We are the 92% Valuing the contribution of research software Neil Chue Hong, FORCE2015 Research Communications and e-Scholarship 13 th January 2015, Oxford Supported by Project funding from
Abstract Reexamining the evidence for research software credit and software as a first class research output In a recent survey conducted by the Software Sustainability Institute of UK research-intensive universities, 92% of researchers said they used research software and 68% said their research would be impossible without software. Yet only 4% of jobs advertised were software related, and we have seen many issues trying to establish career paths that recognise the development of research software as a valued contribution to the community. In this session, I will summarise some of the various initiatives that the Software Sustainability Institute has participated in to raise the profile of software, from public campaigns to better citations, and from training to software journals. Ultimately, I will argue that it is only be reexamining the way that credit is currently given for scholarly communication that we can change the culture around software, by showing that reuse is as significant as novelty or journal impact factor.
Software isn’t special, it’s mainstream Survey of researchers from 15 Russell Group unis conducted by SSI between Aug- Oct respondents covering representative range of funders, discipline and seniority. 69% 92%
This isn’t just about the “traditional” computational sciences Survey of researchers from 15 Russell Group unis conducted by SSI between Aug- Oct respondents covering representative range of funders, discipline and seniority.
And it isn’t just using software, it’s researchers developing software too Survey of researchers from 15 Russell Group unis conducted by SSI between Aug- Oct respondents covering representative range of funders, discipline and seniority. 56% 21%
So what’s the issue? Survey of researchers from 15 Russell Group unis conducted by SSI between Aug - Oct respondents covering representative range of funders, discipline and seniority. Analysis of data from 49,650 grant titles and abstracts published on Gateway to Research covering Analysis of job adverts posted to jobs.ac.uk in 1H % 4% Of UK researchers have had no formal software development training Of jobs advertised in UK universities were software related 77% Of PIs had not included costs for software development in bids 30% Of UK research investment has been spent on research which relies on software … and then there are gender related issues
Career Paths in UK Careers outside academic sector Non-university Research (industry, government etc.) ProfessorPermanent Research Staff Early Career Research PhD students Source: The Scientific Century, Royal Society, 2010 (revised to reflect first stage clarification from “What Do PhD’s Do?” study) UK STEM graduate career paths
How do we cite software? Citing a paper – Via an associated paper – Via a software paper Citing software directly – Using a name – Using a URL – Using a persistent identifier But citation isn’t the problem, contribution is
Attribution and Authorship Which authors have had what impact on each version of the software? Should contribution be collective? Who had the largest contribution to the scientific results? OGSA-DAI projects statistics from Ohloh
Researchers still have universal needs We know we must describe and cite software otherwise we cannot benefit from reuse and refinement But we still need to fix the reward mechanism for non-traditional research outputs – Because otherwise what is my incentive to do this? These slides:
Further reading We are the 92% – It’s impossible to conduct research without software – impossible-conduct-research-without-software-say-7- out-10-uk-researchers impossible-conduct-research-without-software-say-7- out-10-uk-researchers Software Attribution: can we improve the reusability and sustainability of research software –
The Software Sustainability Institute A national facility for cultivating world- class research through software Championing research software across all disciplines Enabling all researchers to access software training Supporting career paths for research software Promoting the importance of software in reproducible research Supported by EPSRC Grant EP/H043160/1