Bio

Jeremy Russell received his PhD in mathematics from Northeastern University in May 2013. He is currently a visiting assistant professor at The College of New Jersey. His research interests are mathematics education and category theory.

Project

The main objective of the project is to demonstrate the potential of using cloud-based homework assignments to diagnose and correct student deficiencies in understanding material taught in certain subjects, particularly mathematics. The goal is to create a scalable product with enough flexibility and power to make it marketable to educators and those doing education research.

The homework assignments themselves are broken down into four phases. The first phase is the evaluation phase, where information about the student’s knowledge of the algorithmic task is gathered. This will vary per individual task, but the idea is create a way for the user to streamline the evaluation. The second phase is the diagnostic phase. Once the student submits answers in phase one, the data will be analyzed, with the main objective being diagnosing what the student is doing incorrectly. Of particular interest is to collect data in hopes of predicting precisely what a student is doing incorrectly based upon the answer alone. The diagnostic phase also includes gathering additional information about the student’s ability to perform each step in the algorithm. The third stage is the correction phase. Here the student is given new questions, which are determined by the diagnostic phase. The student is also given prompts when making mistakes in an attempt to correct errors in understanding. Finally, the student enters the reevaluation phase to determine whether the student can now perform the desired task or requires further diagnosis and correction.

For any given individual task, the above is easy enough to implement. In order to make the approach feasible, one must create a suitable architecture that allows the user to control the algorithmic steps tested, collect data from the students, and analyze student behavior and thought processing with that data. This project sets out to demonstrate that this is possible.