Nearing the end of a five-year study funded by the National Science Foundation, Professors Jamal Abedi and Paul Heckman have learned much about the status of formative assessment in mathematics and are looking forward to the launch of a website with tools for teachers and administrators early next year.
The work is important because assessing student knowledge during instruction—the heart of formative assessment—empowers teachers to address possible deficiencies in student understanding and increase student learning.
In September 2010, Abedi and co-investigator Paul Heckman, associate dean of the UC Davis School of Education, embarked on an ambitious study to assess the state of formative assessment in mathematics in California. The focus on math stems from concern about persistently low statewide achievement in math compared with the rest of the nation, particularly for underrepresented groups of students.
In the first phase of the study, the researchers found a wide range of definitions of formative assessment. “Different people have different understandings. In fact, among assessment experts, teachers, administrators, and testing companies there were different views,” said Abedi.
For example, he explained that teachers might consider homework or quizzes to be formative. Principals might classify information they receive from parents and pass on to teachers as fitting the definition. Test publishers tend to view off-the-shelf standardized assessments aligned with textbooks as the gold standard of formative assessment.
“Measurement experts think any kind of information— including tests, email communications, and meetings with parents—can be put together by teachers to be used formatively,” said Abedi.
To understand the scope of educators’ understanding and to hone in on best practices, Abedi and Heckman, working with the project coordinator, Kimberly Mundhenk and a cadre of graduate student researchers, first sent surveys to all principals in California, about 9,000 in all. They received about 750 responses.
The team then spoke with about 500 district assessment officers involved in curriculum and assessment at districts with large populations of English learners and high levels of poverty. The team asked the districts to identify at least three schools they believed were engaged in good formative assessment practices. From this group, they collected “substantial data” (e.g., sample formative assessments, principal and teacher interviews, and state assessment results) from 122 schools to get a “really good picture” of formative assessment practices at high-need schools.
“We learned that even with the top 122 schools, teachers are not quite sure how formative assessments are done or how the results affect students,” said Abedi.
The consequences, according to Abedi, is that teachers are paying more and more attention to standardized tests and tend to teach to those tests at the expense of assessing student performance during instruction.
“Formative assessments are not part of the high stakes world of testing,” said Abedi.
In the next phase of the research, the team developed in-depth case studies of seven schools. This involved “intense observation,” and more interviews and surveys of teachers and students. The result is a set of “universal formative assessment” components: nearly 250 practices and tools based on their study and extensive review of the research literature.
In the first part of 2015, Abedi and his team will launch a website with the tools for educators to select from and implement. “We will track which tools they pull down to continually refine what is available,” said Abedi.