OASIS: one resource to widen the reach of research in language studies

ABSTRACT Language teachers report having limited direct contact with research and research findings despite generally positive perceptions of research (Borg 2009; Marsden and Kasprowicz 2017; Nassaji 2012). Key reasons teachers give are 1) practical - a lack of time and access, and 2) conceptual - academic papers can be difficult to read (Plavén-Sigray et al. 2017). In turn, researchers worry their research is not (or does not need to be) relevant to pedagogy. This paper reports on one large-scale initiative to bridge some research-pedagogy divides: Open Accessible Summaries in Language Studies (OASIS; oasis-database.org). OASIS aims to facilitate interaction between research and pedagogy, by making research into language learning, language teaching, and multilingualism physically and conceptually accessible. The freely available, one-page OASIS summaries, written in non-technical language, provide information about what the study was about, its importance, what the researcher(s) did and found. Four major international journals now require authors to write these summaries, and many others encourage their authors to write them. We discuss how and why OASIS was established, how the initiative has been received, and how summaries have contributed to professional development activities. We also and highlight some of the challenges we have encountered and discuss future directions.

take-up of OASISas a resourceamong journal editors, authors, and downloaders, indicates to us an apparent need that it is worthy of attention.
Research findings can be difficult to access for people outside academia. For example, Marsden and Kasprowicz (2017) surveyed language teachers and teacher educators and found they had very little or no direct engagement with research (doing, reading, or discussing it). Any engagement was largely indirect, through magazines and journals associated with professional organisations, in which there were very few (on average 12%) references to international peer-reviewed research. However, participants also reported generally positive opinions towards research (for similar findings, see Borg and Liu 2013;Nassaji 2012;Sato and Loewen 2019). The main reported reasons for low engagement tend to fall into three broad categories: a lack of physical access, conceptual access, and time (as well as, potentially a lack of perceived need on the part of researchers and/or potential users of research).
While open access is gaining traction, with some funders encouraging or requiring open access publishing (e.g. the NWO in the Netherlands, UKRI in the UK) and most journals now with open access options, the vast majority of articles in applied linguistics is still behind paywalls. For example, of articles published in 2019, 2020 and 2021 (in an issue or online) in five journals in the field of language learning and teaching (i.e. Language Learning, Language Teaching Research, Studies in Second Language Acquisition, TESOL Quarterly, and The Modern Language Journal), 86% are behind paywalls (in contrast, in psychology, Hardwicke et al. 2022, found that half of 250 articles published between 2014 and 2017 were openly available). Improving rates of open access publication could help to increase engagement in research for all stakeholders, and in particular it could address global inequalities in the knowledge economy.
However, open access cannot address other issues, such as the challenges caused by low levels of conceptual access. Articles usually need to be long and technical. Plavén-Sigray et al. (2017), analysing abstracts published between 1881 and 2015, found that the readability of science reporting has been steadily decreasing (see also Vergoulis et al. 2019;Wen and Lei 2022), arguing that 'lower readability implies less accessibility, particularly for non-specialists, such as journalists, policy-makers and the wider public' (6). On the one hand, some argue that conceptual accessibility is a barrier for which there is, in fact, no desirable solution (e.g. Osborne 2013), as understanding research requires expertise and those who have that expertise already have access to the publications. Nevertheless, plain language summaries have been found to enhance non-experts' understanding of individual research findings (e.g. Kerwer et al. 2021).
The other major barrier to accessing research is reported to be time. Nassaji (2012) found that over 80% of EFL and ESL teachers reported a lack of time being a limiting factor in engagement with research. Similarly, Marsden and Kasprowicz (2017) found that 88% of their teacher respondents reported not having sufficient time to engage with research.
How does OASIS address these challenges? In terms of physical access, OASIS summaries are free at the point of access. In terms of conceptual access, authors of summaries try to avoid technical terms, but if jargon is unavoidable, authors are encouraged to explain it (see the OASIS-genre guidelines). In terms of addressing a lack of time, OASIS summaries are a single page, giving enough space to include some contextual and methodological detail (more than an abstract), yet short enough to read in a few minutes. Each summary provides information about the rationale of the study ('what this research was about and why it is important'), the methodology ('what the researchers did'), the findings ('what the researchers found'), and a section to add any nuances or considerations ('things to consider'). The one-page format necessarily means leaving out methodological detail and theoretical nuance, but they are provided in the full peer-reviewed article to which the summary is linked via a clickable DOI.
In sum, the OASIS initiative aims to engage with researchers and journals, so as to make accessible summaries part of the publication process, and to engage with language professionals, so as to facilitate critical reflection that is stimulated by research (as one source of information) becoming part of their continuing (professional) development. These long-term goals require continued awareness raising and ongoing activities to both promote and investigate usage.
Part 2: ongoing ambitions and activities for OASIS After an initiating and establishing phase, OASIS now focuses on continued engagement with authors of the summaries and users of the summaries.

Engaging with authors of the summaries
Journal buy-in is crucial. Four major journals require every author to submit an accessible summary to OASIS (Language Learning, TESOL Quarterly, The Modern Language Journal, and Studies in Second Language Acquisition). In addition, Language Teaching Research operates an optout system and has a 45-50% uptake rate. 2 Thus, OASIS already has a steady and sustainable stream of summaries of research findings. Together these five journals are responsible for about two thirds of the summaries currently online. The remainder are written by (a) the OASIS team and collaborators to launch the initiative, (b) students as part of their course work, or (c) authors from other collaborating journals. 11 other collaborating journals invite, suggest, or recommend (in, respectively, increasing degrees of strength of conviction) summary writing to their authors. Uptake from these approaches varies and is considerably lower than Language Teaching Research's optout model. For example, ReCALL and System have both been part of OASIS from the start, with editors who are strong supporters of the initiative, and approximately one sixth to a third of their articles published in 2021 and 2022 have associated summaries on OASIS (ReCALL 32%, System 17%).
If weaving accessible summaries into the fabric of academic publishing in language studies is deemed to be desirable, then several lines of activity must be pursued. First, we envisage that uptake from journals will continue to steadily increase over time, as general awareness increases. Second, our experience shows that an increase in the number of journals that require authors to write accessible summaries would have a large and very efficient impact. Third, OASIS is currently diversifying its scope to cover more subdomains in language learning and teaching, language use, and multilingualism. For example, we have added the Journal of Language, Identity, and Education to our collaborating journals, which focuses less on pedagogy and more on learner and teacher identity (e.g. OASIS Summary of Tsai, Straka, and Kimura 2022a, 2022b).
Author driven summary writing is crucial, not only for the long-term sustainability of the initiative but also to help fidelity to the original research article and author approval thereof. However, whilst a lot of the authors of summaries do adhere closely to the OASIS author guidelines, other authors can overestimate what terms are 'accessible' to non-specialist readers (defined as 'educated, but nonexpert' in our guidelines) and considerable editing can be necessary. Some such editing has been undertaken by collaborating journals and some is done by OASIS, currently mainly by Alferink whose post is largely funded by a grant from the UK government's Department for Education and a University of York ESRC Impact Acceleration Account (held by Marsden and Bolibaugh). Future funding is needed to take this need into account, and/or more involvement from journal editors is likely to become necessary (e.g. TESOL Quarterly has an Associate Editor [Crowther] dedicated to editing the summaries before they are sent to OASIS).
In addition, researchers, including editors, need to know what is in it for them, beyond satisfying any sense of epistemological responsibility to share the knowledge they gain. With this in mind, the publishers of System (Elsevier) found that articles with associated OASIS summaries had more downloads (Rachel Conway, personal communication, 19th of May 2020). This finding is in line with evidence from other metascience into the impact of open research practices, such as the finding that articles with associated open data were cited more often than articles without (Colavizza et al. 2020) and that articles with pre-prints had a higher citation rate than those without (Fraser et al. 2020). Similarly persuasive arguments include findings from Kerwer et al.'s (2021) study that examined comprehension of articles in the Journal of Social and Political Psychology and found that comprehensibility was higher for plain language summaries than for abstracts, which, in turn, led to more knowledge acquisition and higher credibility. Thus, researchers who make their research findings more accessible may experience increased engagement with their work. Similar metascientific studies are needed for OASIS summaries. For example, bibliometric studies of download and citation patterns of articles summarised on OASIS versus those that are not (whilst accounting for journals, authors, and institutions in the analysis) would be welcome.

Engaging users of the resource
To date, OASIS has attained good reach, both in terms of quantity of downloads and geographical reach. There have been over 50,000 downloads, meaning that on average each summary is downloaded about 32 times. Every summary is tagged with metadata in several categories, drawing on field-specific ontologies. Categories include, for example, area of research, participant type, language being learned, and age of learner. Users can type any text into the search interface to search the database. This searches any tags as well as the summary text. Alternatively, they can use the tags to refine the results. About half (47%) of our downloaders self-identify (via short questions asked every so often at download 3 ) as language teachers and a fifth (23%) are (also) teacher educators or curriculum developers, professional development providers (2%), parents (2%), test developers (1%), policy makers (1%), and media/journalists (1%). These figures suggest that OASIS reaches audiences beyond the academic sphere, and these users are probably new to being able to access research from peer-reviewed journal publications, though this remains to be empirically tested. Other users include students (37%) and academics (15%), a point that we discuss below under Challenges and Future Directions.
In terms of geographical reach, most downloaders are in the USA and UK, but there are pockets of users in China, Japan, and the Netherlands and there have been visits from people in 182 countries. We are further nurturing international reach by diversifying the number of languages that summaries can be written incurrently, we have summary templates in 14 languages (see for example, Chap, Chan, and Wright 2022 written in Khmer) and have recently added a videoed 'signed' version in American Sign Language. Users can follow OASIS on Twitter where we announce new summaries or could until recently sign up to our monthly newsletter to receive links to all new content in one email. We have been able to see the effects of these awareness-raising activities as downloads increase shortly after each newsletter is issued. However, the bigger the database has grown, the more unwieldy these information streams have become. To address this, we have recently launched a personalised subscription service to replace our monthly newsletter, whereby users can register to be notified when summaries arrive that relate to their own areas of interest (by using the metadata attached to each summary to match this to the areas that the user registered an interest in).
To gauge attitudes to and use of OASIS, we ran two surveys (for which approval was given by the ethics committee of the Department of Education, University of York). The first survey received 56 responses following an email to our mailing list and tweets from our twitter account. Of these respondents (who were, inevitably, self-selected), most were very positive about the summaries. There was wide consensus that the information included in the summaries is useful and that the summaries are easy to understand. 77% agreed or somewhat agreed that OASIS summaries made it easier for them to access research findings. In addition, respondents reported that they had shared their positive opinions with others: 75% said they had recommended OASIS to a friend or colleague and 55% had shared a summary (for example with a colleague or on social media). We also asked how downloaders use and react to summaries on OASIS. A third (34%) said something they had read on OASIS had changed their mind. Perhaps one reason why this was not higher was that humans have a natural tendency to look for evidence to support their existing practices, which would naturally lead to confirmation bias (Wason 1968). However, and more encouragingly, 46% said they had tried something new because of something they had read on OASIS. About half of our respondents also said they had discussed OASIS summaries with other people, either informally or in an organised setting.
A larger sample of 220 language teachers who were introduced to OASIS summaries during a professional development course, and most of whom had not encountered OASIS summaries before, were similarly positive. 83% (somewhat or strongly) agreed that OASIS summaries are easy to understand. 89% (somewhat or strongly) agreed that OASIS summaries have enough information to understand the research. 87% (somewhat or strongly) agreed that OASIS summaries make it easier for them to access research findings. 78% (somewhat or strongly) agreed that OASIS summaries help them develop new ideas. In addition, 88% indicated that they intended to keep using OASIS summaries after the course had finished.
OASIS clearly provides unfiltered, direct access to research findings for anyone wishing to use the database. However, OASIS is also a resource for those wishing to provide filtered (or brokered) access, whereby the information flow is managed by intermediaries with a particular audience and/or goal in mind. For example, Indrarathne (2020) has used OASIS summaries in professional development with teachers in four South Asian countries who do not normally have access to research resources. Also, The National Centre for Excellence for Language Pedagogy (https:// ncelp.org/), an initiative established at the University of York (2018-2023) with and for foreign language teachers in secondary schools in England, used OASIS summaries as a core part of a professional development programme and to inform the creation of a language-driven curriculum and pedagogy in French, German, and Spanish for beginner-low-intermediate classroom learners. Even though the budget included funds to cover some of the teachers' time to allow them to engage in research-informed development activities, there was still a need to select research in order to orient them towards the studies with the most likely potential relevance to the teachers' context and problem spaces. This was especially needed given the lack of primary research carried out in this context and teachers' lack of experience with research or learning and teaching theory. Whilst it is fairly uncontentious that some selection of research will be necessary for brokers of research to achieve a particular goal or purpose, the balance of conditions in which such brokering, exchange, and collaboration happens is fragile: care needs to be taken to avoid raising expectations about the potential usefulness of research (Lightbown 1985) or reverting to a 'one-size-fits-all, technical research fix' approach. Creating opportunities for teachers to exercise professional judgement in evaluating relevance and to reflect on potential implications for their own practice, is challenging to achieve, and research on teacher development using the summaries is needed.
Ultimately, both unfiltered and brokered access are needed. But we emphasise that the unfiltered access is critical to enable a (more) neutral access to research findings from a wide range of perspectives on a wide range of topics. To ensure such breadth of research coverage, widescale buy-in is needed from a broad range of journals and researchers.

Challenges and future directions
OASIS is theory neutral and is not an arbiter of what constitutes good quality research. As an external measure of quality and for practical purposes to constrain the database to fit our capacity, OASIS currently only accepts summaries of articles in journals that are indexed in certain citation indices (Arts & Humanities Citation Index (A&HCI), Social Science Citation Index (SSCI), or Science Citation Index Expanded (SCIE)). At present, journals on the Emerging Sources Citation Index are not eligible for OASIS given that between 2015 and October 2021 only 895 of over 8000 currently active journal titles migrated from it to one of A&HCI, SSCI, or SCIE (Green 2021). Our solution to quality assurance is imperfect, relying on peer review and journal citation systems and rankings, yet alternative approaches are not forthcoming. For example, individual article metrics are not available at the point when OASIS summaries appear online (i.e. typically when the original article is first published online). We will continue to evaluate the acceptance criteria for OASIS as different quality assurance mechanisms evolve.
Another challenge is that OASIS inevitably carries forward the same biases as those found in academic publishing at large, as well as reflecting biases in terms of expertise and interests of the initial project team. It is incumbent upon us to be aware of those biases and to seek to redress the balance. As a first step, we have received funding to write summaries about languages and topics and from geographical areas that are currently underrepresented on the databasewhich resulted in summaries including on topics such as learning endangered languages, black queer language use, and Indigenous language policy, from regions such as Papua New Guinea and Afghanistan, and about languages such as Kala and Ojibwe. We are also inviting additional journals with different foci to collaborate to make the database more representative of the broad field of language studies.
One concern might be that for some people reading accessible summaries might replace reading the original article, thereby reducing direct engagement with research findings. This is a valid concern, and this is one reason that a direct digital link to the original article is always provided in the summary via the article's DOI. However, given that direct engagement is generally low (for example among educators, Marsden and Kasprowicz 2017) and a tentative finding from System that articles with summaries have more downloads (Rachel Conway,Senior Publisher,Elsevier,personal communication,19th of May 2020), we would argue that the overall increase in engagement is a positive phenomenon. Now, with a sizable database of summaries and a steady stream of new content, the time is ripe to explore effective ways of packaging OASIS summaries for specific purposes. With a view to promoting autonomy and individualised usage, the summaries are all licenced under CC BY-NC-SA 4.0 licences, such that they can be reproduced and adapted for non-commercial purposes (if the source and any adaptations are acknowledged). In practice, this means that we warmly welcome contacts from diverse groups working at interfaces with those outside academia. We are receiving a growing number of such approaches, and we also actively seek out such collaborations, making it clear we encourage communities to do their own brokering and selection.
Our first five years have established OASIS as a rich resource. Going forward, we hope that language studies communities will consider how to build on this foundation and ensure that the valuable content is exploited in meaningful and responsible ways. Notes 1. We gratefully acknowledge more recent additions to the OASIS project team, including Cylcia Bolibaugh and David O'Reilly, at the University of York, and the support of many other collaborators listed at https://oasisdatabase.org/about. 2. Innovation in Language Learning and Teaching also uses an optout system, but as they have only recently joined as a collaborating journal not enough data is available to provide an uptake rate. 3. Questions about which country a user is from, their role (e.g. teacher, student, university academic), are asked at the point of download from the same IP address every month.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work was supported by British Academy; Economic and Social Research Council, and Department for Education, UK government.

Notes on contributors
Inge Alferink is the project manager for OASIS at the University of York. She holds a PhD in linguistics and has a broad interest in multilingualism, language learning and education, and knowledge exchange.
Emma Marsden is a professor of second language acquisition at the University of York. She is founder and principal investigator of Open Accessible Summaries in Language Studies (OASIS) and Instruments and Data for Research in Language Studies (IRIS), as well as the National Centre for Excellence for Language Pedagogy (2018 -2023). She has also been involved in various advisory roles, including the government's Department for Education's expert subject advisory group. She has published on a range of topics including open science and second language learning and education.