Experimental

Using Natural Language Processing to Inform Science Instruction (Collaborative Research: Linn)

This project takes advantage of language to help students form their own ideas and pursue deeper understanding in the science classroom. The project will conduct a comprehensive research program to develop and test technology that will empower students to use their ideas as a starting point for deepening science understanding. Researchers will use a technology that detects student ideas that go beyond a student's general knowledge level to adapt to a student's cultural and linguistic understandings of a science topic.

Partner Organization(s): 
Award Number: 
2101669
Funding Period: 
Thu, 07/01/2021 to Mon, 06/30/2025
Full Description: 

Often, middle school science classes do not benefit from participation of underrepresented students because of language and cultural barriers. This project takes advantage of language to help students form their own ideas and pursue deeper understanding in the science classroom. This work continues a partnership among the University of California, Berkeley, Educational Testing Service, and science teachers and paraprofessionals from six middle schools enrolling students from diverse racial, ethnic, and language groups whose cultural experiences may be neglected in science instruction. The partnership will conduct a comprehensive research program to develop and test technology that will empower students to use their ideas as a starting point for deepening science understanding. Researchers will use a technology that detects student ideas that go beyond a student's general knowledge level to adapt to a student's cultural and linguistic understandings of a science topic. The partnership leverages a web-based platform to implement adaptive guidance designed by teachers that feature dialog and peer interaction. Further, the platform features teacher tools that can detect when a student needs additional help and alert the teacher. Teachers using the technology will be able to track and respond to individual student ideas, especially from students who would not often participate because of language and cultural barriers.

This project develops AI-based technology to help science teachers increase their impact on student science learning. The technology is aimed to provide accurate analysis of students' initial ideas and adaptive guidance that gets each student started on reconsidering their ideas and pursuing deeper understanding. Current methods in automated scoring primarily focus on detecting incorrect responses on test questions and estimating the overall knowledge level in a student explanation. This project leverages advances in natural language processing (NLP) to identify the specific ideas in student explanations for open-ended science questions. The investigators will conduct a comprehensive research program that pairs new NLP-based AI methods for analyzing student ideas with adaptive guidance that, in combination, will empower students to use their ideas as starting points for improving science understanding. To evaluate the idea detection process, the researchers will conduct studies that investigate the accuracy and impact of idea detection in classrooms. To evaluate the guidance, the researchers will conduct comparison studies that randomly assign students to conditions to identify the most promising adaptive guidance designs for detected ideas. All materials are customizable using open platform authoring tools.

Using Natural Language Processing to Inform Science Instruction (Collaborative Research: Riordan)

This project takes advantage of language to help students form their own ideas and pursue deeper understanding in the science classroom. The project will conduct a comprehensive research program to develop and test technology that will empower students to use their ideas as a starting point for deepening science understanding. Researchers will use a technology that detects student ideas that go beyond a student's general knowledge level to adapt to a student's cultural and linguistic understandings of a science topic.

Lead Organization(s): 
Award Number: 
2101670
Funding Period: 
Thu, 07/01/2021 to Mon, 06/30/2025
Full Description: 

Often, middle school science classes do not benefit from participation of underrepresented students because of language and cultural barriers. This project takes advantage of language to help students form their own ideas and pursue deeper understanding in the science classroom. This work continues a partnership among the University of California, Berkeley, Educational Testing Service, and science teachers and paraprofessionals from six middle schools enrolling students from diverse racial, ethnic, and language groups whose cultural experiences may be neglected in science instruction. The partnership will conduct a comprehensive research program to develop and test technology that will empower students to use their ideas as a starting point for deepening science understanding. Researchers will use a technology that detects student ideas that go beyond a student's general knowledge level to adapt to a student's cultural and linguistic understandings of a science topic. The partnership leverages a web-based platform to implement adaptive guidance designed by teachers that feature dialog and peer interaction. Further, the platform features teacher tools that can detect when a student needs additional help and alert the teacher. Teachers using the technology will be able to track and respond to individual student ideas, especially from students who would not often participate because of language and cultural barriers.

This project develops AI-based technology to help science teachers increase their impact on student science learning. The technology is aimed to provide accurate analysis of students' initial ideas and adaptive guidance that gets each student started on reconsidering their ideas and pursuing deeper understanding. Current methods in automated scoring primarily focus on detecting incorrect responses on test questions and estimating the overall knowledge level in a student explanation. This project leverages advances in natural language processing (NLP) to identify the specific ideas in student explanations for open-ended science questions. The investigators will conduct a comprehensive research program that pairs new NLP-based AI methods for analyzing student ideas with adaptive guidance that, in combination, will empower students to use their ideas as starting points for improving science understanding. To evaluate the idea detection process, the researchers will conduct studies that investigate the accuracy and impact of idea detection in classrooms. To evaluate the guidance, the researchers will conduct comparison studies that randomly assign students to conditions to identify the most promising adaptive guidance designs for detected ideas. All materials are customizable using open platform authoring tools.

Building a Flexible and Comprehensive Approach to Supporting Student Development of Whole Number Understanding

The purpose of this project is to develop and conduct initial studies of a multi-grade program targeting critical early math concepts. The project is designed to address equitable access to mathematics and STEM learning for all students, including those with or at-risk for learning disabilities and underrepresented groups.

Lead Organization(s): 
Award Number: 
2101308
Funding Period: 
Thu, 07/01/2021 to Mon, 06/30/2025
Full Description: 

A critical goal for the nation is ensuring all students have a successful start in learning mathematics. While strides have been made in supporting at-risk students in mathematics, significant challenges still exist. These challenges include enabling access to and learning of advanced mathematics content, ensuring that learning gains don’t fade over time, and providing greater support to students with the most severe learning needs. One way to address these challenges is through the use of mathematics programs designed to span multiple grades. The purpose of this project is to develop and conduct initial studies of a multi-grade program targeting critical early math concepts. The project is designed to address equitable access to mathematics and STEM learning for all students, including those with or at-risk for learning disabilities and underrepresented groups.

The three aims of the project are to: (1) develop a set of 10 Bridging Lessons designed to link existing kindergarten and first grade intervention programs (2) develop a second grade intervention program that in combination with the kindergarten and first grade programs will promote a coherent sequence of whole number concepts, skills, and operations across kindergarten to second grade; and (3) conduct a pilot study of the second grade program examining initial promise to improve student mathematics achievement. To accomplish these goals multiple methods will be used including iterative design and development process and the use of a randomized control trial to study potential impact on student math learning. Study participants include approximately 220 kindergarten through second grade students from 8 schools across three districts. Study measures include teacher surveys, direct observations, and student math outcome measures. The project addresses the need for research developed intervention programs focused on advanced whole number content. The work is intended to support schools in designing and deploying math interventions to provide support to students both within and across the early elementary grades as they encounter and engage with critical mathematics content.

The Impact of COVID on American Education in 2021: Continued Evidence from the Understanding America Study

This study will build upon the team's prior research from early in the pandemic. Researchers will continue to collect data from families and aims to understand parents’ perspectives on the educational impacts of COVID-19 by leveraging a nationally representative, longitudinal study, the Understanding America Study (UAS). The study will track educational experiences during the Spring and Summer of 2021 and into the 2021-22 school year.

Award Number: 
2120194
Funding Period: 
Mon, 03/01/2021 to Mon, 02/28/2022
Full Description: 

The COVID-19 epidemic has been a tremendous disruption to the education of U.S. students and their families, and evidence suggests that this disruption has been unequally felt across households by income and race/ethnicity. While other ongoing data collection efforts focus on understanding this disruption from the perspective of students or educators, less is known about the impact of COVID-19 on children’s prek-12 educational experiences as reported by their parents, especially in STEM subjects. This study will build upon the team's prior research from early in the pandemic. Researchers will continue to collect data from families and aims to understand parents’ perspectives on the educational impacts of COVID-19 by leveraging a nationally representative, longitudinal study, the Understanding America Study (UAS). The study will track educational experiences during the spring and summer of 2021 and into the 2021-22 school year. The team will analyze outcomes overall and for key demographic groups of interest as students and teachers return to in-person instruction during 2021. This RAPID project allows critically important data to continue to be collected and contribute to continued understanding of the impacts of and responses to the pandemic by American families.

Since March of 2020, the UAS has been tracking the educational impacts of COVID-19 for a nationally representative sample of approximately 1,400 households with preK-12 children. Early results focused on quantifying the digital divide and documenting the receipt of important educational services--like free meals and special education servicesafter COVID-19 began. This project will support the continued targeted administration of UAS questions to parents about students’ learning experiences and engagement, overall and in STEM subjects, data analysis, and dissemination of results to key stakeholder groups. Findings will be reported overall and across key demographic groups including ethnicity, disability, urbanicity, and socioeconomic status. This project will also produce targeted research briefs addressing pressing policy questions aimed at supporting intervention strategies in states, districts, and schools moving forward. Widespread dissemination will take place through existing networks and in collaboration with other research projects focused on understanding the COVID-19 crisis. All cross-sectional and longitudinal UAS data files will be publicly available shortly after conclusion of administration so that other researchers can explore the correlates of, and outcomes associated with, COVID-19.

Supporting Elementary Teacher Learning for Effective School-Based Citizen Science (TL4CS)

This project will develop two forms of support for teachers: guidance embedded in citizen science project materials and teacher professional development. The overarching goal of the project is to generate knowledge about teacher learning that enables elementary school citizen science to support students' engagement with authentic science content and practices through data collection and sense making.

Lead Organization(s): 
Award Number: 
2009212
Funding Period: 
Wed, 07/01/2020 to Sun, 06/30/2024
Full Description: 

Citizen science involves individuals, who are not professional scientists, in authentic scientific research, typically in collaboration with professional scientists. When implemented well in elementary schools, citizen science projects immerse students in science content and engage them with scientific practices. These projects can also create opportunities for students to connect with their local natural surroundings, which is needed, as some research has suggested that children are becoming increasingly detached from nature. The classroom teacher plays a critical role in ensuring that school-based citizen science projects are implemented in a way that maximizes the benefits. However, these projects typically do not include substantial guidance for teachers who want to implement the projects for instructional purposes. This project will develop two forms of support for teachers: (1) guidance embedded in citizen science project materials and (2) teacher professional development. It will develop materials and professional development experiences to support teacher learning for 80 5th grade teachers impacting students in 40 diverse elementary schools.

The overarching goal of this project is to generate knowledge about teacher learning that enables elementary school citizen science to support students' engagement with authentic science content and practices through data collection and sense making. Specifically, the study is designed to address the following research questions: (1) What kinds of support foster teacher learning for enacting effective school-based citizen science? (2) How do supports for teacher learning shape the way teachers enact school-based citizen science? and (3) What is the potential of school-based citizen science for positively influencing student learning and student attitudes toward nature and science? Data collected during project implementation will include teacher surveys, student surveys and assessments, and case study protocols.


 Project Videos

2021 STEM for All Video Showcase

Title: Teacher Learning for Citizen Science

Presenter(s): Patrick Smith, Sarah Carrier, Goforth Goforth, Meredith Hayes, Jill McGowan, & Lindsey Sachs


How Deep Structural Modeling Supports Learning with Big Ideas in Biology (Collaborative Research: Capps)

This project addresses the pressing need to more effectively organize STEM (science, technology, engineering, and mathematics) teaching and learning around "big ideas" that run through science disciplines. Unfortunately, finding ways to teach big ideas effectively so they become useful as knowledge frameworks is a significant challenge. Deep structure modeling (DSM), the innovation advanced in this project, is designed to meet this challenge in the context of high school biology.

Lead Organization(s): 
Partner Organization(s): 
Award Number: 
2010223
Funding Period: 
Sat, 08/01/2020 to Wed, 07/31/2024
Full Description: 

This project addresses the pressing need to more effectively organize STEM (science, technology, engineering, and mathematics) teaching and learning around "big ideas" that run through science disciplines. This need is forcefully advanced by policy leaders including the National Research Council and the College Board. They point out that learning is more effective when students organize and link information within a consistent knowledge framework, which is what big ideas should provide. Unfortunately, finding ways to teach big ideas effectively so they become useful as knowledge frameworks is a significant challenge. Deep structure modeling (DSM), the innovation advanced in this project, is designed to meet this challenge in the context of high school biology. In DSM, students learn a big idea as the underlying, or "deep" structure of a set of examples that contain the structure, but with varying outward details. As learners begin to apprehend the deep structure (i.e., the big idea) within the examples, they use the tools and procedures of scientific modeling to express and develop it. According to theories of learning that undergird DSM, the result of this process should be a big idea that is flexible, meaningful, and easy to express, thus providing an ideal framework for making sense of new information learners encounter (i.e., learning with the big idea). To the extent that this explanation is born out in rigorous research tests and within authentic curriculum materials, it contributes important knowledge about how teaching and learning can be organized around big ideas, and not only for deep structural modeling but for other instructional approaches as well.

This project has twin research and prototype development components. Both are taking place in the context of high school biology, in nine classrooms across three districts, supporting up to 610 students. The work focuses on three design features of DSM: (1) embedding model source materials with intuitive, mechanistic ideas; (2) supporting learners to abstract those ideas as a deep structure shared by a set of sources; and (3) representing this deep structure efficiently within the model. In combination, these features support students to understand an abstract, intuitively rich, and efficient knowledge structure that they subsequently use as a framework to interpret, organize, and link disciplinary content. A series of five research studies build on one another to develop knowledge about whether and how the design features bring about these anticipated effects. Earlier studies in the sequence are small-scale classroom experiments randomly assigning students to either deep structural modeling or to parallel, non modeling controls. Measures discriminate for the anticipated effects during learning and on posttests. Later studies use qualitative methods to carefully trace the anticipated effects over time and across topics. As a group, these studies are contributing generalized knowledge of how learners can effectively abstract and represent big ideas and how these ideas can be leveraged as frameworks for learning content with understanding. Two research-tested biology curriculum prototypes are being developed as the studies evolve: a quarter-year DSM biology curriculum centered on energy; and an eighth-year DSM unit centered on natural selection.

How Deep Structural Modeling Supports Learning with Big Ideas in Biology (Collaborative Research: Shemwell)

This project addresses the pressing need to more effectively organize STEM (science, technology, engineering, and mathematics) teaching and learning around "big ideas" that run through science disciplines. Unfortunately, finding ways to teach big ideas effectively so they become useful as knowledge frameworks is a significant challenge. Deep structure modeling (DSM), the innovation advanced in this project, is designed to meet this challenge in the context of high school biology.

Lead Organization(s): 
Partner Organization(s): 
Award Number: 
2010334
Funding Period: 
Sat, 08/01/2020 to Wed, 07/31/2024
Full Description: 

This project addresses the pressing need to more effectively organize STEM (science, technology, engineering, and mathematics) teaching and learning around "big ideas" that run through science disciplines. This need is forcefully advanced by policy leaders including the National Research Council and the College Board. They point out that learning is more effective when students organize and link information within a consistent knowledge framework, which is what big ideas should provide. Unfortunately, finding ways to teach big ideas effectively so they become useful as knowledge frameworks is a significant challenge. Deep structure modeling (DSM), the innovation advanced in this project, is designed to meet this challenge in the context of high school biology. In DSM, students learn a big idea as the underlying, or "deep" structure of a set of examples that contain the structure, but with varying outward details. As learners begin to apprehend the deep structure (i.e., the big idea) within the examples, they use the tools and procedures of scientific modeling to express and develop it. According to theories of learning that undergird DSM, the result of this process should be a big idea that is flexible, meaningful, and easy to express, thus providing an ideal framework for making sense of new information learners encounter (i.e., learning with the big idea). To the extent that this explanation is born out in rigorous research tests and within authentic curriculum materials, it contributes important knowledge about how teaching and learning can be organized around big ideas, and not only for deep structural modeling but for other instructional approaches as well.

This project has twin research and prototype development components. Both are taking place in the context of high school biology, in nine classrooms across three districts, supporting up to 610 students. The work focuses on three design features of DSM: (1) embedding model source materials with intuitive, mechanistic ideas; (2) supporting learners to abstract those ideas as a deep structure shared by a set of sources; and (3) representing this deep structure efficiently within the model. In combination, these features support students to understand an abstract, intuitively rich, and efficient knowledge structure that they subsequently use as a framework to interpret, organize, and link disciplinary content. A series of five research studies build on one another to develop knowledge about whether and how the design features bring about these anticipated effects. Earlier studies in the sequence are small-scale classroom experiments randomly assigning students to either deep structural modeling or to parallel, non modeling controls. Measures discriminate for the anticipated effects during learning and on posttests. Later studies use qualitative methods to carefully trace the anticipated effects over time and across topics. As a group, these studies are contributing generalized knowledge of how learners can effectively abstract and represent big ideas and how these ideas can be leveraged as frameworks for learning content with understanding. Two research-tested biology curriculum prototypes are being developed as the studies evolve: a quarter-year DSM biology curriculum centered on energy; and an eighth-year DSM unit centered on natural selection.

Paving the Way for Fractions: Identifying Foundational Concepts in First Grade (Collaborative Research: Newcombe)

The goal of this project is to investigate the extent to which individual differences in informal fraction-related knowledge in first-grade children are associated with short- and longer-term fractions and math outcomes, and to see whether there is a causal link between level of informal fraction-related knowledge and the ability to profit from fractions instruction that directly builds on this knowledge.

Lead Organization(s): 
Partner Organization(s): 
Award Number: 
2000424
Funding Period: 
Mon, 06/01/2020 to Fri, 05/31/2024
Full Description: 

Although fractions represent a crucial topic in early childhood education, many students develop only a tenuous grasp of fraction concepts, even after several years of fraction instruction that is aligned with current standards. The goal of this project, led by a team of researchers at the University of Delaware and Temple University, is to answer important questions about the informal understandings of fractions young children have before they come to school and what their relations are to fraction learning in more formal instructional settings. Proficiency with fractions dramatically increases the likelihood of students succeeding in math, which in turn increases participation in the STEM workforce. Importantly, large individual differences in fraction understandings are apparent at the start of fractions instruction in the intermediate grades. Early fraction misunderstandings cascade into more severe math weaknesses in later grades, especially when instruction may shift abruptly from whole numbers to fractions. There is a critical need to understand the roots of individual differences that arise before formal instruction takes place. Young children possess important informal fraction understandings before they come to school, but the range of these abilities and their role in formal fraction learning and development is not well understood. The goal of this project is: a) to investigate the extent to which individual differences in informal fraction-related knowledge in first-grade children are associated with short- and longer-term fractions and math outcomes; and b) to see whether there is a causal link between level of informal fraction-related knowledge and the ability to profit from fractions instruction that directly builds on this knowledge. The findings from the project hold promise for informing early childhood educators how fractions can be incorporated in the first-grade curriculum in new and meaningful ways. Though the findings should be beneficial to all students, the project will specifically target members of groups underrepresented in STEM fields, including ethnic and racial minority and low-income students.

The project design includes both an observational study and an experimental study. The observational study will: (1) document individual differences in informal fraction-related knowledge in first grade; (2) determine concurrent relations between this informal knowledge and general cognitive and whole number competencies; and (3) examine whether informal fraction-related knowledge at the beginning of first grade uniquely predicts math outcomes at the end. The experimental study will explore the extent to which first graders' informal and formal fraction concepts can be affected by training. The researchers will test whether training on the number line, which is continuous and closely aligned with the mental representation of the magnitude of all real numbers, will help students capitalize on their informal fraction understandings of proportionality, scaling, and equal sharing as well as their experience with integers to learn key fraction concepts. Together, the synergistic studies will pinpoint the role informal fraction knowledge in learning key fraction concepts. All data will be collected in Delaware schools serving socioeconomically and ethnically diverse populations of students. Primary measures include assessments of informal fraction knowledge (proportional reasoning, spatial scaling, equal sharing), executive functioning, vocabulary, whole number knowledge, whole number/fraction number line estimation, formal fraction knowledge, and broad mathematics achievement (calculation, fluency, applied problems).

Leveraging Simulations in Preservice Preparation to Improve Mathematics Teaching for Students with Disabilities (Collaborative Research: Cohen)

This project aims to support the mathematics learning of students with disabilities through the development and use of mixed reality simulations for elementary mathematics teacher preparation. These simulations represent low-stakes opportunities for preservice teachers to practice research-based instructional strategies to support mathematics learning, and to receive feedback on their practices.

Lead Organization(s): 
Partner Organization(s): 
Award Number: 
2009939
Funding Period: 
Fri, 05/01/2020 to Tue, 04/30/2024
Full Description: 

The preparation of general education teachers to support the mathematics learning of students with disabilities is critical, as students with disabilities are overrepresented in the lower ranks of mathematics achievement. This project aims to address this need in the context of elementary mathematics teacher preparation through the development and use of mixed reality simulations. These simulations represent low-stakes opportunities for preservice teachers to practice research-based instructional strategies to support mathematics learning, and to receive feedback on their practices. Learning units that use the simulations will focus on two high leverage practices: teacher modeling of self-monitoring and reflection strategies during problem solving and using strategy instruction to teach students to support problem solving. These high-leverage teaching practices will support teachers engaging all students, including students with disabilities, in conceptually sophisticated mathematics in which students are treated as sense-makers and empowered to do mathematics in culturally meaningful ways.

The project work encompasses three primary aims. The first aim is to develop a consensus around shared definitions of high-leverage practices across the mathematics education and special education communities. To accomplish this goal, the project will convene a series of consensus-building panels with mathematics education and special education experts to develop shared definitions of the two targeted high leverage practices. This work will include engaging with current research, group discussion, and production of documents with specifications for the practices. The second aim is to develop learning units for elementary mathematics methods courses grounded in mixed reality simulation. These simulations will allow teacher candidates to enact the high leverage practices with simulated students and to receive coaching on their practice from the research team. The impact of this work will be assessed through the analysis of interviews with teacher educators implementing the units and observations and artifacts from the implementations. The third aim will be to assess the effectiveness of the simulations on teacher candidates? practices and beliefs through small-scaled randomized control trials. Teacher candidates will be randomly assigned to conditions that address the practices and make use of simulations, and a business as usual condition focused on lesson planning, student assessment, and small group discussions of the high leverage practices. The impact of the work will be assessed through the analysis of baseline and exit simulations, measures of teacher self-efficacy for teaching students with disabilities, and observations of classroom teaching in their clinical placement settings.

Leveraging Simulations in Preservice Preparation to Improve Mathematics Teaching for Students with Disabilities (Collaborative Research: Jones)

This project aims to support the mathematics learning of students with disabilities through the development and use of mixed reality simulations for elementary mathematics teacher preparation. These simulations represent low-stakes opportunities for preservice teachers to practice research-based instructional strategies to support mathematics learning, and to receive feedback on their practices.

Lead Organization(s): 
Partner Organization(s): 
Award Number: 
2010298
Funding Period: 
Fri, 05/01/2020 to Tue, 04/30/2024
Full Description: 

The preparation of general education teachers to support the mathematics learning of students with disabilities is critical, as students with disabilities are overrepresented in the lower ranks of mathematics achievement. This project aims to address this need in the context of elementary mathematics teacher preparation through the development and use of mixed reality simulations. These simulations represent low-stakes opportunities for preservice teachers to practice research-based instructional strategies to support mathematics learning, and to receive feedback on their practices. Learning units that use the simulations will focus on two high leverage practices: teacher modeling of self-monitoring and reflection strategies during problem solving and using strategy instruction to teach students to support problem solving. These high-leverage teaching practices will support teachers engaging all students, including students with disabilities, in conceptually sophisticated mathematics in which students are treated as sense-makers and empowered to do mathematics in culturally meaningful ways.

The project work encompasses three primary aims. The first aim is to develop a consensus around shared definitions of high-leverage practices across the mathematics education and special education communities. To accomplish this goal, the project will convene a series of consensus-building panels with mathematics education and special education experts to develop shared definitions of the two targeted high leverage practices. This work will include engaging with current research, group discussion, and production of documents with specifications for the practices. The second aim is to develop learning units for elementary mathematics methods courses grounded in mixed reality simulation. These simulations will allow teacher candidates to enact the high leverage practices with simulated students and to receive coaching on their practice from the research team. The impact of this work will be assessed through the analysis of interviews with teacher educators implementing the units and observations and artifacts from the implementations. The third aim will be to assess the effectiveness of the simulations on teacher candidates? practices and beliefs through small-scaled randomized control trials. Teacher candidates will be randomly assigned to conditions that address the practices and make use of simulations, and a business as usual condition focused on lesson planning, student assessment, and small group discussions of the high leverage practices. The impact of the work will be assessed through the analysis of baseline and exit simulations, measures of teacher self-efficacy for teaching students with disabilities, and observations of classroom teaching in their clinical placement settings.

Pages

Subscribe to Experimental