Measuring software technology impact

These slides are for a 13 July 2017 webinar of the Training and Outreach working group of the NCI Informatics Technology for Cancer Research (ITCR). In them, I discuss the impact of software in research, and different efforts to measure it. Software is becoming more important in research, both in terms of software that is part of a single research project, and in terms of software that enables many projects. The role of software gives rise to a large set of interrelated challenges, and some of these, such as incentives, citation, metrics, publication, communities, career paths, and catalogs are closely tied to the impact of specific software. While there are many items that can be measured to understand the impact of specific software packages, those that are easiest to measure are not those that provide the most insight. New methods for measuring software's impact can be created, but it is very hard to bring them into widespread use. Alternatively, we can use measures that we already have in place for papers. This talk will discuss these issues, where we are today, and where we might go next to better measure the impact of research software.