Source: self-made based on the data located in the corporate portals of the televisions and their Twitter accounts.
10.4185/RLCS-2020-1439
Artículo
COMPARATIVE STUDY OF THE CENTRALITY AND USE OF TWITTER OF REGIONAL TELEVISIONS IN THE ELECTORAL DEBATES OF M-26
ESTUDIO COMPARATIVO DE LA CENTRALIDAD Y USO DE TWITTER DE LAS TELEVISIONES AUTONÓMICAS EN LOS DEBATES ELECTORALES DEL 26 M
Belén Galletero-Campos1
Ana María López-Cepeda1
Arturo Martínez-Rodrigo1
1University of Castilla-La Mancha. Spain
ABSTRACT
Introduction. The electoral debates on television are consistent with previously agreed conditions, within the legal canons established by the LOREG. However, in the current communicative environment, there exists a parallel debate on the second screen which is not subjected to any control. This paper analyzes the conversation generated on Twitter during eight debates, broadcasted by the regional television corporations during the campaign of May 26, 2019.
Methodology. Using the hashtag provided for each TV entity, an analysis of the main nodes interaction and the activity of the official Twitter profiles of televisions during the meetings has been performed.
Results and conclusions. The results indicate that only in some cases televisions manage to be relevant in the parallel debate on Twitter and that they hardly use the potential of the social network to generate conversation. On the other hand, equity from TV corporations is subjected to their own deontology, when they mention to different political parties on Twitter.
KEYWORDS: electoral debates; Twitter; television; regions; centrality.
RESUMEN
Introducción. Los debates electorales televisivos se ajustan a unas condiciones previamente pactadas, dentro de los cánones legales establecidos por la LOREG. Sin embargo, en el actual entorno comunicativo se produce un debate paralelo en la segunda pantalla sujeto a menor control. En este trabajo se analiza la conversación generada en Twitter durante ocho debates emitidos por las televisiones autonómicas durante la campaña del 26 de mayo de 2019.
Metodología. Mediante el hashtag proporcionado para cada uno de ellos, se ha realizado un análisis de los principales núcleos de interacción y de la actividad de los perfiles oficiales de las televisiones durante los encuentros.
Resultados y conclusiones. Los resultados indican que sólo en algunos casos las televisiones logran ser relevantes en el debate paralelo en Twitter y que apenas utilizan el potencial de la red social para generar conversación. Por otro lado, la equidad a la hora de mencionar a partidos políticos y candidatos queda supeditada a la propia deontología de cada ente.
PALABRAS CLAVE: debates electorales; Twitter; televisión; comunidades autónomas; centralidad.
Correspondence.
Belén Galletero Campos. University of Castilla-La Mancha. Spain.
belen.galletero@uclm.es
Ana María López Cepeda. University of Castilla-La Mancha. Spain.
ana.lopezcepeda@uclm.es
Arturo Martínez Rodrigo. University of Castilla-La Mancha. Spain.
arturo.martinez@uclm.es
Received: 08/09/2019.
Accepted: 10/09/2019.
Published: 30/04/2020.
This article is part of the work carried out within the framework of the research project (RTI2018-096065-B-I00) of the State Program R + D + I aimed at the Challenges of Society of the Ministry of Science, Innovation, and Universities (MCIU by its acronym in Spanish), State Research Agency (AEI by its acronym in Spanish), and the European Regional Development Fund (ERDF) on “New values, governance, financing, and public audiovisual services for the Internet society: European and Spanish contrasts”.
How to cite this article / Standard reference
Galletero Campos, B., López Cepeda, A. M. & Martínez Rodrigo, A. (2020). Comparative study of the centrality and use of Twitter of regional televisions in the electoral debates of M-26. Revista Latina de Comunicación Social, (76), 97-119. https://www.doi.org/10.4185/RLCS-2020-1439
CONTENTS
1. Introduction. 1.1. Objectives. 2. Regulation of information in electoral campaign. 2.1. Election campaign and social networks. 3. Method. 3.1. Sample characteristics. 3.2. Procedure for extracting tweets and design of sociograms. 3.3. Calculation of centrality metrics 3.3.1. Degree centrality. 3.3.2. Betweenness centrality. 3.3.3. Eigencentrality. 4. Results. 4.1. Traditional audience, Twitter conversation and evolution of the hashtag. 4.2. Node behavior and centrality measures. 4.2.1. Central televisions: Telemadrid, Castilla-La Mancha Media, and La 7 Castilla y León. 4.2.2. The intermediate case: Aragón TV and Canal Extremadura. 4.2.3. Televisions without relevant activity: Televisión Canaria, La 7 (Murcia) and TPA (Asturias). 5. Discussion and conclusions. 6. Bibliographic references.
Translation by Paula González (Universidad Católica Andrés Bello, Venezuela).
1. Introduction
The attendance of several electoral appointments during 2019 has maintained an interest in political information. In this context, the relevance of conducting electoral debates has gained weight, favored by the emergence of a greater number of political options and, therefore, a scenario of greater competition to transmit electoral messages. The first electoral debates in Spain took place in 1993 but it has been in the last calls when they have consolidated as one more event within the campaign if we talk about the general elections, but not in regional calls. In May 2019, several regional televisions used this format for the first time, which raises the opportunity to carry out for the first time a comparative analysis not limited to the large national networks. Factors such as the spectacularization of politics (Padilla-Castillo, 2015) and the emergence of a more competitive scenario with new concurrent formations, have favored televisions betting on this format and the data confirms that the public has responded. The two national electoral debates between the leaders of the main formations – the one of TVE broadcasted on April 22nd and the one of Atresmedia broadcasted the following day – top the ranking of the most-watched broadcasts during the month of April, with a 38.5% and 27.2% of share respectively*, without taking into account the accumulated audience in simultaneous broadcasts on other channels. In regional televisions, the reached audiences range from the most modest results of the Murcia television, with a 2.5% screen share, to 8-9% recorded by the Castilian-Manchego television, that of Asturias or Telemadrid**.
* Data extracted from the monthly audience report of the consultancy Barlovento Comunicación.
** Data released in a press release by Castilla-La Mancha Media.
Current television faces the challenge of keeping the audience’s attention on television content, but, at the same time, transferring it to social networks, in such a way that communities that comment on the contents seen are articulated (Gómez Aguilar, Paniagua Rojano, and Farias Batlle, 2015). It is about trying to make profitable the effort in the production of these spaces maximizing their reach, something that can be achieved with self-promotion messages prior to the broadcast of the debate but also leading the conversation on other platforms.
According to Harboe’s definition (2009: 7), social television is that which “uses communication technology to connect viewers, in order to create shared experiences at a distance around television content”. Therefore, the emergence of social networks has led the social audience to join the traditional number or percentage of viewers of a program, that which actively participates in the viewing experience through interaction on a second screen (Lochrie and Coulton, 2012; Giglietto and Selva, 2014) – whether it be horizontal with other members of the public, or vertical with the transmitter- simultaneously with the broadcast (Iñesta Fernández, 2015). Some users’ motivations to establish these interactions are to communicate their impressions about a broadcast, search and share information, experience the sensation of a shared viewing, be curious about others’ opinions, and recommend programs (Han and Lee, 2014).
In this sense, Twitter is an ideal platform to amplify the scope of the televised broadcast, as it is a public access network that favors the viral transmission of information (Sequera Díaz, 2013). Besides, the use of hashtags or tags prevents dispersion and favors a structured conversation. The question is: who dominates that conversation during the broadcast of these spaces? The methodologies and indicators to determine the influence of profiles and hashtags are varied. One of the most used tools in the past to measure popularity in this network was the Klout factor, obtained from followers, degree of conversation, and influence of followers (Serrano-Puche, 2012). However, the company that provided this metric from its algorithms suspended the service in May 2018.
Since 2013, a certain scientific production has been located aimed at analyzing, through case studies, the behavior of the so-called social audience, although the works are usually focused on entertainment formats, either in prime time (Saavedra Llamas, Rodríguez Fernández, and Barón Dulce, 2015; Rodríguez-Vázquez, Direito-Rebollal, and Lago-Vázquez, 2016), talent shows (Quintas-Froufe and González-Neira, 2014) live broadcasts (Congosto, Deltell, Claes, and Osteso López, 2013a) or fiction (Deltell Escolar, Claes, and Osteso López, 2013). A common conclusion to these studies is the relationship between emotional and/or controversial content and greater participation of this type of audience. Political communication, however, is not very far from this component of emotions and feelings, as some authors have pointed out (Rúas-Araújo, Puentes-Rivera, and Dapena-González, 2015), both in terms of speeches of political representatives as in the voting decision. Thus, Twitter’s predictive capacity in electoral behavior has been tested in numerous studies at national (Congosto and Aragón, 2012) and international levels (Tumasjan, Sprenger, Sandner, and Welpe, 2010; Asur and Huberman, 2010; Larsson and Moe, 2011; Tjong and Bos, 2012).
More recent are the contributions of Lago Vázquez, Direito Rebollal, Rodríguez Vázquez, and López García (2016) about the consumption and following of the electoral campaign in the general elections of 2015 by the millennial generation, based on a qualitative approach, and the one of Vergeer and Hans Franses (2015), who apply content analysis to broadcasts and tweets. As Campos-Domínguez (2017) concludes, the analysis of the debate generated on Twitter, and, in particular, the mapping of political conversations, remain in recent years as a focus of study in the international scientific environment. However, they have not been applied during the broadcast of autonomous political debates. This comparative study of the eight debates will allow detecting differences in the impact they had on the social network.
1.1. Objectives
The main objective of this research is to analyze what role regional television stations played in the conversation that was generated on Twitter during the regional electoral debates, held in May 2019.
Framed in this purpose, the following secondary objectives are detailed:
Check the temporal evolution of the conversation around the debate label.
2. Regulation of information in electoral campaign
Information and persuasive communication in the electoral campaign are subject to a series of regulations that define the role of the media, in order to maintain neutrality and fair treatment of all political formations. In the regulation of information in electoral campaigns, collected in the LOREG (Organic Law 5/1985, of June 19th, of the General Electoral Regime), four levels of control can be differentiated according to the media in question: a. A maximum level, applicable to public radio-TV, which requires respect for political and social pluralism, for equality, proportionality, and information neutrality; b. An intermediate level, referring to private television, media in which the principles of pluralism and equality must be respected in every case and the principles of proportionality and information neutrality only in debates, interviews, and electoral information; c) An attenuated level, which applies to privately owned broadcasters, which should only respect the principles of equality and pluralism; d) A minimum level for the written press, only subjected to the right of rectification (Holgado González, 2017, pp. 469-470).
In a way, it can be said that this regulation has become obsolete because it hardly contemplates issues such as the informative treatment of digital media or tools such as social networks, which can no longer be considered an incipient phenomenon. The prohibition of publishing electoral polls in the media five days before the elections or spreading propaganda during the so-called reflection day, at a time when information spreads quickly through the internet and social networks, are good samples of a legal framework that is outdated. For this reason, in recent years several voices have been raised in favor of a new regulation of the communication of electoral campaigns on the internet (Campos Domínguez, Valera Ordaz, and López García, 2015, p. 1623). As Holgado González points out, the fact that written media can hold debates by transmitting a signal over the internet, that citizens echo through social networks polls published in foreign media, to name a few examples, highlights the effectiveness of regulations that may not be well suited to new circumstances and therefore have to be revised. (2017, p. 461).
Precisely, one of the criticized aspects is the absence of televised electoral debates in the regulations (Holgado González, 2017), despite its appeal and importance in the electoral campaigns of any country (Téllez, Muñiz, and Ramírez, 2010; Mateos-Crespo, 2017; Navarro Marchante, 2019). The mention of these spaces only appears in the legislation of five autonomous communities: Andalusia, Murcia, Basque Country, Castile and León, and Catalonia. These shortcomings mean that the organization of these meetings depends, almost exclusively, on the will of the candidates. Although there is no obligation nor prohibition to hold electoral debates when they are scheduled it is necessary to comply with minimum obligations: 1. They must notify the Central Electoral Board in advance, five days before the event (Rúas Araújo, Fernández Holgado, and Alén Amil, 2018, p. 162); 2. The publicly owned media and private televisions must respect the principles of neutrality, equality, proportionality, and informative pluralism, while radio stations- which can only broadcast debates but not organize them - are only subjected to the principles of pluralism and equality (Fernández de Casadevante Mayordomo, 2016, p. 212); 3. Finally, the media is obligated to compensate the candidates of political formations excluded from debates against their will (Holgado-González, 2017, pp. 475-476).
2.1. Election campaign and social networks
In order to verify the extent to which social networks have been incorporated into televisions’ electoral coverage plans, these documents have been requested from the seven public entities analyzed (La 7 Castilla y León, which is privately owned, is not required to present this document to the Provincial Electoral Board). Although they all allude to the possible holding of an electoral debate, only two expressly mention social networks in the plan, that of Extremadura and Castilla-La Mancha. In the proposal of the Extremadura entity, it is indicated that “in the social networks of Canal Extremadura, campaign specific-contents will not be generated so as not to interfere in the distribution of time allocated to political formations”. Another model is that of the Castilian-Manchego corporation, which states that “the digital media of Castilla-La Mancha Media will reproduce on its website and social networks exactly the same informative content broadcasted by the Autonomous Television”. These are different ways of acting, although in both cases the approaches are lax enough to give freedom when designing activity patterns in the retransmission of electoral content.
Regarding self-regulation, although there are several televisions that mention social networks in their stylebooks - Corporació Catalana de Mitjans Audiovisuals, Corporación Radio e Televisión de Galicia, Corporació Valenciana de Mitjans de Comunicació, and Euskal Irrati Telebista - only the Basque corporation has a specific section about social networks in campaign time. It states that “the use of social networks must be subjected to the same criteria of independence, impartiality, and informative balance. EiTB will promote citizen participation and debate with candidates through social networks and will encourage formats that favor interaction with the audience”. It should be noted that Basque television has not been considered in this study because no regional elections were held in this community in May 2019.
In the international field, the case of the BBC in the United Kingdom stands out, which incorporates specific manuals of the latest electoral appointments (local, state, and European elections). They contain a section referring to the use of social networks that indicates that the corporation’s personnel must avoid compromising the impartiality of the BBC by expressing their own points of view. Although there is a section related to the debates, only the need for prior consultation with the Main Policy Advisor to consider the issues to be addressed and the type of audience that will be present on the set is highlighted.
Both the inequalities in the norms and codes of use of social networks and the discordant voices regarding the current legislation make it relevant to observe the behavior of televisions in electoral campaigns in order to verify the practices they adopt. This work aims to contribute to the debate about the adequacy of a regulation that the current communicative context has overcome in many aspects.
3. Method
In recent years, the exponential growth and massive use of social networks have aroused great interest from different branches of knowledge to research the relationships between the different actors involved in a given scenario, as well as the nature of those relationships (Roig-Vila, Mondéjar, and Lorenzo Lledó, 2015). Indeed, social networks interconnect people who share interests or social relationships of different kinds, such as kinship, use of a common language, personal and professional collaborations, exchange of information or products, or simply collaboration between different social groups.
Although the technology used to create social networks is relatively new (Rueda Ortiz, 2015), social interaction networks are primary and have always existed. In this sense, and from the broader concept of the term, a network can be considered as a collection of entities and their relationships. In that network, the entities that are connected are called nodes or vertices, and the connections between the nodes are called links. Therefore, and more specifically, a social network is formed when there is an interaction, directly or indirectly, of some people with others, with institutions or other groups. The analysis of these social networks is carried out by applying a set of mathematical methodologies that are framed in the field of graph theory. Thus, network analysis allows us to visualize large amounts of complex interrelationships through maps, graphs or sociograms of connected symbols, and be able to calculate on those maps, concrete measurements of the size, shape, and density of the network, as well as the relevance of each node.
3.1. Sample characteristics
To monitor the conversation generated during each of the eight televised debates, the labels promoted by the televisions to promote the special program have been used. The spaces were broadcast within the electoral campaign of the regional elections of May 26th, between May 14th and May 21st. IB3 (Balearic Islands) broadcasted on May 21st has been excluded, since the chain did not use a hashtag, which prevented the gathering of the specific traffic of the conversation around the debate. On the other hand, it has been taken into account in the sample La 7 (Murcia), which maintains an indirect management, and the Television of Castile and Leon, privately managed, with the intention of detecting if there are differences regarding public entities, such as less control or a different use of social networks.
By filtering the label, all the tweets generated 24 hours before the broadcasting of the television space and up to 24 hours later have been captured, extracting in each case the interactions generated, since, according to the objectives, what is relevant is the exchange and the relationship established between different nodes or Twitter accounts. Once the complete corpus was formed, with a global volume of 96,512 interactions or links between 13,823 nodes or accounts, each case has been individually analyzed and compared.
Table 1.
Source: self-made based on the data located in the corporate portals of the televisions and their Twitter accounts.
3.2. Procedure for extracting tweets and design of sociograms
Once the sample was formed, the extraction of tweets and subsequent analysis was carried out through the commercial software Node XL, which provides a set of tools programmed in Visual C ++ language that allows the analysis of social networks through the formation of graphs or sociograms, using the program Microsoft Excel as the basis for the analysis of the data that make up the network. The collection of tweets was carried out using the Twitter programming interface so that the tweets containing the selected hashtag were imported into NodeXL in the form of two spreadsheets, one corresponding to nodes and one corresponding to links. The first provides information on the Twitter accounts involved in the debate; the second shows, in the form of name pairs, the origin and destination of the message, along with other additional attributes and information about that relationship. In this sense, it is important to highlight that the interrelations produced on Twitter have a directed nature, that is, tweets, mentions or retweets are shown in Node XL as a pair of actors, corresponding to the origin and destination of the message, respectively. Finally, once the data was processed, the mathematical model that represents the relationships between the different actors of the sociogram was formed, and the centrality metrics on the generated graph were calculated.
3.3. Calculation of centrality metrics
According to the objectives of this study, what matters is not the characteristics of each node or actor in itself, but in what way it relates to others in the network. In this sense, data scientists, mathematicians, and sociologists have collaborated over the years to create methodologies that allow calculating network analysis metrics, creating numerical bases that make it possible to compare and analyze behavioral patterns and study their evolution through time. For some authors, the diffusion of a message is proportional to the connectivity of the users, that is, “the higher the value of the centrality calculated for a node, the more important is that node for network connectivity and, therefore, for the diffusion of information” (Arcila, Barredo, and Castro, 2017, p. 69). In this sense, centrality measures have become more relevant in recent years, since they allow us to reflect how important (central) a node is within the network based on objective and quantifiable criteria. As other previous studies have corroborated, the number of followers or profiles followed is less decisive in the diffusion of messages than the activity of users and their position in the mentions network (Toledo Bastos; Galdini Raimundo and Travitzki, 2013).
In this work, to analyze which are the most relevant actors (nodes within the network) of the conversation, different measures of centrality have been calculated: the degree of centrality, betweenness centrality, and the eigencentrality for each case under study.
3.3.1. Degree centrality
The degree of centrality is the number of links (mentions) that are connected to a node (Twitter profile). Thus, two types of centrality degrees can be distinguished: the in-degree centrality is defined as the number of mentions that other actors in the network make about a specific account; out-degree centrality is defined as the number of mentions that an account produces. It is, therefore, a measure of the activity of an actor within the network.
3.3.2. Betweenness centrality
Betweenness centrality is a measure that quantifies how important an actor is within the debate when keeping the different nodes (users) of the conversational network communicated. It can be said that there are key nodes in the network, which serve as a bridge between different groups, so that, if they did not exist, these groups of users would be isolated from the network. Therefore, betweenness centrality quantifies how important a vertex is within the debate when keeping the different nodes of the network communicated. From a methodological point of view, betweenness centrality quantifies the frequency or the number of times a node acts as a ‘bridge’ when two nodes of the network communicate with each other through the shortest path. In other words, the higher the betweenness centrality value is, the more important is the node as an element of connection between different users.
3.3.3. Eigencentrality
This measure is based on the premise that mentioning a popular profile is more relevant than mentioning a solitary, isolated, and of-little-relevance node within the network. This metric takes into account not only the mentions of a profile but also the popularity of the users who mention it. Therefore, a high coefficient of eigencentrality means that that node is in turn connected to many nodes that are also influential. Some algorithms such as Google’s Page Rank use a variant of this measure to calculate the relevance of web pages on the internet.
4. Results
4.1. Traditional audience, Twitter conversation and evolution of the hashtag
In the global comparison of the traffic generated by each of the eight hashtags, it is verified that the results achieved in the debates of the different autonomous communities are very unequal. Table 2 details the following parameters: the set of mentions made by the account of each television, number of accounts mentioned by the profile, the interactions (links) that were generated using the hashtag, and the number of actors involved in the conversation.
Table 2.
Source: self-made. Audience data transmitted by Castilla-La Mancha Media and sample collected through the Twitter programming interface.
The data indicate that a larger television audience does not always mean a greater traffic of tweets during electoral debates: the televisions that reached a greater share quota (Canal Extremadura, Castilla-La Mancha Media, and Television of the Principality of Asturias) do not correspond to those which generated a more active debate on Twitter, except in the case of Castilla La Mancha Media that appears well-positioned in both parameters. On the other hand, it is noted that, in coincidence with other mentioned studies, that a greater number of followers in the profile does not translate into greater traffic during the debates. Only Madrid´s public television is at the top in the number of followers, number of interactions, and number of accounts mentioned. Instead, televisions such as that of the Canary Islands, with a very notable number of followers, did not act as a dynamizer of the encounter, using the hashtag only to promote the space, remembering the day and time of broadcasting. The underutilization of this channel translates into a very low number of interactions, indicating that there was no intense and participatory dialectic during its broadcast.
Regarding this, a significant fact is that there seems to be a correlation between the number of accounts mentioned by the television´s official profile and the amount of traffic generated by the hashtag. In other words, when the television is active using the hashtag and mentioning more actors, it manages to involve more accounts in the conversation and also enhances its interrelation. This is confirmed in the case of Telemadrid, Castilla-La Mancha Media, and La 7 de Castilla y León, televisions that, as will be seen in the graphs, occupied a relevant place in the conversational network of Twitter.
Source: self-made from the collected sample.
Graphic 1. The comparative graph shows the traffic of tweets from six hours before the debate until 24 hours after it.
One of the advantages of exploiting the Twitter channel as a diffusion medium for this type of encounters is that, once the broadcasting of the space is over, the television can also continue the discussion and subsequent analysis on the social network, not only from the set, as it is usual. The comparative graph of the use of the hashtag between the eight televisions before and after the debate clearly shows that the encounter held in Telemadrid was the subject of comments several hours after its broadcast. Specifically, it can be seen that it had an interest-peak early in the morning, coinciding with radio political gatherings and morning television magazines. Thus, Madrid´s television manages to keep its debate relevant and generating conversation the day after, something that also happens, although more tenuously, with the hashtag of Castilla-La Mancha Media. In the rest of the cases, the traffic is basically located parallel to the broadcast.
4.2. Node behavior and centrality measures
4.2.1. Central televisions: Telemadrid, Castilla-La Mancha Media and La 7 Castilla y León
The hashtags of public televisions in Madrid and Castilla-La Mancha are the ones that generated the most interactions and, in both cases, the graphs show that they occupied a central role in the conversation, which is more polarized in the case of the Castilian-Manchego debate. Thus, it is appreciated how each node or actor interacts with a group of followers who are isolated from the rest. Some of these main actors are the presidential candidates, whose profiles are among the top ten positions of the centrality measures, with Paco Nuñez (PP) and Emiliano García Page (PSOE) being the best positioned. Another aspect that stands out is that the account of the candidate of Ciudadanos, Carmen Picazo, has much more relevance than that of her training, which descends to the sixteenth position in betweenness centrality, something that does not happen in the other political parties.
In this case, the activity of the accounts @ccooagentesjccm and @forestales_ccoo stands out when mentioning accounts (out-degree), where they occupy the first and second place. Both profiles dealt with the proposals on the environment that the electoral programs of all parties collected and did activism through a joint conversation under the hashtag #AgentesMedioambientales mentioning an average of 50 users per tweet. The television of Castilla-La Mancha revitalized the debate through its corporate account and also the profile @cmm_noticias, account that was even more relevant when articulating the conversation of the different communities, occupying the second position in betweenness centrality and the first in eigencentrality.
In the Telemadrid debate, the measurements show the relevance of the account of Isa Serra, candidate of the Unidas Podemos formation, which occupies the first place in betweenness centrality and eigencentrality, being the second place in both measures for the account of the candidate of Ciudadanos, Ignacio Aguado, and the third for the account ofTelemadrid, which indicates that it played an important role in connectivity between users and the dynamization of the conversation. Besides, it was television that managed to involve more accounts in the conversation about the debate, with 7,303 users who intervened in some way in the parallel debate. In this case, the density with which the conversation is shown in the graph reveals the high number of links generated, a symptom that it aroused a very high interest among the social audience.
Source: self-made from the collected sample.
Illustration 1. Graph corresponding to the hashtag #ElDebateMadridVota.
Source: self-made from the collected sample.
Illustration 2. Graph corresponding to the hashtag #ElDebateCMM.
A third significant case is that of the television of Castilla y León, which achieved 15,198 interactions. Several main nodes are located in the graph: the corporate accounts of PSOE Castilla y León, and Partido Popular Castilla y León, and their respective candidates, Luis Tudanca (PSOE) and Alfonso F. Mañueco (PP), and Pablo Fernández (Podemos Castilla y León) and Francisco Igea (Ciudadanos Castilla y León). In these two examples, the accounts of their respective parties did not reach much notoriety and the conversation revolved around the leaders, indicating that PP and PSOE better capitalized the debate on the network by moving the hashtag with the accounts of candidate and party.
The graph also shows the central role of the entity to consolidate the conversation. The corporate profile ranks fourth in betweenness centrality and fifth in eigencentrality, so it had a significant position as a bridge between different groups that without it would be isolated by dividing the network into different subnetworks. It is the candidate of Podemos who is furthest from the central core of the conversation and who would be left out without the role of moderator/bridge that plays the account of Castilian and Leonese television.
Source: self-made from the collected sample.
Illustration 3. Graph corresponding to the hashtag #ElDebateCyLTV.
The three cases are a sign of good use of Twitter to channel the debate on the second screen and confirm the centrality achieved by the regional television account (public in Madrid and Castilla-La Mancha; private in Castilla y León) during the broadcast of their debate.
As for the activity of the television´s accounts, it can be said that they maintained a high level of interactions: 119 interactions and 26 accounts mentioned in the case of Telemadrid; 130 and 17 accounts mentioned in the case of Castilla-La Mancha Media; and 223 and 16 accounts mentioned by La 7 Castilla y León. In Table 3 you can see the distribution of mentions to political parties and candidates made by the televisions on Twitter.
Table 3.
Source: self-made based on the data collected in the sample.
The data confirm less attention to Podemos by Telemadrid (13 mentions between candidate and formation, compared to 19 obtained by the Partido Popular), and some fairly equitable dynamics of mentions in the case of the other two televisions, where differences between candidates and political parties were barely produced. Among those mentioned by the Castilian and Leonese television is not the Vox formation, although the candidate of the political formation did not participate in the debate.
On the other hand, Telemadrid did not only mention the ten accounts listed in the table; it also mentioned fourteen accounts that did not belong to political parties, among which were nine journalists or commentators of the chain who participated in the broadcast. These data highlight that a commitment was made to stimulate dialogue in social networks and even to verify live data since other mentioned accounts were Maldito Dato and Maldita hemeroteca, and that of its founder, Clara Jiménez Cruz. Despite this, compared to the 26 mentions made, the Telemadrid account was mentioned by 927 people, which would indicate that there was no communicative exchange between the entity and the viewers, but that Twitter was basically used as a content broadcasting channel.
4.2.2. The intermediate case: Aragón TV and Canal Extremadura
Aragón TV and Canal Extremadura show less activity and similar strategies, with 11 accounts mentioned in the first case and 10 in the second. Aragonese television falls to the tenth position in betweenness centrality and sixth in eigencentrality, so it is less relevant in terms of user connectivity and importance in the debate. In this case, the core of the conversation is clearly Javier Lambán (PSOE). Other important actors were Luis María Beamonte (PP) and Arturo Aliaga (PAR), while in Podemos and Ciudadanos the relevant accounts were those of the parties.
Source: self-made based on the data collected in the sample.
Illustration 4. Graph corresponding to the hashtag #DebateATV26M.
Source: self-made based on the data collected in the sample.
Illustration 5. Graph corresponding to the hashtag #ExtremaduraDecide.
In the case of Extremadura, television is in the first place in both parameters, resulting in the central core of the conversation, although its hashtag only produces 5,984 links, a figure far behind from the three previous cases that exceeded 15,000 interactions. In this sense, one aspect to consider is that it is not the same to register a high degree of centrality in a conversation of just a thousand interactions than in one that achieved the participation of many very active users.
When reviewing the diffusion dynamics on Twitter of the official account @cextremadura during the debate, it is verified that 74% of the tweets did not incorporate any mention. Besides, not all candidates (7) who participated in the televised debate were mentioned. The account mentioned once Guillermo Fernández Vara, of the PSOE; Juan Antonio Morales, from Vox; and Cayetano Polo, of Ciudadanos, also mentioning the accounts of their respective political formations. On the other hand, there are no mentions of three candidates who do not have a Twitter profile - José Antonio Monago (PP), Pedro Lanzas (Extremadura Unida), and Laura Márquez (Actúa) – nor of Irene de Miguel (Podemos) or her formation, although her account, @irenirima, was a central node in the parallel debate, as confirmed by her second place in the number of mentions received (in-degree), the third in betweenness centrality, and the seventh in eigencentrality. In fact, the graph reflects that both Podemos and the Partido Popular are somehow located outside the core of the conversation (Canal Extremadura), only establishing connections with their followers.
Meanwhile, Aragon TV only mentions candidates’ accounts, not political parties, and does so in an equitable manner, with two mentions to each of the eight leaders who participated in the television encounter.
Table 4.
Source: self-made based on the data collected in the sample.
4.2.3. Televisions without relevant activity: Televisión Canaria, La 7 (Murcia), and TPA (Asturias)
In these three cases, the television’s activity on Twitter was very low, which resulted in a much lower number of interactions than that obtained with the other hashtags. The Televisión del Principado de Asturias and Televisión Canaria adopted the strategy of only mentioning themselves. In the case of Murcian television, La 7, there is a greater number of mentions - 10 accounts - but the corporate account did not use hashtags during the debate and only used it as a prior claim. The media opened a thread explaining the rules, blocks, and dynamics that were to be followed and issued a reminder before starting. Despite not releasing tweets, it achieved a higher volume of traffic than the other two cases, with 4,413 interactions.
Although the graphs are not significant, the case of the #debateTPA hashtag, of the Asturian debate, is shown as an example, where you can see a scattered and poorly connected conversation, in which the activity of Adrián Barbón, candidate for the presidency by the Partido Socialista, Podemos Asturies, and the account of the Asturian Socialist Federation barely stand out. Although the account of the Asturian entity was at the center of the debate, as evidenced by the measurements in the ranking of betweenness centrality (2nd place) and eigencentrality (3rd place), the total traffic that the hashtag managed to generate was very small, with 1,572 interactions. This comes to corroborate that being central is not everything to achieve relevance if it is not accompanied by many actors talking about the broadcasted content.
Source: self-made based on the data collected in the sample.
Illustration 6. Graph corresponding to the hashtag #debateTPA.
On the other hand, with the hashtag #RTVCDebate 2,225 interactions were generated but the profile of the Canary Islands’ television, in this case, occupies lower positions in the centrality ranking (9th place in betweenness centrality and 7th in eigencentrality), so it was an irrelevant actor. As already noted, the television released a single tweet announcing the debate. The central core of the conversation during this broadcast was in the account @podemoscanarias, first place in the ranking in the two mentioned parameters and also in the number of times mentioned (in-degree).
5. Discussion and conclusions
In 2019, the proximity television broadcasting entities have opted for the debates, a public service initiative that requires a production effort, both in technical and human resources. However, their broadcasts have reaped mixed results. Share data show that the traditional audience varies in different regions; at the same time, it is confirmed that not all televisions have exploited these spaces on Twitter the same way. Madrid’s television, for example, achieved twenty times more interactions than the Asturian; that of Castilla-La Mancha, fourteen times more. In the light of the data, it can be affirmed that there is a relationship between the television’s use of the hashtag and the communicative exchange achieved in the social network since those televisions that mentioned more accounts managed to involve more users in the debate and jointly produce more interactions for the proposed tag.
Despite the possibilities of interaction, the social audience, a priori, present, it is barely taken into consideration by most regional television, using the corporate profile on Twitter, in most cases, as a platform for self-promotion of the special program or “to release original content and not as a communication channel in itself” (Rodríguez-Vázquez et al., 2016). The results highlight that there is currently no guideline on the diffusion dynamics of these contents on the second screen and each entity chooses to develop its strategy, which ranges from not making use of the hashtag proposed during the broadcast of the debate (La 7), to adopting a very active role in the broadcasting of space, becoming the central node of the conversation (Castilla-La Mancha Media, Televisión de Castilla y León, and Telemadrid). One fact to note is that Castilian and Leonese television, privately owned, is not only among the ones that most dynamized the debate on the network, but has been equitable in the number of mentions of candidates, although this debate was broadcasted in deferred, which may have influenced the Twitter-use routines. The least even turns out to be Canal Extremadura.
On the other hand, it is also perceived that not all political candidates occupy the same space in the digital debate. Some of them do not even have a Twitter profile (José Antonio Monago) which translates into an advantage for other candidates, who are able to concentrate the leadership in social networks during the broadcast. In this sense, the profiles of Isa Serra (Podemos) and Ignacio Aguado (Ciudadanos) stand out, who managed to concentrate great prominence around the Telemadrid debate. Both occupy the first positions in betweenness centrality and eigencentrality, which indicates that they were key points to connect users and that, in turn, the profiles with which they interacted were relevant. In this sense, it is as important to prevail in the televised dialectical exercise as in its digital replica. Some studies have confirmed that the candidates that gain influence in social networks are those capable of creating a synergy between traditional media and these new channels (Karlsen and Enjolras, 2016). Besides, social networks favor a campaign that is more personal and less framed in the acronym of the party, an issue that is more decisive when choosing representatives of territorial organizations.
An aspect that attracts attention is that in the ranking of profiles that most mentioned other accounts (out-degree) the first places are usually occupied by, in most cases, users who are not public figures, that is, they correspond to private individuals’ profiles who are, however, very active on Twitter and achieve notoriety in the conversation. This is the case, for example, of Brujastur, which with just 85 followers occupies the first position in accounts mentioned with #ElDebateMadridVota; or that of Anais Gallina, leading the same parameter for #ElDebateCyLTV. In fact, although this is not the case, the media have already echoed false accounts created by political parties or even marketing agencies to monopolize networks in these events*.
The current scenario suggests that it would be appropriate to agree on certain dynamics of the use of social networks for these types of television spaces that are not part of conventional programming and are subject to a specific regulation. It is noted that electoral coverage plans hardly mention social networks, obviating the existence of these new channels that, however, are essential to amplify the number of potential viewers of the program and reach the segment with voting capacity that is furthest away from conventional television, that of young people (Lago-Vázquez et al., 2016). The precautions to remain in the territory of impartiality may be deterring the entities when trying to lead the debate on Twitter, thus losing the opportunity to become relevant actors in every media. Another cause may be that audiences are still taken into account only from traditional measurements and, therefore, management teams concentrate their efforts on earning share points, without taking advantage of their Twitter accounts as a claim towards content.
* The traps of political parties on Twitter, Cadena Ser, Irene Dorta, 04/21/2019.
The results show that a renewal of the existing regulation is necessary, in a flexible manner and combined with self-regulation and co-regulation, to be able to adapt to the change that digitalization and technologies entail (Muñoz-Saldaña and Gómez-Iglesias, 2013), mainly in the coverage of special interest events such as debates or interviews by the media, especially those of public ownership, under a strong suspicion of politicization in Spain (Soengas-Pérez and Rodríguez-Vázquez, 2014; Humanes, Fernández -Alonso, 2015; Marzal-Felici and Zallo-Elguezabal, 2016; Soengas-Pérez, Elías-Pérez, López-Cepeda, 2018). An example of good planning is that of Telemadrid, which demonstrates that it is possible to adopt an active strategy of diffusion and broadcasting of these spaces in social networks. The ways to do it without departing from the neutrality framework are, either through systematic retransmission incorporated into the electoral coverage plan, or through the specific development of this issue in the style books. Both self-control tools are susceptible to change and renewal, while new technological challenges are being raised.
Bibliographic references
AUTHORS
Belén Galletero Campos: Belén Galletero Campos is a professor at the Faculty of Communication of the University of Castilla-La Mancha. She is part of the research team of two national R + D projects and has directed research projects with companies and entities such as the Women’s Institute of Castilla-La Mancha or the NGO Plan Internacional. She is part of the Mediacom research group, oriented to research on the structure of communication in the autonomous community. She has conducted research stays at the Centro de Estudos de Comunicação e Sociedade of the Universidade do Minho (PT) and has collaborated in the Re/media.Lab - Laboratório e Incubadora de Media Regionais project (Universidade da Beira Interior, PT). She is a member of the Advisory Board of Radio-Television Castilla-La Mancha on behalf of the University of Castilla-La Mancha.
belen.galletero@uclm.es
H-index: 2
Orcid ID: https://orcid.org/0000-0002-9549-9507
Google Scholar: https://scholar.google.es/citations?user=pRNNxaQAAAAJ&hl=es
Ana María López Cepeda: Ana María López Cepeda has a degree in Journalism from the University of Santiago de Compostela (USC), a Law Degree from the Universidad Nacional de Educación a Distancia (UNED), and a Ph.D. in Communication and Journalism from the University of Santiago de Compostela (USC). She is currently a professor at the University of Castilla-La Mancha (UCLM) where she teaches subjects on Media Structure, Media System, and Public Communication Policies. She is a Main Investigator of the Sociology of Communication Research Group at the UCLM. As a researcher, her areas of specialization are public communication policies, audiovisual policies, copyright and journalism, the structure of communication groups, and public service audiovisual media. In the professional field, she worked as a journalist and editor in written press and communication cabinets.
ana.lopezcepeda@uclm.es
H-index: 5
Orcid ID: http://orcid.org/0000-0001-8328-9142
Google Scholar: https://scholar.google.es/citations?user=cJm7LFIAAAAJ&hl=es
Arturo Martínez Rodrigo: Arturo Martínez Rodrigo is a Professor Assistant Doctor in the Faculty of Communication of the University of Castilla-La Mancha (UCLM). He is a Technical Telecommunications Engineer and a Graduate in Audiovisual Telecommunications Systems from the University of Castilla-La Mancha (UCLM) and a Doctor from the UCLM with an Extraordinary Doctorate Award in the area of Engineering. He is specialized in data processing and analysis with special emphasis on the statistical design of automatic learning models and advanced classifiers. He has conducted research stays at the University of Surrey (UK), at the Department of Mechanical Sciences, and at the Intelligent System Laboratory at the Universidade do Minho (PT). He is the author of more than 50 scientific papers in this field and an expert in the study of nonlinear dynamics and its application to the world of data, as can be derived from his research and state projects in which he participates.
arturo.martinez@uclm.es
H-index: 11
Orcid ID: https://orcid.org/0000-0003-2343-3186
Google Scholar: https://scholar.google.es/citations?user=mJ1GX94AAAAJ&hl=es&oi=ao
Ana María López Cepeda. University of Castilla-La Mancha. Spain.
ana.lopezcepeda@uclm.es
Arturo Martínez Rodrigo. University of Castilla-La Mancha. Spain.
arturo.martinez@uclm.es