Diversity, Equity, and Inclusion in Academia

Preface: I have never sat on an admissions committee, and my thoughts are based on conversations I have been able to observe, mostly on Twitter. These aren't necessarily representative of my department, and I can't speak to any specific university. This is a compilation of overarching themes I've observed from a variety of (primarily USA-based) academics. My thoughts are obviously going to be skewed by who I do (and, more importantly, don't) know.

---

Diversity and Inclusion. It’s almost as much of a buzzword in Comp Sci as “Machine Learning” right now. Personally, I tend to lump the two together without parsing them as two separate concepts most of the time, so I’d like to take a second to break the two up.

Diversity: Who is in the room? Do the experiences that people bring to a project represent all backgrounds? Notably, are there people (and not just one “token person”) who have the life experience and perspective.

A soccer metaphor: You don’t want a soccer team full of forwards. You need some midfielders, defenders, and goalkeepers. A roster of 11 forwards won’t ever win a game.

Inclusion: Once a “diverse” group of people are part of your cohort, are they part of your community? Do they feel safe to voice their opinions?

Soccer metaphor continued: So you have a balanced roster of forwards, midfielders, defenders, and goalies. First, you need to practice together so you can work as a team. But more than that, by working with forwards, defenders and goalies can learn what to look for when facing an opposing forward, so that when it’s game time, they’re prepared for the challenges ahead.

Features are not equal to experiences

There are a plethora of experiences I have had as a woman in Computer Science, both positive and negative. I’ve bonded with labmates as self-proclaimed “bad-ass women of science.” I’ve been (to my knowledge) the only woman in the room as a professor told our class to “never get married, because women just take your money and ruin your lives.” These experiences shape who I am, and I am able to relate to many other women through these experiences.

Because I check the “woman” box, it is assumed I bring that family of experiences to the table, and this is the diversity I bring. Because we have an obsession with quantification, we tend to represent people– and their human experiences– as this vector of characteristics, in a sense.

Diversity stems from experiences, not from features. Often, our experiences are shaped by our features, but they are not solely our features.

Our obsession with quantification

The first step in solving a problem is acknowledging it exists. To a large extent, Computer Scientists in the United States recognize that the ~80% of computer scientists in the US who identify as men is not representative of the ~50% population of male-identifying people in both the US and world. And this is a great first step.

As companies and university departments push to “be diverse,” whatever that means, we want to quantify how diverse we are: “Half of our students are female.” “10% of our team self-identifies as queer.” “20% of our students are people of color.” It goes on.

I have three knee-jerk responses here:

• We can’t really quantify inclusion that well, and quantifying our diversity only gets us so far.
• This reduces peoples’ human experiences to the labels we slap on them.
• If we do accept considering people as a collection of features, we tend to miss many unique experiences that are correlated with the intersectionality of these labels.

In our diversity-hacking, we tend to ignore the intersectionality of many of the labels we slap onto people. For example, a 1976 lawsuit against General Motors Corporation, led by Emma DeGraffenreid and other former GM employers alleged employment discrimination in the workplace. In GM’s defense, they claimed that they had both black employees and female employees, so they clearly did not discriminate. However, looking at the intersectionality of these labels, the company had disproportionately few black women hired at GM at the time. (Yes, this was 1976, but this is one of the more famous examples on the importance of intersectionality. Modern examples can certainly be found as well.)

When we solely aim to “be diverse,” we try to choose how to measure diversity in the lowest-cost, lowest-impact way possible. This gives companies and universities an easy out. As a result, Diversity and Inclusion initiatives in tech often disproportionately benefit white women of higher socioeconomic status. See this article for some examples of people who get neglected by the convenience of focusing on the most convenient “minority group” in tech.

Now don’t get me wrong. Being able to set quantifiable goals for oneself is generally a good thing. Not all quantification of D&I initiatives is bad, but I remain critical of clinging to them for dear life.

How academia discriminates (a very incomplete list)

• Barrier to entry: As one example, parents (especially single parents) are often discouraged from entering academia because of the possible pay cut as a graduate student, sometimes excessive hours expected, and struggle to find childcare when conference travel is required.

• Selection process: When admissions decisions are filtered down based on a resume, it doesn’t encapsulate potential; it captures (utilization of) past opportunity. We’re missing a part of the admission pipeline that shows potential and interest. Instead, we incentivize “accomplishing diversity” in the cheapest and most convenient way possible.

Other resources

• David Bindel’s summary of Cornell’s efforts.
• AINow Institute’s work on discrimination in AI

An aside: some of the experience I bring to the table here

In my personal experience, I grew up in a town where many people return after college, if they go. A four-year university experience wasn’t expected of us like I have been able to see in my past life as a youth pastor just outside Washington, D.C. I started at my liberal-arts college as a math major because 1. we didn’t have engineering and 2. it was the only subject in high school that I continually found intriguing. We didn’t have any Computer Science classes at my high school. My parents don’t work anywhere near tech. When I started undergrad, I picked the cheapest school possible in state because my parents said they would pay for my graduate degree out of state (this speaks wonders to my financial privilege that my parents would have been able to help me out in that way) if I stayed in Florida for undergrad. I took my first Comp Sci class because it was required for the math major. I thought I was terrible at programming (which at the time, I equated to Computer Science) because I had never programmed before. After a month or so, I promised myself I would never take another CS class. In large part, this was because I felt alone in the class; I was the only woman. It seemed like everyone else knew what they were doing when I was lost. Thankfully, I found a wonderful advisor who encouraged me to apply to graduate programs after seeing how much I enjoyed research, but only after two years of avoiding the subject entirely. Mentorship from my undergraduate advisor is the only reason I’ve survived in the field

In the two years I spent avoiding computing entirely, I was able to do two internships as a youth pastor: one just outside D.C. in a predominantly wealthy suburb, and one in a rural North Carolina town. My students in D.C. seemed to be expected to do x, y, and z to boost their resumes to get into top universities. I honestly got burnt out just observing them for a summer. In North Carolina, it was a big deal when students went to a four-year university. Tech wasn’t particularly emphasized, since there weren’t any major tech companies in the area. People seemed to get jobs or degrees they were familiar with. In most cases, these young people weren’t shown that careers in tech were a viable option. By and large, they didn’t have the role models to show them a path to tech.