Neil Heffernan has long worked to help students receive the benefits of individualized instruction while reducing the time it takes to deliver it. That’s part of what inspired him and his wife, Cristina, to found ASSISTments in 2003, a free math learning program that provides students with instant feedback on problems and gives educators data about student performance.
“As a platform provider, my job is to help teachers fool their kids into thinking they can pay more attention to them than they really can,” says Heffernan, a computer science professor at Worcester Polytechnic Institute. “No teacher, be it an adjunct professor, or a person like myself, has enough time to look at everything.”
Heffernan is famous in the educational world for his efforts in this regard. He’s been featured in The New York Times Magazine and has received multiple grants to develop and improve educational software. As of late, Heffernan, a one-time middle school math teacher, has turned to crowdsourcing to pursue his goal of providing widespread computer-assisted individualized feedback. Crowdsourcing has influenced many aspects of society, including food reviews and encyclopedia writing, but it has yet to be fully utilized for education. “I believe that crowdsourcing now has the potential to transform education,” Heffernan says.
He’s currently spearheading work on educational crowdsourcing with his team at WPI and recently published a paper at the ACM Learning @ Scale conference that examines how crowdsourcing has an impact on student achievement.
“It has been shown in multiple studies that expert-created on-demand assistance, such as hint messages, improves student learning in online learning environments,” the paper notes. “In the 2017-18 academic year, 132,738 distinct problems were assigned inside ASSISTments, but only 38,194 of those problems had on-demand assistance.”
Solving simple problems
To improve the problem-assistance ratio dramatically, Heffernan’s team launched a tool that allowed teachers to create on-demand assistance for a problem as they assigned it. These “assists” consist of helpful hints as well as short video tutorials, and are designed to mimic the kind of assistance a one-on-one math tutor might provide. Heffernan’s team then encouraged a select group of vetted teachers to make their assists available to other teachers who use ASSISTments.
“We found that teachers inside ASSISTments had created 40,292 new instances of assistance for 25,957 different problems in three years,” the paper notes. “We also conducted two large-scale randomized controlled experiments to investigate how on-demand assistance created by one teacher affected students outside of their classes. Students who received on-demand assistance for one problem resulted in significant statistical improvement on the next problem performance. The students’ improvement in this experiment confirmed our hypothesis that crowd-sourced on-demand assistance was sufficient in quality to improve student learning, allowing us to take on-demand assistance to scale.”
This paper won a “Best Paper” award at the conference. Heffernan believes it has major implications for education. “The textbook publishers don’t want to hear this story, that this random motley crew can succeed,” he says.
Though ASSISTments focuses on math, Heffernan believes the same principles that make that program can be applied to other subjects, such as physics, and even eventually to areas such as grammar. He’s also exploring programming written feedback for students that would work similarly to Google’s Smart Compose, which offers suggestions as you type. In Heffernan’s system, teachers would input a grade and would be offered a suggested comment based on the grade range. It’s all part of his effort to save faculty time spent on small tasks.
“The job of any faculty member is so insanely difficult I want to make technologies that help them triage in simple cases,” Heffernan says. “We want to help the teachers in the easy cases so they can spend more time on the hard cases.”
