Introduction

Background

Social interaction is crucial for scientific advancement and may be facilitated by various communication channels [1]. The emergence of digital communication tools, mainly social media services such as Twitter (rebranded as “X” since 2023; https://twitter.com/), Facebook (https://www.facebook.com/), Instagram (https://www.instagram.com/), and TikTok (https://www.tiktok.com/), has revolutionized all aspects of society. This change in how instructors and researchers access and share knowledge can lead to increased social capital [2].

The dissemination of scientific findings, which aims to promote scientific thought and knowledge beyond traditional academic settings [3], is impacted by advancements in social media and current digital communication environments. This is exemplified by the numerous channels used, such as social networking services, and the emergence of innovative formats, including participatory, interactive, and visual content (such as videos or infographics).

The potential of Web 2.0 technologies and transmedia narratives (stories transmitted through multiple media and platforms) can support dialogues between academia and society, foster collaboration among peers, and enhance the visibility, identity, and digital reputation of researchers and research groups [4]. Successful scientific outreach is essential for the research process and requires creativity in promoting and communicating scientific knowledge to increase public understanding of specific issues [5].

The shift towards digital platforms for academic and scientific communication has led to a convergence with social media, resulting in the development of new communication models for transmitting information in various ways. Furthermore, these platforms promote open science, which involves sharing the results of university and academic center research with the general public [6].

The increasingly important role of social media in science communication presents a significant challenge for various stakeholders, including research groups, individual researchers, organizations, journalists, and scientific communicators such as bloggers [7]. The primary issue is the need to utilize resources effectively to produce content that serves an informative function in the current digital environment and caters to the consumption patterns prevalent in this context. However, these resources can also foster social settings that encourage collaboration, stimulating the interest of the general public and social actors enthusiastic about scientific knowledge shared by researchers and instructors.

The digital environment has had a major impact on online interactions, with the potential to replace traditional forms of social interaction. However, the academic community has not emphasized effective communication techniques for the dissemination of scientific findings [8]. Furthermore, social networks have not been fully utilized for this purpose, as the discourse on science communication has primarily focused on how science journalists, communication departments, and other scientific personnel use various social network services in settings such as universities and research centers.

Objectives

This study aimed to investigate the methods that educational researchers in Spain have used to promote the dissemination of scientific knowledge on Twitter/X as a platform and to compare these practices with those of science influencers (e.g., scientist bloggers) [9]. The study’s specific objectives included the following: (1) to assess the Twitter/X activity of educational researchers in Spain; (2) to determine the nature of the relationships on Twitter/X among educational researchers and scientific influencers; (3) to identify the primary thematic areas addressed by researchers and influencers in Spain; (4) to evaluate the level of impact of the posts published by researchers and influencers on Twitter/X; and (5) to identify the variables that may have influenced the posts’ impact.

Methods

Ethics statement

Ethics approval was not required because we only used data from published work.

Study design

This article is rooted in a quantitative investigation that utilized Twitter/X data, which was subjected to content analysis, sentiment analysis, network analysis, topic analysis, and statistical analysis.

The present study adopted a methodological approach based on the communication and scientific dissemination project in educational matters in Spain through social networks (the Comscienciaeduspain Project, No. FCT-20-15761), executed in Spain [10]. This research significantly influenced the study design, data collection procedures, and analysis.

Data sources and measurement

In the present study, the Elsevier Scopus and Clarivate Web of Science databases were used to collect a comprehensive dataset comprising 6,737 articles published between 2016 and 2020 by Spanish educational researchers. This dataset was used to identify 210 primary authors whose Twitter/X accounts matched their names. These authors were subsequently categorized as researchers using the sample examined in this work (Fig. 1) [9]. Additionally, 38 science influencers based in Spain were included for comparison [10].

At a methodological level, the following steps were carried out (Suppl. 1). First, a statistical analysis was conducted, including a Pearson correlation analysis and regression tree modeling, to measure the Twitter/X presence of researchers and identify variables affecting the impact of their posts. Second, to identify influential users and hubs as well as examine the formation of communities within the educational field, Gephi ver. 0.9.4 (Gephi Consortium) was used to perform a network analysis. Third, text-mining techniques based on natural language processing were employed to identify the primary thematic areas addressed on Twitter/X by researchers and influencers. Finally, the Kardashian index (K-index) was used to calculate the impact within Twitter/X of scientific researchers and influencers [10].

Variables

The study analyzed the following variables: (1) the number of registered users (researchers) on Twitter/X and their influence on the platform; (2) the impact of these users on their posts and the degree of interaction; (3) the homophily and associations among researchers on the platform; (4) the types of association observed among researchers; (5) the differences between researchers and influencers and their impact on the platform; (6) the thematic axes addressed by researchers and influencers; (7) the impact of Twitter/X posts published by researchers and influencers; and (8) the factors that influenced the impact of these users’ posts.

Bias

No selection bias was expected, as all target articles and journals were included.

Study size

For a descriptive study, an advance estimate of sample size was not needed.

Statistical analysis

Descriptive statistics were used to show the results.

Results

Presence on Twitter/X

Few of the researchers were present on Twitter/X (4.4%), which highlights the limited involvement of educational researchers. Between 2016 and 2020, the 210 researchers of interest in the current study shared a total of 185,020 posts on Twitter/X. Of these posts, 38.7% were original content created by the researchers, and the remaining 61.3% were retweets of posts from other users. The data were collected by analyzing the researchers’ Twitter/X posts.

The network analysis of the 210 researchers (Fig. 2) indicated a low degree of homophily, with minimal interconnectivity among the researchers concerning their interactions and connections. Despite possessing standard academic and research profiles, many of these researchers were not actively engaged on Twitter/X.

Numerous clusters were found through the results of the network analysis, with the top 10 accounting for a substantial portion (30.7%) of the retweets. The modularity value of 0.828 signifies high cohesion within the network and supports the inference that researchers tended to form distinct groups within the broader network based on their interests, affiliations, or research topics. This value is appropriate for social sciences-oriented studies [11].

In addition, the network analysis of the 210 Twitter/X accounts associated with researchers highlights two noteworthy aspects: multiple distinct groups within the network and the significant role of media references.

The existence of distinct groups within the network suggests that educational researchers tend to form clusters on Twitter/X based on their shared interests, affiliations, or research topics. Notably, only one account was responsible for a substantial portion of the network traffic (10%), while the remaining accounts contributed less, each below 3%. This suggests the presence of diverse communities or subnetworks within the broader network of educational researchers considered by this study.

Prominent references to news media were found within the network. The eigenvector values, which indicate the most influential nodes, highlight the importance of profiles such as @ diaryeduca (eigenvector, 1), @xavierbonal (eigenvector, 0.95), @el_pais (eigenvector, 0.88), @DiariEducacio (eigenvector, 0.79), and @eldiaryes (eigenvector, 0.78). These profiles belonged to educational portals, news media, or other researchers. Their high eigenvector values suggest their significant impact on how information flowed within the network: educational researchers actively engaged with and amplified content from these influential profiles, including news sources and digital media platforms.

The network analysis of the science influencers’ retweets supported several significant findings (Fig. 2). The network was cohesive, with users following individuals who shared similar content and dissemination patterns. Notably, strong connections were observed in how science influencers collaborated with significant news media, suggesting mutual interests in collaboration among these entities.

Science influencers have connected extensively with prominent scientific journalists from leading Spanish news outlets (Table 1). Fig. 2 depicts the eigenvector values for several prominent influencers in the scientific field, including journalists and scientific entities such as @pmarsupia (eigenvector, 0.99), @angency_sinc (eigenvector, 0.99), @FECYT_ciencia (eigenvector, 0.98), @LluisMontoliu (eigenvector, 0.97), @javisalas (eigenvector, 0.97), @Calvoroy (eigenvector, 0.96), and @aecomcientifica (eigenvector, 0.95).

The current investigation has identified a tightly knit group of influential individuals who exhibited similar characteristics and worked closely with prominent news outlets. Additionally, the research has uncovered distinct subgroups within this network, highlighting interconnectedness. The eigenvector scores have emphasized the prominent roles of certain individuals, primarily journalists and scientific institutions, who exerted significant influence in disseminating scientific information and shaping the intricate relationships within the network.

Thematic axes addressed on Twitter/X

The survey results indicated various subjects that researchers and science influencers have explored on Twitter/X. Table 1 indicates that a relatively small percentage of the researchers’ posts were focused on academic endeavors such as educational initiatives, research projects, conferences, and other scholarly pursuits.

The 210 accounts’ meager usage of educational metadata on Twitter/X suggests a failure to classify and arrange content about the analyzed entities (Table 2). Hashtags, also known as tags, are crucial in this procedure. The hashtags utilized by the 210 accounts bore no connection to educational topics and concentrated on current events, personal matters, or political issues instead. The data in Table 2 indicate the constrained potential of this resource in facilitating users’ access to information about the researchers.

Fig. 3 depicts the principal topics covered by the researchers. The posts collected from the 210 Twitter/X accounts indicate how these topics were framed and focused on specific educational contexts. Limited exposure is apparent regarding themes related to the researchers’ academic work or research.

The Twitter/X users affiliated with the researchers have generated the most retweeted posts, which aligns with our previous findings. However, most of these posts were not directly linked to the field of education (Table 3) [12–16].

Between 2016 and 2020, the 38 influencers posted 60,000 tweets, of which 61.9% were original posts; the remaining 38.1% were retweets. This distribution is the inverse of the researchers included in this study, who primarily retweeted content from others. In contrast, the influencers were primarily content creators, and despite collectively having only one-tenth the number of accounts of the researchers, they have published 300% more posts focusing on their content.

The influencers’ content demonstrated a heightened focus on conveying scientific knowledge rather than discussing personal topics (Fig. 3). Influencers frequently utilized hashtags in their posts, with 38,365 (10.4%) containing hashtags. Popular hashtags included #Science (1,095 times), #Research (380 times), #innovation (207 times), #impact (203 times), #science (187 times), #research (81 times), and #transference (36 times), with capitalization resulting in different counts. Many of these posts highlighted outreach conferences for all audiences and websites dedicated to the dissemination of scientific findings, such as Naukas, with 1,854 hashtags related to these topics. Furthermore, the influencers combined scientific topics with elements of popular culture and events, as evidenced by hashtags like #StarWars (64 times), which received the most retweets among all their posts (Table 4) [17–21].

Impact of the Twitter/X posts

In the K-index analysis, no accounts associated with the researchers had a K-index greater than 5, suggesting a minimal impact. Additionally, the analysis revealed accounts that could be classified as nano-influencers because their posts had limited dissemination and reach, with a maximum K-index value of 0.85 and an average K-index of 0.0015. These accounts averaged 1.73 retweets and 4.89 likes per post, with median and mean follower counts of 1,224 and 2,699, respectively (Fig. 4).

The influencers in this study had an average K-index value of 0.78, comparable to or lower than the values observed for the 210 accounts associated with the researchers. While the K-index values for these influencers may be lower, it is essential to consider that influencers with a more extensive follower base generally experienced greater diffusion. As a result, the influencers had an average of 13.61 retweets per post, 27.28 likes per post, and 51,540 followers per account (with a median of 34,775 followers). Despite their significant follower count, the influencers generally had a less significant impact on their followers than the researchers did, but their follower count ensured the wide dissemination of their posts and a larger average number of subsequent shares (Fig. 4).

Variables that affected the impact of the posts

The Pearson correspondence analysis and regression tree yielded several noteworthy observations. For the researchers, the relationship between the K-index and the number of retweets for a post was significant (0.914). In contrast, the impact of “likes” (0.005), the total number of published posts (0.002), the number of followers (–0.005), and the other accounts that the researchers followed (0.004) were relatively insignificant or negligible. Several variables had a statistically significant correlation with an influencer’s K-index value, including the number of retweets (0.661) and “likes” received by a post (0.477). On the other hand, the impact of the number of published posts (–0.025), the number of followers (0.009), and the number of followers of the influencer’s influencers (–0.003) were found to be statistically insignificant.

From the regression tree analysis (Fig. 5), the number of retweets was determined to be the sole variable that significantly affected researchers, whereas the number of retweets and followers also influenced K-index values. Those who had posts with more retweets and a substantial follower base tended to have higher impact indices as well. Consequently, having more followers could result in a more significant impact and broader dissemination of information to an audience, both for influencers and for researchers associated with accounts that did not meet the criteria of being an influencer.

Discussion

Key results

Within the study’s scope, the levels of involvement and participation of researchers and influencers on Twitter/X had noticeable differences. Whereas the influencers tended to have more prominence and dynamism, fostering strong connections with like-minded users and collaborating with other scientific and social agents to promote scientific knowledge, the researchers often lacked an effective communication strategy, which hindered their impact on movements like open science or social appropriation.

Interpretation

The researchers had a minimal impact on shaping the attitudes and behaviors of other Twitter/X users. They could not create networks and informal learning channels, which hindered the development of their social capital and limited the exchange of knowledge. Our statistical analysis showed that the science influencers included in the study had a more significant influence than the researchers on their respective followers. This disparity may be attributed to the influencers’ thematic consistency on Twitter/X, which often extended beyond their academic or research work and included various topics. Our findings suggest that the researchers have not used Twitter/X effectively to enhance their communication’s real-time and interactive nature.

Limitations

This study was limited in focusing on specific researchers from Spain in the education field who have published in journals indexed in Scopus and Web of Science between 2016 and 2020. Additionally, it compared these researchers with specific influencers recognized in other studies related to this investigative topic.

Implications

Based on this study’s findings, fostering transparency among academic and research participants is crucial for encouraging them to share information beyond their current efforts. Additionally, providing comprehensive mentorship and assistance to researchers in creating and implementing communication strategies on digital media would strengthen the unity of scientific advocates and enhance the effectiveness of their messaging, ultimately attracting larger audiences. It is essential to improve the media education of researchers from the early stages of their academic careers and establish mechanisms for institutional recognition of individuals within the case population who take on leadership roles in disseminating knowledge generated in Spanish universities and research centers. Recognizing and appreciating their efforts would further motivate and encourage effective dissemination of scientific information.

The above suggestions present a few of the many ways researchers can enhance their utilization of social media platforms to ensure that scientific research is disseminated effectively there. However, broadening the scope of research topics is crucial because these areas are expected to become increasingly relevant in the coming years.

Conclusions

The researchers had a minimal impact on disseminating scientific information on Twitter/X because they tended to share only a few original posts and mainly retweet topics unrelated to instruction or research. These researchers did not use Twitter/X as a tool for scientific communication, which limited their ability to make new connections outside their existing social and academic circles. Promoting informal learning that covers diverse topics and knowledge levels is crucial for encouraging engagement and collaboration.

Notes

Conflict of Interest

No potential conflict of interest relevant to this article was reported.

Funding

This work was supported by the Spanish Foundation for Science and Technology, funded by the Spanish Ministry of Science and Innovation (No. FCT-20-15761).

Data Availability

Dataset files are available from Zenodo.

Dataset 1. The raw data (available at: https://zenodo.org/records/10827329)

Dataset 2. The coded data for analysis (available at: https://zenodo.org/records/10864405).

Supplementary Materials

Fig. 1.

Flowchart showing the selection process for the researchers and science influencers.

Fig. 2.

Networks formed by (A) researchers and (B) science influencers on Twitter/X.

Fig. 3.

Dendrogram with the main topics addressed by (A) researchers and (B) science influencers.

Fig. 4.

Retweets and follower metrics (the Kardashian index) among (A) researchers and (B) science influencers.

Fig. 5.

Regression tree of variables affecting the Kardashian index values of (A) researchers and (B) science influencers.

References

• 1. Fernández-Navas M, Alcaraz-Salarirche N, Pérez-Granados L. Status and problems of qualitative research in education: divulgation, research and access of university teaching staff. Educ Policy Anal Arch 2021;29:46. https://doi.org/10.14507/epaa.29.4964. Article
• 2. Barashkova AL, Vorob’ev IV, Shavaev AA, Zapolskaya AN. New methods of science popularization in the social media: modern trends and communications. Proceedings of the 2019 IEEE International Conference on Quality Management, Transport and Information Security, Information Technologies 2019 Sep 23–27; Sochi, Russia. IEEE; 2019. p. 463–5. https://doi.org/10.1109/ITQMIS.2019.8928354. Article
• 3. Zaragoza JC, Roca D. The Youtuber movement in Spanish scientific dissemination. Prism Soc 2020;(31):212-38.https://revistaprismasocial.es/article/view/3942.
• 4. Harmatiy O. Science coverage: what does the audience want and really need? Exploring media consumption in Ukraine. J Creat Commun 2021;16:97-112.https://doi.org/10.1177/0973258620981799. Article
• 5. Galvão T, Tavares CA, Fernandes M, Moreira de Sousa JC, Noll M. Case report: scientific dissemination as a mediator of the research consolidation process: proposals for institutionalization [version 1; peer review: awaiting peer review]. F1000Res 2021;10:886. https://doi.org/10.12688/f1000research.55655.1. Article
• 6. Wang Y, Fikis DJ. Common core state standards on Twitter: public sentiment and opinion leaders. Educ Policy 2019;33:650-83.https://doi.org/10.1177/0895904817723739. Article
• 7. Sanz-Lorente M, Guardiola-Wanden-Berghe R. Communicate science. Hosp Domic 2019;3:173-83.https://doi.org/10.22585/hospdomic.v3i2.57. Article
• 8. Dudo A, Besley JC, Yuan S. Science communication training in North America: preparing whom to do what with what effect? Sci Commun 2021;43:33-63.https://doi.org/10.1177/1075547020960138. Article
• 9. Pérez-Rodríguez AV, González-Pedraz C, Alonso-Berrocal JL. Twitter as science communication tool in Spain. Main agents and communication networks. Comm Pap 2018;7:95-112.https://doi.org/10.33115/udg_bib/cp.v7i13.21986. Article
• 10. Said-Hung E, Martínez J, Brändle G. [The study of scientific communication and dissemination in educational matters in Spain. The Comscienciaeduspain Project] In: Said-Hung, E, Arriba D, editors. [Communication and scientific dissemination through social networks: approach from the educational field in Spain]. Editorial Tirant Lo Blanch; 2023. p. 17–38.
• 11. Lantz B. Machine learning with R: expert techniques for predictive modeling. 3rd ed. Packt; 2019.
• 12. @othbcn. Stories from #CatalanReferendum: in small village Sant Iscle they hid the ballot box and started playing dominoes [Internet]. Twitter/X; [posted 2017 Oct 2; cited 2024 Jun 17]. Available from: https://x.com/othbcn/status/914750829080580096.
• 13. @xavierbonal. [From the UAB, we investigated Confinement, childhood, and young people in Catalonia. Can you answer this survey if you live with sons and daughters between 3 (P3) and 18 years? https://t.co/DM0yvnXo4o (Catalan and Spanish). Maximum diffusion please!] [Internet]. Twitter/X; [posted 2020 Jun 1; cited 2024 Jun 18]. Available from: https://x.com/xavierbonal/status/1243237163199803392.
• 14. @lamagiadelaef. [Here is an ESCAPE ROOM GUIDE with links to fake creators, digital locks, and enigmas... so you can create one in your own classroom. https://t.co/67SCXkVm5K #escaperoom #gamificacion] [Internet]. Twitter/X; [posted 2020 Feb 8; cited 2024 Jun 18]. Available from: https://x.com/lamagiadelaef/status/1226264575256989698.
• 15. @xavierbonal. [Perhaps it will annoy more than one, but it is how I think education is being treated in this pandemic. Disregard the right to education https://t.co/syxrvSXzWr through @el_pais] [Internet]. Twitter/X; [posted 2020, Jun 1; cited 2024 Jun 18]. Available from: https://x.com/xavierbonal/status/1267344631181778944.
• 16. @estrella_dura. [Today, a new financial instrument in the European Parliament has been voted to support long-term unemployed. VOX has voted against!! So are the partners with whom PP and Citizens agree whenever they can. They are patriotic and turn their backs on the working class.] [Internet]. Twitter/X; [posted 2019, Oct 10; cited 2024 Jun 18]. Available from: https://x.com/estrella_dura/status/1182320759928246272.
• 17. @aberron. [How do you explain to children the importance of soap against viruses? I do not know the author, but from here, I make the wave (it has come to me by WhatsApp).] [Internet]. Twitter/X; [posted 2020 Mar 12; cited 2024 Jun 18]. Available from: https://x.com/aberron/status/1238566869021442049.
• 18. @pmarsupia. [“Wash your hands well.” This is the advice that you have heard many times in recent days. However, why is hand washing with soap so effective against the coronavirus? Behind this advice, there is a fascinating science! (THREAD/)] [Internet]. Twitter/X; [posted 2020, Mar 8; cited 2024 Jun 18]. Available from: https://x.com/pmarsupia/status/1236766433452924928.
• 19. @jmmulet. [The coronavirus epidemic arrived, and the Chinese government set up two standard hospitals at full speed, not a megacenter of acupuncture or traditional Chinese medicine. You do well?] [Internet]. Twitter/X; [posted 2020 Feb 2; cited 2024 Jun 18]. Available from: https://x.com/jmmulet/status/1223901552068497408.
• 20. @Irreductible. [Awesome... Red-green tomato separator O_O] [Internet]. Twitter/X; [posted 2017, Aug 28; cited 2024 Jun 18]. Available from: https://x.com/Irreductible/status/902147747943501824.
• 21. @Irreductible. [Today, thanks to vaccines, is a historic day. August 25: They declare Africa free of polio.] [Internet]. Twitter/X; [posted 2020 Aug 25; cited 2024 Jun 18]. Available from: https://x.com/Irreductible/status/1298237941756628992.

Figure & Data

References

Citations

Citations to this article as recorded by