Open science in language learning research: IRIS and OASIS

Paquot, Magali;Marsden, Emma;Plonsky, Luke;Thompson, Sophie
(2019) EUROCALL 2019 — Location: Louvain-la-Neuve, Belgium (28.August.2019)

Files

EuroCALL_Abstract_ST-LP.docx
  • Open Access
  • Microsoft Word XML
  • 13.69 KB

Details

Authors
  • Author
  • Marsden, Emma
    Author
  • Plonsky, Luke
    Author
  • Thompson, Sophie
    Author
Abstract
IRIS is at the forefront of the open science movement. It is an open, highly searchable repository, holding materials and data used for research into language learning and teaching. IRIS holds more than 1,500 materials. These include different kinds of tests for a wide range of language features in a variety of languages, such as, grammaticality judgement tests, elicited imitation tasks, teaching intervention activities, questionnaires, software scripts, sound and video files, among many more types of materials. IRIS now also hosts 73 sets of data from all types of studies, such as written production data from an investigation into the use of Twitter by language students (Kelly, 2019) and test scores from a computerised adaptive testing study (Mizumoto, Sasao, & Webb). The areas of research covered by IRIS encompass all the subdomains of research in the language sciences and language education. This includes a large number of materials that are highly relevant for investigating computer assisted language learning and complexity, the conference themes. For example, 67 entries are specifically tagged with the label computer assisted language learning, 33 with task-based, 73 interaction, 53 attitude and motivation questionnaires, 45 with complexity. The main purposes of IRIS, in line with those of open science more generally, are to improve the quality, systematicity, validity and reliability of research, as well as facilitate the research process and reduce re-invention of the wheel (Marsden, Mackey, & Plonsky, 2016). This is important because, for example, only about 1 in 400 published articles in the field of second language research are replication studies. Findings are more likely to be replicated when materials were shared or authors overlap (Marsden, Morgan-Short, Thompson & Abugaber, 2018). Yet, very low proportions of materials are available to facilitate replication and improve its quality. Support for IRIS is considerable, 36 top journals in the field support IRIS by encouraging their authors to share materials and data through the repository, as do national and international professional associations such as The American Association for Applied Linguistics, The British Association for Applied Linguistics and International Association of Teachers of English as a Foreign Language (IATEFL). A number of journals are currently awarding Open Science Badges to articles with open materials or data. However, this endeavour towards increased transparency requires a more concerted effort from across the field. Only 1 in 6 articles have materials uploaded on IRIS, and although the percentage of materials available on IRIS has increased over the last 5 years, (12% in 2013, 23% in 2017) much still needs to be done (Marsden, Thompson & Laflair, in progress). By embracing open access and transparency as a field, we can overcome the lack of replication studies on language learning and teaching research and extend the generalizability of findings to different settings, demographics, target languages, modalities, and research areas. This paper seeks to advance these efforts and to further argue for the place and value for IRIS—and materials sharing, more generally, in the CALL community.
Affiliations

Citations

Paquot, M., Marsden, E., Plonsky, L., & Thompson, S. (2019). Open science in language learning research: IRIS and OASIS. EUROCALL 2019, Louvain-la-Neuve, Belgium. https://hdl.handle.net/2078.5/91403