UC Berkeley’s Data Science Undergraduate Studies, or DSUS, advertises itself as being committed to diversity and inclusion, and as being the “future of data science”. In fact, School of Information Dean Jennifer Chayes recently said at a Campus Conversation: “We need this marriage of the qualitative and the quantitative.”
DSUS is failing at all of these goals. If this is the future of data science, I want out.
I worked as a coordinator for the Data Scholars Program, which its goal is to support students from underrepresented groups in data science by creating community, and offering mentoring and networking opportunities, as well as academic support. From my experience, very few of the more than 1000 students majoring in data science are BIPOC. And in 2020, women made up just 40% of the graduating class. The current makeup of this program is a clear indicator of how important the work of Data Scholars is and how much more DSUS needs to do to make its program an accessible and welcoming space to all students.
Data Scholars shifted from a space of inclusivity with exciting potential to one that, when I left, received no budget allocated from DSUS and survived off of a skeleton crew.
Since I left DSUS in August 2020, the program has been chipped away at and watered down. Three semester-long seminars of 25 or more students were reduced to one. The three instructors dedicated to supporting Data Scholars turned into one graduate student researcher. Four staff-members working part-time for Data Scholars fell to two. Data Scholars is one of the few programs in DSUS that specifically focuses on supporting underrepresented students, or even equity at all, and must receive the attention and funding it deserves.
Despite its intention, I even saw inequities being perpetrated within DSUS itself, suggesting that the program’s leadership needs more training on how to effectively support marginalized communities. During my five semesters at DSUS, I witnessed at least four staff members leave due to feeling overworked, under-resourced and disrespected. I saw students outside the Human Contexts and Ethics program, or HCE, — undergraduate and graduate — being treated as inferior by their supervisors. I also saw students subjugated to heavy favoritism, which more often than not seemed to reflect existing social inequities in terms of race, gender and disability.
I experienced being overworked myself. I felt pressure to volunteer myself for extra work so that students needing support could get it. While this may not be unusual for an academic environment, it is directly in opposition to the “supportive community” Data Scholars claims to build and certainly does not imply that students are “the heart of our program,” as DSUS suggests.
I also worked on the HCE student team as a graduate student researcher and student team lead for two and a half years. HCE’s mission is to “examin[e] the dynamics of technology development and human life in a datafied world,” a crucial goal in data science pedagogy.
While I worked with the HCE team in our capacity as curriculum developers, we were routinely ignored by professors in the electrical engineering and computer sciences, computer science and data science departments. When we did manage to get meetings with people who had the power to change curriculum and structure, we were criticized and seemed to have been not taken seriously. The administration above us seemed to allocate far fewer resources to us than other DSUS student teams. HCE actually receives almost all of its funding from external grants, like the Responsible Computer Challenge Grant.
If the staff at DSUS took the time to engage with HCE curriculum, they would know that in training the future generations of data scientists, ethics and an understanding of implicit racism are crucial. We are not computers. We are humans. The technology we build reproduces our own conscious and unconscious beliefs and the powers of the institutions and structures we live within. We are all influenced by bias, prejudice, racism, sexism and many other forms of violence.
DSUS needs to take concrete, accountable steps toward creating deliberately antiracist pedagogy. This should include increasing funding and resources for Data Scholars and the HCE courses and student team, as well as hiring its own Equity and Inclusion Officer. DSUS must commit to working with the HCE team to revise curriculums, public-facing materials including the annual National Workshop on Data Science Education and internal practices in order to center the experiences of marginalized students. People in all positions in DSUS must also educate themselves about antiracist pedagogy practices, engage in affirmative action practices and recognize that there is no one-solution-fits-all: Making spaces truly welcoming for all students means reimagining spaces from the ground up.
Students must make a conscious effort to pressure the department into making these changes as well; speak out in classrooms, forums and other academic spaces when instructors and classmates perpetuate injustices. Finally, I urge members of DSUS and Computing, Data Science, and Society’s leadership to take responsibility for the continued failure to make good on its promises of inclusion and equity.
DSUS does not do justice to students, and the university needs to put its money where its mouth is.