This work-in-progress innovative practice paper describes an approach and presents preliminary results of an effort by the NSF-funded Computing Alliance of Hispanic-Serving Institutions (CAHSI) to build research capacity of faculty at HSIs and students from underserved populations. A key factor in our nation's ability to innovate solutions to grand challenges and compete in a technology-enhanced world that rapidly changes is the involvement of individuals with different perspectives, experiences, and disciplinary knowledge. Diversifying representation in research cannot be achieved without involvement of HSIs, which enroll significant numbers of minoritized students in U.S. higher education. This paper describes a CAHSI-Google Institutional Research Program (IRP) that builds research capacity through partnerships between computing doctoral-granting CAHSI institutions and computing non-doctoral granting CAHSI institutions. This paper describes the IRP and its well-defined process to support faculty as they develop and refine research ideas and submit competitive proposals for funding through the IRP that includes a collaboration plan outlining coordination mechanisms and student professional development efforts.

Google's CS Research Mentorship Program (CSRMP) cultivates pursuit and persistence in the computing research trajectory for students from historically marginalized groups through virtual career mentorship from industry professionals, a peer community, and just-in-time resources. Since 2018, 287 Google mentors have engaged 1,018 students from 247 institutions in the U.S. and Canada. The program employs socioemotional support and advocacy to navigate systemic barriers by validating students' intersectional identities in order to improve outcomes in core constructs for students: self-efficacy, sense of belonging, research skills, motivation to pursue graduate school and research careers, and intersectional capital. Evaluation outcomes from 400 matched respondents (68% response rate) indicate that CSRMP affects positive, statistically significant change in those constructs that largely persists across demographic subgroups. 80% aim to pursue computing research careers, and significantly fewer students are undecided about their future career. We were also able to identify disaggregated learnings: Black, Indigenous, and Latinx students are significantly less likely to submit to a research conference, and students from Historically Marginalized Groups (defined within) are significantly less likely to apply to a CS graduate program. We discuss key design elements of the program, how the findings are informing future iterations, and the potential for the model to scale.

This panel brings together personnel from the University of Washington Seattle, the University of Massachusetts Amherst, and Google, who have helped shape the National Center for Women & Information Technology's (NCWIT) Extension Services for Graduate Programs (ES-Grad) a program designed to provide resources and strategies to enact systemic change in doctoral computing programs. Panelists will discuss barriers to recruiting and retaining more diverse students in doctoral computing education, as well as strategies and initiatives that can be replicated to foster equitable and inclusive PhD programs.

In this experience report, we describe the rationale and need for immersive research experiences (IREs) in computer science (CS) that are designed to foster an inclusive community that encourages pursuit of graduate education for undergraduate women. Google's exploreCSR supports institutions across the US to execute IREs in computing throughout the academic year. We describe the program design and framework, the evaluation model, and present outcomes from two years of implementation across 29 institutions, with 1,983 (92% female) student participants collectively. The unique features of the program are that it aligns goals, measurements, and best practices across a national network of hands-on, localized IREs, resulting in peer communities and a sizable sample of undergraduates who identify as women and/or African-American/Black, American Indian/Alaska Native/Native American, Hispanic/Latinx, and/or Native Hawaiian/Pacific Islander (AAHN). We discuss recommendations for effective IRE programming based on our evaluation and the features found to be particularly salient for AAHN women. The contribution of this work is in describing how a national initiative for IREs builds community and creates conditions known to support persistence of women in computer science.