Exploring Process Improvement Decisions to Support a Rapidly Evolving Developer Base

Software engineers learn through experience and training to optimize their software development process through a dynamic cycle of assessment, planning, enactment, and evaluation. When needed, software process improvement (SPI) frameworks (SPIFs) fill in experiential and educational gaps. However, because of the needs that drove their original development, existing SPIFs assume either a specific set of situational criteria or enough software engineering (SE) expertise to select and tailor SE best practices as needed. Outlier projects involving novice software developers or projects with unique concerns are left to find their own path to a quality software product. With such scenarios becoming more prevalent as computing becomes pervasive across society, there is a growing need for flexible, accessible SPIFs that allow decision-makers of all skill levels to efficiently self-drive their SPI efforts according to their individual goals and constraints. This paper outlines how we, as a research community, can and must meet this need by addressing the underlying theoretical, empirical, and practical challenges.