Faculty Advisor or Committee Member

Joseph E. Beck, Reader

Faculty Advisor or Committee Member

Ivon Arroyo, Advisor




Common goals of Educational Data Mining are to model both student knowledge as well as student affect. While research continues along these lines of gathering data and building models of students' changing knowledge and affect states, little is being done to transform this collected (raw) data into meaningful entities that are more relatable to teachers, parents and other stakeholders, i.e. people who are not researchers. This research has entailed the iterative design and development of Teacher Tools, created with input from teachers and other experts. Teacher Tools is a web application designed as part of the MathSpring.org Intelligent Tutoring system --the component that teachers interact with, to set up classes as well as analyze resulting data from their students. In our study, we redesigned the existing version of MathSpring's Teacher Tools in three iterations, based on feedback gathered during each of those phases. The feedback captured from the first iteration clearly suggested for multiple design level changes with respect to math content organization, the interface, and the complexity level of the existing performance reports. Responses to Prototype I during the second iteration, designed on the basis of responses from the first iteration, were met by teachers with mixed to positive responses regarding usability and understandability. Experts at this point suggested further areas of improvement from a usability standpoint, which resulted in Prototype II of the Teacher Tools. Prototype II was then subjected to a third and final improvement iteration; this one was well received by a new set of 10 math teachers and other experts, who thought that Prototype II was very useful to them, in general. Teachers were able to appreciate the use they could give to these Teacher Tools to understand their students better, as well as guide future action plans that would alter their teaching based on information about their students' behavior, performance and affect of their students. In summary, we have created a software product for teachers that supplements the MathSpring tutoring system, which summarizes rich information from data logs into visualizations and other representations. These Teacher Tools have proved useful to teachers in Middle Schools in Massachusetts, who claim they are ready to use this information to change their teaching plans.


Worcester Polytechnic Institute

Degree Name



Computer Science

Project Type


Date Accepted



Restricted-WPI community only


Mathematics Education, Emotion Charts, Mastery, Teacher Tools, Learning Technologies, Teacher Dashboards