From February 2020 through June 2022, Digital Promise supported middle school English language arts teachers across the country in using Project Topeka tools and resources to teach argumentative writing. As a technology-based tool, teachers were able to continue using it as schools turned to online learning in response to the COVID-19 pandemic.
Through three successive implementation waves (winter 2020, fall semester 2020, and school year 2021–22), Digital Promise researchers conducted exploratory, mixed methods research on the nature of argumentative writing instruction; the ways in which teachers used an automated essay scoring and feedback tool and standards-aligned instructional resources in their instruction; and student growth under Project Topeka. Digital Promise also convened 16 teacher leaders in person in summer 2022 and virtually during fall semester 2022 to dive deeper into their teaching practices with and perceptions of AI as an instructional tool, based on their experiences with Project Topeka. The voices and expertise of those teacher leaders are infused throughout this research series.
Project Topeka: Word from the Wise (Practitioners) to the Wise (Policymakers, Changemakers, Entrepreneurs) on the Promise of an AI-driven Instructional Tool synthesizes the key findings across this research series including teachers’ use of Project Topeka tools and resources, the benefits and challenges to their writing instruction, and the extent to which students benefited from Project Topeka. The paper poses some provocative reflections about the implications of AI tools for writing instruction.
Teaching Partner, Grading Assistant, Substitute Teacher: Three Approaches to Teaching Middle School Writing with Artificial Intelligence draws on teachers’ interviews (n=27) to generate an initial theoretical framework of how teachers approach AI—as a teaching partner, as a grading assistant, and as a substitute teacher—to then analyze the extent to which students’ writing scores and the types of students served are associated with teachers’ approaches to leveraging AI, drawing on student scores (n=1,373 students) and teacher surveys (n=18). (Full article submitted for journal publication and is under review. Please see the EdSurge article, Teaching Partner, Grading Assistant or Substitute Teacher? for an overview.)
Automated Essay Scoring in Middle School Writing: Understanding Key Predictors of Students’ Growth and Comparing AI- and Teacher-Generated Scores and Feedback draws on writing scores of 3,233 7th- and 8th-grade students associated with 51 teachers to examine relationships between writing opportunities, aspects of instruction, and student scores. Additionally, based on an in-person convening of 16 teachers who used Topeka, we compare and contrast teachers’ scoring and feedback to the AI scoring and feedback for a random sample of 160 essays, and we identify themes that emerged from their reflection on the process and the differences.
A Descriptive Analysis of Teachers’ Writing Instruction Using Project Topeka, An Automated Essay Scoring Tool explores whether having an essay scoring tool that provides immediate feedback and scores would lead to changes in teachers’ writing instruction and students’ writing opportunities. The study draws from teacher interviews and surveys to understand how teachers typically teach writing instruction and if and how their writing instruction changed using Topeka.
Also check out these EdSurge articles highlighting the research and teachers’ experiences using Project Topeka and perspectives on instructional AI tools:
Project Topeka Resources