REUs: Research Experiences for Undergraduate Students
From Anita Borg Institute Wiki
From the proceedings:
Panelists: Joan M Peckham (University of Rhode Island), Teresa Dahlberg (UNC Charlotte), Sheila Humphreys (UC Berkeley), Daniela Stan Raicu (DePaul University), Ingrid Russell (University of Hartford), Harriet Taylor (National Science Foundation)
Researchers are sometimes daunted by the specter of bringing undergraduate students into their labs. Experienced undergraduate research mentors will talk about the challenges and benefits of research programs for undergraduates. This will include the design of student projects, training in team work, ethics, problem solving, and the dissemination of research. Other topics include program logistics and the benefits to faculty and their institutions for hosting REUs.
Joan introduced everyone. Harriet from NSF couldn't make it, but is enthusiastic about REUs and wants to encourage people to check it out. She is the Program Director of the [link http://nsf.gov/crssprgm/reu/ NSF CISE REU Program]
Each panel members had several minutes to talk.
Teresa - Practical issues about running an REU site when you have to do everything else as a faculty member
Logistics of the UNC Charlotte site
- the goal of the site is to "engage as many faculty as possible"
- this lessens the burden on individuals
- diversity of research interests and approaches is beneficial to students
- divide and conquer
- had 14 and then 19 students split between four labs
- one faculty leader per lab, either senior or very experienced with undergraduates
- have a graduate student as a research mentor
- 3-6 undergrads in each lab
- the second year, had assistant professors and more graduate students as well
The structure helps foster collaboration among faculty, since sharing students is a good way to share research. It is also good for the undergraduates, because they are a part of a real lab and can participate in meetings and parties, but also have peers. Being part of a team is good because the undergraduates are all at different levels. Some might be engaging in research, some might be developing the software that supports research. Teresa tries to get undergraduates to return and see the progression from doing support to doing research directly.
The REU sites also had activities across the labs with all the undergraduates, with a weekly seminar.
How do you get the faculty interested?
- REUs groom PhD students. All three senior undergraduates enrolled in the graduate program after their REU
How do you keep the "good stuff alive throughout the school year"?
- have the undergrads develop "outreach based on their research"
- pair REU site with STARS Leadership Corp
- STARS gets undergraduates doing outreach to the middle schools and high schools.
- [It makes the students] "Figure out how [they're] going to tell a younger student about what you did and get them excited about it"
- might result in students going out to schools giving presentations, or giving demos at open houses and stuff
Lane, a student in the visualization lab, had a few comments:
- "Came in as a math major"
- the program "created opportunities"
- now Research Assistant and computer science major
- It is important to keep it going after the summer program ends - "Needs to be a continuation"
- REU provided a research project to present and the confidence that goes with it
- it "makes you want to give back"
Ingrid - Sources of funding for undergradute research
University of Hartford is not an REU site, but has several sources of funding for undergraduate research programs.
REU supplements
- For existing NSF projects it is "relatively easy to get an REU supplement"
- grants are through NSF/CCLI
- easy to apply, not full review
- "not a large amount but you don't have overhead" "no request for indirect cost"
- usually Ingrid has used 2-3 students for projects related to AI class
- eg) web crawler, ml techs for automatic tagging
- REUs don't support faculty time, just student
- REUs require students to be US citizens
State Sources: "Connecticut Space Grant Consortium"
- in this case NASA-funded
- "has to be relevant to NASA" "but I think it's not very hard to make a case for that"
- college has to be part of a consortium
- this can fund travel for faculty and students and faculty time
- not a lot of money but makes it possible
- need initial contact with some NASA centers
- potential to lead to NASA internships for students
- NASA requires students to be US citizens.
Corporate R&D Labs
- in this case, United Technology Research Center and Pfizer Global research and development
- university has a center "U of H Engineering Applications Center" that tries to get corporate sponsors
- Ingrid goes through this center rather than approaching companies directly
- Benefits of corporate approach for undergraduates [slide]
- real-world applications
- dealing with clients
- visits to corporate facility and labs
- multidisciplinary: problem domain and team
- potential for summer internships
- very good pay
- opportunities for international students: employment is under the auspices of a university program
With corporate projects, it tends to start with a project and require recruitment of students to the project. In the other two, it tends(?) to be the other way around. Corporate project also have NDAs and confidentiality issues.
Example project with Pfizer - voice acoustics as biomarker. "Can speech signals be used to monitor the response to treatment: Parkinson's disease, depression, . . . ?"
Jan mentions that another option for research programs is to implement them as part of a curriculum to get teaching credit.
Daniela - How do you bring it to publication?
- Daniela is PI of a medical informatics REU
- started in 2005
- runs for 10 weeks during the summer
- bring students from around the country
Info [slide]
- 26 students
- 46% women
- 15% first generation college students
- 73% students from outside DePaul and Northwestern
- 40% of the students were sophomores and juniors"
At DePaul 26% students are female, so the REU has been a successful draw, perhaps because the project illustrates the application of abstract computer concepts to practical problems like medicine.
- DePaul provides the CS and Northwestern the med
- "first significant research experience for most of the 26 students"
Student Projects need to have: [slide]
- clearly defined objectives and a limited knowledge set required for success
- indicate why the project is iportant
- should be manageable within the period of the REU program
- should also expose students to all phases of research
Dissemination of results
- present results midway (lit review and implementation ideas) and at the end (results)
- "Every other week" project reports
- allows mentors to check on progress and provide help if necessary
- "even if the students were not prepared to present at a conference" get them to write it in the form of a conference paper.
- encouraged to release as open source software
Undergraduate conferences [slide]
- annual Argonne Symposium for Undergraduates in Science, Engineering and Math
- Grace Hopper Celebration of Women in Computing
- Chicago Area Undergraduate Research Symposium (CAURS)
- DePaul science and math day
Other ways to spread results:
- encourage them to continue the project at home institutes as honours theses and senior projects
- nominate them for the CRA Outstanding Undergraduate Research Awards
Benefits [slide]
- working with ugs can rejuvenate one's research ideas
- provide gs the opportunity to mentor ug students
- 'home-grown' PhD students for your research lab
- establish an undergraduate research culture at your own institution
- for students it's great for portfolio
```Sheila - Data on the long-term results of the program
- has had modest funding since 1991
- instigated by a group of African-American students concerned about the pipeline
- diversity is the focus
- only allowed to attend once
- commissioned an external review
- "had a chance to see longer range outcomes"
- "interested in what the ultimate impact might be"
- First REU alumnus appointed to Berkeley
- 9 or 10 PhDs out of the REUs
Copies of the report were available, will probably be if you contact her. Some highlights:
- 25 respondents out of 30 randomly selected
- 60% African American or Hispanic
- 30% women
- 12 students had parents with high school education, two with only grade school
- only 2 from Berkeley
- 70% had no research experience (this is changing because there are so many REU programs now)
- 90% "felt they had learned the context", how to fit their tiny project into larger research goal and how benefit society
- 2/3 able to connect research to coursework
- 3/4 talking to faculty more
- 40% continue connected research
- "REU projects greatly increase students' confidence about their ability to do research"
- 100% said their confidence had been increased in various measures
- aspirations heightened
- In response to an open-ended question about the lasting impacts
- 22/25 said increased confidence, learned how to conduct research and enjoyed it
- 75% still in touch with faculty, grad studnet
- faculty surveyed indicated
- "satisfaction with the program"
- saw program as "serving a broader range of students ".
- "helping these students to achieve in various ways"
- " they would continue this program whether or not there was REU funding"
"Increasing diversity through research" - things like REUs are a way to increase diversity that is "not at all controversial" and do not face the resistance of some other diversity programs
- the program and sustainability of the program has been improved through th
- Intel funds many schools for undergraduate programs
- research can be more sustained during a year with local students
Questions
Q: What about undergraduates who aren't actually CS majors, that might be interested in interdisciplinary work? A: Jan - No problem, she uses art students, for example, for the graphics projects. Daniela - has some biology, psychology students as well
Q: The "Clare Boothe Luce Program" has begun sponsoring undergraduate research programs and can be another source of funding.
Q: Undergraduates can be apprehensive about REUs. How should they be promoted? A: Teresa suggests Lane comment Lane - "you get more involved with your faculty, and the other students as well---you're going to feel you belong." Tell them that even a student without much of a background can do it, there's lots of help. Get success stories out there Lisa [another student] - found out about it from a poster . . . now has students ask her about it. The advisors had put the posters up and convinced her to apply Jan - sounds like students and then get faculty are the best pushers Virginia Tech - created Facebook groups for the students that participated. The pictures and videos of the fun stuff there are
Q: Logistics of bringing in outside students: insurance? Housing? A: Shiela - you have to worry about it, but it's not that bad. Intermediate staff can be helpful Jan - graduate programs can get involved once they're convinced it's hepful
Q: Undergraduate wonders what are the requirements and application process A: Shiela - There are REU files in the sponsor's areas at this conference Online application
- suggest student 3.0 GPA but not necessarily everywhere
- motivation for grad school and potential for research
- letters of recommendation
Jan - NSF website has the info for the NSF REUs Teresa - STARS alliance table has one Audience - deadlines vary widely. Government labs tend to be earlier, SITCIE Spec Int Group in website has intro
Q: How long is the research program? Jan - varied, depends on location. Have recruited anywhere from 8-10 weeks, local students can be year-round