Can student learning improve using teacher crowdsourcing in online learning platforms such as ASSISTments?

As part of the Teachers for Research and Feedback program, ASSISTments teachers are matched with researchers to improve math education (and the ASSISTments platform) for many other teachers and students. Learn about one such research project below. If you're interested in becoming a Teacher for Research and Feedback, or to hear about upcoming opportunities, learn more here.

This project was covered in The Hechinger Report, in an article entitled PROOF POINTS: A crowdsourcing approach to homework help. Read more here.

RESEARCHER: Thanaporn (March) Patikorn, Worcester Polytechnic Institute

QUESTION WE WANTED TO ANSWER: 

Can student learning improve using mathematical hints and complete explanations crowdsourced from teachers in online learning platforms such as ASSISTments? 

HOW DID WE TRY TO ANSWER THIS:

We had teachers write thousands of hints (a series of suggestions to help solve the problem, that do not give the answer) and complete explanations (full explanation of how to solve the problem, including the answer) for thousands of problems embedded in ASSISTments.

We then randomly delivered hints and explanations to students and recorded their ability to correctly answer not only the given problem, but subsequent problems on the same topic. 

We call this feature TeacherASSIST in ASSISTments. 

WHAT DID WE FIND:

The randomized control trial showed that if a student had just been randomly assigned to receive a TeacherASSIST message (a hint or explanation as described earlier), they were reliably less likely to ask for help again. In other words, students who receive these hints/explanations are more likely to persevere and solve the next problems without needing any additional assistance. To confirm the findings, we ran the same experiment again in the following academic term and found the same results with slightly larger effects. In response to the findings, Dr. Neil Heffernan, ASSISTments Founder and Computer Science Professor at Worcester Polytechnic Institute, said, “This shows that kids felt more inclined to make attempts on problems after they’ve been given assistance. It appears that students’ propensity and willingness to attempt to solve new problems is a leading indicator of success.”

Dr. Heffernan went on to say, “We already had proof that ASSISTments, even without these messages, drastically increases student learning. This result just adds to the main effect, making ASSISTments a more proven and effective tool.”

HOW CAN THIS HELP TEACHERS:

This study gives teachers further evidence that the use of ASSISTments and in particular, ASSISTments with the use of the TeacherASSIST feature, can drastically improve student math learning. Teachers also need to know that they can use ASSISTments to create their own hints and explanations. These are then delivered to their students, as needed, while students are working on their math problems.  Especially in these times of distance or remote learning, it is really important for students to get immediate feedback, and support.

ASSISTments teacher and Ambassador, Andrew Burnett writes about the impact that TeacherASSIST has had in his class on his blog. He says, “Earlier this year at the open house/back to school night I was explaining ASSISTments, immediate feedback, and the explanation videos to parents.  When I was done a parent raised his hand and said, ‘Well this explains the conversation I had last night with my son.  When he had finished his homework I asked him how it went.  My son said ‘there were a couple of problems that I had trouble with but Mr. Burnett helped me.’ To which I replied ‘did you talk on the phone with Mr. Burnett or was he here? How did he help you.'”  Bingo!!  This is exactly what I was hoping would happen.  I helped my student when he needed it the most.”

Thea Durling supported the TeacherASSIST project for two years. In her blog on the professional development benefits of doing research with ASSISTments, Thea says that her participation gave her “(1) a clearer sense of the scope of [my curriculum and (2) forced me to slow down and put words to the why of what I was doing, so that I could teach it to my students.” 

LINK TO MORE INFORMATION ON THE STUDY AND FINDINGS

The above is a synopsis of the research and its conclusions. If you want to see a video that details the full study, click here.  

The full paper is here Thanaporn, P. & Heffernan, N. T. (2020) Effectiveness of Crowd-Sourcing On-Demand Tutoring from Teachers in Online Learning Platforms. Learning at Scale 2020.

Read more from Thea Durling, an ASSISTments teacher who participated in this project, on her professional development experience here.

See press coverage on this research in The Hechinger Report here.