News Repertoires and Political Information Efficacy: Focusing on the Mediating Role of Perceived News Overload

In an era of news abundance, people increasingly feel uncertain as to whether they possess adequate information to participate in politics. However, previous research has not paid attention to the issue of such efficacy. To fill the gap, this study examines (a) how different types of news repertoire are associated with political information efficacy and (b) how perceived news overload mediates the relationship. The findings show that people have three distinct types of news repertoires, such as commentary-oriented, TV, and social media news repertoires, and those who consume news via a commentary-oriented news repertoire tend to have a higher level of political information efficacy, while the use of TV and social media news repertoires does not show significant effects on political information efficacy. In addition to that, the perceived news overload negatively mediates the link between all three types of news repertoires and political information efficacy. Two-wave original survey data were used for analyses.


Introduction
We are living in an era of news abundance. Given the critical role of news in raising democratic citizens who are wellinformed and politically engaged (e.g., Delli Carpini & Keeter, 1996;de Vreese & Boomgaarden, 2006), it can be inferred that the way people cope with the exponential increase in news supply may affect citizens' political lives.
When bombarded with excessive news, individuals naturally feel overloaded (Holton & Chyi, 2012;Song et al., 2017), and a common approach to deal with such overload is tailoring individuals' daily news repertoires. A news repertoire is a combination of different news sources people regularly rely on to follow public issues in the hybrid news environment (Yuan, 2011). The composition of individuals' news repertoires affects the quantity and quality of news they consume every day and eventually their political perceptions and behavior.
Among the many types of citizen competency that are related to the kinds of news repertoires and perceived overload, individuals' political information efficacy is worthy of attention. Political information efficacy refers to individuals' level of "confidence" in their possession of sufficient information to participate in the political process (Kaid & Postelnicu, 2005). These days, news consumers are compelled to sort through diverse layers of news sources, ranging from professional news organizations to unknown social media users. Accordingly, individuals may feel uncertain as to whether they are getting adequate information from the right kinds of news sources. Despite the importance, extant research has not paid the necessary attention to political information efficacy in the context of the high-choice news environment.
Thus, in this study, we examine (a) how the different types of news repertoire that people use are associated with their level of political information efficacy and (b) how perceived news overload mediates the relationship. In doing so, we aim to offer further insights into how the current news settings influence individuals' news consumption patterns, their political ability, and eventually the health of democracy.

News Repertoires and Political Information Efficacy
With technological development, news platforms and news content providers via such platforms proliferate. The low cost of producing and disseminating news in the internetbased platforms lowered the threshold to enter the news business (Nie et al., 2010). Now, the lines between news creators and news consumers have become blurred, thus increasing the amount of news flow (Holton & Chyi, 2012). Even traditional news media, such as TV and newspapers, are diversifying the news disseminating channels, for example, to websites, social media, and mobile devices (Ju et al., 2014), adding more layers of news choices. Thus, Hermida (2010) defined today's situation as an "ambient" news environment where news content is omnipresent.
Users simply cannot consume all the news stories provided, and, instead, form their own daily news repertoires to manage their regular news habits. It has been noted that an investigation of the effect of an individual information source is meaningless in the high-choice media environment (Diehl et al., 2019;Kim, 2016) because people use news from multiple platforms every day, and such multitasking is more common among the young generation (Purcell et al., 2010). Thus, people need to put in more effort when they choose which media to use (Schwartz, 2004). The concept of a "repertoire" in the context of media consumption stems from the concept of a "channel repertoire," which was first used to explain how an individual or household chooses the set of TV channels it watches regularly (Heeter, 1985). The concept was developed into the media repertoire, and studies have identified the composite of media platforms or content types used together (Hasebrink & Popp, 2006;Jeffres et al., 2004;Kim, 2016;Schrøder & Steeg Larsen, 2010;Taneja et al., 2012;Van Rees & Van Eijck, 2003) or factors that predict the choice of each repertoire (Kim, 2016;Neuendorf et al., 2001).
News repertoires have become the focus of recent studies and constructs of the repertoire have been distinguished (Edgerly et al., 2018;H. Lee & Yang, 2014;Strömbäck et al., 2018;Taneja et al., 2012). For example, news users were found to be clustered into traditional and internet-based news repertoires (Choi, 2016) or into news avoiders, emerging news seekers, and traditional news seekers (H. Lee & Yang, 2014). A study by Bos et al. (2016) revealed four kinds of repertoires including minimalists, news omnivores, public news consumers, and popular news consumers. Similarly, Strömbäck et al. (2018) identified five news repertoires, including minimalists, public news consumers, local news consumers, social media news consumers, and popular online news consumers. These cases show that revealed repertoires encompass various aspects of news choices such as the size of news repertoires (e.g., news omnivores and minimalists), content preferences (e.g., public news consumers, popular news consumers), and platform preferences (e.g., traditional, social media).
Despite the heightened interest in the composition and predictors of news repertoires, the political consequences of having distinctive news repertoires have been relatively understudied. A few related studies found that the level of political knowledge was the highest among those who had a traditional media repertoire, followed by internet-only news, and tabloid newspapers (Kim, 2016). A study by Diehl and colleagues (2019) showed that those who use news from multiple platforms (i.e., having a large size news repertoire) are more likely to be politically active citizens, and Schrøder and Steeg Larsen (2010) noted that the social media news repertoire is a positive predictor of online and offline political participation.
Another notable consequence of news repertoire choice is individuals' level of political information efficacy because people's news repertoire is related to how people choose, combine, and process news. Political information efficacy is defined as "the voter's confidence in his or her own political knowledge and its sufficiency to engage the political process (to vote)" (Kaid et al., 2007(Kaid et al., , p. 1096. While the concept is closely related to political efficacy, particularly internal efficacy, it is different in that its focus is more on confidence in the information. Political efficacy is citizens' perceived ability to influence political process (Campbell et al., 1954) and is frequently divided into external efficacy (i.e., the extent of belief about governmental responsiveness to citizens' demand) and internal efficacy (i.e., the extent of belief about one's competence to understand and participate in politics) (Acock et al., 1985;Niemi et al., 1991).
While political efficacy is a strong predictor of citizens' political behavior, notably voting, the perceived lack of information or knowledge has also been found to function as an obstacle to participating in politics (Kaid et al., 2000). In other words, the perceived level of their own knowledgeability is also critical in determining individuals' willingness to participate in the political process (Kaid et al., 2007;Kaid & Postelnicu, 2005;Tedesco, 2007), in addition to the absolute level of information or knowledge (Bartels, 1996).
Studies have revealed that news use via various information sources results in different levels of political information efficacy because users process each source of information in a different way (Kaid et al., 2007). In particular, the use of internet information sources, such as online political messages (Tedesco, 2007(Tedesco, , 2011, was associated with high political information efficacy. Interpreting the findings that exposure to online political advertisements yielded a higher level of political information efficacy among young adults compared to exposure via TV, Kaid and Postelnicu (2005) explained that the different effects may stem from the availability of further informationseeking allowed in each information environment. On the internet, users can seek additional information sources, thus increasing political information efficacy. The presence of interactive features in the web environment was also noted as a factor that contributes to the increase in information efficacy in that such interactivity facilitates further engagement with the related political information (Tedesco, 2007). Given this, it is reasonable to infer that different news repertoires are likely to have different effects on users' political information efficacy. Thus, we propose the following research questions.

RQ1:
How are the news repertoires composed? RQ2: How is each news repertoire associated with the level of political information efficacy?

Repertoire and Political Information Efficacy
In examining the relationship between news repertoires and political information efficacy, the role of perceived news overload needs to be considered. People feel overloaded when the volume of information is beyond their capacity to process within a given time (Beaudoin, 2008;Meyer, 1998). Overload is not a new phenomenon. It is a natural response to excessive information because people's cognitive capacity is limited. As early as 1970, the term "information overload" has been used by Toffler (1970) to describe how excessive information relates to cognitive as well as sensory over-stimulation.
In the fast-paced digital environment, people experience an ever greater level of overload (Holton & Chyi, 2012;Ji et al., 2014;Nordenson, 2008;Song et al., 2016;York, 2013). The number of TV channels has increased, and internet sites, social networking sites (SNSs), and blogs are disseminating abundant information (S. K. Lee, Kim, & Koh, 2016). News has become ubiquitous, and consumers participate in not only news consumption via multi-platforms but also news mediation and distribution, which demands continuous processing of news, unconsciously, or consciously (Robinson, 2011). Moreover, the "push" notification system of mobile news exacerbates the news overload (Fox et al., 2007;Schmitt et al., 2018) and makes people feel helpless and feel as if they have little informational control (Jackson & Farzaneh, 2012).
The reaction to overload has been frequently noted. People feel stressed, confused, distracted, anxious, frustrated, and pressured when they are overloaded (Edmunds & Morris, 2000;Eppler & Mengis, 2004). According to the person-environment fit model, people feel stress when they find an imbalance between environmental demands and coping abilities (Lazarus & Folkman, 1984). The gap stimulates overload, eventually resulting in strain (Ayyagari et al., 2011;Ragu-Nathan et al., 2008).
In addition to such psychological reactions, poor decision-making, failure of task completion, and even avoidance may also accompany news overload (Eppler & Mengis, 2004;Fox et al., 2007;Lang, 2000;Schmitt et al., 2018). When overload results in an "avoidance" approach, it naturally limits individuals' confidence that they possess sufficient information and knowledge about public issues to participate in politics (i.e., political information efficacy).
It is also likely that different news repertoires may result in a different level of news overload and eventually a different level of political information efficacy. Indeed, the type of news content and platform influences the extent of overload (Holton & Chyi, 2012;Lang, 2000;A. R. Lee, Son, & Kim, 2016). When information is delivered extremely quick (Lang, 2006;Lang et al., 2015) or content is complex and highly dense, people feel more overload (Lang, 2009). For instance, text requires more cognitive effort for processing than visual or audio information (Savolainen, 2007). In particular, online platforms such as the internet, e-readers, and Facebook have been noted to generate more news overload (Eveland & Dunwoody, 2002;Holton & Chyi, 2012;Nordenson, 2008). When the level of perceived news overload was compared between traditional and new media, news consumption through SNSs, mobile news, and search engines was a positive predictor of overload, while TV news use was negatively associated with the level of overload (Chen & Chen, 2019). This may be because of the fact that it is not necessary to spend time in retrieving the content while consuming TV news (Edmunds & Morris, 2000).
Summing up the argument above, we expect that the level of news overload functions as a significant mediator linking news repertoires and political information efficacy, while the strength and the direction of effects may vary depending on the type of news repertoire. Thus, the following question is posed.

RQ3:
How does the level of perceived news overload mediate the association between the type of news repertoire and political information efficacy?

Procedure and Participants
To answer the research questions, we conducted a two-wave online survey in South Korea using online panels of a professional survey company (Macromill Embrain). The company randomly invited survey participants via email among their 1.2 million online panel members, who agreed to participate in such research. The quota-sample approach was employed so that the final respondents reflect the proportion of gender, age, and residential districts of the overall Korean population, following the examples of previous survey research (e.g., Gil de Zúñiga et al., 2017). The average age of the sample was 42.55 years and 50% of the respondents were females, reasonably reflecting an average age (41.7 years) and the proportion of gender (100.5 males, per 100 females) of the South Korean population in 2018. The first wave of the survey was conducted on March 14 to 20, 2019. A total of 3,956 panel members clicked the survey link, and 1,264 participants completed the survey, resulting in a 32% completion rate. The second wave of the survey was conducted on April 23 to 30, 2019, among the respondents of the first wave survey. In total, 874 people participated in the second wave survey, resulting in a 69% retention rate.

Key Variables
News repertoires. Variables used to determine different news repertoires were based on the questions asking how many days per week respondents used news from various channels on an 8-point Likert-type scale (0 = never; 7 = almost every day) in the T1 survey. News overload. The extent of respondents' perceived news overload (M = 21.86, SD = 6.29, Range = 0-40, Cronbach's α = .84) was measured in the T1 survey using a total of 10 questions that were used in previous studies (e.g., Ji et al., 2014;A. R. Lee, Son, & Kim, 2016;Song et al., 2017). On a 5-point scale (0 = strongly disagree; 4 = strongly agree) respondents provided their level of agreement if they feel they "encounter too much news compared to their limited time," "receive too much news that they cannot deal with," "are burdened to process all of the news," "can process all the news information effectively," "are distracted by too much news," "receive too many news notifications when they work," "think only a small part of the news they encounter is relevant," "encounter too many meaningless news stories," "have difficulties in finding necessary news," and "are overwhelmed by the amount of news information they encounter." Political information efficacy. The variable was measured using questions that were asked in the extant research (Kaid et al., 2007;Tedesco, 2011). Respondents provided their level of agreement with the following statement on a 5-point scale: "I think myself well-qualified to participate in politics," "I think that I am better informed about politics and government than most people," "I feel that I have a pretty good understanding of the important political issues facing our country," and "If a friend asked me about the presidential election, I feel I would have enough information to help my friend figure out who to vote for." Political information efficacy was measured in T1 (M = 6.99, SD = 3.67, Range = 0-16, Cronbach's α = .90), as well as in T2 (M = 7.20, SD = 3.69, Range = 0-16, Cronbach's α = .91).

Findings
To answer the research questions and test the hypothesis, we conducted principal component analysis (PCA), regression analysis, and bootstrapping techniques. RQ1 was related to the composition of news repertoires. We conducted factor analysis with Varimax rotation on a total of 11 news userelated questions, and the results showed that three constructs of news repertoires successfully loaded on 11 items (with an eigenvalue greater than 1). These three factors together explained 60.83% of the total variance (Table 1).
The first repertoire consisted of four items, all of which were related to the use of news sources that focus on the offering of in-depth interpretation and commentary of public issues, such as printed newspapers, magazines, podcasts, and political talk shows. This construct was named the "commentary-oriented news repertoire" (M = 4.46, SD = 5.33, Range = 0-28, Cronbach's α = .68). The next repertoire, "TV news repertoire" (M = 13.35, SD = 6.63, Range = 0-21, Cronbach's α = .83), loaded on three items: network TV, cable TV, and general programming TV. Items such as news use via SNS, mobile messenger, YouTube, and blogs (M = 13.26, SD = 8.39, Range = 0-28, Cronbach's α = .73) made up the "social media news repertoire." To examine the association between each news repertoire and political information efficacy (RQ2), we specified an ordinary least squares (OLS) regression model. To predict the changes in the extent of political information efficacy depending on the type of news repertoire respondents relied on, the researchers conducted lagged autoregressive analysis following the examples of previous studies (e.g., Shah et al., 2005;Valenzuela & Bachmann, 2015). Table 2 shows that among the three different news repertoires identified in RQ1, the commentary-oriented news repertoire (β = .10, p < .01) was the only one that predicted the level of political information efficacy in T2, controlling for political information efficacy in T1 as well as other control variables.
The next question (RQ3) was about the mediating role of the perceived news overload in linking the use of each news repertoire and political information efficacy. As was revealed in RQ1, only the commentary-oriented news repertoire was found to have positive direct effects on political information efficacy (B = .08, CI = [.0409, .1183]). However, all three news repertoires were found to have negative effects on political information efficacy when respondents felt over- . Thus, the findings show that the level of perceived news overload negatively mediates the association between the type of news repertoire and political information efficacy (Table 3).

Discussion
This study revealed that (a) individuals have distinctive news repertoires, such as commentary-oriented, TV, and social media new repertoires; (b) among the three types of repertoires, those who relied on the commentary-oriented news repertoire showed a significantly higher level of political information efficacy; and (c) news use via any of the three repertoires can limit political information efficacy when people experience news overload.
The finding that printed newspapers, news magazines, news podcasts, and political talk shows constitute a single news repertoire, termed the commentary-oriented news repertoire in this research, is worth noting. While they are seemingly unrelated, the commonality of these items may lie in the offering of a more in-depth and interpretive analysis of news events. While the interpretative nature of newspapers and magazines is well accepted, such role of news podcasts and political talk shows may be debatable because many of them are entertainment oriented. However, recent studies show that the motivation for listening to/watching news podcasts and political talk shows is to obtain good analyses and commentaries. For example, political talk shows, particularly when it is entertaining, contribute to the feeling of being better informed and having internal political efficacy (Roth et al., 2014). In terms of news podcasts, it was also found that one of the main motivations for listening to them is to know diverse perspectives, particularly from the outsider sources of news or opinions, in addition to the motivations of convenience and entertainment (Newman et al., 2019).
The positive role of the commentary-oriented news repertoire in enhancing political information efficacy reflects that, amid the fast and abundant news cycles, what people need is not mere fragmentary news information but "perspectives." It should be admitted that it is arguable whether the components of the repertoire (e.g., newspapers, news magazines, news podcasts, political talk shows), always deliver good  perspectives. However, those sources are better at reporting issues with a more in-depth explanation of backgrounds and with expert commentaries compared to other sources. The findings reflect the argument of Stephens (2014) that we need "wisdom journalism" that offers original and challenging perspectives. When individuals are provided with good interpretations and perspectives, they may feel confident that they are consuming necessary information to be used for political behavior. Nonetheless, it should be noted that those who have commentary-oriented news repertoires use related news sources less frequently. For example, the data of this study showed that users of commentary-oriented news repertoires consumed related news sources an average of 1.12 days per week, while those who used TV news repertoires were found to use related news sources 4.45 days per week and social media news users consumed 3.32 days per week. Thus, the findings show that those who use commentary-oriented news repertoires consume news less frequently but have a higher level of political information efficacy. The findings indicate that the quantity of news consumed or the frequency of news consumption does not necessarily correlate with the level of political information efficacy.
Relatedly, the insignificant role of social media news repertoires is also noteworthy. During the past decade, social media have risen as a major source of news and its positive role in raising good citizens (Choi, 2016;Gil de Zúñiga et al., 2012) has been frequently noted. In the early days of political information efficacy research, it was noted that the interactive information environment tends to increase individuals' political information efficacy (Tedesco, 2007). However, ironically, the use of social media news sources may lead users to feel that they are less equipped with necessary news information. While the scope of interactivity has been enormously improved in the social media news environment, excessive interactivity may not function as a positive factor in enhancing users' information efficacy.
The insignificant role of social media use in promoting political information efficacy may also be related to the quality of political information circulated within social media. An analysis of Telegram messages between 2015 and 2020 found that so-called junk news (i.e., low-quality news) sources were more consumed and yields more user engagement than the news produced by mainstream media (Knuutila et al., 2020). Moreover, it was revealed that social media use leads users to visit fakes news sites (Nelson & Taneja, 2018), and cyberbots systematically disseminate fake news on social media for political purposes (Lazer et al., 2017). Thus, it is likely that the prevalence of misinformation and junk news lowers the level of confidence that the user possesses enough information to participate in politics.

Conclusions and Limitations
In this study, the significant role of news repertoires was examined to determine the extent of people's political information efficacy for the first time. The findings showed that the commentary-oriented news repertoire is the only kind of repertoire that may affect political information efficacy positively. In particular, it is worth noting that the commentary-oriented news repertoire was a positive and strong predictor, while key predictors that have long been known to be associated with citizen competence, such as education, income, and political interest, were not significant predictors ( Table 2). The findings suggest the unique role of new repertoires in determining and regulating individuals' political information efficacy.
The study also revealed the negative ramifications of perceived news overload, regardless of the kinds of news repertoires. It even constrains the originally positive influence of the commentary-oriented news repertoire. The findings indicate that "overload" is indeed an important factor to understand the current news environment and its influence; to be better informed and to become more competent Note. Demographics, political characteristics, and political information efficacy in T1 were included in the analyses as control variables. LLCI = lower limit confidence interval, ULCI = upper limit confidence interval.
as a citizen, the amount of "news use" itself does not matter. What is more important is how wisely individuals choose and build their own news repertoires among abundant news sources so that they do not feel overloaded but get the necessary information. In this sense, the phenomenon of "news overload" needs to be further studied in the context of its political effects.
The study called our attention to the importance of news overload, but it has limitations. While we conceptualized and measured the level of perceived news overload based on previous studies, more recent studies show "overload" is a multifaceted concept. For example, Chen and Chen (2019) noted that news overload consists of six concepts: stress, reluctance, confusion, worry, overwhelming, and beyond one's capability to process. Thus, a more detailed and delicate approach to news overload is needed in future studies.
We also focused on political information efficacy as an important aspect of citizen competence in the fast-changing, high-choice news environment. In this study, we shed light on the concept, which was intensively studied in the early days of the internet and then ignored, but we did not examine how political information efficacy is associated with the actual exercise of citizenship, such as political participation, notably voting. Future research is required to answer such questions. In particular, we need to examine whether "confidence" in the possession of political information is always a positive factor. With the increasing spread of false information, it is also likely that we see citizens who are obviously "misinformed" but are very "confident" that they have sufficient knowledge to participate in the political process. Thus, we need to take into account those variables in examining the more detailed role of political information efficacy.
While we revealed three different types of news repertoires, we cannot claim that those three repertoires are exhaustive. While our results are meaningful in that they bring attention to the importance of different news repertoires in determining information efficacy, we only focused on the 11 frequently used news sources in South Korea. Thus, there may be other meaningful repertoires that should be examined concerning political information efficacy. In dealing with the issue, many studies do not differentiate between content preference and platform preference. In addition, one study notes that the size of the repertoire is an important aspect as well (Diehl et al., 2019). While admitting that the preferences are mixed and sometimes indistinguishable, a more sophisticated approach is required in future research to conceptualize and measure news repertoire.
Despite such limitations, we contributed to a better understanding of the current news environment and its implications on citizenship as well as the healthy functioning of democracy by focusing on new concepts, such as news repertoires, news overload, and political information efficacy. The findings of this study will serve as another starting point to further investigate the ever-changing news environment.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.