Skip Navigation
Funding Opportunities | Search Funded Research Grants and Contracts

IES Grant

Title: My Science Tutor: Improving Science Learning through Tutorial Dialogs (MyST)
Center: NCER Year: 2013
Principal Investigator: Ward, Wayne Awardee: Boulder Language Technologies Inc.
Program: Cognition and Student Learning      [Program Details]
Award Period: 4 years (6/1/2013-05/31/2017) Award Amount: $3,499,713
Type: Efficacy and Replication Award Number: R305A130206

Co-Principal Investigators: Ronald Cole, Brandon Helding

Purpose: Many students in the United States are not proficient in science and do not demonstrate competence in challenging science content. Many schools and districts are moving toward inquiry-based science curricula that allow students the opportunity to use inquiry, experimentation, and analysis to actively construct ideas. However, even with inquiry-based science curricula, there are students who do not master science content. The current project will test the efficacy of a computer-based intelligent tutoring system for third-, fourth-, and fifth-grade students called My Science Tutor (MyST), which was developed and pilot-tested under an earlier IES grant (Improving Science Learning Through Tutorial Dialogs). Researchers will compare MyST to both human small-group tutoring and business-as-usual during this current project. Students in the business-as-usual condition will not receive any additional tutoring either from MyST or a human tutor.

Project Activities: This project involves two studies: a replication study aimed at confirming the results found in the pilot study; and an efficacy study conducted as a fully powered randomized control trial. The replication study will include students in 63 classrooms, randomly assigned at the classroom level to one of three conditions: one-on-one MyST tutoring; small-group human tutoring; or business as usual. The efficacy study will include 84 classrooms (42 in Year 2 of the project and 42 in Year 3) randomly assigned to one of two conditions: one-on-one MyST tutoring; or business-as-usual.

Products: The final product of this project will be evidence of the efficacy of MyST to impact students' science knowledge. Peer-reviewed publications will also be produced.

Structured Abstract

Setting: This study will be conducted in elementary schools from three districts in Colorado.

Sample: Approximately 2,200 students in grades three, four, and five will participate in this study.

Intervention: My Science Tutor (MyST) is a computer-based intervention designed to supplement classroom instruction using the Full Option Science System (FOSS) curriculum. FOSS is a non-textbook based science curriculum focused on allowing students the opportunity to actively construct ideas through inquiry, experimentation, and analysis. MyST uses a life-like 3D character named Marni who asks open-ended questions and responds to students' speech. Marni and the student have a shared dialog about science which allows MyST to continuously assess students' knowledge on a particular topic. MyST uses character animation, automatic speech recognition, dialog modeling, and language and speech generation. Multimedia content is also integrated into the intervention during 20-minute tutorials aligned with FOSS content which are intended to provide the scaffolding for students to think and talk about science with Marni.

Research Design and Methods: This project involves two studies. The first is a replication study of the pilot study from the original IES funded Development and Innovation grant. For this study, 63 classrooms in third, fourth, and fifth grades will be randomly assigned to one of three conditions: one-on-one MyST tutoring; small-group human tutoring; or business as usual. The second part of this project is a two-cohort efficacy study. In each of Years 2 and 3, 42 third-, fourth-, and fifth-grade classrooms will be randomly assigned to receive either one-on-one MyST tutoring or business as usual. Students will be pre-tested and then receive one academic year of tutoring or business as usual. Students in both groups will be post-tested.

Control Condition: The replication study has two control conditions: a small-group human tutoring condition and a business-as-usual control condition. The efficacy study uses a business-as-usual control condition.

Key Measures: To assess science knowledge, students will complete the Assessing Science Knowledge (ASK) assessments, which are part of the FOSS curriculum, at both pre-test and post-test. Additionally, the Measures of Academic Process (MAP) will be used as pre- and post-tests, and the Transitional Colorado Assessment Program (TCAP) tests will be used as post-test for fifth-graders (the TCAP is only administered in the spring of fifth grade). Finally, student and teacher surveys will be used to examine perceptions of tutoring and problems with the MyST program.

Data Analytic Strategy: To examine the efficacy of MyST, researchers will construct a hierarchical linear model within a structural equation modeling framework. The analyses will nest students within classrooms. Additionally, the models will account for change within grade (pre- and post-test) and across grade (third, fourth, and fifth grade).

Related IES Projects: Improving Science Learning Through Tutorial Dialogs (R305B070434)