Exploring the Impact of Management and Process on Academic Research Software

Principal investigator

Dr Sorrel Harriet

Affiliations

Rationale

Research is increasingly software dependent across all stages in the project life cycle, from the acquisition and management of data to its analysis and presentation. Academics across all disciplines are increasingly engaging in software development to varying degrees. At the same time it is recognised that the quality and openness of the software has implications for the reproducability, integrity and impact of the research. Effort is therefore being made to support academics in adopting best practices with respect to the development and sharing of research software.

Objectives

This study seeks to map the organisational structures and management processes adopted by academics and RSEs on small-large scale research software projects. The primary goal will be to gain a deeper understanding of common problems linked to process and develop recommendations that are sensitive to different organisational structures. Such recommendations could also serve as a useful reference to research councils who may have difficulty evaluating the robustness of proposals involving substantial software and/or technical product (STP) outcomes, particularly when planning and process-related documentation may only be presented at the end of a project. An attempt shall also be made to quantify the relative success of past projects and to look for process similarities between those which are perceived to be most successful.

A secondary goal will be to encourage a culture of openness around imperfection and experiential learning in the academic research software context.

You can read about the motivation behind this research via Medium, as well as via Sorrel’s SSI application video.

Call for participation

The study is to be conducted in 2 phases, both of which rely on outsider participation. Further information for participants is presented here.

Phase 1a: Case study interviews

The aim here is to conduct a series of anonymised case study interviews with a broad cross-section of academic software ‘teams’. The word ‘team’ is used in a loose sense as the team need not be a traditional group of developers working on a project simultaneously. The team could consist of a single developer and senior academic who is supervising the project; or it could be a succession of developers who have worked on the same project at different stages in its life cycle. The only constraints are that the project was funded over a finite period and included software or technical products as primary outputs. Examples of questions that may be asked at interview are available here.

Phase 1b: Stakeholder survey

This short electronic survey seeks to obtain some quantitative data from interview participants and from key stakeholders identified during the interviews. It can be conducted during or after the interviews. Examples of questions which will appear in the survey are available here.

Phase 2: Working group

Once the interview data has been collated, it will be reported and a working group established to derive recommendations from it. The working group is likely to meet during late Spring/Summer 2020, allowing time either side for the acquisition of case study data and formalising the output of the working group.