Purpose: The purpose of this project is to develop an intelligent tutoring system intended to promote students' understanding of fractions, an area that many students find difficult to master. The HALF (Helping At-risk and Low-achieving students in Fractions) system will present learning problems in conjunction with virtual manipulatives and videos designed to anchor to-be-learned concepts within already-familiar topics. After students attempt to solve these new problems, the system will diagnose gaps in student understanding and provide feedback to teachers intended to support individualized instruction and practice in basic concepts of fractions.

Project Activities: During the initial years of the project, the research team will follow an iterative design process to develop, test, and redesign the HALF system. In addition, the team will develop a webinar presentation for teachers to support their use of the software package. In the last year of the project, the team will conduct a between-subjects experiment to test the effectiveness of the new intervention with a commercially available software program and real (i.e., not virtual) manipulatives.

Products: At the end of this project, the research team will have developed two main products: (1) an intelligent tutoring system, HALF, designed to provide instruction to students in fractions; and (2) a professional development webinar intended to increase teachers' familiarity with and understanding of the software package. Articles describing the research outcomes of the development and testing process will also be prepared.

Structured Abstract

Setting: The participating schools will be selected from the Warren County Public Schools District in Bowling Green, Kentucky and the Williamson County Schools District in Franklin, Tennessee.

Population: The sample includes over 300 students in Grades 5–6 who are at-risk and low-achieving. In addition, approximately 10 teachers will participate in the project.

Intervention: The HALF system will integrate anchored instruction, adaptive computer-based instruction, teacher-led instruction based on student performance data, and small group problem solving. The anchored instruction will be a five-minute video—viewed by the whole class—where a realistic problem-solving situation is presented. This video provides a meaningful context for developing and applying skills in the use of fractions to solve authentic problems. The HALF computer-based adaptive instructional system will use instructional principles from both direct instruction and deliberate practice, and will follow a specific set of instructional procedures designed to develop competent levels of performance for all learners. As students interact with the system, it will collect and analyze performance data. The system will provide this feedback to teachers, who will then be able to provide individualized and small-group instruction.

Research Design and Method: The three-year project involves both the development and testing of the intervention. In Year 1, researchers will employ iterative design, where they will utilize fidelity checklists and student assessment data during the process to determine if components of the intervention are operating as intended. In Year 2, researchers will focus on re-designing the intervention using analyses of qualitative data collected from field observations, interviews with students and teachers, and teacher journals. In Year 3, the research team will conduct a between-subjects pilot study comparing students and teachers using the HALF intervention with students and teachers in the control condition. Random assignment is at the classroom level. This pilot evaluation will help determine the promise of the intervention and the feasibility of implementing HALF with fidelity in math classrooms.

Control Condition: The control condition will consist of four teachers and their classrooms. The teachers will use a commercially available fractions intervention that fully integrates physical manipulatives into instruction.

Key Measures: Student assessments on math performance will be drawn from the large-scale evaluation of the Rational Number Project curriculum. Topics covered in this assessment include basic fraction concepts, fraction equivalence, fraction order, addition and subtraction, estimation, and number line. Additionally, the team will gather data using fidelity checklists, computer log files, student and teacher interviews, audio and video recordings, and observations.

Data Analytic Strategy: Quantitative data from the pilot study will be analyzed using descriptive statistics and a mixed regression model, including calculating effect sizes for the treatment and control groups. Qualitative data will be coded by two researchers, allowing for inter-rater reliability, which will allow for categories to emerge from the data.

Products and Publications

Journal article, monograph, or newsletter

Mendiburo, M., and Hasselbring, T. (2014). Technology's Impact on Fraction Learning: An Experimental Comparison of Virtual and Physical Manipulatives. Journal of Computers in Mathematics and Science Teaching, 33(2): 209–231.

Mendiburo, M., Hasselbring, T., and Biswas, G. (2014). Teaching Fractions With Technology: What Type of Support do Students Need as They Learn to Build and Interpret Visual Models of Fractions Ordering Problems?. Journal of Cognitive Education and Psychology, 13(1): 76–87.

Proceeding

Mendiburo, M., and Biswas, G. (2011). Virtual Manipulatives in a Computer-Based Learning Environment: How Experimental Data Informs the Design of Future Systems. In Proceedings of the Artificial Intelligence in Education: 15th International Conference (pp. 510–512). Berlin: Springer-Verlag.

Mendiburo, M., Biswas, G., and Hasselbring, T. (2012). Design-Based Research in Practice: A Technology-Based Classroom Experiment That Explores how Students use Virtual Manipulatives to Order Groups of Fractions. In Proceedings of the 10th International Conference of the Learning Sciences (ICLS 2012): The Future of Learning, Volume 2, Short Papers, Symposia, and Abstracts (pp. 573–574). Sydney, Australia: International Society of the Learning Sciences.

Mendiburo, M., Sulcer, B., and Hasselbring, T. (2014). Interaction Design for Improved Analytics. In LAK '14 Proceedings of the Fourth International Conference on Learning Analytics and Knowledge (pp. 78–82). Indianapolis, IN: ACM.

Mendiburo, M., Sulcer, B., Biswas, G., and Hasselbring, T. (2012). Interactive Virtual Representations, Fractions, and Formative Feedback. In Proceedings of the 11th International Conference, ITS 2012 (pp. 716–717). Chania, Crete, Greece: Springer-Verlag Berlin Heidelberg.