Software research needs reuse focus

Why this is here: The author noted a picture of “The Isolator Helmet”—an invention designed to eliminate distractions—as a reminder that simplistic solutions to complex problems often fail in the real world.

Tom Zimmerman at UC Irvine hosted a forum on Developer Experience, where speakers discussed the need for more reusable software research. The event included a panel on the SPACE framework and questions about the relevance of “activity” metrics in the age of AI. The author argues current productivity measures, like PR throughput, are often misleading without careful statistical modeling.

He advocates for “reusable research”—projects that impact theory, data, or practice beyond initial publication. This requires proactive effort after publication, including adapting findings to new contexts and translating them for broader audiences. Intervention studies, built with real people and systematic theory testing, are particularly valuable for generating reusable knowledge.

The author shares his team’s work on a Code Review Anxiety intervention, which focused on self-efficacy rather than skill deficits, informed by direct conversations with developers. He believes soliciting research questions from practitioners is crucial for relevant and impactful work, even if it feels risky to open science to outside input. The work continues as he seeks connections with organizations prioritizing open science in tech.