A Qualitative Analysis of Open-Access Publishing-Related Posts on Twitter

1 Introduction

Open-access (OA) publishing is an aspect of scholarly communication that elicits a lot of reaction among stakeholders in the scholarly community for many reasons. First, debate on OA interfaces with ethics, economics, and governance/democracy (accessibility and equitability of information resources) as a result of which the scholarly publishing system adopts various OA models, including the gold, hybrid, green, bronze, and diamond models. Also, high article processing charges, embargoes and other publisher restrictions, the subscription-based model, high paywalls, and free peer reviews, which attract unbalanced profitability to the “big publishers,” have made OA and the scholarly communication discourse particularly controversial (Fuchs & Sandoval, 2013; Hunter, 2018). The scholarly communication system is also highly controlled due to funder mandates and restrictive publisher policies (Baldwin & Pinfield, 2018). The existence of pirate libraries and predatory publishing also adds to an already complicated situation. In the knowledge economy, where information is categorized as a public good, its production, access, and supply cannot but be of very high value and interest to stakeholders.

Second, involved in the controversy of OA to scholarly resources are various communities of practice (Bacevic & Muellerleile, 2018; Day, Rennie, Luo, & Trucker, 2020; Fuchs & Sandoval, 2013; Moore, 2017). Scholars are both creators and users; publishers are investors, marketers, and policy marketers; and institutions, taxpayers, and government are the major spenders, funders, and sponsors, as well as policymakers. Fuchs and Sandoval identify stakeholders with diverse viewpoints: librarians, editors, publishers, academics, associations, labors, and policymakers. The various interest parties are somewhat suspicious of the contributions and benefits of other vested parties (Odlyzko, 2014; Tennant et al., 2016). Bulock (2018) illustrates this distrust by exploring how traditional scholarly publishers like Elsevier are involved in the very profitable OA publishing venture and have also added other solutions like indexing and abstracting services, analytics, and workflow. With such business arrangements, they usually create a lock-in for the client-institutions, thereby introducing an unhealthy dominance in the scholarly community.

Moore (2017) explores this complexity further by delving into the genealogical, historical, and cultural development of the OA movement. He explains that OA items are boundary objects that have a variable nature, with a definition determined by the community of practice specifically engaged in considering it at any point in time. Thus, the meanings, motivations for, values attached to, and routes to OA may vary according to the context in question. Hunter (2018) argues that the growth and transformation of the OA paradigm are hindered by being “patchier and slower than expected” as it did not go as planned by the earlier proponents due to cultural inertia, various interests, and resistance to change. This is part of what has made the OA an interesting and debatable concept/movement despite its introduction into scholarly communication over 20 years ago (Mering & Hoeve, 2020).

According to Vlokhoven (2019), quality does not necessarily interface with openness. Pinfield (2015) asserts that, in terms of quality, researchers continue to be uninterested, suspicious, and skeptical of OA publishing. Skepticism about the rigor of peer review in OA, particularly when payment is involved, is also highlighted as a major concern of faculty by Scott, Murphy, Thayer-Styes, Buckley, and Shelley (2023). Bacevic and Muellerleile (2018) debate OA as a moral good and a product of the digital moral economy. Other debates and areas of disagreement about OA are the motivations behind publishing in OA mega journals (Fuchs & Sandoval, 2013; Wakeling et al., 2017), concerns about article processing charges (Asai, 2020, 2021), viewpoints about predatory journals (Salehi, Soltani, Tamleh, & Teimournezhad, 2020), and the workability of Plan S (Haug, 2019).

Rowlands, Nicholas, Russell, Canty, and Watkinson (2011) note that social media is widely used by scholars for a variety of research-related purposes, such as collaboration, conferencing, research discovery, and dissemination. A scholar with a Twitter handle can engage in activities such as following other scholars or being followed, tweeting or retweeting posts, and hash-tagging. Klar, Krupnikov, Ryan, Searles, and Shmargad (2020) identify Twitter as a major platform for sharing and disseminating research work in present times, unlike in the past, when only journals, press releases, or university media were the available options. Both authors and publishers utilize Twitter to call attention to their scholarly works and research outputs. Taylor and Francis (2022) identify other benefits of scholarly engagement with Twitter, including connecting with other academics, crowdsourcing ideas, disseminating research, and enhancing research impact and author visibility.

Twitter and other social media are regularly used to measure author and publication metrics (Altmetrics) (Ladeiras-Lopes et al., 2022). Chandrasekar et al. (2020) explore the use of Twitter by academics in the field of urology and report that ownership of a Twitter account is highly associated with high H indices and research productivity. Twitter has become an emerging space for academics to engage in conversations among themselves, engage with the public, exchange opinions, and analyze ideas on a global and open platform. In summary, Twitter is a very important meeting point for scholars for various activities that promote research scholarship and even scholars’ well-being.

Lachmar, Wittenborn, Bogen, and McCauley (2017) explain that Twitter has very good use for analyzing comments because of the weak personal links between Twitter users, which enables them to be open to other users, thereby providing a more uninhibited and unbiased exchange of information. Since OA is an important concept and movement in scholarly communication, this article aims to explore the thematic direction of the concept based on Twitter discourse. The goal is to understand the nature of and development of OA in recent times.

2 Research Question

What are the perspectives and concerns regarding OA publishing as expressed on the Twitter (X) platform?

3 Materials and Method

A sample of 3,257 original tweets or posts with the hashtag #openaccess, which were posted on Twitter between May 1 and May 7, 2022, were successfully retrieved using Twint, a python-based open-source intelligence tool from the Git repository. The output was exported in CSV format to Excel for further analysis. Twitter posts related to discussions about OA were only considered for analysis as most of the comments with the hashtag #openaccess were adverts inviting users to check their OA articles out. The exclusion criteria included advertising of publications, events, and posts in other languages but English. After eliminating Twitter posts not directly related to OA, a final dataset of 67 tweets was obtained. Twitter data retrieved included usernames, the text of the tweet, and the date and time of the tweet. Since Twitter is a public platform that also allows for anonymity, informed consent has been conveniently waived as only accounts set as public were used.

4 Data Analysis

Each of the 67 selected posts was allocated a unique identification number. Subsequently, each of the posts was descriptively analyzed using qualitative content analysis methods. The two researchers studied each of the posts to understand the underlying meanings and assumptions embedded in the texts. The researchers, who are both librarians, worked separately to identify the themes in the posts. After identifying the themes independently using an inductive open coding approach, the two results were reconciled to identify the distinct themes that emerged from the exercise. Thereafter, the posts were categorized into the adopted themes. Where posts had more than one theme, it was allocated to the most dominant team or categorized into more than one theme.

5 Results

6 Discussion

Pinfield (2015) conducted a textual analysis on 589 articles, blogs, emails, and other sources published on OA using a textual analysis tool. After visualization, the major themes were identified and clustered. The first cluster is composed of research, author, policy, institution, library, and repository, while the second cluster is composed of journal, OA journal, article, publishing, APC (article processing charge), and subscription. A merging of the two clusters generated eight major themes including Green-Gold relationship, evidence base for OA, research and researchers, policy, repositories, journals, institutions, and impact. These themes are very similar to the themes identified in the present study.

APCs are fees charged by publishers to authors as publication fees to cater for publishing charges to publish in OA journals. The trend is that authors pay APCs. Authors who are not affiliated with institutions that provide funds to publish in major journals, or who do not have grants, have fewer publishing options (Pilato and Tran, 2020). Based on the characteristics of the discussions surrounding APCs evident in the Twitter dataset analyzed, it is clear that this theme constitutes a significant portion, approximately 25%, of the discourse on #openaccess. The nature of the posts within this theme reveals multifaceted perspectives. First, there are calls to action for more affordable APCs, indicating a widespread concern among Twitter users regarding the accessibility of publishing in OA journals. Second, there are justifications provided for lower APCs or arguments advocating for a more transparent understanding of the true cost of publishing an article. This reflects a nuanced dialogue surrounding the economic aspects of scholarly publishing and the need for fairness and equity. Third, there is discourse around subsidization or waiving of APC charges for scholars in low-income countries. This indicates the intersectionality of access to research and the global disparities that exist in funding and resources. These findings resonate with existing literature, which shows the significance of APCs as a barrier to the adoption of OA, particularly among researchers from low- and middle-income countries (Vervoort, Ma, & Bookholane, 2021). Scott et al. (2023) identified high APCs as a major limitation to the uptake of OA, aligning with the sentiments echoed in the Twitter discussions. Moreover, Burchardt (2014) and Scott (2018) argue that APCs pose a threat to the principles of academic librarianship and scientific integrity. The discussions on Twitter not only corroborate these findings but also emphasize the urgent need for novel approaches to address equity and diversity in research publishing, as suggested by Fontúrbel and Vizentin-Bugoni (2020) and Mekonnen et al. (2022). Therefore, the prevalence of discussions on APCs emphasizes their significance as a focal point in the discourse on OA. By implication, there is a need to address economic barriers to ensure inclusivity and accessibility in scholarly communication.

The comments of the posters on repositories show the value placed on repositories. OA repositories are digital archives that host research outputs from scholars on subject repositories (such as AfricArxiv) or institutional repositories and provide free and unrestricted access to researchers – this refers to the green OA model. The analysis of Twitter discourse reveals that approximately 16% of #openaccess tweets center around repositories. The posts generally present positive sentiments toward their role in academic communication. Users express appreciation for the contribution of repositories to free access and discoverability of research outputs, emphasizing their value to both scholars and institutions. AfricArxiv’s efforts to enhance the discoverability of African language resources and the value of repositories such as DOAJ and academic library repositories are significant among the tweets. The prominence of these posts aligns with existing literature, which emphasizes the growing importance of repositories in scholarly communication as shown in studies by Bashir, Gul, Bashir, Nisa, and Ganaie (2021), Idiedo, Omigie, and Ebhomeye (2023), Maggio, Artino, and Driessen (2018) and Torres-Salinas (2020). However, challenges such as poor utilization and depositing habits among researchers in developing countries (self-archiving), as noted by Nobes and Harris (2019), and reservations among faculty about using repositories, as highlighted by Scott et al. (2023), suggest another side to repository use and access.

About 15% of the posts are expressive of the negative notions about paywalls. The comments describe paywalls as unaffordable, inequitable, untenable, and oligopolistic. Similarly, Day et al. (2020) reflect on paywalls that introduce restrictions as a disservice to the public and to researchers who invest their resources to make the research available in the first place. Fontúrbel and Vizentin-Bugoni (2020) explain that these restrictions are responsible for pirate options such as SciHub. James (2020) notes that the paywall creates inequality, reduces the public value of scientific knowledge, and results in the use of pirate OA options, which can be regarded as a form of civil disobedience. Odlyzko (2014) refers to the big deal packages of publishers as having a destructive effect on the role of libraries and the ethos of OA. Knoche (2020) posits that the power and dominance of capitalist academic publishers (through commodification of knowledge) affect the progress of OA. In effect, the negative perception of posters is also reflected in the literature.

About 15% of the tweets are based on support initiatives to OA. The tweets identify initiatives for OA such as advocacies, persistent identifiers, Open Access Tracking Project data, consortia funding, academy’s database boycott, publisher negotiations, partnerships/collaborations, open infra projects, and talks. These findings signify a growing awareness and engagement within academia regarding the importance of OA. They reflect concrete actions undertaken to address challenges related to access and affordability. Many of these ideas have been suggested and recommended for a more equitable, affordable, and sustainable OA by researchers. The actions include publisher negotiations (Hunter, 2018), advocacies (Lukens, 2021), boycotts (Heyman, Moors, & Storms, 2016), and tools such as oaDOI (Bosman & Kramer, 2018) and funder mandates (the UK-SCL) (Baldwin & Pinfield, 2018). This theme confirms Fuchs and Sandoval’s (2013) opinion that the discourse surrounding OA, though pivotal for academia’s future, faces a dearth of defined visions, prompting a call for innovative social approaches.

As OA operates on many models which open it up to contextualization, conceptualizations, and redefinitions, and because of the evolving nature of OA with respect to technologies, policies, and globalization (Hunter, 2018; Moore, 2017), and the sometimes-controversial nature of OA, there is the likelihood of active players in the movement to attempt to re-emphasize the meaning of OA to their publics. Approximately 12% of the analyzed tweets shed light on the definition of OA by emphasizing the fundamental principles of OA and prompting reflection on the ambiguity surrounding the term “open” in scholarly communication, thereby advocating for clear definitions and standards. Many recent research papers have conspicuously and purposely provided clarification on the background, nuances, and contexts of OA (Fuchs & Sandoval, 2013; Guédon, 2017). The implication of this sense-seeking is that the OA concept is still at an evolutionary stage.

OA benefits were espoused by the tweet posters as advantageous, acceptable, free, and a public good. These benefits are also supported by a large corpus of literature. The citation advantage of OA publications (Lewis, 2018; Piwawar et al., 2018), public good (King, 2020; Wang, Pourang, & Burrall, 2019), and diffusion of knowledge (Teplitskiy, Lu, & Duede, 2017) are benefits of OA that are well researched.

Some of the tweets harvested indicate displeasure with OA policies for being subject to change quite often. While some others announced changes in their OA policies. Baldwin and Pinfield (2018) explain the complex operating environment of the scholarly communication system as riddled with academic tensions, bewildered researchers, sophisticated workflows, increasing costs, and compounded by restrictive publisher policies. Policies are necessary to regulate such a system, but the tweets indicate an abundance of conflicting and complicated policies. Current OA literature reflects the presence of studies on funder policies (Huang et al., 2020), publisher policies (Ellison, Koder, Schmidt, Williams, & Winchester, 2019), and institutional policies (Melero, Melero-Fuentes, & Rodríguez-Gairín, 2018). All of which reflect the complex OA system.

7 Limitations

A major limitation of the study is the sample size of the tweets. First, the data source was restricted to Twitter posts alone, excluding posts and viewpoints from other academic social media platforms like Facebook and LinkedIn and stakeholder reports. Second, data collection was limited to a 7-day period during the first week of May, preventing generalization beyond this specific timeframe.

Also, the data collection focused solely on public accounts, potentially excluding opinions and perspectives from users with private accounts. Due to careful checking, despite harvesting 3,267 posts, only 67 tweets were identified as relevant to the study’s objectives, resulting in a relatively small sample size that may limit the generalizability of the findings. While efforts were made to ensure a representative and diverse selection of tweets, the constrained sample size may not fully capture the breadth of trends and sentiments available across all relevant online platforms.

Future research should aim to address these limitations by employing larger and more comprehensive datasets from multiple social media platforms, including both public and private accounts, over an extended period, thereby allowing for a more robust examination of OA perceptions expressed through social media platforms.

Alternative research methods and data-gathering techniques could also provide more depth to the study. Personal interviews and focus group discussions could serve this purpose, ensuring that richer and more balanced findings, than a one-sided view of users of Twitter (X).

8 Conclusion

Article processing charges, repositories, paywalls, support initiatives, re/definition of OA, benefits of OA, and OA policies were the common themes for reflecting on OA in the first week of May 2022. The comments reflect the evolving nature of the OA terrain. The comments also reflect that there are tensions in the scholarly community over how OA is managed based on the tone and contexts of the posts, especially those connected to APC, paywalls, support initiatives, definitions of OA, and OA policies. The discourse on OA is dominated by scholars, particularly those affiliated with institutions. While this study threw up interesting findings on developments in OA based on data derived from Twitter, the findings should be treated cautiously due to the small sample size and short time frame. Future studies could expand the analysis across a longer time frame and possibly incorporate other avenues of scholarly communication.

Another dimension could investigate how themes emerging from a variety of scholarly communities vary and converge (a form of comparative analysis), for instance, comparing viewpoints from peer-reviewed literature, institutional policies, and stakeholder reports. Through this approach, a more comprehensive and robust assessment of the evolving perspectives and practices of OA publishing can be arrived at.