Can Educational Assessment Improve Teaching?
What Would it Take to Design an Assessment System Around a Goal of Improving Teachers’ Knowledge and Skills?
How would I design an assessment or assessment system if my goal was to improve teaching? My colleagues, Chris Domaleski and Leslie Keng, and I met recently with state assessment leaders who are thinking about reforming their assessment system.
The Center for Assessment routinely facilitates these sorts of meetings and we are known for persistently pressing our clients to clarify the problem or issue they are trying to address. We want to establish how the assessment results will be used to serve a specific purpose – this discussion becomes the first step in outlining a theory of action for the assessment program.
When one of the state leaders indicated his goal for a reformed assessment is to help teachers get better at teaching, I nodded and said, “Well sure, teachers often play a mediating role if our ultimate goal is improving student learning.” But when he went on to define improved teaching quality as having teachers develop a better understanding of rigorous content and the skills necessary for helping students meet these more rigorous expectations, I then realized this was a different goal than we typically consider in our assessment design.
Placing a Greater Emphasis on Teacher Improvement When Developing a Theory of Action
Identifying goals, anticipated outcomes, and intended uses is the first step in creating a theory of action. Theories of action for assessments designed to support student learning almost always include some sort of teacher action, such as interpreting assessment results to identify student learning needs in order to tailor instruction to meet those needs.
In our world of student assessment, the stated outcomes of many assessment programs are focused on improved student achievement or other student outcomes. Unfortunately, most large-scale assessment designs are rarely able to support such goals, but that challenge has been discussed in many other CenterLine posts already.
Therefore, considering how to design an assessment system to improve teaching quality really got me thinking. How would we design an assessment differently if our goal was to help teachers learn more about teaching, their content area, and their students compared with an assessment designed for measuring student learning?
We cannot meet the goal of improved teaching through assessment alone. I would make a similar argument for other ambitious educational goals, but clearly, designing an assessment to improve teaching quality must include strong linkages to other aspects of the educational ecosystem.
First Steps Toward Improved Teaching Should Focus on Curriculum
I have relied on Elmore’s (2003) seminal “instructional core” to frame related discussions of educational reform. He posited the only way to improve student learning at scale was to raise the level of the content offered to students, improve the quality of instruction related to that content, and increase student engagement with that content. Elmore emphasized that all three components must be addressed before reformers can hope to see gains in student learning. However, addressing the instructional quality component of the core requires focusing more specifically on the connections among curriculum, assessment, and instruction.
I do not see how to meet the goal of having teachers develop a better understanding of rigorous content and the skills necessary for helping students meet these more rigorous expectations unless the assessments are situated in a high-quality curriculum and/or well-specified learning progressions. For example, Lorrie Shepard and her colleagues (2018) persuasively described how learning progressions serve as the foundation for both instruction and assessment, enabling teachers to gain greater knowledge of student learning and teacher practices.
A rich curriculum, based on a sound theoretical conception of learning, helps provide a picture of “meaningful content” and helps sequence learning activities so students have a legitimate opportunity to pursue rigorous learning goals. (I refer to “curriculum” here even though I know many might prefer “learning progressions”. I’ll save that argument for another day.) For now, either choice must provide a defensible framework for assessment design and, more importantly, a lens for interpreting the assessment results.
Once Curriculum is Established, How Do We Then Design Assessments to Improve Teaching?
This foundation in content and curriculum is critical to help teachers learn about student learning and get smarter about teaching. But, the mechanism for how teachers will actually get better at teaching is still implied and not explicit.
This is a vexing problem. It is one thing to provide assessment information tied to rich curriculum, but it is another thing to help teachers improve the ways in which they foster student learning. Even beyond situating the assessments in strong curriculum, we need to find ways for teachers to learn from assessments to improve their teaching.
Designing assessments, such as rich performance tasks, to support teachers’ examination of student work is a sensible step. Teachers will need support, generally in collaborative groups (e.g., Professional Learning Communities or PLCs) to make sense of their students’ products. Such examinations of student work will help teachers develop a nuanced understanding of how students progress from novice to more complex domain competence. This type of collaborative interaction can also support instructional improvements as teachers learn from one another. For example, we can imagine a conversation: “How did you instruct your students so they were able to provide that level of work?” There are likely assessment types other than performance tasks that can support increases in teacher learning, but having access to actual student work seems to be a critical factor.
Sounding a familiar refrain, we can provide useful information and foster teacher collaboration, but those do not cause improved teaching practices. Earlier this year, I offered some thoughts about how to rethink teacher evaluation and supervision. I pushed for a coherent connection between instructional coaching of teacher practices and examinations of student learning. I am not suggesting we need to rely on teacher evaluation as the mechanism for assessment to support improved teaching – and certainly not the types of state-controlled teacher evaluation prominent during the Race to the Top era. However, instructional coaching, mentoring, and peer feedback, common to more supportive educator-development systems, must be part of a theory of action with a goal of having assessment results support improved teaching.
Pursuing Ways to Improve Teaching Quality Using Assessments
I think I answered the question that started this post of whether assessments can improve teaching. No, assessments cannot improve teaching. Perhaps a better question to ask is, “what types of assessments, situated in what type of system, are likely to support improvements in teaching quality?” As you can see, I am struggling with these ideas and I will continue to share thoughts as I continue to think about this challenge.
I welcome others’ thoughts about how to move the needle on this complex, but critical, effort in 2020. We invite collaboration and will be happy to publish guest posts on CenterLine related to this topic.