9:41 p.m. A different kind of homework
Sophia Eristoff, a sophomore majoring in materials science and biomedical engineering, is sitting in a quiet corner of the Gates Center, finishing up her online assignment for tomorrow’s class: CS 15–110 Introduction to Computer Science. Designed for non-majors, the course is a requirement for Eristoff — one she’s glad to fulfill. She is already seeing the ways in which knowledge of Python programming will make her a better engineer.
Eristoff and her classmates find the work in 15–110 to be quite challenging, but they are making good progress while having an educational experience only possible at CMU. This semester, 15–110 is a “blended course,” combining in-class meetings with CMU’s own educational technology that is designed to improve and accelerate student learning.
Tonight, Eristoff is watching videos and working through practice problems using software that lets her know when she is making the right-and wrong-choices. The immediate feedback is a hallmark of CMU’s Open Learning Initiative (OLI), which has dozens of high-quality online courses in topics from computer science to English. OLI also tracks how students respond to each lesson and problem and then uses that data to continuously improve the course. Blended classes like 15–110 use OLI modules to help Eristoff learn more at a faster pace.
Such blended courses are a hallmark of the Simon Initiative, started at CMU in 2013 by President Subra Suresh. The Simon Initiative aims to transform higher education instruction through CMU-led advances in learning science and its applications. Three key projects support these efforts: LearnLab, DataLab, and OLI.
This initiative takes its name and inspiration from Herbert A. Simon, the Nobel Prize-winning CMU computer scientist who pioneered Artificial Intelligence and was fascinated by the capacity of computation to support the full range of human cognition.
Through the Simon Initiative, Carnegie Mellon is re-engineering courses so that the educational experience for its own students — and students around the world — is more effective, efficient, and personalized.
For Eristoff, the most important thing about the course is that, by the time she gets to class tomorrow, she already will have worked with the material being discussed.
“Normally lectures go way too fast, but with this class, I’m seeing stuff for the second time. Sometimes, when I’m confused in class, I’ll go back to the modules to remind myself of what I figured out the night before.”
- Sophia Eristoff, sophomore materials science and biomedical engineering major
2:37 a.m. Servers never sleep
As she worked her way through the online modules earlier in the night, Sophia Eristoff’s responses went to a DataLab server, where they were recorded on the 15–110 course dashboard. This data provided Eristoff and her instructor with immediate feedback on the progress of the class.
While most of the campus sleeps, a collection of servers in the Andrew machine room in Cyert Hall processes data from online tutors every second, creating and growing existing datasets based on student actions. This room is the central location of DataLab, the world’s largest open and free repository of learning data, which is available for learning scientists all over the world to use in their research. This data has been collected by CMU since 2005, reflecting more than 200 million observations of learning activities from nearly a quarter of a million students, including Eristoff.
“My CMU faculty colleagues are scientists; they trust data. And the data shows that there is a better way for us all to teach our students than through lectures alone.”
- Ken Koedinger, professor of human-computer interaction and psychology
“It’s all about the data,” says Ken Koedinger, professor of human-computer interaction and psychology, co-coordinator of the Simon Initiative, and director of LearnLab, the scientific arm of the Simon Initiative. As one of the world’s leading learning scientists, Koedinger has shown that well-designed technology can dramatically improve student learning outcomes.
Rather than recycling a textbook or broadcasting a lecture, Koedinger and his colleagues take full advantage of the software’s capacities to actively engage students in learning the material, collect data on their performance, and provide feedback every step of the way.
This approach works: studies have shown that in Carnegie Mellon’s OLI courses, students can master twice as much course material in half the class time, and there is evidence that they retain more material.
“One of the axioms of learning science is that what you think you know about how students learn is usually wrong,” Koedinger says. “The data tells you the truth.” Koedinger cites one of his own misconceptions: like most math teachers, he had firmly believed that students found solving algebra word problems harder than working with the equations; the data showed that exactly the opposite is true.
In 15–110, every keystroke, mouse click, and answer from Eristoff and her classmates is fed back in many directions — to the student, the instructor, and the course designer, and stored in DataLab for learning researchers. This is the Simon Initiative at work.
10:30 a.m. Class time
Before today’s meeting of 15–110, Eristoff’s instructor, Dilsun Kaynar, assistant teaching professor of computer science, reviews the course dashboard, which displays student results from the homework. She can drill down to see where each student is struggling, or view the class as a whole. She learns that although her students are doing well with conditional statements and with loops, they are struggling with more sophisticated exercises that combine the two concepts. She knows that students will benefit from some in-class activities that apply these concepts in a step-by-step way.
Today, students also will have the chance to do practice problems together in class. For Eristoff, this opportunity to see her classmates’ work is particularly useful. Although she is a solid programmer, she often has extra lines of code that could be eliminated if she’d thought to use a different function.
11:22 a.m. The students’ experience
Students recognize the difference that this hybrid instruction brings them, and they want more of it. Gifts to the Simon Initiative Innovation Fund will support more technology-enhanced learning opportunities at CMU and beyond.
“[OLI] allows me to take more responsibility for my learning … it really de-stresses the whole situation because you have this other opportunity to learn that’s then incorporated into the classroom.”
- Lilah Buchanan, senior psychology major
Two students who completed CS 15–110 discuss their course experience with Norman Bier, executive director of the Simon Initiative and director of OLI.
1:07 p.m. The professor’s experience
After class, Dilsun Kaynar meets with her colleague, Marsha Lovett, director of CMU’s Eberly Center for Teaching Excellence and Educational Innovation. Lovett is also a co-coordinator of the Simon Initiative and a teaching professor of psychology. She helps CMU faculty and grad students to be better teachers and use evidence-based methods and technologies in the classroom.
Over the past academic year, the Eberly Center reached nearly one-third of all CMU instructors through its programs and one-on-one services, including 435 consultations on teaching and learning across all seven schools and colleges.
3:53 p.m. From CMU to the world
The Simon Initiative leadership team meets to discuss ongoing projects and plans — working on specific new courses, training for colleagues, or finishing projects for the Global Learning Council, a group of university, industry, and philanthropic leaders chaired by CMU President Subra Suresh that focuses on learning technologies that achieve improved educational outcomes.
The Simon Initiative is currently focused on three overarching goals:
1. Build a learning engineering ecosystem at Carnegie Mellon to support faculty to improve student learning outcomes. This approach is currently being applied in three areas of fundamental importance: writing and communication; statistical reasoning; and computing.
2. Create an open source software “backbone” to link the state-of-the-art tools developed at CMU (such as OLI) to each other and to tools developed elsewhere so that learning engineers anywhere in the world can have a toolbox that has been demonstrated to improve learning outcomes.
3. Adapt the world-leading methods and tools developed at CMU and make them widely and freely available so that any instructor, anywhere in the world, can create his or her own modules and courses customized for any group of students.
Richard Scheines, dean of the Dietrich College of Humanities and Social Sciences, is faculty lead of the Simon Initiative and an enthusiastic proponent of designing technology-enhanced learning courses across the university.
“We’re creating a one-stop shop for professors from any discipline to come in, learn what’s available, and get help implementing these tools in their courses; they can collect the data, analyze the data, and improve the system as it goes forward.”
- Richard Scheines, dean of the Dietrich College of Humanities and Social Sciences; faculty lead, The Simon Initiative
4:35 p.m. Coffee in Gates
Sophia Eristoff puts in almost two hours nightly on homework for 15–110, a time commitment she considers reasonable. Next semester she will move beyond the intro courses of her first year at CMU and take materials sciences courses and a biomedical engineering lab. She’s on track to graduate a semester early and hopes, upon completion of her bachelor’s degree, to immediately start a graduate program in engineering and then a research career.
Thanks to 15–110 and the power of technology-enhanced learning promoted by the Simon Initiative, Eristoff has a strong foundation in programming, learning that was supported with the most effective technology around.
Help more students experience the power of effective learning technologies by supporting the work of the Simon Initiative.