Lauren Michael: The Missing Millions Democratizing Computation and Data to Bridge Digital Divides and Increase Access to Science for Underrepresented Communities
Download as PPTX, PDF0 likes212 views
The document discusses barriers to access in computation and data for underrepresented communities, emphasizing the urgency to expand and diversify support for researchers to address global issues by 2030. It highlights the importance of innovative NSF operations, inclusive access, and integrated investments in data, software, and community engagement to bridge digital divides. Recommendations include experimenting with grant processes and fostering partnerships to achieve systemic change in addressing disparities in technology access and utilization.
![Findings
1. There are substantial barriers
to access.
2. Accessibility = Access + Ability
3. Racial, gender, and other forms
of underrepresentation in data
and computing need to be
studied, socialized, and
addressed.
4. Insufficient engagement with
underrepresented institutions.
5. It is computation, and
software, and data.
“I am hopeful that this will create change. Too
often, agencies have these conversations and still
do business in the same way—with no change.
There needs to be a change in who gets funded
and who doesn’t. Most review panels haven’t
heard of many of the HBCUs.”
“Doing broader impacts is work that requires
expertise. Example of an NSF panel where
everyone was able to assess scientific merits but
no one had expertise in broader impacts. There
should be a seat on review panels for broader
impacts.”
“Career paths with data and computing [need to
be] clear and widely evident—especially for
community colleges and HBCUs.”](https://image.slidesharecdn.com/4nrplmichael-230330202813-5f91473b/85/Lauren-Michael-The-Missing-Millions-Democratizing-Computation-and-Data-to-Bridge-Digital-Divides-and-Increase-Access-to-Science-for-Underrepresented-Communities-5-320.jpg)
![Findings (cont.)
6. More diverse fields and
disciplines need support.
7. Facilitation successes need
expansion.
8. Distributed/edge CI
capabilities are expanding.
9. The community is open to
experimentation in NSF
operations.
10. Systemic change is needed.
“I am hopeful that this will create change. Too
often, agencies have these conversations and still
do business in the same way—with no change.
There needs to be a change in who gets funded
and who doesn’t. Most review panels haven’t
heard of many of the HBCUs.”
“Doing broader impacts is work that requires
expertise. Example of an NSF panel where
everyone was able to assess scientific merits but
no one had expertise in broader impacts. There
should be a seat on review panels for broader
impacts.”
“Career paths with data and computing [need to
be] clear and widely evident—especially for
community colleges and HBCUs.”](https://image.slidesharecdn.com/4nrplmichael-230330202813-5f91473b/85/Lauren-Michael-The-Missing-Millions-Democratizing-Computation-and-Data-to-Bridge-Digital-Divides-and-Increase-Access-to-Science-for-Underrepresented-Communities-6-320.jpg)
![Paths: Access+Ability
1. Ensure inclusive access
- 1A Expand investments in apprenticeships, internships and training grants
- 1B Promote the full spectrum of data and computing resources to researchers.
- 1C Expand investments in easy-to-use software, data and services.
- 1D Ensure investments integrate for frictionless workflows across data, software
and compute capabilities.
- 1E Explore investments in compute and data that don’t require command-line.
- 1F Develop entry-point mechanisms for K-12, undergraduates, and graduate
students to use data.
“Key is frictionless research computing and data.
Look at the entire pathway that data flow. Friction
points will be technical and cultural. More generally,
[reduce friction in] how we integrate data, compute
on data, share data, and store data—all have
multiple elements associated.”
Shorter-Term
Longer-Term](https://image.slidesharecdn.com/4nrplmichael-230330202813-5f91473b/85/Lauren-Michael-The-Missing-Millions-Democratizing-Computation-and-Data-to-Bridge-Digital-Divides-and-Increase-Access-to-Science-for-Underrepresented-Communities-8-320.jpg)
![Paths: NSF Operational Experiments
2. Innovate with experiments in NSF operations
- 2A ”Block grants” to HBCU, TCUs, HSIs and other minority-serving institutions and
consortia.
- 2B Experiment with solicitations and review processes that give greater weight to
broader impact.
- 2C Increase flexibility and broader leeway to program officers for RAPIDs, EAGERs,
RCNs, CC*.
- 2D Foster an “investment portfolio mindset” to include more ”high-risk, high
reward” initiatives aimed at reaching the ”missing millions”.
- 2E Support mechanisms and integrative adaptation to incorporate lessons learned.
“Experiment with reviews that give
greater weight to broader impacts
[relative to scientific merits].”
Shorter-Term
Longer-Term](https://image.slidesharecdn.com/4nrplmichael-230330202813-5f91473b/85/Lauren-Michael-The-Missing-Millions-Democratizing-Computation-and-Data-to-Bridge-Digital-Divides-and-Increase-Access-to-Science-for-Underrepresented-Communities-9-320.jpg)
Lauren Michael: The Missing Millions Democratizing Computation and Data to Bridge Digital Divides and Increase Access to Science for Underrepresented Communities
- 1. The Missing Millions: Democratizing Computation and Data to Bridge Digital Divides and Increase Access to Science for Underrepresented Communities Alan Blatecky (RTI), Damian Clarke (Alabama A&M), Joel Cutcher-Gershenfeld (Brandeis), Deborah Dent (Jackson State), Rebecca Hipp (RTI), Ana Hunsinger (Internet2), Al Kuslikas (AIHEC), Lauren Michael (MS-CC, Wisconsin) NSF #2127459 We acknowledge the traditional owners of the lands where Washington DC, Maryland and Virginia now stand -- the Pamunkey, Chickahominy, Upper Mattaponi, Rappahannock, Monacan, and Nansemond. We pay respect to their Elders past, present and emerging.
- 2. https://www.nsf.gov/pubs/2020/nsb20222/nsb20222.pdf https://www.nsf.gov/nsb/committees/vision2020cmte/NSB-missing-millions-figure-063021.png 2023 NCSES Diversity and STEM report: https://ncses.nsf.gov/pubs/nsf23315/report The Missing Millions
- 3. Motivating Question and CI Implications • Given the broad applicability of computing and data to virtually all fields and disciplines, access to CI plays a significant role in engaging the missing millions to address a growing global competition gap. • There is urgency in addressing these matters. The pace of change with technologies is ever-accelerating, and the COVID-19 pandemic has thrown into sharp relief social disparities and digital divides. How can NSF significantly expand, diversify, and support the development of new cohorts and communities of researchers to address pressing research, social, and global issues in 2030? From the Congressional Research Service: https://sgp.fas.org/crs/misc/R46613.pdf
- 4. 88 individuals across 15 focus groups + 6 interviews: • Researchers/scholars and leaders across scholarly domains • Early career scholars • TCU & HBCU Leaders, Faculty and IT staff • Other higher ed CI leaders and staff • Govt. and Center/Laboratory CI staff • Digital teaching and fabrication ecosystems • Industry experts https://www.rti.org/publication/missing-millions/fulltext.pdf Study Methods
- 5. Findings 1. There are substantial barriers to access. 2. Accessibility = Access + Ability 3. Racial, gender, and other forms of underrepresentation in data and computing need to be studied, socialized, and addressed. 4. Insufficient engagement with underrepresented institutions. 5. It is computation, and software, and data. “I am hopeful that this will create change. Too often, agencies have these conversations and still do business in the same way—with no change. There needs to be a change in who gets funded and who doesn’t. Most review panels haven’t heard of many of the HBCUs.” “Doing broader impacts is work that requires expertise. Example of an NSF panel where everyone was able to assess scientific merits but no one had expertise in broader impacts. There should be a seat on review panels for broader impacts.” “Career paths with data and computing [need to be] clear and widely evident—especially for community colleges and HBCUs.”
- 6. Findings (cont.) 6. More diverse fields and disciplines need support. 7. Facilitation successes need expansion. 8. Distributed/edge CI capabilities are expanding. 9. The community is open to experimentation in NSF operations. 10. Systemic change is needed. “I am hopeful that this will create change. Too often, agencies have these conversations and still do business in the same way—with no change. There needs to be a change in who gets funded and who doesn’t. Most review panels haven’t heard of many of the HBCUs.” “Doing broader impacts is work that requires expertise. Example of an NSF panel where everyone was able to assess scientific merits but no one had expertise in broader impacts. There should be a seat on review panels for broader impacts.” “Career paths with data and computing [need to be] clear and widely evident—especially for community colleges and HBCUs.”
- 7. “The challenge with broadening impacts is that it has to be part of the proposal, but that is usually the end of things. There is no accountability. We know that if we start a proposal with the technical aspects, the reviewers will never get to the community aspects. So, we start proposals with broader impacts. Most don’t, but if you start with broader impacts, that then becomes the first thing to be discussed in a review.” “The National Science Board should be engaged in this conversation. There should also be town halls on the missing millions hosted by professional societies. This will need a broad coalition.”
- 8. Paths: Access+Ability 1. Ensure inclusive access - 1A Expand investments in apprenticeships, internships and training grants - 1B Promote the full spectrum of data and computing resources to researchers. - 1C Expand investments in easy-to-use software, data and services. - 1D Ensure investments integrate for frictionless workflows across data, software and compute capabilities. - 1E Explore investments in compute and data that don’t require command-line. - 1F Develop entry-point mechanisms for K-12, undergraduates, and graduate students to use data. “Key is frictionless research computing and data. Look at the entire pathway that data flow. Friction points will be technical and cultural. More generally, [reduce friction in] how we integrate data, compute on data, share data, and store data—all have multiple elements associated.” Shorter-Term Longer-Term
- 9. Paths: NSF Operational Experiments 2. Innovate with experiments in NSF operations - 2A ”Block grants” to HBCU, TCUs, HSIs and other minority-serving institutions and consortia. - 2B Experiment with solicitations and review processes that give greater weight to broader impact. - 2C Increase flexibility and broader leeway to program officers for RAPIDs, EAGERs, RCNs, CC*. - 2D Foster an “investment portfolio mindset” to include more ”high-risk, high reward” initiatives aimed at reaching the ”missing millions”. - 2E Support mechanisms and integrative adaptation to incorporate lessons learned. “Experiment with reviews that give greater weight to broader impacts [relative to scientific merits].” Shorter-Term Longer-Term
- 10. Paths: Data/Software 3. Elevate data and software investments to the level of computing investments - 3A Support broader experiments and impact in usability of services with a focus on reproducibility, interoperability, extensibility and sustainability . - 3B Pioneer innovations in open and consistent sharing and reuse of data and software, especially for the next generation of students, researchers and scientists. - 3C Balance investments in innovative technology with investments in sustained services. - 3D Ensure that useful data can be found and utilized by anyone, anywhere. “Decisions about what is funded and prioritized should reflect interests in a democracy—not limited to the views of academics at the front of their field. It is not the newest thing—shiny objects—what about sustainability of technologies, usability of technologies?” Shorter-Term Longer-Term
- 11. Paths: Expanded Scope and “Edge” Engagement 6. Enable technologies and communities that enhance capabilities at the “edge” - 6A Invest in CI and community laboratories at the “edge”. - 6B Explore investments in edge data, software, and compute interfaces that are easily accessible (GUI, Apps). - 6C Build ubiquitous compute and data, widely distributed and accessible. “Closing a digital divide implies movement toward equal access. Beyond equal access is equitable access—people needing more help get more help.” Shorter-Term Longer-Term
- 12. 8. Accelerate the broader impacts of CI investments - 8A Document past NSF awards that have achieved broader impacts with technology adoption. - 8B Provide technical assistance to prospective PIs on effective ways to address broader impacts. - 8C Approach broader impacts with the same level of methodological rigor given to intellectual merits. “The challenge of with broadening impacts is that it has to be part of the proposal, but that is usually the end of things. There is no accountability. We know that if we start a proposal with the technical aspects, the reviewers will never get to the community aspects. So, we start proposals with broader impacts. Most don’t, but if you start with broader impacts, that then becomes the first thing to be discussed in a review.” Shorter-Term Longer-Term Paths: Racial, Gender Underrepresentation and Broader Impacts
- 13. Paths: Social and Multi-Agency Sustainability 10. Bridge across directorates and federal agencies 11. Foster Consortia and Partnerships 12. Engage with NSF Leadership “The problems in higher education are very siloed—you have to get rid of the silos. The same is true with government agencies. NSF needs to collaborate more with FAA, NOAA, NASA, NIH, DOD, OSTP—cyberinfrastructure is cross-cutting. Otherwise, you will not have the resources you need—you will only be talking to yourself.” Shorter-Term Longer-Term
- 14. •Manish Parashar, National Science Foundation •Matthew Rantanen, Southern California Tribal Chairman’s Association Your Panelists
- 15. Read the Report Image Source: https://www.cio.com/article/3296703/how-to-find-and-implement-emerging-technologies-as-a-cio.html https://www.rti.org/publication/missing-millions/fulltext.pdf
Editor's Notes
- #10: We then asked study participants for a vision of success toward specific recommendations for the NSF, which were organized into 12 “Paths”. These are highly intersectional, where many of the 12 overlap with one another. For the sake of time, I’ll highlight some specific recommendations from some of these, and referencing later sessions in the workshop.
- #11: For each Path, we organized specific recommendations in terms of short-term priorities and longer-term strategies. Captured in the first Path, we heard repeat requests for investment toward improving Access, Awareness, and Ability, including Coordinated education of the research community around existing CI investments and priorities Interfaces and entry points to data and workflows that are more accessible to early users, including those at the K-12 and undergraduate levels And, training and professional development programs for underrepresented communities
- #12: We found that those we gathered input from were highly supportive of high-risk, high-reward experimentation in NSF operations and opportunities specifically geared to engage the Missing Millions. For example, Block grants investing in institutions and consortia that engage underrepresented minorities (you’ll get to hear perspectives and examples from minority-serving campuses and referencing consortia they're building). Funding opportunities with increased weight on broader impact and even with the specific objective of reaching the missing millions.
- #13: Another of the strongest sentiments from participants was the frustration that the most significant CI investments (by their perception) have gone toward ‘big HPC’, - with less prioritization on easy-to-use interfaces to accessible data and methodologies (software) that are necessary to on-ramp future researchers from diverse scholarly domains and underrepresented groups. There was also a recognition that sustained investments in the most successful services are necessary, in balance with incentives to innovate on new interfaces.
- #14: There was recognition that underrepresented voices and scholarly domains need to be engaged and explicitly included so that NSF and CI-specific funding priorities reflect their interests, And that this means NSF needs long-term strategies for including these voices not only in the funded research portfolio, but at other layers of NSF operations and decisions.
- #15: And so, these prior Paths intersect to address the observation that a specific set of research problems have traditionally driven the most significant investments in CI, whereas scholarly domains with significant and direct social impact (a lot of the social sciences, humanities, public health) are not only underrepresented, but have greater draw and representation across diverse communities. There were sentiments that NSF needs to adopt a clean-sheet approach to developing CI priorities that address more representative research and education priorities.
- #16: Relatedly, our 6th finding captures the idea of engaging the missing millions by physically bringing CI *to* underrepresented communities, with distributed and accessible modes of participation not just for institutions of higher ed, but for their surrounding communities and *through* these institutions. - Examples mentioned in the report include some you’ll hear about in the next session - OSG, NRP - which our keynote speaker also referenced earlier this morning.
- #17: We also heard multiple assertions that underrepresentation, and factors contributing to it or designed to address it, need to be prioritized as a specific area of study, engaging and developing underrepresented participants and institutions to contribute to these areas.
- #18: Path #8 represents recommendations to support the ability of CI projects to plan for and achieve broader impacts, beyond enhancing the weight and rigor of review, such as: Curating and promoting examples of projects with successful broader impact in technology adoption Providing services to help projects develop and execute on their broader impact strategies
- #19: Reflective of the social impact priorities of underrepresented communities, we even heard calls for studying the social and environmental impact of computing investments, specifically, as a critical component to understanding their societal value.
- #20: And the last few Findings really address the processes by which NSF can achieve and sustain these impacts, by engaging relevant partnerships not just across directorates, but across agencies that fund research and education, given the cross-cutting criticality of cyberinfrastructure. Our participants also recognized the need for NSF to gather input from existing consortia that engage and represent underrepresented communities, and to support the development of new consortia where there are gaps in representation. You’ll get to hear about a new CI-driven consortium of minority-serving institutions, tomorrow. And finally, a recommendation to begin regular dialogue between the advisory bodies to the OAC, CISE, and NSF, in compliment with the prior.