Data Science is a vast area of practice. As the field continues to be defined and redefined, there are a variety of career pathways to explore. I am taking a “Perspectives of Data Science” course this semester at Indiana University as part of my master’s curriculum. It is an exploratory course that will help me focus my interests and strong points in this huge field.
According to an April 2016 report “The State of Data Science”, there were 11,400 data scientists working for 6,200 companies belonging to the job networking site, LinkedIn. The top skills of these professionals included data analysis, R, Python, data mining and machine learning. If you asked 10 data scientists to define what they do, you’d probably get at least 10 different answers. Surely this confusion can be sorted out with data?
Today I delve into the report Analyzing the Analyzers by Harris et al. to see if the data science profession can be sorted out a bit more clearly and quantitatively. We don’t presume there is one type of “doctor” or one type of “engineer” so why should there be only one type of data scientist? Next week, I’ll talk about my personal preferred career pathway.
The 2012 Harris survey had 250 data professionals from around the world. Four categories were defined as a result of these responses. Survey respondents used a list of 22 generic skills to self-identify into the four groups : data businessperson, data creative, data developer or data researcher. Figure 3-2 shows the list of generic skills color coded by skills groups: business, machine learning/big data, math/operations research, programming and statistics.
I like to think of Figure 3-3 from the Harris survey as the approximate percentage of time each data scientist group performs each of these skills in their jobs. Even though the self-reported data may have some biases, I think these groups are a great starting point for a discussion about the types of data scientists.