Report: Support Needed for Wealth of Research Software Created by ARC Grants

A report commissioned by the ARDC estimates that nearly half of the Australian Research Council (ARC) grants from 2010 to 2019 resulted in software production, highlighting the need to strengthen support for research software and its discovery, reuse and sustainability.

The report, Unearthing Research Software, was delivered as part of the ARDC’s program to implement a national agenda for research software. Using machine learning techniques, it estimates that approximately 47% of 13,784 ARC grants between 2010 and 2019 gave rise to software infrastructure, tools or code of some kind. The percentage is also estimated to have gone up during the decade as the volume of grants increased.

High and rising levels of ARC-funded software production are estimated across disciplines and organisations, including in the humanities, arts and social sciences (HASS), where 25% of research grants created software despite the domain’s unique nature and methodologies. No significant relationship was found between funding amounts and the probability that the research will produce software.

Eva Maxfield Brown and Nic Weber, authors of the report, said, “The amount of software produced by ARC funding is substantial and increasing. This validates the need identified by previous qualitative and survey-based studies ^{[1] [2] [3]} to address the undersupply of research software support, the difficulty of discovering and reusing software, and the demand for socio-technical infrastructures to meaningfully sustain software.”

In light of the abundance of research software created through ARC grants, the report calls for better documentation, discoverability and management of research software. Specifically, it recommends that:

1. software production descriptions (SPDs) are required in grant applications to ensure the software is not invisible or repetitive
2. grant awardees are required to report software as an outcome to foreground the software’s role in the creation of the knowledge and for future SPDs to be evaluated against it
3. a national research software census is conducted for researchers and support professionals to identify and document software, helping ARC understand where future funding should be directed for software sustainability
4. needs for research software engineering (RSE) and support in the workforce are identified to ensure ARC-funded software is maintained.

Dr Tom Honeyman, who led the research software program at the ARDC, said, “With this report, we hope to provide the first ever estimates of the scale of the pipeline of software production arising from ARC funding overall. The estimate that a wealth of software is created through ARC grants across domains and organisations bolsters our case for systemic change in supporting the discovery, development and maintenance of research software while giving us a clear sense of the scale of the challenge.”

Read the report in full.

The ARDC is funded through the National Collaborative Research Infrastructure Strategy (NCRIS) to support national digital research infrastructure for Australian researchers.

References

1. Barker, M., & Buchhorn, M. (2022). Research Software Capability in Australia. Zenodo. https://doi.org/10.5281/zenodo.6335998 Jump back
2. Ram, K. & Howison, J. (2003). Research Software Visibility Infrastructure Priorities Report. https://doi.org/10.5281/zenodo.8404846 Jump back
3. Stevens, F. (2022). Understanding How Researchers Find Research Software for Research Practice. https://doi.org/10.5281/zenodo.7340033 Jump back