How do legacy and digital media curate coronavirus content. An assessment of newsletters from the USA and four European countries

This study focuses on the curation of newsletters specialized in Covid-19 news from the mass media of France, Germany, Spain, the United Kingdom, and the USA. The newsletters of 21 media outlets –15 traditional and six digital natives– were studied during the peaks of the pandemic in April and November 2020. The study follows an evaluative method of curation quality, based on the analysis of the following parameters: number of curated contents, time range, origin, sources –according to organization type and morphology–, authorship, sense-making techniques, and hyperlink informational function. The results identify the main characteristics of news curation in these newsletters and their differences from general newsletters, as well as the sources used, and also make it possible to establish curation quality standings and to identify differences between countries and among legacy and pure digital news outlets.


Introduction
The Covid-19 pandemic has had a significant impact on the media, both in the news consumption habits of audiences, and in the organization of the media's professional routines and the type of information they offer to their readers.
News consumption increased considerably in 2020, particularly during the first months of the year. The public turned to the media, particularly traditional media, more than they had so before, given the need for up-to-date and more reliable information than that available from other sources such as social media. This took place against a background of uncertainty typical of a situation of a health crisis that was also characterized by a proliferation of misinformation (Casero-Ripollés, 2020;Nielsen et al., 2020;Rosenberg; Syed; Rezaie, 2020).
As a result, during the pandemic, the media encouraged remote work and other changes in newsrooms, and there emerged, with greater or lesser success, several news products addressed to meeting the specific news needs of their audiences in times of pandemic (Newman, 2021;Gupta, 2021). One example of this approach to the new "user-centric" news product that emerged during the pandemic is the newsletters that specialize specifically in coronavirus news. These were launched by different media worldwide against the background of health crisis and information overload (Silva-Rodríguez, 2021) and are the subject of this study.
Although a long-standing service of the digital news media, newsletters have revived and are experiencing a period of great expansion, as shown in recent studies (Newman et al., 2019;Newman et al., 2020). The last few years have seen a clear surge of the interest of readers in electronic newsletters, while email continues to play a major role in people's news habits (Fagerlund, 2016;Newman, 2020). The validity and topicality of newsletters are such that they have even been described "as an icon of a new way of conceiving and 'doing' the media today" (Santos; Peixinho, 2017). Some keys to this success may lie in the appreciation now felt by audiences for "finite editions" (Suárez, 2020), in this era when digital newspapers and the timelines of social media such as Twitter are seen as a continuous and interminable source of news (The New York Times, 2019). Other reasons could be their ability to channel communication between subscribers and authors (Isaac, 2019; Santos-Silva; Granado, 2019), and to regularly offer a selection of valuable content from the available information circulating on the Internet, thus making them a very appropriate channel for news curation (Rojas-Torrijos; González-Alba, 2018).
Thorson and Wells (2015, p. 31) define curators as "active selectors and shapers of content working under conditions of content abundance", with the task of receiving, filtering, reframing, and remixing messages. These authors propose a framework that includes Journalistic, Social, Personal, Strategic, and Algorithmic curation, in which the news flows are curated respectively by media, social media, individuals, strategic actors -such as politicians, corporations, governments, and interest groups-and computer algorithms.
Our research fits in the first one of these five categories, journalism and news media. In this context, Guallar and Codina (2018, p. 783) define journalistic content curation as: "a complex set of activities that include: 1) search and monitoring, 2) selection, 3) analysis and verification, 4) management and publishing, and 5) characterization or sense-making of the information published online. This involves the 6) dissemination of such products through digital platforms".
This research is focused on how news is being curated in this new "iconic" product of current journalism while viewing curation as a significant and essential element of most of the journalistic newsletters currently being produced. It can, consequently, be classed along the lines of the media's own line of research on news curation (Cui;Liu, 2017;Codina, 2018), and not of the news curated by the public on their social media networks (Bruns, 2018;Masip; Ruiz-Caballero; Suau, 2019).
As yet, however, specific research on newsletters and their prominent presence in current digital journalism has not been very extensive. Notwithstanding, some studies in recent years on digital news media newsletters can be cited.  (2021), to which this current study may also be added.
The goal of this research is therefore to analyse the use of news curation by new newsletters specializing in Covid-19 that have been introduced in the media of several countries during the pandemic.

Method
The research entailed descriptive and evaluative methodologies that include expert analysis and content analysis techniques (Strauss; Corbin, 1990; Creswell, 2009; Morales-Vargas; Pedraza-Jiménez; Codina, 2020). The CAS (Curation Analysis System), described in Guallar et al., (2021b) and used in other research (Guallar et al., 2021a) was followed. This method uses an analysis system based on parameters and indicators that are divided into two major dimensions: Content and Curation. Table 1 shows, for each dimension, the parameters observed and the indicators for evaluating them. A binary (0-1) indicator scoring system was established to evaluate a characteristic's absence (0) or presence (1). Only in the Quantity indicator, of the parameter "Number of contents", was scoring multiple (0-3) so that characteristic could be assessed according to the following scale: negative (0); average (1); positive (2); very positive (3). In this methodology, the quality criterion used when assessing newsletters curation is based on the number of indicators present in any given parameter: for instance, when assessing the sources according to organization type, a newsletter will have a score between 0 and 4 according to the presence of any of these four different sources: official sources, corporate sources, media, and citizens. [1] The  Newman et al., 2020;Mayhew, 2020;Tobitt, 2020). However, the scrutinization of traditional and digital media showed that not all of them offered their readers newsletters that specialized in monitoring the pandemic. For example, a newspaper as consolidated and influential as France's Le Monde did not publish a specific newsletter about the crisis. The authors, therefore, decided to study those media that were indeed issuing a newsletter about coronavirus, regardless of whether these had the largest audiences, the deepest tradition or influence, or were traditional or digital. The final selection is shown in Table 2.
The newsletters of 21 media from the 5 selected countries were therefore studied. Fifteen of these are published by traditional media and six by digital natives.
The study focused on the two peak moments in which the pandemic most affected Europe and the United States in 2020: in April, the peak of the first wave, and in November, in the second. One newsletter was observed every week (not all media offered daily newsletters and periodicity fluctuated greatly). The newsletter issued on Fridays (or the day closest to Friday) was chosen. In some cases, media that published a newsletter in April stopped doing so when the first wave subsided and did not republish from November onwards (Politico and BuzzFeed in the United States, El País in Spain, Die Welt in Germany, and The Guardian in the United Kingdom). elDiario.es, meanwhile, began issuing their newsletter on a weekly basis in November and published it on Sundays; 5 newsletters were studied. A total of 149 newsletters were therefore analysed: 84 from April and from 21 media, and 65 from November and from 16 media.
To validate the reliability between the researchers, previous to starting the analysis all observers were first trained to conduct newsletter codification according to the literature (Guallar et al., 2021b). After the analysis, each of the three observers codified a sample of newsletters individually and an intercoder reliability test was performed using Krippendorff Alpha methodology (Krippendorff;2004). The results were α=0.791, α=0.8106, and α=0.885 respectively, indicating a high degree of consistency between observers.. Table 3 shows some general aspects of the newsletters analysed. Note that in every case but one, the titles of the newsletters contain the word "coronavirus", which is therefore considered the main identifying term of the content, and that newsletter issue frequency is mainly daily, with a few exceptions of twice a day, three times a week and weekly.

Analysis of the Content dimension
The results obtained following the CAS method are presented below and the data set may be consulted at: The results start with the parameters of the Content dimension.

Quantity of curated content
This research found that the newsletters evaluated that publish the most curated contents include around 30 informative items. Using this value as a benchmark, this indicator was evaluated using the following scoring system: number of contents curated 1-10: one point; 11-20 contents: 2 points; more than 20: 3 points.
In this indicator, in which no significant differences were observed between the two periods studied, six newsletters obtained the top score of 3 (38% of the total), closely followed by four more (with scores above 2.6). Slightly fewer than half of the newsletters analysed, therefore, lie in the upper range in terms of numbers of curated contents. This, together with the fact that only two newsletters scored lower than 1, and that a large group of nine obtained scores of between 1 and 2, shows that the newsletters analysed were by and large well curated.
Specifically, the newsletters with the most curated content and the highest score (3)

Time range of curated contents
Four criteria are considered in this parameter: -Hindsight or non-time specific information: published in previous months or years, or when the content is non-time specific. -Recent information: from the last few days and weeks. -Current information: from the last 24 hours. -Real-time information: live or constantly updated.
The newsletter is a product with its own personality that offers readers a more personalized experience Seven newsletters scored very highly (from 3.1 to 4 points) in several time ranges. The top score (4) was obtained by only one newsletter, Spain's elDiario.es, during the first wave. In the second wave, its score fell to 3.2, thus yielding an overall average of 3.6, which was also the highest for all the newsletters in this study. Next came five of the United States' newsletters of the six studied (The Boston Globe, BuzzFeed, The New York Times, Politico, The Washington Post) and another Spanish newspaper, Nació Digital. The four possible time ranges were used, to a greater or lesser extent, in all these me- dia, thus meaning that the curation quality offered for this parameter was high. In the lower positions in the time range parameter, five newsletters made use of up to three ranges, six newsletters of up to two, while there were three more newsletters that used just one time range. The time range used in all cases by all the newsletters is Current information (published in the last 24 hours), combined, depending on the case, with information from one or more than one time range, including information updated in real-time.

Origin of the curated content
The study analysed whether the curated contents were produced by the media outlet itself or were external. Results show that eight media outlets always incorporate both types of content in all their coronavirus newsletters. These media are the United States' newspapers BuzzFeed, The Boston Globe, The New York Times, and Politico, the French publications Le Figaro and Le Parisien, The Guardian from the United Kingdom, and elDiario.es from Spain. In the second wave, the US news outlet Vox also incorporated them both, even though it did not do so for all its newsletters in the first wave, because, paradoxically, in one newsletter (from 10 April) it did not include any of its own contents. Likewise, four other newspapers sometimes (but not always) used external content together with their own. It should be clarified that in many of the cases of combining in-house contents with external material, there is less of the latter, as will be observed in greater detail below in section Analysis of external sources. Lastly, eight media exclusively used content published by their own outlet and none from external sources, an aspect that might suggest a low level of curation quality. These newsletters, which may be considered to fall short in this aspect, are the German Berliner Zeitung and Frankfurter Allgemeine Zeitung, the British Daily Telegraph and The Times, and the Spanish ABC, El Confidencial, El Mundo and El País.

Information sources according to organization type
Four categories were considered: official sources (from public administration bodies), corporate sources (from companies, associations, and other private organizations), the media, and private citizens. One newspaper, the Spanish elDiario.es, obtained the top score for this indicator, which shows that in all its newsletters, it constantly used the four types of sources. In second place was Politico, from the United States, which always used three types of sources. Following closely behind were Vox, BuzzFeed, and The Guardian, which used (although not always) three types of sources, and then came Le Figaro and Le Parisien, which used two kinds of sources. There follows a group of five newspapers that used two sources, albeit not in all their newsletters and to different degrees. Last comes a final group of eight media that used only one single type of source. Here, the only source consulted was the media, which is the only type always used in all the newsletters. The other three types of sources are fairly evenly distributed, generally at low levels, in some newsletters.

Information sources according to their morphology
Four types of sources were considered: web pages, blogs, social media and secondary sources. The highest scores, in the best cases, were between 2 and 3 and obtained by the newsletters of elDiario.es and Politico (only elDiario.es scored 3 points in the first wave) and by Vox and The Washington Post in the first wave, although their scores fell ostensibly in the second. The newsletters of the outlets ABC, BuzzFeed, El Mundo, Le Figaro, and The Daily Telegraph always used two types of sources. The most frequently used types of sources, meanwhile, were web pages, in all the newsletters. In some newsletters, these are combined with social media and blogs, while there were very few secondary sources.

Authorship
Most of the newsletters analysed are authored by their own editors, which may mean one person or several people. This is generally associated with curation quality, which is also apparent in the other parameters. The list of authors is shown in sample analysed no personalized authorship was given but was rather ascribed to the media outlet in general. This generally coincides with newsletters featuring less journalistic input, except for BuzzFeed and Le Parisien in the first wave.

Sense-making techniques
This section looks at the techniques for adding value to curation (known in the content curation literature as sense-making; for example, Deshpande; 2013; Guallar et al., 2021b). It considers the following techniques: summarize (informative technique); comment (opinion or interpretive technique); quote (based on giving a verbatim quote to the original content); and storyboarding (which consists of bringing together items from different formats through narration).
Four media outlets scored well in sense-making and obtained 3 out of 4 possible points for their newsletters: only one outlet, The Boston Globe, obtained this score in the two waves studied, while the others obtained it in one (BuzzFeed and The Washington Post, in April, and elDiario.es, in November). These media, together with Politico, are the only outlets that used (in all or several of their newsletters) more than two sense-making techniques. At the other extreme, three media, ABC, El Confidencial, and El Mundo, from Spain, use no technique and publish newsletters with fully automated content without adding further value to the headline and the link.
The most commonly used technique in all the newsletters except for the three that use none is Summarize, which is therefore confirmed as the essential technique for journalistic curation in newsletters. The other three techniques also appear in the newsletters analysed, with variations depending on the newsletter. Comment is used widely in elDiario. es, Nació Digital, BuzzFeed, Politico, and The New York Times, which are newsletters with a more personal editorial approach. Quote is apparent in Vox, The Washington Post, The Boston Globe, and Le Figaro; while Storyboarding is used in all of the following: elDiario.es, Nació Digital, The Boston Globe, BuzzFeed, Frankfurter Algemeine, Süddeutsche Zeitung, and The Guardian.

Link function
This section looks at the journalistic or informative function of each hyperlink or link to curated content within the newsletter. Based on Cui and Liu (2017) and Guallar et al. (2021b), it distinguishes the following categories: Unmodified; Describe or Sourcing curation (the hyperlink describes or summarizes the content); Contextualizing curation (the link is used to contextualize or document information); Interpreting curation (the text of the link is the author's interpretation or opinion); Cite source; Cite author; and Call to action.
The four top-scoring newsletters in this section are elDiario. es, Politico, El País, and The Boston Globe, each with scores of over 4 out of 7. They are closely followed by BuzzFeed, Le Figaro, Nació Digital, and The New York Times. The most widespread use of links involves Describing or Sourcing curation, in all the newsletters except for the three Spanish outlets with automated content. There is also some diversity in the use of the links. The next most frequent are Unmodified, Cite the source, and Cite the author. The Interpreting curation option, meanwhile, which indicates more personal curation, in conjunction with the commenting technique, also appears a lot in Politico and Nació Digital (only in the first wave) and in The New York Times.

Frankfurter Allgemeine newsletter
The New York Times newsletter

Analysis of external sources
In answer to RQ3, the external sources used in the newsletters were studied in detail using in-depth qualitative analysis. Our analysis is grouped according to types of media (legacy and pure digital) and within these, by country, with the observation of different trends for each category.

Legacy media
In all cases and in all countries, except for The Guardian in the UK and The New York Times, legacy media offer practically no content that is not produced in-house. These two outlets nonetheless offer a low proportion of external content, in comparison with links to their own content. The Guardian, which only offered its readers newsletters in the first wave, published 13 links to external content (15% of the total). This curated content links mainly to expert medical sources, specialized scientific journals, and institutions such as the WHO, or the websites of health authorities in the United States or the United Kingdom. There were only two links to the media (The New York Times and Reuters news agency).
The New York Times, for its part, offered 22 links to external content in the two periods studied (11% of the total). Most of these were to other media, which even included competitors (such as The Wall Street Journal, The Washington Post, or the Reuters and AP news agencies). There was also a not very significant number of links to specialized medical sources and official organizations. Both these newspapers always link to other web pages and not to secondary sources, blogs, or social media (except just once in the case of The New York Times).
Among the Spanish legacy media analysed, not a single outlet offered external content. In each newsletter, both El Mundo and ABC published a link to their Twitter and Facebook accounts and ABC also included a link to Instagram. El País (the only outlet to publish the newsletter solely in April) included only one link to YouTube and two to Spotify and no more.
The only two French media analysed included very few links to external content (there were three in Le Figaro in all its newsletters, and two in Le Parisien in the first wave and one in the second). These were to official public service websites that offer information on the forms required to avoid confinement or information on where to report cases of abuse. They did not link to other media or to blogs or social media accounts.
The German media also included virtually no external links. There were only two (Die Welt linked to The New York Times once, and once to the German radio station RBB24) in the four media analysed and in both periods. The Süddeutsche Zeitung included two links to the digital formats of its own Jeztz group.

Pure digital
In pure digital media (only present in the USA and Spain) a very different trend and one with constant nuances was observed. In the United States' media, the number of external content in pure digital media is much higher than in legacy media. Politico includes 25% external content in its newsletters while BuzzFeed and Vox notably offer much more external curated than in-house produced content: As regards the dilemma between producing edited or automated newsletters, the former prevails Politico newsletter were also some links to social media sites such as Facebook, Twitter, Instagram, and TikTok. In the second wave, elDiario.es offered fewer external links (44), albeit of the same diverse nature.
In the United States, BuzzFeed linked mainly to external media (newspapers, television networks, agencies, and local media, which are always from the United States). Only 5 of the 40 links to external content were to social media -specifically Twitterwhile the remainder was to the media. Vox, on the other hand, provided a much more varied external content. Its monothematic newsletters provided links to specific contents that specialize in the specific matter at hand, and were to media, official sources, corporate sources, medical websites, databases, and some (few) social media accounts, specifically Twitter. Table 4 shows the final score of the newsletters yielded by our analysis, based on the Curation Analysis System (CAS) (Guallar et al., 2021b). In both waves, the Spanish newspaper elDiario.es clearly obtained the highest score. Following closely behind, in the first wave, were the newspapers Politico, The Washington Post, The New York Times, The Boston Globe, and BuzzFeed from the United States, and Le Figaro from France, and in the second wave, only The Boston Globe. This group of media, the large majority of which are US newspapers, plus a leading Spanish and another French news outlet, can thus be considered the news media that offer the highest quality newsletters insofar as coronavirus news curation is concerned. Note that the scores were higher overall in the first (April 2020) than in the second wave (November 2020), with a few exceptions such as the French newspaper Le Parisien, the score of which improved in its November newsletter, and that in the second wave there were five media outlets that did not publish newsletters.

Newsletter standings
The first group of newspapers considered to offer the best news curation is followed by a second group of media in a halfway position in terms of quality of news curation: The Guardian from the UK and the Spanish El País (only in the first wave), Vox from the United States, the Spanish Nació Digital and the German Süddesutsche Zeitung. At the foot of the table, there are newspapers from all the countries studied (except the USA), the curation quality of which is poor and based exclusively or mainly on a list of links with little or no journalistic contribution.

Discussion and conclusions
In answer to RQ1, the research has made it possible to identify the characteristics of news curation in the coronavirus newsletters of newspapers in the 5 countries studied (France, Germany, Spain, the United Kingdom, and the United States of America).
Most newsletters included from 11 to 30 curated contents per newsletter. In terms of time range, as expected, current information published over the last 24 hours predominated. This was sometimes combined with information from other time ranges, which indicates greater use of non-time specific information (usually regarding health advice) in coronavirus newsletters than in general information newsletters (Guallar et al., 2021a). As regards the sources of information, there are not only differences among newsletters that curate external sources on a supplementary basis to their own, and those that never link to information that is not their own, but also the variety of types of sources used in coronavirus newsletters is larger than in general newsletters, although media sources continue to predominate (Rojas-Torrijos; González-Alba, 2018).
The presence of identified authorships is also significant. This may be due both to an evolutionary trend in newsletters towards the greater presence of signature newsletters and to the fact that they are specialized information services (Silva-Rodríguez, 2021). The use of links for information purposes is, therefore, quite varied even with the predominance of sourcing curation as in other studies (Cui;Liu, 2017); while the most commonly used sense-making technique is summarize, although other techniques such as comment, quote, and storyboarding are slightly more common in coronavirus than in general newsletters. All this suggests that the heralded potential of newsletters as a channel for exploiting and experimenting in content curation (Jack, 2016) is unfolding and that, as regards the dilemma between producing edited or automated newsletters (Fagerlund, 2016), the former are starting to prevail as part of a positive trend that points to greater use of intellectual or author journalistic curation (Guallar et al., 2021a;Silva-Rodríguez, 2021).
It has also been observed that curation in newsletters that specialize in coronavirus shares most of the characteristics observed in previous studies on general newsletters (Rojas-Torrijos; González-Alba, 2018; Guallar et al., 2021a), although there has been a slight improvement in all the indicators observed in comparison with general newsletters: there is a greater presence of authorship, time ranges supplementary to current information, a greater variety of sources and a broader range of sense-making techniques and not just summarize. Some nuances were observed in coronavirus newsletters that suggest slightly higher quality curation than in general information newsletters, in most of the media analysed. This echoes the results published by Silva-Rodríguez (2021).
In answer to RQ2, it was possible to rank news curation quality in the coronavirus newsletters analysed. The newsletter of the Spanish elDiario.es obtained the top score, followed by those of a numerous group of newspapers from the United States (Politico, The Washington Post, The New York Times, The Boston Globe, and BuzzFeed) and the newsletter of the French outlet Le Figaro. The newsletters of all these media can be described as examples of good curation practice and the main benchmarks or models that should be followed to ensure quality curation in newsletters, at least in terms of specialized information on coronavirus.

The Guardian newsletter
The results suggest a slightly higher quality curation in coronavirus newsletters than in general information newsletters Profesional de la información, 2022, v. 31, n. 3. e-ISSN: 1699-2407 11 The bibliography features no references, however, to curation quality standings in media from different countries other than in the recent study by Guallar et al. (2021a) on Spanish newspapers. This study, therefore, presents an initial look at the international situation, which should be verified and confirmed in future studies.
As regards the in-depth study of sources proposed in RQ3, legacy media were found to offer little content from external sources, while pure digital media, led by Spain's elDiario.es and BuzzFeed and Vox from the United States, were the outlets that offered the most, and in some cases even more links to external websites than to their own website. It is also worth noting that all the media analysed linked to the contents of media web pages and, to a much lesser extent, to official and corporate sources (very often associated with health) or citizens. The origin of the contents and the format was, again, most diverse in the pure digital elDiario.es, Politico, Vox, and BuzzFeed (not only websites but also blogs and social media). When they did offer content from media other than their own, it was usually from major international benchmark media, which includes both newspapers and television networks or news agencies. Links to specialized content were, in most cases, to organizations such as the UN, the WHO, or to expert medical websites. Insofar as the use of sources is concerned, the pure digitals were the media outlets with newsletters that most resembled their own model of digital journalism, considered as "bits organizing other bits", because digital media "do not provide [new] bits of information, they organize information that is already available someplace else (within the newspaper or outside it)", as put by Santos and Peixinho (2017, p. 780).
In answer to the two-pronged question RQ4 (differences between countries and differences between legacy and pure digital media), large differences were found among countries. No pure digital media outlets from France, Germany, and the United Kingdom were observed to offer their users specific curated newsletters about coronavirus and only digital media from the United States and Spain did so. Another fact worth mentioning is that in France, a country with major benchmark media such as Le Monde and Libération (to mention just two examples), only two newspapers, both legacy media outlets, saw the health crisis caused by the pandemic as an opportunity to offer their users specialized and personalized content, thus fulfilling a basic characteristic of newsletters (Silva-Rodríguez, 2021; Santos; Peixinho, 2017). If we compare legacy media among countries, the three outlets from the United States stand out from the others in terms of the quality of their curated contents; at the other end of the scale were the Spanish, German and British outlets (with the exception of The Guardian, which performed better than the others).
According to our results, therefore, the United States is the country with more coronavirus newsletters that score in the highest ranking of curation quality in both the first and second waves of the pandemic. The French media (except for Le Parisien in the first wave) also performed well while elDiario.es from Spain was ranked best of all in both periods. In their newsletters, these benchmark media published curated content that provides access not only to their own main news, but also links to a variety of media and institutional, corporate services, and social media, while also offering curation based not on the mere description but, in some cases, on the outlet's own interpretation. This yields a product with a greater personality and gives readers a more personalized experience, thus differentiating it from other media outlets.
At the other extreme are the newsletters from the United Kingdom (except for The Guardian) and Germany, all of which are legacy media and offer worse curated content. Lastly, there are the automated media (such as the Spanish legacy outlets ABC and El Mundo and the pure digital El Confidencial). The aim in these cases is to increase traffic without personalizing the newsletter's contents, a trend that has been observed particularly in legacy more than in pure digital media. Their purpose is informative, but they also follow the strategic end of encouraging subscriptions and visits to the website, as well as micropayments and content downloads, being the goal to win and to retain users (Silva-Rodríguez, 2021). Legacy media, therefore, pursue the beneficial impact brought by greater traffic when calculating the price of their advertising and, consequently, the financing of their business, as other studies have pointed out: El País, for example, is said to "fundamentally pursue a generation of traffic that makes its financing profitable through advertising, since its publications always refer to publications on the medium's website" (Rojas-Torrijos; González-Alba, 2018, p. 192).
In simple terms, legacy media that have made the transition and adapted to the digital environment could be said to be using the benefits of this technology mainly to increase their webpage traffic, while pure digital media also use them to provide users with access to the broad range of opportunities offered by hyperlinks and the Web, including social media. One exemplary model is the case of elDiario.es, as shown in this study and in previous work (Guallar et al., 2021a; Silva-Rodríguez, 2021).
The identification of high-quality curated coronavirus newsletters may foster among media outlets this method of distributing information to their public: it is a user-centric alternative to the hyperinflation of often fake informative products that circulate through social media..
Lastly, some of the achievements and limitations of this study and suggestions for future research should be mentioned. In terms of achievements, this is the first study to have evaluated the quality of news curation in various media in Legacy media offers little content from external sources, while pure digital media offers the most, in some cases even more links to external websites than to their own different countries, based on the application of a system that can be used and replicated in other research work on journalistic curation. The limitations include those of the system used, which could be refined and improved in successive studies. The scoring system selected, forexample, could be called into question: a binary score (0-1) was used in most parameters whereas a scale (0-3) could have been applied to make analysis more precise. Moreover, scores were assigned regardless of the number of elements present. Hence, for example, a point was awarded if a newsletter included external content in that indicator, regardless of whether it offered one or more links from that category. These matters could be nuanced or reviewed in future work.
Finally, as far as future studies are concerned, it has been proposed that the sample studied should be extended to other countries and media, and different types of newsletters (general, specialized) should be compared for the same period of time, in order to analyse news curation in other non-newsletter journalistic products such as the news on the home page of digital media or a specific section, or to perform analysis in newsletters, supplementary to curation, of other important matters, such as their business model or journalistic structure.

Note
The newsletter is a user-centric product that offers an alternative in a market saturated with information products