 1–6. Available at: http://journal.media-culture.org.au/index.php/mcjournal/article/view/220

23.

Hermida

(2010b) Twittering the news: the emergence of ambient journalism. Journalism Practice 4(3): 297–308.

24.

Hermida

Thurman

(2008) A clash of cultures: the integration of user-generated content within professional journalistic frameworks at British newspaper websites. Journalism Practice 2(3): 343–356.

25.

Hooper

Coughlan

Mullen

(2008) Structural equation modelling: guidelines for determining model fit. Electronic Journal of Business Research Methods 6(1): 53–60.

26.

Hwang

(2001) Internet news and its business model in terms of analysis of news users and digital subscriptions. Report no. 07. 1 December. Seoul, Korea: Korea Press Foundation.

27.

Jenkins

(2006) Convergence Culture: Where Old and New Media Collide. New York: New York University Press.

28.

Jeon

(2012) Twitter users’ perception about news production, distribution, and consumption via Twitter. Journal of Korean Society for the Scientific Study of Subjectivity 25: 195–216.

29.

Jönsson

Örnebring

(2011) User-generated content and the news. Journalism Practice 5(2): 127–144.

30.

Kim

. (2007) The future of journalism. Report no. 07-08. Seoul, Korea, 1 December. Gwacheon, Gyeonggi-do: Korea Information Society Development Institute.

31.

Kim

Kang

(2010) Digital news subscription strategy. Korean Association of Newspapers. Available at: http://presskorea.or.kr/upload/%EB%89%B4%EC%8A%A4%EC%BD%98%ED%85%90%EC%B8%A0%20%EC%9C%A0%EB%A3%8C%ED%99%94%20%EC%A0%84%EB%9E%B5.pdf

32.

Kim

(2010) Increasing the global competitive edge of the Yonhap News Agency in the new environment in terms of the strategy of using the Internet. In: The meeting of the Korea News Agency Commission, Seoul, Korea, 8 April, pp. 25–51: Korean Society for Journalism & Communication Studies.

33.

Kim

(2009) Changes in the market of news media and the role of the Yonhap News Agency. In: The meeting of the Korean Society for Journalism & Communication Studies, Yangyang-gun, Gangwon-do, Korea, 18 September, pp. 19–34: Korean Society for Journalism & Communication Studies.

34.

Kotler

(1986) The prosumer movement: a new challenge for marketers. Advances in Consumer Research 13(1): 510–513.

35.

Lee

Lim

(2012) A study of digital subscriptions strategies for Korea’s daily newspapers. Report no. 2012-05. Seoul, Korea, 31 August: Korea Press Foundation.

36.

Leung

Wei

(1999) Seeking news via the pager: an expectancy-value study. Journal of Broadcasting & Electronic Media 43(3): 299–315.

37.

Mueller

Hancock

(2008) Best practices in structural equation modeling. In: Osborne

(ed.) Best Practices in Quantitative Methods. Thousand Oaks, CA: SAGE, pp. 488–508.

38.

Palmgreen

Rayburn

(1985) A comparison of gratification models of media satisfaction. Communication Monographs 52(4): 334–346.

39.

Proulx

Heaton

Choon

MJK

. (2011) Paradoxical empowerment of produsers in the context of informational capitalism. New Review of Hypermedia and Multimedia 17(1): 9–29.

40.

Rayburn

(1996) Uses and gratifications. In: Salwen

Stacks

(eds) An Integrated Approach to Communication Theory and Research. New York: Lawrence Erlbaum Associates, pp. 145–163.

41.

Ritzer

Dean

Jurgenson

(2012) The coming of age of the prosumer. American Behavioral Scientist 56(4): 379–398.

42.

Robinson

(2011) Arts and leisure participation among IT users: further evidence of time enhancement over time displacement. Social Science Computer Review 29(4): 470–480.

43.

Robinson

De Haan

(2006) Information technology and family time displacement. In: Kraut

Brynan

Kiesler

(eds) Computers, Phones, and the Internet: Domesticating Information Technology. Oxford: Oxford University Press, pp. 51–69.

44.

Robinson

(2007) “Someone’s gotta be in control here” The institutionalization of online news and the creation of a shared journalistic authority. Journalism Practice 1(3): 305–321.

45.

Sankaranarayanan

Samet

Teitler

. (2009) Twitterstand: news in tweets. In: Proceedings of the 17th ACM SIGSPATIAL international conference on advances in geographic information systems, Seattle, WA, 4–6 November, pp. 42–51. New York: ACM.

46.

Statistics Korea (2012) E-national indicator: population structure according to the gender and age. Available at: http://www.index.go.kr/egams/stts/jsp/potal/stts/PO_STTS_IdxMain.jsp?idx_cd=1010&bbs=INDX_001

47.

Sundet

Ytreberg

(2009) Working notions of active audiences: further research on the active participant in convergent media industries. Convergence: The International Journal of Research into New Media Technologies 15(4): 383–390.

48.

Swanson

(1987) Gratification seeking, media exposure, and audience interpretations: some directions for research. Journal of Broadcasting & Electronic Media 31(3): 237–254.

49.

Thurman

(2008) Forums for citizen journalists? Adoption of user generated content initiatives by online news media. New Media & Society 10(1): 139–157.

50.

Toffler

(1990) Power Shift: Knowledge, Wealth, and Violence at the Edge of the 21st Century. New York: Bantam Books.

51.

Tsao

Sibley

(2004) Readership of free community papers as a source of advertising information: a uses and gratifications perspective. Journalism & Mass Communication Quarterly 81(4): 766–787.

52.

Van Dijck

(2009) Users like you? Theorizing agency in user-generated content. Media, Culture & Society 31(1): 41–58.

53.

Wheaton

Muthen

Alwin

. (1977) Assessing reliability and stability in panel models. Sociological Methodology 8(1): 84–136.

54.

Wofford

(2012) User-generated content. New Media & Society 14(7): 1236–1239.

55.

Yonhap News Agency (2013) Domestic & world wide subscriber. Available at: http://www.yonhapnews.co.kr/aboutus/4210000000.html