Cornell U researchers preparing math program to 'reverse engineer' students' work
- Computing and information sciences researchers from Cornell University are preparing to present research on software that could help grade math assignments in K-8 at the Conference on Human Factors in Computing Systems, eSchool News reports.
- The research behind the software used a 300-student sample from Metametrics to examine what students were doing right or wrong on addition and subtraction problems, with a goal of reverse engineering students' thought processes to identify what their intent might have been at various stages of coming to a solution.
- The researchers hope the software can serve as a grading solution that provides reports on teaching outcomes for classrooms and areas that may require more teacher attention, though it will initially only work for simple math until expanding to algebra and advanced equations down the road.
While a computer program can effectively determine whether the solution a student arrived at on a math problem — or any other question with a definitive correct answer — is correct, inferring what their thought process was on the way to that solution is more complicated. But it's also more helpful when it comes to assisting students who might be struggling with a concept. This is why, on paper-and-pencil assignments, math teachers in particular have long requested that students "show their work" rather than just writing a solution.
In the digital age, where students are more likely to input their answer directly into the learning platform being used, there may not be space to show that work, or doing so may not come across as effectively as seeing it on paper. While educators can walk through those steps with students one-on-one to figure out what they were thinking, having some idea of that beforehand is likely much more beneficial to both parties. If a program can take data from hundreds of students' paths to a solution on a variety of problems and deduce how others likely went about solving similar problems, it could potentially save educators plenty of time and enable them to offer much more effective feedback.
Solutions that can save time for educators have been an increasing priority for administrators in schools where flipped, blended and project-based learning models have been implemented. While these approaches can save instructional time and allow educators to have more one-on-one opportunities with students, they can also require more preparatory and grading time on the part of educators and contribute to increased likelihood of burnout. If the idea is to ultimately save teachers time while maximizing their efficiency, any program that can potentially lend a hand in routine processes like grading is likely a welcome addition in the classroom.
Follow Roger Riddell on Twitter