Data Science: Where the Boys Are
A friend and colleague asked me about my observation that although there are many women in analytics (see my article, “The STEM Profession that Women Dominate”), that the media and other sources paint the new “Data Science” image as a boys club. In fact, I find that Data Science events are far more male-dominated than other analytics gatherings, and this is even truer for Big Data (“O’Reilly Strata: Deluded About Diversity?”. He and his colleagues wanted to know more, so I put down a few observations…
I think it comes down to the changing face of computer science. At SPSS, I noticed that women’s presence in software development and sales declined over the years. When I entered the company, the heads of both those teams were women, and there were many women on staff, but both roles were later taken over by men and women disappeared from staff. Yet there were always many women in roles that required statistical analysis skills.
In early 2009, I began attending software industry events frequently, and noticed that attendees skewed male, and strongly so. It was around then that I became aware of the decline in women’s participation throughout computer science and software. I also began to encounter a remarkable level of obvious sexist behavior among young software developers. I started to hear stories of male programmers hiding behind screen names and posting strong sexist rants about women in programming.
Later that year the New York Times published the Hal Varian quote about statistician as a “sexy” job and the term “data science” started to come up frequently. But the data scientist was almost always male. Other than Hilary Mason, women data scientists were more or less invisible in the press.
In real life, I see more and more preoccupation with data storage and management, less and less on thoughtful analysis. The more emphasis on programming and databases, the more that the scene becomes a boys’ club. Recently I attended a local “Big Data” event – there were at least 100 people in the room, yet for a few minutes I did not see one other woman. In the end, only about 5% of the attendees were women.
The local Data Science meetup draws about 20% women. I brought an aspiring young female analyst with me to the first meetup of that group. When we entered, we met the organizer of the group. I asked him what his company did. His reply? “We’re bad motherfuckers!” He asked what we did at my company and I assured him that we, also, were bad motherfuckers. The young woman kept her composure, but she did not stay long and I doubt that she cared to return.
Let me contrast that with a conference where I spoke recently – the Association of Public Data Users. Lots of women there, with legit data analysis credentials. Many of them work for the Federal statistical agencies or key private organizations and have serious macho data management skills. But you don’t see them in the tech press. (I was stunned that a [yes, he was male] government official from the open data program spoke to that crowd and seemed to have no respect for them whatsoever – yet I encounter the same arrogance from the tech community all the time.)
“Data science” is now associated with lots of programming, lots of data management. I get calls from recruiters looking for data scientists, and what they want is a programmer, not an analyst. Well, what they want is a unicorn, but mostly a programmer unicorn.
Date: October 17, 2012