Health Measurement of Data-Scarce Software Ecosystems: A Case Study of Apple’s ResearchKit

Abstract

Current methods for measuring open source software ecosystem health are unable to measure the health of young software ecosystems, due to a lack of data. This paper proposes a new method for measuring software ecosystem health. By using a mixed method design with interviews as the primary data source, a health measurement can be performed on data scarce ecosystems. This is applied to ResearchKit, Apple’s SDK to create applications for medical research. The case study shows that the ResearchKit ecosystem is threatened by the outbound links of the third-party software developers. These developers intend to create web-based applications as ResearchKit suffers from a selection bias that makes it unsuitable for most medical research. The interviews exposed an inherent problem that is unrelated to ecosystem size and may not have been found in a traditional health measurement.