Analyzing the Analyzers

Chapter 3. A Survey of, and About, Professionals

How can we, as a professional community, fix this problem? Perhaps by using the tools we know best — data collection, data analysis, and data communication. In mid-2012, the three of us set out to do some old-school data science and constructed a survey of practitioners. But not just any survey. We wanted to ask questions that would help us understand and define sub-groups — not based on years of experience, or academic degrees, or titles — but based instead on how data scientists think about themselves and their work. We didn’t ask about verticals, or pay scales, or org charts. We avoided tool and technique questions about database platforms or favorite statistical or machine learning techniques. Others have asked those questions, and the answers are interesting, but not relevant to our problem.

We created a five-page web survey, taking less than 10 minutes to complete and focusing on five areas: skills, experiences, education, self-identification, and web presence. (Regarding web presence, we asked for those willing to share their LinkedIn, Meetup, and GitHub profiles, so that we could perform additional text analysis. However, due to relatively low response rates and some technical issues, the results were not usable, and will not be reported.) After testing the survey on a small group of friends and colleagues, we shared it broadly, evangelized it to professional Meetups, and posted links on every relevant social network we could ...