Bringing Cognitive Tutors to the Internet: A Website that Helps Middle-School Students Learn Math
Purpose: Students in American schools consistently perform lower in mathematics than their counterparts in many other countries. One possible way to improve math achievement is to augment in-school mathematics learning with after-school activities. However, after-school tutoring is limited and, of the many websites available for math instruction, few are free and few offer guided learning by doing. The researchers propose to develop a website for middle-school mathematics (grades 6-8) where students can practice and develop their mathematics skills with artificial-intelligent software called Cognitive Tutors. This site builds on previously developed curricula involving Cognitive Tutors that has been shown to improve the math achievement levels of high school students over more typical mathematics courses. The researchers seek to make this type of computer tutoring widely available via a free website for students, as well as support for tutors, teachers, and parents who want to provide direction to students using the site. The proposed site will provide step-by-step tutoring on conceptually rich math problems, which should be useful in a range of contexts, including after-school programs, homework, classroom exercises, and libraries.
Project Activities: In developing and organizing content for the website, the researchers will draw heavily on the existing Cognitive Tutor Middle Level Mathematics courses that were developed by Carnegie Mellon University. The website content, which will include a complete 6th- 8th-grade mathematics curriculum, reflects National Council of Teachers of Mathemtics (2000) curriculum, teaching and assessment standards. It will target and develop five content strands across the three-year sequence: numbers and operations, algebra, data analysis, geometry, and ratios and proportional reasoning. The development plan includes iterative content development, system development, and management. After initial development, the research team will complete pilot testing of the new site in after-school programs, and will monitor the feasibility of student, teacher, and parent use of the site. Measures of accomplishment will include data describing the time students spend on the site, number and type of problems completed, and declining error rates while solving problems on the site, as well as successful use of the site by teachers, as measured by the amount of use of the site’s teacher tools.
Products: The primary outcome of this project is the development of a website for mathematics support and published reports.
Purpose: The researchers propose to develop a website for middle-school mathematics (grades 6-8) where students can work with a computer-based tutoring system called Cognitive Tutors. The proposed website will provide content for a complete 6th-8th-grade mathematics curriculum. Researchers will build on the existing Cognitive Tutor Middle Level Mathematics curriculum by developing new problems that emphasize the use of multiple strategies in mathematical reasoning. Site development will also include support materials such as interactive worksheets and quizzes, and automated reporting of student progress and performance for tutors, teachers, and parents who want to provide direction to students using the site. The proposed site will provide step-by-step tutoring on conceptually rich math problems, which should be useful in a range of contexts, including after-school programs, homework, classroom exercises, and libraries.
Setting: The setting for this project is schools in the Pittsburgh area.
Population: The participants in this project are students in grades 6-8 who are enrolled in after-school programs in four middle schools. The team anticipates that approximately 200 students will participate in this development effort.
Intervention: To make the new website most useful, the researchers will extend the existing Middle School Cognitive Tutor problem set so that students are more frequently exposed to the use of multiple strategies in mathematical reasoning. By exposing students to the use of multiple strategies, students can learn that mathematics problems can be tackled in different ways and that ideas like the distributive law can be applied to many different types of mathematical problems. The new tutorial content will also help students recognize and explain the connections between the different strategies. The website will enable teachers to assemble interactive worksheets and quizzes by selecting existing problems from the site. The automated reporting feature of the program will also allow teachers and students to view data and reports on student progress and performance.
Research Design and Methods: The researchers will make the website operational with initial versions of the interactive worksheets and quizzes, and the automated reporting feature in Year 1. Tutors and problems for all three grade levels will be adapted from the existing middle school curriculum. In Year 2, the researchers will make iterative revisions of tools and reports and existing tutors, as well as develop initial multi-strategy tutors, and revise content developed in Year 1 based on analysis of usage data. In Year 3, researchers will make further iterative revisions of prior tutors and develop a second set of multi-strategy tutors. Teachers will review the tutor problems in relevant units of the Cognitive Tutor middle-school math courses, looking for opportunities to extend the problem-solving strategies or representations used. Where necessary, new tutor activities will be designed by the teachers, principal investigators, and the research associate.
Key Measures: Key measures include user satisfaction, use by students, problem completion by students, performance improvement over time, and, successful use by teachers.
Data Analytic Strategy: There will be at least two cycles of real-world use of the developed tutors in actual after-school programs. Problems observed during these sessions will be fixed before the next iteration. The Cognitive Tutor Authoring Tools (CTAT)-built tutors have built-in logging capabilities that contain a detailed record of all student actions with the tutor and the tutor responses to these actions. These log data will be analyzed.
Related IES Projects: Combining Advantages of Collaborative and Individual Learning with an Intelligent Tutoring System for Fractions (R305A120734) and Use of Machine Learning to Adaptively Select Activity Types and Enhance Student Learning with an Intelligent Tutoring System (R305A130215)
Journal article, monograph, or newsletter
Aleven, V., McLaren, B.M., and Sewall, J. (2009). Scaling-Up Programming by Demonstration for Intelligent Tutoring Systems Development: An Open-Access Website for Middle-School Mathematics Learning. IEEE Transactions on Learning Technologies, 2(2): 64–78.
Aleven, V., McLaren, B.M., Sewall, J., van Velsen, M., Popescu, O., Demi, S., Ringenberg, M., and Koedinger, K.R. (2016). Example-Tracing Tutors: Intelligent Tutor Development for Non-Programmers. International Journal of Artificial Intelligence in Education, 26(1): 224–269.