Big Data and Gender

Where does big data fall on in the spectrum of gendered ways of thinking? In that vein, does the focus on a lack of gender diversity in high tech obscure a larger problem, which is a masculine way of thinking?

The following paragraphs are the introduction to a research paper I am writing for an interviewing methods class (Sociology 273F). So far, I have done 8 interviews with big data experts. However, I have not found evidence in favor of my hypothesis which is as strong as I had hoped. This could be because my interviewing schedule was not very good (or could be better). Certainly, my sample is incredibly biased and could be vastly improved to be more representative of high tech, which I would presume would also give me more data in support of my hypothesis. If I choose to continue on with this hypothesis, however, I would choose to include a quantitative component using a big data method (possibly text analysis on tweets with “big data” in them on Twitter?). In any case, here’s the current introduction:

Silicon Valley has a gender problem; the high tech companies themselves admit it. In 2014, Google led the way, publishing their diversity data for the world to see, and the numbers were revealing--70% of their workforce was male. However, their logic was that in order to address the problem, they had to be open about the facts(1). Within a month of the announcement, Yahoo!, LinkedIn, and Facebook all followed suit, revealing very similar diversity statistics. It was as though there was safety in numbers; the statistics were reflective of Silicon Valley as a whole(2). Today, not much has changed. Google’s 2016 diversity data revealed that its workforce is 69% male and 31% female, and within technology specifically, it is 76% male and 24% female (Avalos, 2016).

Despite the relatively static numbers, or perhaps because of them, there has been a growing Women in Tech movement. In fact, just two weeks ago, the Grace Hopper Conference (the leading women in tech conference, and a “celebration of women in computing”) concluded its largest event ever, with 15,000 attendees (a 25% increase from 2015)(3). If it was not true before, women and men are aware of this pervasive problem, and many companies are trying to solve it(4).

When I first proposed that big data was hypermasculine, I thought it was a no-brainer. Silicon Valley is overwhelmingly male. Tech cultures are dominated by displays of hegemonic masculinity, or a way of practicing gender that legitimizes patriarchy and perpetuates its position by subordinating women (Connell, 1995). And all of this is in the context of a new type of masculinity in Silicon Valley--one which glorifies “technical skills and brilliance” over “looks and athletic ability” (Cooper, 2000, p.382).

Given this, I was surprised to find that after conducting four interviews with two men and two women, all of whom would be considered minorities in Silicon Valley, there was almost no use of gendered terms (or gender at all) the description of big data. Some of the more technical terms to describe big data methods are gendered (such as data cleansing, extraction, querying, processing, transforming, analyzing--as actions to be performed on big data as an object, by the big data expert as the subject), but while they came up in conversation to facilitate the description of big data projects, they were in no way the focus of the discussion. However, any lack of gendered speech was betrayed by a clearly gendered way of knowing. Herein lies the danger: the focus on the lack of gender diversity in Silicon Valley obscures the true problem, which is a homogenous, masculine way of thinking and knowing that is pervasive in high tech.

Krista SchnellComment