It is now broadly recognized that women and U.S. racial and ethnic minority groups are acutely underrepresented in STEM fields. Multiple barriers, including the affordability of and access to new technologies and the lack of educational opportunities for diverse students, have led to a digital divide that demonstrates the inequities associated with who have access to data and the skills to extract knowledge from those data. Because of the nature of big data, collective action is urgent and imperative to work alongside the marginalized faculty and student communities who are disproportionately impacted by its use, while finding ways to empower those communities to advocate for their own data.

Data science education at Minority Serving Institutions (MSIs) can narrow the digital divide by ensuring a diverse student population has the knowledge and skills required to address issues of social and cultural relevance. MSIs have proven to be effective at recruiting, retaining, and preparing underrepresented students for the STEM workforce when compared to predominantly white institutions. Our project brought together current and former faculty from Tribals Colleges and Universities (TCUs) and Historically Black Colleges and Universities (HBCUs) to discuss opportunities to empower and support equitable data science education across these institutions and co-create solutions that ameliorate the digital divide. We focused specifically on the application of data science to the environmental sciences, due to the significant growth in large ecological datasets and environmental synthesis projects over the last decade. 

With this project, we aimed to co-create an actionable Recommendations Report for undergraduate environmental data science education at TCUs and HBCUs. Toward this end, the Leadership Committee convened HBCU and TCU working groups with environmental data science resource partners for several meetings and workshops, and solicited broader community feedback on our progress with a feedback form. The Leadership team then incorporated and synthesized the outcomes from these discussions and feedback form into a living Google document and finalized Recommendations Report. 

The published Recommendations Report (https://zenodo.org/record/8231167) contains all of the key findings from our work. In particular, some themes emerged across both groups (HBCUs and TCUs) when the sentiments expressed were analyzed for full or partial alignment. Integrating community in Environmental Data Science (EDS) research was a common aspiration across TCUs and HBCUs. For TCU WG members, additional funding to integrate language and culture with science so that the community and students learn from each other was a recurring theme. For HBCU WG members, the aspiration to integrate the community reflected who should be involved, and included at all levels (institutional, faculty, and student). In addition, HBCU WG members echoed the needs of TCUs for additional and more flexible funding to support their efforts and the needs of their students.

There was also partial alignment for a common theme related to mathematics. Both WGs expressed the need for EDS to be more flexible in the math requirements to prevent these requirements from becoming a barrier to successfully navigating a degree in EDS, as it can be for other STEM fields. For example, there was interest in creating early pathways in data analytics which emphasize data gathering, cleaning, analysis, and visualization. At the same time, data science programs could work with mathematics departments to customize a shorter prerequisite-free mathematics sequence that includes the linear algebra and calculus skills needed for understanding dimensionality reduction and machine learning. For TCUs, providing pathways for students to do data science without strong math backgrounds was identified as an existing strength, implying that they are already doing this. For HBCUs, creating opportunities for early exposure to data science core concepts regardless of math background was reported as an opportunity, or something they could be doing a little more of.

While each institutional group had its own key priorities, as well as its own values around how the work should be done, the Recommendations Report should represent possibilities for future investments in programs that support both institution types. For example, a large investment has been made into HBCUs, both private and National Data Science Alliance funding, to develop and share data science programming, research, and curricula. The possibility of investing in this level of coordination for TCUs was identified as an important future consideration. However, a key step toward this goal is to build relationships with tribes, especially around data governance. 

The overlap in themes identified in the report could be used to hone in on opportunities that make strides in areas benefitting multiple communities. For example, investing in research and innovation through community-research partnerships in data science could benefit both TCUs and HBCUs or promote minority success at other institutions. Furthermore, the overlap in priorities could represent opportunities to invest in broader collaboration across contexts, for example, sharing practices for scale and sustainability, creating and leveraging networks of support across administration, alumni, and other sectors.

Last Modified: 09/09/2023
Modified by: Micaela S Parker