| Popis: |
Presentation at AAAL 2023, as part of the Ethics in applied linguistics research methods: Evidence and action colloquium Growing numbers of applied linguistics journals and funders now recommend or mandate the public archiving of research data. There are multiple reasons to advocate data sharing, including an ethical imperative to maximise the benefit of our participants’ contributions and ensure the veracity of our results. Indeed, beyond the immediate goal of reproducing published findings, data sharing also underpins the “future proofing” of lines of research, by enabling reconceptualisation or reanalysis with new statistical techniques, as well as the development of cumulative, synthetic lines of research (Bolibaugh, Vanek & Marsden, 2021). While maximising the use of public funds and our participants’ time undoubtedly represent ethical goods, little is known about the reality of data sharing practices within the field of applied linguistics, including the extent to which data deposits are useful, reusable and functional, conforming to FAIR principles (Wilkinson et al., 2016) and the extent to which sharing practices maintain public trust. In a pre-registered study, we adapt the methodology of a well-known survey of data quality (Roche et al., 2015) to evaluate the completeness, reusability and alignment with ethical guidelines of datasets that have been archived in the IRIS materials and data repository (an established repository in the languages sciences described by field specific metadata). Data deposits were rated against the data completeness and reusability assessment criteria developed by Roche et al. (2015) that we further adapted to explicitly rate adherence to ethical directives of non-identifiability. Initial findings suggest that there is a wide variability in both the completeness and reusability of archived datasets, with a majority of authors providing minimal contextual information regarding the data themselves beyond the obligatory metadata required by IRIS. We conclude by offering concrete recommendations to improve the quality of data sharing practices in applied linguistics. |
