GK12-Graduate-Student-Application
From TeacherWiki
Professor Heffernan plans awarding the 2011-12 fellowships before April 16th, so if interested, learn more and apply NOW!
Contents |
General Overview
Graduate Research Assistantships are available for students interested in working on Intelligent Tutoring Systems Research and examining how they are used in real school settings.
This National Science Foundation (NSF) GK12 grant funds five Computer Science graduate students for up to two years over the five year life of this grant with stipends of $30,000 per year and a full tuition waiver. Although we would like all students to apply, the federal government requires all applicants to be US citizens or permanent residents. If you do not meet this requirement, Professor Heffernan has other money available for a small number of non-US citizens and foreign nationals. Unfortunately he is not allowed to pay a lower stipend amount for students funded through the NSF.
Research will be done using ASSISTments. The research goal of the ASSISTment Project is to provide a way to give web-based intelligent tutoring to middle and high school students in mathematics and science and is a joint collaboration between Worcester Polytechnic Institute (WPI) and Carnegie Mellon University (CMU). ASSISTments is currently being used by over 7,000 7th and 8th graders in the Worcester Public Schools as well as schools near Worcester and Pittsburgh. WPI and CMU provide this service free while conducting interesting research on what makes for effective intelligent tutoring. For more information see http://www.wpi.edu/research/discovery/ and on the left scroll down to Intelligent Tutoring Systems that Teach and Assess.
See also A Day in the Life of a GK12 Fellow and TV News Spot
Types of Research Done
The types of research we do includes:
- Educational data mining: We examine millions of student's actions for patterns. Educational data mining is to education what bio-informatics is to biology. One of Dr. Heffernan's students, Zack Pardos and he won a $3,000 dollar prize (out of 600 entries) at the ACM Knowledge Discovery and Data-mining CUP competition. Among some of the reasons we won is we created the the first Bayesian Model Tracing method that allow them to model differences in student learning rates. Take a look a the papers at Zach Pardos web page to see series of interesting results that have been pushing the state of the art in tracking student performance and learning. One of Dr. Heffernan's recent PhD students, Mingyu Feng, won Best Paper of the Year at the User Modeling Journal, had her work cited in a national publication and talked about in detail the the US Dept of Educational National Educational Technology Plan. See here for the paper and details. In summary, Dr. Heffernan's students are getting national recognition for their work and interesting jobs when they are done (Mingyu is now a Research Scientist are SRI International.)
- Creating innovative education applications: A second aspect of Dr. Heffernan's lab's work, is developing the most widely used free platform for education of its kind. Usage has been doubling every year, and over 10, million problems have been solved by students. Called ASSISTments, it is used by dozen of school districts as part of their normal math curricula. There are many uses of ASSISTments. Some students are doing nightly homework with ASSISTments and teachers use it classrooms. Most interesting, graduate students have added new features like a parent accounts. Zach Broderick, for instance was the masters student who created the parent notification system and he is now done two randomized controlled trials where some students were assigned to have their parents to get accounts and others were in the control group. He found that parents feel better connected to their student math classroom, and increasing in homework completion rate. Zach alredy has the first study published (http://teacherwiki.assistment.org/wiki/images/b/b8/Parent.notification.pdf) and a second one in submission.
There are lots of other innovations that have been created for classrooms like the Automatic Reassessment and Retraining System or the Data Driven Instruction feature, or our machine learned method of detecting when student are off task. The ASSISTments lab is currently re-writing a large amount of our code using the Google Web Tool kit. So if you want experience doing software engineering on a project that will effect lots of students and teachers, come join Dr. Heffernan's lab.
- Basic Cognitive Science research: Dr. Heffernan and his students have run over 22 randomized controlled studies to determine what is best way to help student learn. His former PhD student, Leena Razzaq’s work has shown that students’ learning can be maximized by first recognizing their level of background knowledge; higher performing students learn more from explanations while students with a weaker math background learn more from our intensive tutoring techniques. IOS Press. 2007. [1]
- Scaling up intelligent tutoring systems to work for millions of students and teachers. They are trying to create something like Wikipedia but for test questions. More info at The_ASSISTments_Platform.
Requirements
NSF requires graduate student fellows to work 15 hours a week helping the school, 10 of which must be at a school. Sometimes you will share your experiences of being a scientist with middle and high school students and inspire them to pursue careers in math and science. Sometimes you will be working with students in the computer lab and then designing new activities to run in the ASSISTment System. Sometimes you will be testing new features that you built for the ASSISTment System with your middle schools students and teachers. The goal is to have a tight relationship between your research goals and your classroom activities. The five hours a week that is not in school, will be spent preparing for your classroom activity. A typical research ASSISTantship (RA) will, with the advice of one of the faculty, come up with new experiments or features that you think will be of research interest, and then implement new ideas in code.
Affiliated Faculty
Affiliated Faculty at Worcester Polytechnic Institute:
- Neil Heffernan: Intelligent Tutoring Systems (Director of PIMSE)
- Janice Gobert: Cognitive Psychology (coPI)
- Joseph Beck: Educational Data Mining
- George Heinemann: Software Engineering
- Ryung Kim: Statistics (Math Department)
- Rob Lindeman: Human Computer Interaction
- Murali Mani: Databases
- Gary Pollice: Software Engineering
- Charles Rich: Intelligent user interfaces
- Carolina Ruiz: Data Mining
- Elke Rundensteiner: Databases
- John Goulet: Mathematics Education (Math Department). Directs the | of Mathematics for Educators program
Faculty at Carnegie Mellon University associated with the ASSISTment Project:
- Ken Koedinger (HCII)
- Brian Junker (Stats)
We have multiple graduate courses in this area, so there is a lot to learn in the classroom as well as in your research project. You can also take classes in the Graduate program in Learning Science and Technologies.
To Apply
- Apply to the computer science department's graduate program WPI-CS or to the new program in Learning Sciences & Technology.
- Send a short description about why this RA is a good match for your background and interests. To make the application process easier, include copies of whatever materials you sent (or will send) to apply for a graduate degree in Computer Science at WPI. Send e-mail (unofficial) copies to Professor Heffernan, Computer Science Department, Worcester Polytechnic Institute, 100 Institute Road, Worcester, MA 01609. All email should be directed to nth@wpi.edu and to ch@wpi.edu.
We use a rolling admission program. In the interest of time, if your materials look strong, we might be able to tell you if you will get this RA, pending your formal acceptance into a WPI graduate program.
Principal Investigator
Neil Heffernan
