September 2014 Vol. 26/No.8
Jane Stout, Director of CRA’s Center for Evaluating the Research Pipeline (CERP), is the PI on a three-year grant recently awarded to CRA by NSF for the amount of $783,975. The project, entitled Promoting a Diverse Computing Workforce: Using National Survey Data to Understand Persistence Across Undergraduate Student Groups, will use data collected from CERP’s Data Buddies program, to understand predictors of student retention in computing, with specific focus on the experiences of underrepresented groups. Data collection for the project will begin during the fall of 2014, and track students’ successes and persistence through 2017.
CERP’s Data Buddies methodology involves semiannual collection of large-scale, cross sectional survey data that measure the experiences of students pursuing computing career tracks from a variety of institution types (PhD granting; terminal Masters granting; liberal arts colleges; women’s colleges; HBCUs; MSIs), and from a wide array of demographic groups (gender; race/ethnicity; first-generation college students; students with disabilities). A particularly striking characteristic of Data Buddies data is that the datasets are large enough to hone in on the experiences of many different student populations – an endeavor that is challenging if not impossible at single-site social science research centers due to underrepresented groups’ inherently small size. This project will focus on broadening participation in computing by measuring theoretically validated predictors of achievement and persistence, then assessing whether those predictors are equally important across student groups and across time.
The project’s research team is well suited to conduct this type of research. Stout has a PhD focusing on social psychological and education theory, extensive experience leading large-scale survey research in education settings and expertise in a host of quantitative analytic methods. CERP’s two research associates with advanced degrees in social science fields, Ama Nyame-Mensah and Heather Wright, will also work on the project. The project’s advisory board is composed of leaders in the computing community who head diversity initiatives for a broad range of student demographics and will play a central role in project oversight: Jamika Burge, Coalition to Diversify Computing (CDC); Richard Ladner, The Alliance for Access to Computing Careers (AccessComputing); Marigold Linton, Society for Advancement of Hispanics/Chicanos and Native Americans in Science (SACNAS); Rebecca Wright, Committee on the Status of Women in Computing Research (CRA-W). CERP’s infrastructure, the basis for data collection, is well established, as evident by its 4-year history of successfully collecting large datasets from students in computing programs.
The project will generate a clear set of strategies and best practices to promote persistence among a broad range of students at the postsecondary level, which will be offered to computing departments across the country. Stout will present findings to computing departments in CRA’s and CERP’s professional networks, and at high-impact professional conferences concerned with education in the computing community, such as the CRA Conference at Snowbird and SIGSCE.
The Computing Community Consortium (CCC) invites proposals for visioning workshops that will catalyze and enable innovative research at the frontiers of computing. Successful activities will articulate new research visions, galvanize community interest in those visions, mobilize support for those visions from the computing research community, government leaders, and funding agencies, and encourage broader segments of society to participate in computing research and education. Past examples can be found here.
Workshop organizers are expected to bring together a group of scientists and practitioners in the area of interest, and to formulate a program that encourages new ideas, innovative thinking, and broad discussion. Workshops can be of varying sizes, typically ranging from 20 to 100 participants. It is important that the participants cover a broad spectrum to ensure full coverage of the area, both in terms of content area representation and employment (academia, industry, research labs, and policy and funding organizations).
Workshops are expected to have a tangible output – for example, a whitepaper (or set thereof) or a workshop report. Workshop outcomes should be targeted to multiple audiences (the research community, science policy groups or funding agencies, the general public), and the deliverables should be tailored for easy dissemination. CCC will help to support both workshop organization and the subsequent generation and communication of the output.
The CCC encourages creative ideas from all segments of the computing research community on topics ranging from the formulation of new basic research areas and technologies to the use of new or existing research ideas and technologies to address important scientific or societal challenges.
Proposals may be emailed to firstname.lastname@example.org at any time. For CCC planning purposes, proposals with start dates prior to September 2015 should be submitted by December 1, 2014. The full RFP and a Visioning Best Practices Guide can be found here. The CCC will host a webinar about Visioning on October 2, 2014 at 1:00pm ET to answer your questions. Please register here.
By Suzi Iacono, Acting Assistant Director of NSF for CISE
It was great to see so many of you during this summer’s CRA Snowbird conference! The talks were great and the side conversations stimulating. To keep the dialogue going, I want to highlight some of our upcoming activities.
CISE Assistant Director Transition and Search Update
As you know, Farnam Jahanian, who served as the CISE Assistant Director (AD) from March 2011 to August 2014, has transitioned to Carnegie Mellon University, where he is now serving as VP of Research. During his tenure, CISE led three Presidential initiatives: the National Robotics Initiative, the National Big Data Research and Development Initiative, and US Ignite. He oversaw the formulation of more than 25 new solicitations, including several collaborative efforts that spanned multiple NSF directorates, federal agencies, and the private sector. Farnam was a strong advocate for how basic research can be uniquely central to an innovation ecosystem that drives global competitiveness and addresses national priorities. His contributions are sure to provide a lasting legacy for the future of our discipline. The next time you see or talk to Farnam, please thank him for his exceptional leadership to the CISE directorate and community.
NSF is actively searching for the next AD for CISE. As is customary at NSF, the Deputy AD serves as the Acting AD in the interim. I want to thank the search committee for their work; I expect there will be an update soon.
CISE Committee of Visitors
CISE is convening a Committee of Visitors (COV) this fall for the divisions for Computer & Communication Foundations (CCF), Computer & Network Systems (CNS), and Information & Intelligent Systems (IIS). (The Division of Advanced Cyberinfrastructure (ACI) recently conducted a COV so their next one will be sometime in the near future.)
NSF relies on the judgment of external experts to maintain high standards of program management, to provide advice for continuous improvement of NSF performance, and to ensure openness to the research and education community served by the Foundation. COV reviews provide NSF with external expert judgments in two areas: (1) assessments of the quality and integrity of program operations and program-level technical and managerial matters pertaining to proposal decisions; and (2) comments on how the outputs and outcomes generated by awardees have contributed to the attainment of NSF's mission and strategic outcome goals.
CISE thanks Jim Kurose from the University of Massachusetts, Amherst for serving as the chair of the COV and the three vice-chairs who are serving for each of the divisions: Salil Vadhan of Harvard University for CCF, Peter Steenkiste of Carnegie Mellon University for CNS, and Nancy Amato of Texas A&M University for IIS.
CISE Advisory Committee Meeting
We are planning to hold the fall meeting for the CISE Advisory Committee (AC) on November 13-14, 2014. The CISE Advisory Committee provides advice and recommendations to NSF concerning support for computing research, education, and infrastructure. More specifically, the AC provides up-to-date information on the state of the field and the many challenges we all face and advice on the impact of NSF policies and programs on the CISE community and works on special issues, for example, by forming ad hoc subcommittees and working groups to carry out studies, as needed.
Over the past few years, the CISE AC has formed ad hoc groups around the issues of Computer Science Education and Workforce Development, CISE Midscale Infrastructure Investments, and Visioning CISE 2025. Their advice on these issues and others of importance to the CISE community has led to several important outcomes. The agenda of this summer’s CRA Snowbird Conference was a testament to that. Several CISE AC members led sessions to engage the broader community in conversations about these issues and, in addition, held a town hall meeting to discuss the growing enrollments in computing courses. CISE is grateful for the active engagement and thoughtful contributions from our AC members.
The meeting this fall is open to the public and CISE is currently working with the co-chairs, David Culler of UC-Berkeley and Fran Berman of RPI, to formulate an engaging agenda. One topic that is certain to be on the agenda is the report from the COV chair and vice-chairs.
Other CISE staff changes
CISE is pleased to announce that Lynne Parker of the University of Tennessee will serve as the Division Director of IIS beginning in January and will be joining as an expert this fall. Chaitan Baru from the San Diego Supercomputer Center at UCSD has recently joined CISE as Senior Advisor for Data Science. We are happy to have Lynne and Chaitan join the CISE team and look forward to working with them.
We also have many new Program Officers joining us, as well as several who have returned to their home institutions. We are also happy to have several new AAAS and Einstein Fellows joining CISE this fall. The work of the rotating Program Officers and Fellows is invaluable to NSF and helps the U.S. to maintain its leadership in scientific discovery and engineering innovation.
New Video Showcases CISE-funded Computer Science Curriculum
A new NSF-produced Science Nation video illustrates an innovative computer science curriculum – Exploring Computer Science, ECS. Created by a team lead by Jane Margolis of UCLA, ESC has broadened the participation of students taking computing courses. Today, more than 2,000 students in the Los Angeles United School District (LAUSD) are learning computer science through ECS each year. Many of these students are African American and Latino. In addition, at a time when the national average of female students who participate in AP computer science is about 19 percent, the LAUSD ECS enrollment is 40 percent female – twice the national average.
We’re always interested in promoting exciting news from our NSF-funded projects. If there’s news that you’d like us to know about, please contact: email@example.com.
This fall, CRA is implementing a major upgrade to the Taulbee survey.
Why a new Taulbee? We began with the goal of developing the long-requested ability for each department to select an individual peer group and compare key results. As we looked into it, we realized that the entire Taulbee process needed improvement. There were too many user-unfriendly aspects to the online interface, too many ways for bad data to slip through, and too many manual steps in the analysis. We wanted to provide peer group reporting, but even more than that, we wanted to improve the timeliness and accuracy of information available to the computing community.
But I rely on Taulbee! Don’t change it! We’re changing the infrastructure, not the substance. This year’s survey will ask for the same data as last year. The same reports will go back to participating departments and be published in CRN.
So what’s new? The underlying survey software and data management are completely different. After reviewing responses to an RFP, we selected a vendor called Peerfocus ( www.peerfocus.com ) that specializes in helping associations conduct Taulbee-like functions of collecting data from individual members and reporting back aggregate results without letting members see each others’ individual responses. Their existing clients include several higher education associations, and their software already had many of the features we required and some we hadn’t thought to ask for. The new platform will also have enhanced security capabilities.
What’s been done so far? Peerfocus has been working with CRA to configure and customize their system to Taulbee’s requirements. Survey entry was beta tested by 12 departments at 11 institutions who re-entered their 2013 Taulbee data; we made a number of fixes and adjustments based on their feedback. (Thanks to the testers at Carnegie Mellon, Drexel, Duke, Georgia Tech, Ohio State, Purdue, Rochester Institute of Technology, University of Maryland Baltimore County, University of Massachusetts Amherst, University of Toronto, and University of Utah.) The new system was demoed at Snowbird in July.
What will the obvious differences be?
If you’re responsible for Taulbee data entry, you’ll see the following differences:
1. Each Taulbee user will have his/her own individual password, not a group password per academic unit.
2. The survey pages will look different (but collect the same data).
3. There will be additional data validation, including the ability to compare to previous year’s data so that the system can question if, for example, you awarded 175 BS degrees last year but report awarding either 18 or 1800 this year.
4. There are administrative features built into the system that mean you should see fewer general messages to all Taulbee users, and more that are applicable to your particular status.
5. There is a section at the end of the survey for feedback on the survey process itself. We welcome your comments and will be very attentive to issues and concerns with the new system so that we can make further improvements.
There will be a user guide to the new system to help make the changeover.
September 15, 2014 - Pdf for data gathering and new user guide available
October 15 – Online survey open
November 19 - Preliminary deadline (salary data needed by this date)
End of December - Preliminary salary report to participants
January 20, 2015 - Final deadline
Early April – Full report to CRA members and participating departments
May - Full results published in CRN
Spring – Benchmarking/peer group comparison feature available
By Jennifer L. Welch
Nancy Amato, Unocal Professor of Computer Science and Engineering at Texas A&M University, has had a banner year. She is the recipient of two prestigious awards for mentoring, the Habermann and the Harrold/Notkin awards, elected to the CRA Board, and will shortly be of CRA-W. She exemplifies teaching, research and service excellence in computing.
Habermann Award: CRA's A. Nico Habermann award is presented annually to someone who has made outstanding contributions aimed at increasing the numbers and/or success of under-represented groups in the computing research community. Nancy received this award in 2014. Her passion is involving undergraduates, especially women and members of under-represented groups, in research. She has accomplished this goal particularly through her work with the CRA-W/CDC's Distributed Research Experience for Undergraduates (DREU) program and the National Center for Women & Information Technology (NCWIT) Academic Alliance.
Nancy has been a key contributor to CRA-W/CDC's DREU program (formerly known as DMP, Distributed Mentor Program). She joined the DMP as Director (2000-2003) and has been Co-Director since 2004. This highly selective program matches promising undergraduate women and undergraduate men from under-represented groups in computing with faculty mentors for a summer research experience at the faculty member's institution. Since 1994, roughly 800 undergraduates from 300 institutions and mentors from 100 host research universities have participated in DREU. Nancy's tenure with DREU oversaw a large scale-up in the program, with applications increasing ten-fold (from about 50 in 2000 to more than 450 in 2014) and participation increasing more than three-fold (from about 20 per year to 60-70 per year). In addition, the program expanded from all women to include men from under-represented groups. One of her early efforts was a successful NSF proposal that funded the program for five years, starting in 2002, with $1.6M. To handle the scale-up, she developed a web-based system for administering the program, which has been adopted by several other programs, including the Grace Hopper Conference. To increase participation further, she encouraged the mentors to provide matching funding. The DMP project, in part, led to the CRA-W being honored with the 2003 Presidential Award for Excellence in Science, Mathematics and Engineering Mentoring for "significant achievement in mentoring women across educational levels". A study conducted in 2011 showed that 39% of the DREU participants attended graduate school in computing as compared to 22% of undergraduates who instead had a comparable, but some other, undergraduate research experience.
Nancy has also contributed greatly to the NCWIT Academic Alliance. NCWIT is a coalition of more than 450 prominent corporations, academic institutions, government agencies, and non-profits working to increase women's participation in information technology. Its Academic Alliance (AA) consists of over 275 colleges and universities, which work together to implement changes in higher education to increase women in IT. Nancy was the co-chair of the AA from 2009 to 2011, and since then has served on the advisory committee. While she was co-chair, Nancy set up the current administrative structure for the AA and drafted the bylaws describing the structure. Drawing on her passion for involving undergraduates in research, Nancy conceived and oversaw the NCWIT AA's REU-in-a-box project, in which the AA worked with the NCWIT social scientists to develop a resource that guides faculty in developing research experiences for undergraduates. The material explains the benefits of such experiences and guides prospective faculty mentors through the process.
Harrold/Notkin Award: The NCWIT Harrold and Notkin Research and Graduate Mentoring Award is given in memory of Mary Jean Harrold and David Notkin, to recognize faculty members who combine outstanding research accomplishments with excellence in graduate mentoring. Nancy received this award in 2014 as the first ever recipient. Nancy is a leading researcher in robotic motion planning, and she and her students invented the algorithmic foundations of sampling-based motion-planning and have applied these techniques to computational biology. She also works in computational geometry and in parallel computing, particularly on parallel algorithms and data structures. Her honors include IEEE Fellow and AAAS Fellow.
Motion planning deals with finding paths to move an object from an initial position to a goal position in some space. Nancy's work has developed probabilistic roadmap methods (PRMs) for this problem. The original PRM (Kavraki, Svestka, Overmars, and Latombe, 1996) used uniform random sampling to construct a compact representation of feasible paths in the space. Nancy and her students proposed several novel PRM variants, creating sampling methods that increase the number of samples in the critical areas of a search space that require more accuracy. These targeted non-uniform sampling methods significantly advanced the state of the art in motion planning and other search algorithms because they make PRM applicable to a much wider set of problems, for which they previously were not feasible. For example, they showed for the first time how to apply PRM methods to navigate narrow passages in motion planning.
Nancy has led the community in showing how to use PRMs to understand molecular motions, and in particular, to simulate protein and RNA folding. Understanding how these molecules fold is a challenging and important problem in biology. Nancy had the key insight to apply PRMs to this problem. Surprisingly, the only substantive change required was to substitute the collision detection check used for robotic applications with a check that favors protein conformations with low potential energy. This approach allows rapid simulation of detailed information in a way not possible before. As a result, Nancy's group opened up a new research area in computational biology. Key graduate student participants in this project were Guang Song (Iowa State University), Shawna Thomas (Texas A&M), and Lydia Tapia (University of New Mexico).
Nancy has also made significant contributions in computational geometry. She and her graduate student Jyh-Ming Lien (George Mason University) introduced a novel technique, called approximate convex decomposition, for partitioning a polyhedron into approximately convex pieces. The result provides similar benefits to those obtained from perfectly convex pieces but can be computed much more efficiently. Essentially any problem that deals with large geometric models, including graphics animations, CAD/CAM, and solid modeling, may benefit from this technique.
Finally, Nancy has made important contributions to parallel computing, both systems and algorithms. She collaborates with Prof. Lawrence Rauchwerger, also at Texas A&M, on a large project developing a parallel C++ library called STAPL (Standard Templates Adaptive Parallel Library) designed to ease the task of parallel programming. Applications developed using STAPL have been used in three Department of Energy Predictive Science Academic Alliance Program centers. On the algorithmic side, together with student Roger Pearce, she developed novel techniques for efficiently processing large graphs, such as social networks, that are "scale-free", a property that creates load balance challenges for parallel processing. Nancy and Roger's method overcomes these challenges and has resulted in experiments that achieved a seventh place rating in the 2011 Graph500 competition and additional experiments that were featured in the 2012 Graph 500 list.
Nancy is a remarkably active and gifted research advisor, both to graduate and undergraduate students. She has built a cohesive research group that has a lot of fun, while being extremely productive in research. Since her arrival at Texas A&M University in 1995, she has graduated 13 PhD students (six from under-represented groups); seven of these students have gone to academic careers, three to research labs, two are postdocs, and one works at a startup company. She has 13 current PhD students (seven from under-represented groups). She has graduated 18 master's students (ten from under-represented groups) and has seven current master's students. She has worked with more than 100 undergraduates and five high school students, the vast majority of whom are women and under-represented minorities.
In 1996, Nancy established and has continuously served as the faculty advisor for Aggie Women in Computer Science (AWICS), an organization devoted to improving the environment, both socially and professionally, for women undergraduate and graduate students in Computer Science and Engineering at Texas A&M. AWICS has been a remarkable success. It has received funding from several companies, sponsors distinguished lectures, organizes a professional development seminar series, administers a peer-mentoring program, and more. AWICS has brought a large number of women to the Grace Hopper Conferences over the years and was one of the first ACM-W chapters.
CRA Leadership Roles: Nancy was elected to the CRA Board of Directors in February 2014 for a three-year term starting in July 2014, and she will begin a three-year term as co-chair of CRA-W in fall 2014.
About the Author: Jennifer L. Welch is Regents and Chevron II Professor of Computer Science and Engineering at Texas A&M University. She received her B.A. from the University of Texas at Austin and her S.M. and Ph.D. from the Massachusetts Institute of Technology. Her research interests are in distributed computing.
By Jane Stout, CERP Director
One hundred eighteen graduate students (n = 75 women, n= 143 men) indicated (a) the degree to which they endorse the stereotype that women are less capable in computing that men; (b) how much they felt they “belong” in computing and (c) their self-efficacy in computing. Men endorsed the negative stereotype to a greater degree than women, p < .01. However, among women, stronger endorsement of the negatively stereotype was associated with a lower sense of belonging and lower sense efficacy in computing, ps < .05; men’s stereotype endorsement was unrelated to their belonging and self-efficacy. These results highlight the importance of fostering a stereotype-free training environment so that women’s self-concept in computing is unconstrained by negative cultural beliefs about their ability.
Note: Stereotype endorsement Stereotype endorsement was assessed by asking students to indicate their agreement with and aggregating the following items: Although some women might be good at computing, women in general tend to be better at other things; there is no doubt in my mind that women are just as talented at computing as men are (reverse scored); Computing fits men’s personalities better than women’s; Computing seems to come more naturally to men than women, using a scale ranging from (1) Strongly disagree – (7) Strongly agree. Belonging was assessed by asking students to indicate their agreement with and aggregating the following items: I feel like I belong in computing; I feel like an outsider in the computing community (reverse scored); I feel welcomed in the computing community; Computing is a big part of who I am; I do not have much in common with other people in computing (reverse scored); I see myself as a computing person, using a scale ranging from (1) strongly disagree – (7) strongly agree. Self-efficacy was assessed by aggregating the following items regarding students’ confidence that they could do the following: Become the go-to person for expertise in your content area; Publish papers as first author in the top journals of your field; Discuss theory with senior members of your field; Win awards for your work; Articulate thoughtful answers to theoretical questions about your work during a presentation; Successfully receive funding for a grant on which you are the Principle Investigator (PI); Become a respected member of the research community in your research area, using a scale ranging from (1) Not at all confident – (5) Extremely confident.
This analysis brought to you by the CRA’s Center for Evaluating the Research Pipeline (CERP). Want CERP to do comparative evaluation for your program or intervention? Contact firstname.lastname@example.org to learn more. Be sure to also visit our website at http://cra.org/cerp/.
By CCC Blog
Inspired by the traditional Japanese art form of origami, researchers from Harvard University and Massachusetts Institute of Technology (MIT) have figured out how to take flat sheets of specialized paper and plastic and make it self-fold into a complex machine that can “get-up and go”.
The robot starts as a flat sheet with embedded electronics, which can then transform autonomously into a functional machine. This is done using shape-memory composites that fold themselves along embedded hinges to recreate fundamental folded patterns. The origami-inspired robot can fold itself in 4 minutes and walk away without human intervention. This demonstrates the potential for both complex self-folding machines and autonomous, self-controlled assembly.
“This is a new approach of making self-folded machines that have computational and controllable function upon completion of the self-folding process,” said Daniela Rus, Director of the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT and one of the authors on the Science report titled “A method for building self-folding machines”. Rus is also a member of the Computing Community Consortium (CCC).
In previous work done by Rus and her colleagues, the self assembly was a single step process and all the joints folded simultaneously under a uniform form. That proved to have some limitations, compared to their new multiple-step design.
“The process produces complex shapes that have the ability to self-assemble using multiple steps. The self-assembling process is partially encoded as computation and partially encoded in the mechanical structure. The resulting mechanisms are dynamic and have function–they can move in controlled ways,” said Rus.
There are many potential uses for these self-folding machines, including search-and-rescue scenarios where they could navigate small tunnels, deploy into space for various forms of exploration, or for self-folding shelters that rapidly assemble in disaster zones. The additional benefit is that these robots can be shipped flat in large quantities and then assembled on-site, making them potentially extremely valuable for our future.
The work was funded by an NSF Expeditions in Computing grant as well as an NSF Emerging Frontiers in Research and Innovation (EFRI) Origami Design for Integration of Self-assembling Systems for Engineering Innovation (ODISSEI) grant. See the Science article for more information.
Photo: Jason Dorfman, CSAIL photographer. An insect-like robot printed and designed using the new process being developed to revolutionize the way robots are developed.