Politainment on Twitter: Engagement in the Spanish Legislative Elections of April 2019

The mediatisation of politics is based on the logic of spectacle. Politainment defines the phenomenon in which political information is trivialised by the hybrid narratives in which it is included and its anecdotal tone, with the aim of reach‐ ing an audience that seeks entertainment rather than information. This phenomenon has reached the digital sphere; the media, political parties, and prosumers are interested in using the new communicative context to expand their audience or become producers of new narrative formulas that act as a loudspeaker for online infotainment policies or discourses. This research examines the engagement obtained by politainment producers on Twitter, a network where debates about televi‐ sion content are concentrated. The article examines the tweets issued by Spanish television programmes that carry out poli‐ tainment. The research focuses on the Spanish general elections held in April 2019 to establish whether this social network acted as a sounding board for television broadcasts and how it contributed to fixing ideas and content. The researchers conducted a content analysis on a sample of 7,059 tweets and 2,771 comments. The results show that the production, promotion, and communication strategies of programmes on Twitter are still scarce and unoriginal. The behaviour of pro‐ sumers is not very creative, active, or interactive, preventing the creation of a debate on Twitter or the construction of a horizontal (user–user) or vertical (user–programme) interaction on the content published.


Introduction
The processing of politics in modern political systems is based on the logistics of show business, which has developed into a television landscape closer to a cartoonish distortion of reality than the broadcast of serious political information.Television politainment has been present in Spain and several other European countries, North America, and Latin America since the 1990s (Conde-Vázquez et al., 2019;Prado & Delgado, 2010;Stark, 1997;Thussu, 2007).It can be defined as a global phenomenon whose development is due to the appear-ance of private television channels, the audiences' interest in sensational information, the approach of politicians to this format, and its profitability (Kovach & Rosenstiel, 2007;Moy et al., 2005;Neijens & Brants, 1998).Politainment can be regarded in different television formats: television news programmes, magazines, and talk shows, in which topics and features of infotainment are introduced (Berrocal et al., 2014;Díaz Arias, 2011;Früh & Wirth, 1997;Gordillo et al., 2011;Ortells, 2015;Pellisser & Pineda, 2014;Zamora-Martínez et al., 2022).Some examples can be seen in emotional and sensational news, in which the voice of the citizen and the journalist is relevant as in live connections, sound effects, and music post-production elements such as slow motion, split screen, and headlines, as well as recordings with dramatic, symbolic and subjective nuances.The classification of political programmes established by Cebrián and Berrocal (2009) distinguishes three genres: spectacularised political formats, politicised magazine programmes, and political info shows.The spectacularised political formats deal with current political affairs by participants from the traditional political-journalistic sphere and incorporate production formulas or elements from other television genres that distort the treatment of political information.The politicised magazine programmes mix trivial sections with political information sections, and the infoshows deal with purely political content from a humorous and entertaining point of view.In addition, its strong impact on the media landscape has transformed politainment into a formula that, according to some, threatens public debate and the democratic system by addressing political information superficially (Blumler, 1992;Moy et al., 2005;Prior, 2005;Sparks & Tulloch, 2000;Valhondo, 2007); although others argue that it improves news recall and fosters public interest (Baum, 2002;Brants, 1998;Ferré & Gayà, 2010;Salgado, 2008;Stockwell, 2004;Taniguchi, 2011;Thussu, 2007).
This kind of content not only occurs during electoral campaign periods but is also seen during so-called permanent campaigns.Likewise, there is a transfer to the digital environment: enticing stories, political scandals, and the proliferation of trivial anecdotes related to show politics that rule the virtual landscape (Berrocal et al., 2022;Dader, 2012).The growth of this journalistic trend of news sensationalism particularly affects social networks, as they are the place where politicians, media, and prosumers converge.Politicians try to use them as loudspeakers to send messages, promote themselves, make their agendas visible, and communicate with their audiences (Berrocal, 2017;Castells, 2010).The audience receives considerable amounts of television content related to the political environment (imitations, memes, and videos); thus, using these channels to obtain information, reply to, reproduce, and spread messages created by others, encourages the development of politainment on the internet.
Twitter's quick, simple, and public nature has made this social network the most successful for informative purposes.Regarding this popularity, politainment formats and contemporary political activity have increased their presence on this network because it concentrates the bulk of the debates on television content, grants greater broadcasting power to television programmes, allows interaction with the audience, and provides details about their preferences and opinions by sending instant messages which relate to the broadcast content (Zamora-Martínez & González-Neira, 2022).
The audience, which in the past, associated the activity of watching television with a particular time of day and to a specific location, can now be informed everywhere and at any time due to the power of social networks.This important change in the audience has led to them now being referred to as prosumers.This term refers to the digital engagement of individuals and is defined as a communicative interaction that shows itself in the form of clicks, likes, shares, comments, suggestions, and other content produced by users (Ballesteros, 2019;Dhanesh, 2017;García Orosa & Alafranji, 2021).The interaction of users by reproducing, voting, broadcasting, or commenting about television content on social networks such as Twitter means that the television programme ceases to be "an outdated product of a single broadcast to become a long-term project" (Sánchez Tabernero, 2008, p. 274), thanks to the "buzz that its broadcast causes, through different means" (Gallego, 2013, p. 20).This has encouraged changes in the ways of producing, distributing, and consuming information and different kinds of content (Riera, 2003).The symbiosis between television and social networks is becoming increasingly noticeable within the media landscape since it reinforces television content, broadens its life beyond the broadcast, and allows interaction with the audience.
In this line of work, this article examines, within the context of the 2019 general elections in Spain, the engagement gained by politainment broadcasts on Twitter, as it is considered the network where political debates about television content are mainly located (Giglietto & Selva, 2014;Halpern et al., 2016).Likewise, this social network was chosen because it is an open platform (Williams et al., 2013) that allows a constant information flow among users without any restrictions (Wu et al., 2011).Thus, the Twitter platform becomes an interesting environment to study the dissemination of information and the strength of the content broadcast spontaneously and uncontrollably (Leonhardt, 2015;Mohr, 2014).
The research pursues the following three objectives (O): O1: To study the broadcast of politainment programmes on Twitter to establish which communication strategies achieve the most engagement in tweets, and to identify the level of this engagement.O2: To examine the behaviour of the prosumer audience's responses in relation to politainment programmes.
O3: To determine when the tweets with the greatest repercussion are broadcast and when the greatest number of comments are issued, as well as to determine if the tweets of the programmes and the comments of the users behave similarly throughout the life cycle of the issue.Under these premises, the research is based upon hypotheses that include new proposals, such as estimating which tweets with a softer frame will achieve a greater engagement, or how the interactivity of prosumers will result in humorous comments.In addition, this research aims to determine when there is greater activity on this network and during what periods the tweets linked to television politainment are mostly spread.The following three hypotheses are detailed below: H1: The profiles of the programmes examined on Twitter achieve a greater engagement in their broadcasts when they include audio-visual documents and mentions, the frame is that of soft news, and they use colloquial language.H2: The audience's response to the tweets with the greatest interaction is characterised by the inclusion of audio-visual documents and emoticons, the use of colloquial language, and critical and humorous intentions.
H3: The tweets of the programmes with the most engagement and the audience's comments mostly take place during the broadcast and are concentrated at the beginning and end of the life cycle of the programme's broadcast.

Sample and Tool
To find out the engagement obtained by politainment programmes on Twitter, the content analysis method has been used, as it has been proven a very useful tool for systematically, objectively, and quantitatively analysing communicative messages (Berelson, 1952;Riffe et al., 1998;Wimmer & Dominick, 2010).The analysis samples of the study are made up of 7,059 tweets and 2,771 comments, collected from March 28th-April 28th 2019, during the run-up to the Spanish general elections.The selected dates include 15 days of the pre-campaign and 17 days of the electoral campaign, including both election day and the day of reflection.The reason for including both dates is due to the dissemination that politainment programmes make on Twitter during these days.On the other hand, only tweets about politics were analysed in this article, discarding those that talked about other topics or guests of the programme during the month of monitoring.The tool to download tweets from Twitter accounts was the paid version of Twitonomy.
The selection of politainment programmes comprises all those television programmes broadcast during the pre-campaign period and the electoral campaign.It has been shown that political information received a spectacular treatment either through the narrative style, the election, and the treatment of sources and themes, or due to the technical characteristics of post-production.The sample includes programmes that are entirely politainment and others that include specific sections belonging to one of the three sub-genres mentioned in the theoretical introduction.In total, 18 infotainment programmes were studied.All of them were programmes that were broadcast during those given dates: Los Desayunos (Televisión Española), La Mañana (Televisión Española), La Noche en 24 Horas (Televisión Española), Espejo Público (Antena 3), El Hormiguero 3.0 (Antena 3), Cuatro al Día (Cuatro), Todo es Mentira, El Programa de Ana Rosa (Telecinco), Ya es Mediodía (Telecinco), Mi Casa es la Tuya (Telecinco), Salvados (La Sexta), El Objetivo (La Sexta), La Sexta Noche (La Sexta), Al Rojo Vivo (La Sexta), El Intermedio (La Sexta), Más Vale Tarde (La Sexta), and Liarla Pardo (La Sexta).

Procedure
The research was carried out in four different stages.In the first stage, the 253 tweets with the highest engagement were selected from the total sample of 7,059 based on the interactions obtained: the sum of the retweets, likes, and comments.Figure 1 shows that 50% of the total interactions were concentrated in the most engaging 253 tweets.
In the second stage, a quantitative index was built to measure the level of engagement based on the average between the number of comments, the number of retweets, and the number of likes of three groups (high, medium, and low), with which the different variables that make up the analysis template of politainment programmes were crossed.In the third stage, the sample of the prosumers' comments-comprised of 2,771 tweets-was analysed.To do so, the number of comments that appear most frequently in the 253 tweets with the most engagement was considered, with 13 being the most common number detected.This is sought to establish a fixed value of comments in order to analyse and standardise the study.The value of 13 comments is represented in Figure 2 which is a probability distribution figure to determine which population slot is adjusted and which corresponds to a lognormal distribution.
Finally, the fourth stage refers to the temporary study of the programmes' tweets.This was done by sizing the time variable to carry out the study in all cases simultaneously and to ignore the different broadcast frequencies.To do this, a period was defined, which the authors of this research call the "life cycle of the broadcast" (Figure 3).It includes ranges from the start of a programme (N) to the start of the next programme (N + 1).By doing so, the tweet is located by deleting the time variable, and the analysis is homogenised for the different durations and time slots.

Content Analysis
The study required the design of two template models, one for analysing the tweets with the most engagement issued by the politainment programmes, and another for studying the comments of the social audience.In total, the programme template comprises 22 categories and 143 variables, while the prosumers template comprises 21 categories and 107 variables.The template is made up of three blocks.The first section focuses on the record and scope data, a basic dimension when coding, which includes the variables "date," "time," "URL," "programme," "number of comments," "number of retweets," and "number of likes."In this first part, the politainment programme template also includes the "broadcast period" category to identify the tweets that were issued during the pre-campaign and electoral campaigns.For the section designed to study the social audience, we included the category "user account," which aims to identify the prosumers making the comments.
The second block refers to the formal and descriptive aspects of the tweet.This section includes two different parts: the first refers to textual complements that encompass the categories "links," "mentions," "emoticons," and "hashtags"; the second refers to audio-visual attributes that address the type of "document," "main character," and "space."Regarding the differences in this block, related to the social audience template, the variables "link to other tweets" and "link to social networks" were included in order to provide a more complete view of the comments published by the prosumers.In the case of the programme template, the visual section is further explored.Three variables were formulated for politainment programmes: "spontaneous photography," "posing photography," and "selfie"; for the social audience, the variables were "image" and "linked image." The third block, entitled content aspects, brings together the categories of "information," "news topic," "context frame," "predominant frame," "language," "bias," "intentionality," and "feedback."The first two categories belong to the template of politainment programmes, and the last concern the social audience.The "information" category includes the variables: "hard news" (breaking public affairs involving political leaders, fast-moving current affairs, and information about disruptions or problems that are presumably important for citizens); "soft news" (sensationalist information that refers to events); "soft interviews" (interviews where decontextualisation, personalisation, and emotionality predominate, as well as a combination of hard and soft questions); "serious interview" (current affairs news addressed through the statements given by the interviewee); "whistleblower report" (exposes events affecting a certain community in which indications of illegality are confirmed); and "political parody" (political issues narrated through humour, dramatisation, and criticism).The "news topic" refers to issues such as election campaigning, political corruption, campaign politics, politician, and party as news.The "context frame" encompasses the variables proposed by Iyengar and Kinder (1987) called the "episodic frame," centred on the individual, on a particular event or incident which is not discussed in depth, and the "thematic frame," dedicated to important information of a general nature or problems in public life.The "predominant frame" includes the variables of "morality," "politics," "human interest," "attribution of responsibility," "sensationalism," "conflict," "humour," "public framing," "conjecture," and "consensus."The "language" category is subdivided into "formal language," which emphasises posts that are carefully and rigorously written, and "colloquial language," which highlights tweets that introduce grammatical errors, misspellings, and inaccuracies in the choice of vocabulary (e.g., crutches, swear words, and colloquial expressions).The "bias" category is organised into three sub-categories: "positive," "negative," and "not applicable."The "positive" variable refers to publications with friendly or favourable connotations towards the topic or the character addressed, the "negative" variable is precisely the opposite of the previous case, and "not applicable" refers to the absence of bias in the tweet.The "Intentionality" is divided into the variables "informative," "humorous," "critical," "self-promotion," and "other."Finally, the feedback contemplates the subcategories established by Sunstein (2009): "information cascade," when a new issue is detected in the comment (arguments, evidence, or data) different from what was commented in the programme's tweet.The "cascade of conformity" will prove when there is approval regarding what has been published, and "group polarisation" will be decided when the comments discuss and refute the content.The authors participated in the analysis as coders, with an agreement index using Holsti's formula (1969) ranging between 0.90-0.96.Disagreements were resolved by a review of the differences between the three coders.

Analysis of the Tweets of Politainment Programmes With the Highest Engagement on Twitter
After the study was carried out, it was observed that the greatest number of tweets made by the prosumer audience of the politainment programmes was concentrated during the pre-campaign (75.89%) and not during the official election campaign (24.11%), as would be expected.This phenomenon could be due to the fact that the pre-campaign period is already very active in the framework of modern political communication and the user's interaction decreases as time goes by due to the fatigue caused by the lifespan of the electoral activity.Concerning the metrics of the tweets, the main action was to like content (31.49%), followed by retweeting it (18.31%),while commenting was the least preferred (5.27%).This leads us to affirm that there is a low interaction level, considering that comments are the element of interaction related to a greater engagement with the social audience (Ballesteros, 2019).Likewise, the time slots that stand out for registering the most important engagement with the social audience are prime time, midday, and morning (Figure 4).
In relation to the textual research carried out, it is noted that those that manage to achieve the greatest engagement with the social audience are the hashtags of the topic of the day (36.63%), the programme itself (27.57%), and the main character (26.75%).It is also important to take into consideration the links that redirect the user to the URL of the channel where the programme is broadcast which enables the audience to follow it live (49.64%), the links to videos of the television schedule (37.96%), and finally, the mentions of political profiles (40.31%) and famous personalities (21.43%).The emoticons do not manage to achieve a great impact in the tweets of the television shows.Only 33.20% of the messages that include emojis encouraged the audience to interact, which leads us to affirm that using emoticons is not a successful strategy.
On the other hand, considering the most successful audio-visual aids to encourage engagement with the audience, screenshots (26.88%), videos (26.09%), and videos with links (21.34%) stand out as the strongest resources in this environment (Figure 5).This last variable is known in Twitter's jargon as Twitter Cards since they allow one to see a preview of multimedia information.The analysis highlights the success achieved by audio-visual documents, compared to texts, which may be due to the algorithms used in Twitter's platform that favour audio-visual content or to the prosumer's choice to interact with these tweets because such content can be understood faster than text.De Vries et al. (2012) and Jamieson (2006) state that social network users are more attracted to posts with greater vividness, that is, those in which sight and hearing come into play.
The topic "political and/or party as news" is the one that manages to achieve the greatest impact in the tweets of the politainment programmes.Also popular were those that contained soft news (77.86%) and those written with formal language (85.35%).The pseudoinformative nature of most of the programmes is the key to understanding the success of the formal language in the selected sample.Less popular were the ones that have an informative intention (43.08%) and those including criticism (36.36%), in line with the current affairs narrative of a large number of the politainment programmes studied.Finally, the episodic context frame (99.60%) and the sensationalist (22.13%), human interest (18.97%), and public frames (17%) are the ones that arouse the most engagement in the programme tweets, strengthening the idea that soft is what generates more interest and interaction, thanks to its ability to highlight the most striking facts.There is no evidence regarding the bias applied since the highest engagement is distributed similarly between the positive bias (37.94%) and the negative bias (37.15%) adopted by the tweet.

Behaviour of the Audience in Twitter Politainment Programmes
Regarding the results of the analysis of the user response, it is observed that the social audience comments mostly under total (44.60%) or partial (21.42%) anonymity: some users employ pseudonyms, fake names, and hidden identities; others only give part of their name and an impersonal photograph, which helps them to interact more and be less inhibited when responding (Figure 6).These results show that the official accounts of politainment programmes on Twitter have a community of strangers on whom much of the active participation relies.They are also in line with the social and communicative reality of Twitter: at the same time as they affect digital engagement, anonymity makes the social audience feel less inhibited, safer from being discovered (Domínguez & Hernán, 2010) and, therefore, more involved.Being real implies taking responsibility for what is expressed, which makes them take a more cautious approach.Also noteworthy is the low number of times that the prosumers make use of the retweets (30.71%) and comments (18.98%), compared to likes (63.75%; as seen in Figure 7) with which they intend to leave proof of having seen or read the tweet, thus supporting what is stated in it (Gerodimos & Justinussen, 2015).This trend shows that the users are comfortable being passive in this environment (Muñiz et al., 2017).However, when the user participates with a comment, it is mostly to start a discussion about the tweets (18.48%), described by Sunstein (2009) as group polarisation, or to reaffirm the majority message (13.75%), which the same author refers to as a cascade of conformity.
On the other hand, the comments stand out for their lack of elaboration, with links (3.28%), hashtags (12.20%), emoticons (17.48%), and mentions (40.03%) in the comments being quite unusual user responses.Audio-visual documents are also rare, as they are only included in 9.42% of the comments.This could be due to the adult public towards which the politainment spaces studied are oriented, as these resources are more common among young people (Azuma & Maurer, 2007;Baron & Ling, 2011;Thurlow & Brown, 2003).
In relation to the study of the audience's responses, the frequent use of colloquial language in the comments (62.36%) stands out, suggesting that the users seek to express themselves in an open way, similar to the spoken language (Figure 8).The two intentions in which the responses are included are critical (37.28%) and informative (30.71%).The behaviour, in this case of the social audience, is due to the public nature of the message or the desire to gain some benefit or advantage from the rest of the community, either in the form of a comment, retweet, or like.The episodic setting (92.71%), based upon the private life of individuals or certain events, is the most common, together with the public setting (53.48%), in line with the news set by the programme or the moment.On the other hand, the bias of the messages is ambiguous since very similar percentages are found regarding the positive ( 25

Temporal Analysis of Programme Tweets and User Comments
Regarding the temporal analysis that was carried out regarding the tweets of the programmes and the comments, it is clear that the politainment programmes achieve their highest engagement (62.65%) through the tweets issued during the broadcast schedule, especially within the first hour after the posting of the tweet (52.8%).This behaviour, which is observed in the first section of the life cycle of the broadcast, is modified depending on the time slot to which each programme belongs.
In this sense, it is observed that the programmes broadcast in prime time and at night accumulate their greatest interactivity during the first part of the life cycle, while the programmes broadcast from morning to afternoon shows a greater dispersion (Figure 9).These results show how greater interactivity could be achieved in politainment programmes according to the time slot.For example, during midday, after lunch, and afternoon slots, a significant interaction is seen during the last part of the life cycle, that is, in the time before the broadcast, which could be used to encourage engagement with the social audience.Taking a closer look at the behaviour of the programmes and users during the broadcasting period of the television space, we see that different patterns are detected depending on the time slot.The clearest trends are the linear behaviour during the broadcast period in prime-time programmes, a greater incidence in the first section of the period in the morning programmes, and a greater incidence in the final section in after-lunch television shows.These behaviours are reflected both by programmes and users (Figure 10).

Conclusions
This research presents new results within the field of active audiences, such as the study of the timing of the social audience's activity depending on the broadcast schedule.These results can be used as a basis for future research and the action of television managers since it will allow them to design a digital strategy that favours engagement in today's highly competitive television market.
The study partially supports the three hypotheses raised at the beginning of this research.The first stated that the profiles of the programmes examined on Twitter achieve a greater engagement in their broadcasts when they include audio-visual documents and mentions, the frame is that of soft news and when they use colloquial language.The findings show that politainment programmes achieve greater engagement when they include audio-visual documents, contain soft news, and introduce mentions, but not when they use colloquial language.In this context, in most cases, the tweets with the highest engagement are written in formal language.This may be due to the pseudo-informative nature of the programmes examined and the fact that social audiences expect politainment programmes to have a coherent style between what is said on television and what is expressed on Twitter.
The second hypothesis stated that the response of the social audience related to the tweets with the greatest interaction is characterised by the inclusion of audio-visual documents and emoticons, the use of colloquial language, and critical and humorous intentions.The objective is considered partially fulfilled because the audience avoids using emoticons and audio-visual documents, and they do have a humorous purpose when responding, as it was presumed, but they do normally use colloquial language and critical intent.
Finally, the third hypothesis is related to the fact that the tweets of the programmes with the most engagement and the users' comments occur during the television programme broadcast and are concentrated at the beginning and the end of the life cycle of the broadcast of the programme.In this case, the results show, in the first place, that politainment programmes obtain the greatest engagement through publications issued at the same time as the programme is broadcast; however, the largest number of comments take place after the programme is broadcast.Secondly, it is observed that the trend is to gather the majority of tweets and comments during the beginning of the life cycle of the broadcast, while the concentration that takes place at the end of it is quite insignificant.
This research supports the results of recent work in which it is observed that the social audience does not behave as creatively, actively, and interactively as might be expected (Lin & Chiang, 2019;Rodríguez et al., 2017), in this sense, no political consequences were perceived given the low activity of the social audience.In the same way, in the action of the profiles of the programmes examined, inconsistent and not very innovative production, promotion, and communication strategies are also observed (Coromina et al., 2020;Franquet et al., 2018;Sequera, 2013).These actions are not adapted to the platform's characteristics, and the resources it offers are not used productively, nor is engaging content created to generate a greater interaction with users.Apart from that, it is evident that spectacularised political information broadens its audience significantly when placed on Twitter and that soft, sensational, personalistic news topics that contain audio-visual elements produce a greater engagement.This deserves significant attention from the academy due to the complexity of its consequences: on the one hand, the approximation of politics and its

Figure 8 .
Figure 8. Examples of comments using colloquial language.Note: The translation of the tweets is as follows (a) "There @Albert_Rivera small cookers!!, The only one who has cooked what he has.Nothing like working hard", (b) "Every morning when I go to the toilet I remember a lot about casado and many like him and even more when I am constipated YOU CAN'T IMAGINE HOW I SHIT ON THEM."

Figure 9 .
Figure 9. Evolution of tweets/programme's comments during the life cycle by broadcast slot.
Engagement of the tweets according to the time slots.
Figure 5. Engagement of the tweets related to audio-visual aids.
.51%) and negative (27.86%) biases related to what is being said.Classification of the user account according to its typology.