Transforming Teachers' Instructional Practice: The Role of Formative Assessment Data From Homework

• In Event: 54.028 - Efficacy of an Online Mathematics Homework Intervention: Rigorous Research Transforming Knowledge to Action

Abstract

Objective
This presentation shares findings from teacher interviews and observations comparing instructional practices of treatment and control teachers. This research investigated the impact of the ASSISTments online homework platform on teachers’ instructional practices.

Perspective
The study is informed by the research literature on student homework and online tools (e.g., Feng, Heffernan, & Koedinger, 2009), teacher feedback to students (e.g., Azevedo & Bernard, 1995), and teachers’ use of formative assessment to inform instruction (e.g., Bennett, 2011; Brookhart, 2007). Increasingly educators are urged to employ data-driven practices. Yet, data have been largely absent in teachers’ modal practice of reviewing mathematics homework and planning daily instruction. This study tested the use of formative assessment data from homework to examine teachers’ adaptations in practice.

Methods
Interviews and observations consisted primarily of qualitative or descriptive data. Standard analytic procedures were followed to code and analyze the qualitative data. Observations also included numeric data.

Data Sources
Seventh-grade mathematics teachers were interviewed at the end of their first year of implementing ASSISTments. In all, 33 treatment and 19 control teachers were interviewed with a standardized protocol, representing 61% of participants. Interviews were coded using line-by-line coding for themes (Charmaz, 1995).

Classroom observations were conducted at the end of the second year of implementation. A total of 23 treatment teachers were observed each twice and 15 control teachers were observed once, all with a brief post-observation interview. This sample represented 45% of participants. We used a highly structured observation tool that included both descriptive narrative and coding for dimensions of practice. Observation data were analyzed for patterns within and across the two conditions (Miles, Huberman, and Saldana, 2014).

Results
Two broad questions framed the inquiry:

1. How do teachers’ homework practices change when they use an online homework platform that includes data on students’ homework performance?

2. What other impacts on instruction occur when teachers have formative assessment data?

Treatment teachers shifted from a random or comprehensive review of homework that characterized the practice of control teachers to a purposeful selection of homework problems. While both groups of teachers spent roughly the same amount of time reviewing homework (24% of class time), treatment teachers reviewed substantially fewer problems (9.5 compared to 3.2) in more depth and focused on common errors rather than answers. Further, treatment teachers used the data reports on homework to plan and revise instruction for the day, targeting concepts and students needing reinforcement.

Significance
Findings from this study illustrate the knowledge to action concept in practice. At the teacher level, a tool providing formative assessment data on individual and whole class homework performance had the power to transform both homework practice and instructional practice more broadly. The study provides strong evidence that when teachers have a highly efficient and accessible data tool, and high quality professional development, they can make significant changes in practice informed by data.

Authors

• Janet C. Fairman, University of Maine

• Bethany Jorgensen, Cornell University