This research falls within the scope of analyzing the characteristics and forms of content curation in digital journalism and media, focusing specifically on a very particular and distinctive type of journalistic product: the so-called live or real-time coverage.

These journalistic products are characterized by providing continuous updates throughout an event as it unfolds, thereby covering a time span directly tied to the event, which can last several hours or even days. For this study, the Live documents from two informationally significant events have been selected: the two election days held in Spain in 2023 (the municipal and regional elections on May 28 and the general elections on July 23). This selection is based on the assumption that the undeniable informational importance of both events may potentially enable or facilitate a varied and meaningful use of the content curation practices under investigation.

The main objective of this study is to analyze the characteristics of content curation in Live News within digital press outlets. In line with this objective, four research questions are proposed:

The unit of analysis in this study is each Live News article. As the material for analysis, eight Spanish digital newspapers among the most widely read have been selected, ensuring an equal representation in the sample between legacy media and pure digital outlets. The analyzed media include four legacy outlets —El País, El Periódico, El Mundo, and La Vanguardia— and four pure digital —El Confidencial, elDiario.es, El Español, and Okdiario.

As mentioned earlier, the selected Live News articles cover two election days in Spain: the municipal and regional elections on May 28, 2023, and the general elections on July 23, 2023. Data were collected live from the media websites during the two election days and the following days and were subsequently processed using the CAS method, as detailed below. To study curated content, each Live News article was examined for links—whether in the text, an image, or embedded content (Figure 1)—to materials independent of the Live document itself (Figure 2).

The research method employed follows a descriptive and evaluative approach based on a multiple case study and expert analysis techniques (Morales-Vargas et al., 2020), specifically using the CAS method (Curation Analysis System). This method, detailed in Guallar et al. (2021b), incorporates various previous contributions on content curation, such as those of Deshpande (2013) and Cui and Liu (2017). It has been used in other studies analyzing journalistic curation (Guallar et al., 2021a; Guallar et al., 2022; Cascón-Katchadourian et al., 2022; Lopezosa et al., 2023a), which will be discussed in the final section.

The CAS method is based on a system for analyzing content curation, organized into a series of parameters and indicators grouped under two main dimensions: Content and Curation. Table 1 provides an overview of the parameters and indicators for each dimension. Specific details about each parameter and indicator will be presented in the Results section.

This method employs a binary scoring system for indicators (0-1), evaluating the absence (0) or presence (1) of an indicator. Following the original CAS method proposal, this study assigns equal weight to each indicator (1 point), without prioritizing some indicators over others. For example, in the case of temporal ranges, a retrospective curated content is assigned the same value as a current one. However, the CAS method introduces an exception to the binary scoring for the parameter Quantity of Content, where the scoring is multiple (0-3) based on the following scale: poor (0), fair (1), good (2), very good (3).

It is important to note that the scoring system is based on the premise that “the existence of greater variety in the use of each element that constitutes curation—from source types to sense-making techniques—will result in higher curation quality and, by extension, in higher journalism quality offered to the public” (Guallar et al., 2021b).

Additionally, it should be clarified that decimal scores may appear in the media’s results because two Live News documents are analyzed for each outlet, corresponding to the two election days. For example, if a media scores 2.5 on a specific indicator, this could be because one document received a score of 2 and the other a score of 3.

The resulting database used in this study can be accessed at: https://dataverse.csuc.cat/privateurl.xhtml?token=b8e9beb9-0210-454c-8b3f-7e9cc0122928

The results obtained from applying the CAS method to the Live News articles from the two 2023 election days in Spain across the eight selected digital newspapers are presented below, in the following order: Content Dimension, Curation Dimension, and Overall Data for the analyzed media outlets.

The main achievement of this work can be identified as being the first to analyze in detail the curation of content in Live News, demonstrating the significant level reached during election days.

In fact, studies on Live News typically focus on television and radio (Guribye and Nyre, 2016; Ilan, 2021) and on new live-streaming technologies (Liang, 2012), with very few studies referencing Live News tangentially, as they mostly address new journalistic formats (Thorsen and Jackson, 2018) and emerging forms of news production and presentation on the web (Ekström et al., 2013).

The following summarizes the main conclusions of this study. Regarding RQ1, the curation of content in Live News, at least in the case of major news events like election days, is characterized by the following: The typical profile of curation in Live News during an election day is that of a signed journalistic piece (with the exception of some legacy media), featuring a large amount of curated content, far exceeding what is usual in other news. The content predominantly falls within the current time range (the last 24 hours), along with some real-time and timeless content (such as tags for politicians). It mainly curates content from media categories and websites, with some links to official sources and the social network X. It frequently uses summarization and citation techniques, and its links predominantly feature description or sourcing curation and calls to action, with less frequent citation of authors and unchanged links.

On the other hand, and addressing RQ2, a ranking of curation quality in these journalistic pieces has been established, where a legacy newspaper, El País, and two pure digital outlets, El Confidencial and El Español, stand out in the top positions. However, all the analyzed media generally receive good scores. From the discussion presented earlier with previous studies on newsletters, the significant level of curation achieved in these journalistic products is evident.

Regarding RQ3, when comparing the curation in Live News with studies on curation in newsletters, which include both media and independent newsletters (Guallar et al., 2021a; Guallar et al., 2022; Cascón-Katchadourian et al., 2022; Lopezosa et al., 2023a), similarities and differences are observed. Among the similarities, the media with the best scores in Live News achieve a score similar to the best newsletters. Additionally, both types of documents predominantly curate content from media outlets and web formats, with little content from secondary sources.

The observed differences are more significant. First, in the Content dimension, the number of curated contents is much higher in Live News than in newsletters. There is also, logically, a greater presence of real-time content in Live News compared to newsletters, except in some specialized ones (e.g., coronavirus). Independent newsletters mainly curate external content, while media outlets focus on their own content, with nuances between legacy and pure digital outlets, which will be discussed later. Furthermore, newsletters feature more content from citizens and social networks.

In the Curation dimension, some differences in authorship are observed, which is always present in independent newsletters, and with different circumstances in media newsletters. In terms of techniques, there is a greater presence of Comment in independent newsletters and Summarize in the other products analyzed, with a significant use of Quote in Live News. Regarding the function of the links, Live News shows more use of contextualizing curation, call to action, and quoting the content source, while newsletters feature more sourcing curation and unchanged links. The greater use of calls to action in Live News compared to newsletters can be attributed to the constant updating nature of these documents, which often provide brief summaries of other news (many times from the same outlet). The call to action serves to encourage the reader to click on these other news pieces from the same media, thus generating more web traffic. In the rest of the functions, the patterns are similar to those of independent newsletters.

Overall, the Live News analyzed from election days stand out due to the higher volumes of curated content in each document, thereby achieving a significant richness in content curation. It is important to highlight that this places them at levels similar to, or even surpassing, newsletters, which have until now been considered the archetypical product of curation. However, as observed, there are significant differences among newsletters from media outlets, specialized ones, and independents.

Finally, regarding RQ4, on the curation of Live News in legacy media and pure digital outlets, the evaluation using the CAS system does not show major differences, but some nuances can be observed. Thus, legacy media more frequently curate traditional information sources, such as media outlets and official sources, while pure digital outlets are more open to citizen and corporate sources. Additionally, as seen in previous studies (Guallar et al., 2022), there is more curated content from external sources in pure digital outlets, although in both cases, their own curated content predominates. Furthermore, in digital natives, the personal authorship of the Live News is always present, whereas legacy media either do not include it or do so partially, possibly due to the greater weight of the corporate brand in traditional media compared to digital natives.

Although it is true that our work uses the CAS method, which has been tested and academically evaluated in several studies, primarily on newsletters, this represents an initial approach to confirm the viability of this analytical method for other journalistic products. As such, this work is not without limitations. These limitations include the difficulty in assigning different weights to the indicators of the same parameter, as all are given the same value (1 point), which does not fully capture the possible differences in value among them. Another limitation relates to the analysis of the optimal (or non-optimal) use of curation, as CAS identifies curation techniques but does not assess their effectiveness. Additionally, limitations are observed in the analysis of the topic, as the news analyzed in this research only correspond to election days, and it is unknown what the curation quality level of this journalistic piece would hold true for other events.

In this regard, and based on the limitations identified, future studies could seek ways to assess the quality of content to determine, for example, if the summary is well done, if the commentary adds value, etc. It would also be of interest to count and study the updates or segments of each Live News item to determine their number and the percentage of curation relative to the total updates. Furthermore, to address the thematic limitations, it would be interesting to analyze Live News from various types of events, as well as to expand the analysis to media outlets from other countries.