Relationship Among the Attributes of World Countries and Their Coverage in Tweets of International News Agencies: 2010–2016

Background/objectives: To examine the factors which can influence the presence of world countries in news tweets of international news agencies. Methodology: The study draws upon the World Systems Theory, which categorizes the world into Core, Semi-Periphery, and Periphery with respect to economic, political, and communication relationships. We attempt to study and compare the coverage of Core, Periphery, and Semi-Periphery countries in tweets of international news agencies. Scholars argue that certain attributes of a country make that country more newsworthy for international news agencies and also these factors contribute significantly in making these countries more prominent on the digital landscape of twitter. We used the method of content analysis of purposively selected tweets of four international news agencies; AFP, AP, Reuters, and Xinhua about the 15 sample countries, including Pakistan, for the period of 7 year from 2010– 2016. Findings: We found that there are significant differences in the coverage of world countries in tweets of international news agencies. Core and Semi-Periphery countries are given more coverage in international news tweets. Similarly, Core and SemiPeriphery countries are more retweeted and liked by the followers of international news agencies. Finally, we found that GDP is not the sole determinant of countries portrayal and their sharing on Twitter by the international news agencies and their followers. Novelty/ improvements: If a country, including Pakistan, wants to increase its sharing on Twitter, the country should develop its information sector and internet penetration should be accelerated.


World System Theory and International News
Theoretically, to date, a well-defined theory regarding international news flow on Twitter is absent. However, the world-system approach to international news flow and portrayal of countries on traditional media is borrowed. World System Theory has been previously used by Ref. [33] to study the countries mentions and prominence on twitter. In Ref. [34], Wallerstein defined a world-system as one in which there is extensive division of labor. He categorized nations into three categories; core nations (originally comprised of Western Europe and later expanded to include North America and Japan), periphery (Latin America, Africa, Asia, the Middle East and Eastern Europe, etc.), and semi-periphery (India, China and Japan, etc.) [35]. Several studies on international news flow and portrayal of world countries have used world system theory as theoretical foundations [36][37][38][39][40][41][42]. It has also been tested in the digital age for studying the information flow concerns [43][44].
In [45] categories of core, periphery, and semi-periphery countries to classify sample countries. In Ref. [45], Chase-Dunn et al. studied the phenomenon of economic globalization over the past two centuries. On the base of world trade data, they classified the countries into three categories of core, periphery, and semi-periphery countries. Firstly, we conducted a pilot study of randomly selected one year from 2010 to 16 to find out the mentions of world countries in tweets of sample international news agencies. Then most mentioned five Core countries (United States, United Kingdom, Russia, Japan, and Israel), five Semi-Periphery countries (South Korea, China, Iran, India, and Turkey), and five Periphery countries (Libya, Egypt, Syria, Pakistan, and Afghanistan). Countries were categorized into Core, Semi-Periphery, and Periphery on the basis of previous world system studies [46].

Country Attributes and International News
A country attributes influence its coverage in the international news. According to [47] world system variables of GDP, levels of exports and population are the key predictors of international news coverage. They also found that the negative valence of a nation gets more coverage prominence in the international news.
In Ref. [33], Wu et al. studied the factors influencing the countries mentioned on Twitter. They studied the three kinds of factors which may influence the countries' mentions on Twitter. In his excellent work on international communication, Chang [36] studied the coverage of world countries in the news of Reuters. He found that Core nations are more prominent in the coverage of international news agency. However, Semi-Peripheral and Peripheral nations have to pass through different filters including determining events, context, internal attributes, and international interaction to become prominent in the news coverage of international news. His model presented world-system position and determining events as the primary filters for the international news coverage. The present study also takes its roots from this model. We also extend this scholarship in the context of digital media. We aim to study the relationships among the filters of world system status, news determinants and attributes of countries for the study of news tweets of international news agencies.
In Ref. [48], Wu investigated the influence of systemic determinants on international news coverage in 38 countries. Systemic factors include traits of nations, magnitude of interaction and relatedness between nations and logistics of newsgathering. Multiple regression is implemented to assess 9 systemic determinants in each country in the world. In spite of some variation, trade volume and presence of international news agencies were found to be the 2 primary predictors of the amount of news coverage. In this way, we can argue that trade and economy is one of the main determinants of international news. By gaining theoretical support from their findings, the researchers also examined the relationship of a country attributes of population, area, freedom of expression and index, GDP, political instability with their mentions, retweets, favorites, replies and shared portrayal. These attributes data were collected from different sources. Population and GDP data were collected from the World Bank. Population data were collected from The World Bank, area data from CIA website, Freedom of expression and political instability data were collected from world governance indicators of World Bank, and GDP data were collected from World Development indicators of World Bank.
In the present study, country attributes of population, area, internet users, voice and accountability index, political stability index, and GDP were considered to measure their relationship with the countries' mentions, retweets, favorites and shared portrayal via tweets of international news agencies. The date of population was collected from the databank of World Bank. Internet users' data were retrieved from The Global Economy website and from Internet Live Statistics website. Data of countries area were collected from the American CIA library. GDP data were also collected from the World Bank, and The Global Economy website. In the World Governance Indicators, Voice and Accountability Index reflects perceptions of the extent to which a country's citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free media. The index ranges from −2.5 to 2.5. Similarly, in the World Governance Indicators political Stability and Absence of Violence/Terrorism measures perceptions of the likelihood of political instability and/or politically-motivated violence, including terrorism. In this way, we build an argument that there are differences and imbalances in coverage of Core, Periphery, and Semi-Periphery countries by the international media, particularly international news agencies.
After reviewing the literature on international news flow and world system theory in the context of digital media, we attempt to continue the effort of [33] for building a comprehensive theory about the news flow on twitter. With this theoretical and conceptual support, we hypothesize the following statements.
H1: There would be significant differences in the presence of Core, Periphery, and Semi-Periphery countries in tweets of international news agencies.
H2: There would be significant differences in the portrayal of Core, Periphery, and Semi-Periphery countries in tweets of international news agencies.
H3: There would be significant differences in the sharing of Core, Periphery, and Semi-Periphery countries via tweets of international news agencies.
H4: A country attributes (population, area, GDP, political stability, freedom of expression, and internet penetration) will more likely to determine its mention, retweet rate, portrayal and shared portrayal in tweets of international news agencies.

Methodology
The study uses the method of content analysis and secondary data analysis. Firstly, the researchers performed the content analysis to study the presence and portrayal of the countries in news tweets. Secondly, we used secondary data about the attributes of countries to measure their relationship with the variables of study. Lastly, one-way ANOVA test was applied to compare the coverage and portrayal of core, periphery, and semi-periphery countries. The correlation and regression was applied to measure the relationships of country attributes and their coverage in international news agencies.
The Universe of this study is tweets of international news agencies and population was the all countries of the world. We selected 15 countries purposively (details are explained earlier) and Twitter accounts of four international news agencies; AFP, AP, Reuters, and Xinhua. Firstly, because these four agencies have highest twitter followers. Secondly, these agencies have been studied in international news flow studies; Reuters, AP [49][50][51][52][53], AFP [54][55], and [56]. Study evident that most of the U.S journalists depend on global news services for getting the content of international news [57]. We selected 75,932 tweets related to the sample countries from the official twitter accounts @Reuters @AP @AFP and @XHNews. Tweets were retrieved from twitter API during the month of July, 2017. Unit of Analysis for this study was defined as tweet of sample news agencies, mentioning the name of any sample countries. Only English language tweets were coded. Moreover, only the text of the tweets was coded. Images and hyperlinks were not coded nor followed. Only the tweet by the selected accounts was coded. @Replies to the selected accounts were also excluded.

Coding Procedures
In this study, the tweets were collected from the Twitter API live streaming. All tweets were retrieved during the months of May & June of 2017. Due to the slow rate of downloading, it took two months to retrieve all tweets of selected news agencies from year 2010 to 2016. Then tweets were stored in the form of PDF documents. Furthermore, after the selection of sample countries, tweets were coded manually by searching the name of sample country. Three coders were selected to code the content. These coders were graduated in Mass Communication & Media Studies & their medium of instruction was English. They were provided three weeks training about the code book and coding instructions. Intercoder reliability was obtained 0.82 by Cohen Kappa. Furthermore, validity of the coding sheet was ensured through expert opinion.

Portrayal
The three categories of portrayal are defined as: Positive, if a tweet creates positive image of the mentioned country on human perception; neutral, if a tweet creates neither positive nor negative image of the mentioned country/nation on human perception and negative, if a tweet creates negative image of the mentioned country on human perception. For making data measurement at ordinal level and to calculate shared portrayal, positive was assigned code +1 and neutral was assigned 0 and negative was assigned −1 code. We also quantify the number of replies, favorites, and retweets to the selected tweets for the analysis of world countries sharing on twitter.

Retweets, Favorites, and Shared Portrayal
In this study [58], we quantify the number of retweets, replies, and favorites of news tweets about the selected countries. Retweet is an essential feature of twitter which amplifies the message of international news agencies. As Choi found online opinion leaders are still influential in spreading news content on Twitter. Here, in this study, it is argued that if a country is tweeted positively by international news agencies and further it is retweeted

Indian Journal of Science and Technology
Vol 13(08), DOI: 10.17485/ijst/2020/v13i08/149958, February 2020 more by the followers of these agencies then the collective impact and shared portrayal of the tweet will also increase in the positive direction. However, if a country is tweeted negatively by an international news agency, and it is more retweeted and favorited by its followers then it will create a negative shared portrayal of that country. So, we developed a formula to measure shared portrayal as follows. Shared portrayal = Portrayal × (Number of Replies + Number of Retweets + Number of favorites) Here, valence denotes the portrayal of country-issue network. Which is valued as +1, 0, and −1. Shared portrayal was calculated by using SPSS and putting variables to the abovedefined formula.

Findings and Discussion
We found that international news agencies are using twitter effectively for the distribution of news (Table 1). In the previous studies on international news flow, scholars found Reuters and AP more influential [59][60]; however, we found Xinhua, Chinese news agency is tweeting more than other news agencies (Table 1). It shows that the social media is assisting to change the global patterns of international news distribution. Now, Xinhua, a news agency of semi-peripheral country, is also competing for the dominant world news agencies; Reuters, AP, and AFP on Twitter. Moreover, Xinhua also has 3rd more twitter followers. However, here, we cannot undermine the fact that China is now also moving towards core countries due to its technological, economic and political advancement.
Furthermore, we found that as previous literature on traditional media claim that there are significant differences in news coverage of developed and underdeveloped countries () we also note that core and semi-periphery countries are given more coverage (44.6%) and (36.7%) respectively as compare to periphery countries (18.7%) ( Table 2). It reveals that international news agencies are still cultivating dominant structures of global news distribution in the social media age. They are using their tweets for establishing the dominance of powerful countries in other world countries. Hence, we find support to confirm H1 that "There would be significant differences in the presence of core, periphery and semi-periphery countries in tweets of international news agencies".
Previous studies evident that international news agencies are instrumental and they promote the global inequality [60][61]. It is claimed that international news agencies portray developed world positively and presents a distorted and negative image of underdeveloped countries [62][63][64][65][66][67]. Within the paradigm of world system approaches, it is argued that core, semi-periphery, and periphery countries are portrayed differently in international news by the global media [68][69].
In the present study [3], we also find to support the argument of world system theorists. In case of news tweets, we found that there are significant differences in the portrayal of core, periphery, and semi-periphery countries ( Table 3). Core and semi-periphery countries are covered more positively in tweets of international news agencies. On the other hand, periphery countries are covered more negatively in tweets of international news agencies (Table 3). Therefore, on the contrasting to previous study, which found twitter as change agent in international news distribution, we found it invalid in case of tweets of international news agencies. Global news agencies are reproducing the traditional news flow imbalances among the developed and underdeveloped countries rather than changing it. Therefore, we find support to confirm H2 that developed countries would be portrayed positively and underdeveloped countries would be portrayed negatively in tweets of international news agencies.
Twitter has different dynamics of news distribution as well as consumption [14][15][16][17][18]. Replies, Retweets, and favorites are essential features of the twitter. These are considered the influential feature which allows twitter users to receive as well as propagate international news instantly [14]. Mass media provides major topics on Twitter and influential Twitter users also propagate these major topics [70]. In this study, we found that there are significant differences in the mean of replies, favorites, and retweet rate of core, periphery, and semiperiphery countries (Table 4). Moreover, core countries are more favorites and retweeted than semi-periphery and periphery countries by the followers of international news agencies (Table 4). However, interestingly semi-periphery countries are less retweeted and  replied than periphery countries (Table 4). In this way, we found a little evidence to claim that the tweets of international news agencies and their sharing is changing the traditional hierarchies of world countries in their propagation on twitter. Furthermore, in this study, we introduced the concept of shared portrayal for the study of collective impacts of tweets of international news agencies. We calculated the shared portrayal of countries according to the formula given in methodological section. We studied the collective effect of a tweet, retweets, favorites, replies, and portrayal. Because @replies and favorites allows the twitter users to interact with the news tweets [14][15][16][17][18]. Secondly, retweet multiplies the magnitude of that tweet [71]. More importantly, portrayal determines the direction of that magnitude either it is positive for a country or not. We found that there are significant differences in shared portrayal of Core, Periphery, and Semi-Periphery countries via tweets of international news agencies (Table 4). Core and Semi-Periphery countries are valued more positively than periphery countries by international news agencies and their twitter followers (Table 4). So, our findings do not support the argument that twitter is altering the traditional information and portrayal imbalances among the nations. We found an empirical evidence to argue that not only the tweets of international news agencies, but also the followers of the international news agencies are reproducing the traditional structure of international news imbalances among the nations. Twitter followers of international news agencies are also taking part in this propagation. Hence, we found support to confirm H3 that there would be significant differences in the sharing of Core, Periphery, and Semi-Periphery countries via tweets of international news agencies. In this way, tweets of international news agencies are also becoming instrumental and establishing their monopoly on the agenda of twitter users about the developed and underdeveloped countries.
Finally, the study attempts to explain the relationship between different attributes of countries and their sharing via tweets of international news agencies. As it was evident that certain attributes of countries like area, population, freedom of expression, internet users, political stability, and GDP are the predictors of a countries mentions and portrayal in international news coverage [36]. We also consider these attributes to explain their relationships with countries' mentions, retweets, favorites, @replies and shared portrayal via the tweets of international news agencies. We found that the population of a country negatively correlates with its retweet rate and geographical area of a country has positive correlation with its mentions of a country in news tweets of international news agencies (Table 5). Interestingly, freedom of expression of a country does not matter in determining its mentions, retweets, favorites, replies, and shared portrayal (Table 5). However, Internet penetration and political stability are significantly correlated with the shared portrayal of countries (Table 5). GDP is also a determining factor for the mentions, favorites, and shared portrayal of a country in news tweets of international news agencies. Therefore, it is argued that the GDP of a country has a strong influence on the decision of media editors and also on twitter users' decision about the international news consumption and sharing except retweeting (Table 5). In this way, on the contrary, to the argument of [33] those periphery countries are also becoming a core in mentions of countries on twitter. But in case of news tweets, periphery countries are still in the periphery in mentions, favorites and shared portrayal ( Table 5).
The retweet rate, favorite rate and @replies to the tweets of international news agency. Interestingly, we found that not only the international news agencies, but also their followers are spreading inequality and imbalances among the developed and underdeveloped countries. Developed countries are more retweeted, liked, and commented than underdeveloped countries by the followers of international news agencies. In the present study, the major reason of these imbalances is not only GDP, but also the internet penetration and political stability of the country (Table 5). Underdeveloped countries are mostly shared negatively due to the conflicts and political instability. Here, it is noteworthy to mention, that most of the conflicts in underdeveloped countries had a direct or indirect link with the military interventions of developed countries. For example, in Syria, Egypt, and Afghanistan, the developed countries like Russia, U.S, and Israel are stake holders. Therefore, it is claimed that international news agencies are instrumental in portraying underdeveloped countries negatively. In sum, we find partial support for the confirmation of H4 that "a country attributes (population, area, GDP, political instability, freedom of expression, and internet penetration) will more likely to determine its mention, retweet rate, portrayal and shared portrayal in tweets of international news agencies. " Hence, study reveals that GDP is not the sole determinant of countries portrayal and their sharing on twitter by the international news agencies and their followers.

Conclusion
We found that communication imbalances among the Core, Semi-Periphery, and Periphery exist in the tweets of international news agencies. Core countries are covered more and positively and on the other side, Semi-Peripheral and Peripheral countries are covered less, and negatively. However, Semi-Periphery countries like China are growing towards positive coverage in international news. Chinese news agency Xinhua is also competing the dominance of western news agencies effectively to influence the image of the China in global news flow. We also studied the retweet rate, favorite rate and @replies to the tweets of international news agency. Interestingly, we found that not only the international news agencies, but also their followers are spreading inequality and imbalances among the Core, Semi-Periphery, and Periphery countries. Core countries are more retweeted, liked and commented upon than Semi-Periphery and Periphery countries by the followers of international news agencies. Indian Journal of Science and Technology Vol 13(08), DOI: 10.17485/ijst/2020/v13i08/149958, February 2020 Finally, it is concluded that information flow and communication imbalances exists on the digital platform of twitter. Tweets of international news agencies are reproducing the traditional world hierarchies in distribution of international news rather than changing or replacing it. Although the social media is playing an important role of alternative media, yet it has several limitations. It is not as effective as assumed in changing the structures and patterns of international news distribution. In the modern information societies, there is a need to formulate the global communication policy to create and disseminate balanced world view in social media platforms generally and on news tweets specifically. Moreover, if a country wants to create and influence its positive image on the Twitter, Moreover, if a country wants to increase its sharing on the Twitter, the country should develop its information sector and internet penetration should be accelerated. These findings provide valuable insights to harness Pakistan's information and digital policies as a strategic goal.

Limitations and Future Recommendations
Findings of our study should be viewed in the context of several limitations. Firstly, only limited numbers of countries were selected due to the methodological and time constraints. Furthermore, these countries were selected on the base of randomly selected one-year pilot study. We performed the countries' mentions study of the year 2011. Later on, during the study variations were observed among the mentions of countries depending upon the context of international issues. Therefore, in future studies this phenomenon should be studied with more sample countries along with time series design. Secondly, only mainstream Twitter accounts of sample news agencies were selected. Now, international news agencies like Reuters and AP also have their different twitter accounts for different regions. These twitter accounts should also be studied and coverage and sharing of world countries in these regional accounts can also be compared with their coverage and portrayal in mainstream twitter account of international news agencies.
Thirdly, we only focused on English language tweets. The area can be explored further in native languages. It can also be compared with the language and discourse of mainstream twitter accounts of international news agencies. The scholarship can also be extended to study and compare the role of these news tweets in influencing the news agenda of regional news outlets. Lastly, due to the methodological constraints, we only focused on the number of retweets, @replies, and number of favorites. We adopted the quantitative technique; we do not include the content of retweets, quote tweets, and @replies. Moreover, the authenticity of the followers of international news agencies is also a limitation of this study. APPENDIX 1. Descriptives of sample core, periphery, and semi-periphery countries in news tweets of international news agencies from 2010 to 16