The Power of Linking Assessments of Problem Solving to Instruction

By: Jeff Heyck-Williams

Earlier this summer, I had the pleasure of working with eight amazing teachers from Two Rivers as they developed performance tasks to assess problem solving skills.  Building on the work of Jill Clark, who has been developing our assessments for critical thinking and problem solving for the last couple of years, Kai Blackwood, Helen Gasperetti, Jazmin Heartfield, Kathryn Mancino, Maria Nguyen, Laura Sparks, Mo Thomas, and Julia Tomasko each embarked on a process that culminated in the creation of eight unique assessments of problem solving.  Their tasks spanned the grades from preschool through eighth grade, and crossed all subject areas.  As part of our larger project to assess the transfer of critical thinking and problem solving skills, these assessment tasks were designed to be content neutral and get to the heart of the cognitive skills students need to be effective problem solvers.

We began this work with a couple of basic assumptions about problem solving.  First, we believe that problem solving is a broadly applicable skill that can be cultivated as a thinking routine across all disciplines.  Second, we believe that we can gain meaningful insight into a student’s ability to apply problem solving skills through short performance tasks.   Building on these assumptions, we tackled the problem of how to create rubrics that defined the construct of problem solving and of how to develop tasks that would provide evidence that students are able to transfer the skill.

What has become clear through the course of this work is the tight connection between assessment, instruction, and learning.  We have been deeply influenced by the work of Grant Wiggins and Jay McTighe in their seminal work, Understanding by Design.  So we understood the concept of backwards design and have regularly applied it to our work in building students’ conceptual understanding within our approach to project-based learning.  However, the work of creating assessments of critical thinking and problem solving has shed light on the applicability of backwards design in the development of cognitive skills.  Specifically we have learned the value of defining what we mean by problem solving, of developing tasks that accurately assess that skills, and of teaching students the thinking routines to develop these skills.

DEFINING PROBLEM SOLVING

When we first embarked on this work, we realized the importance of defining what critical thinking and problem solving look like in student work. Teachers needed concrete examples of how to make problem solving come to life in the classroom and for students to truly develop problem solving skills.

With this in mind and building on the work of Catalina Foothills School District in Arizona and Laura Greenstein’s work in Assessing 21st Century Skills, we created a general rubric of problem solving.  Through use in classrooms and multiple revisions, we have attempted to define problem solving by identifying five separate core components of the problem solving process: 1. Identifying what Is Known About a Problem; 2. Defining the Problem; 3. Generating Possible Solution Strategies; 4. Applying Problem-Solving Steps; and 5. Evaluating Solutions.

This rubric has given a definition for teachers to explore what problem solving can look like in their classroom.  More importantly, it provides concrete language around which to talk to students about how they can improve their problem solving skills.

DEVELOPING TASKS THAT ASSESS PROBLEM SOLVING

However, simply defining the construct of problem solving for teachers and students isn’t enough to help them see how it can live and develop in their work.  By creating tasks that assess problem solving, teachers deconstructed our definition as they applied it to the work of their students.  They recognized that students problem solve all of the time, but uncovering how they used the components in their problem solving was less clear.

​By creating content neutral tasks and then defining the right sets of questions that would elicit evidence of student thinking, teachers began to understand how they could help students get better at thinking through a problem by highlighting how they approached each component defined in our rubric.  We realized early on that each component of the rubric needed specific language that would point to specific “look fors” as evidence that students were accomplishing that step in the process effectively.  Then in response to each of those components, teachers needed to ask targeted questions that would uncover student thinking related to those “look fors.”  For example, when determining whether or not a student is accomplished in “Generating Possible Solution Strategies,” we have to ask the student to explain not only what strategy they will use, but what steps they will use to implement the strategy.  Without that specific cue, we may not elicit much evidence about whether or not students can effectively generate solution steps.

Thus the creation of tasks that assess problem solving allowed teachers to have a clear and consistent understanding both of what we mean by problem solving but also what problem solving can look like in student work.  It gives the teachers a clear target at which to aim as they plan experiences for students that will develop their problem solving skills.  

TEACHING STUDENTS THINKING ROUTINES

What was most clear through this work was the importance of making thinking visible.  If we are not able to see and to hear what students are thinking either through writing or their verbal comments, then we are not able to assess where they are in their thinking or to provide them with quality feedback to improve their thinking.

Thus we have worked closely with the ideas of Ron Ritchhart, Mark Church, and Karin Morrison in their book Making Thinking Visible.  Specifically we have seen the power of developing common thinking routines in our classrooms and across our school that allow students to develop consistent approaches to critical thinking and problem solving.  By having clear but adaptable components of problem solving for example, students are able to effectively work through each component improving their overall approach to solving problems in the future.  

With problem solving, we use a K-W-I structure adapted from Linda Torp and Sara Sage’s Problems as Possibilities: Problem-Based Learning for K-16 Education.  Applying this structure in classes from math and science to larger projects, students identify what is KNOWN about the problem,  WHAT they need to find out, and IDEAS for solutions.  Each of these components map directly back to three of the components of our problem solving rubric.  By reinforcing the K-W-I as a regular thinking routine, the structure becomes a habit of mind that students draw upon whenever they are faced with a problem, even when they are not explicitly asked to fill out a K-W-I chart.  Thus the thinking routine transfers to any setting and any problem that the student will face in and out of school.      

LINKING ASSESSMENTS OF PROBLEM SOLVING TO INSTRUCTION

Planning backwards from first defining problem solving, to developing assessment tasks, and finally to teaching regular thinking routines, we are building a system which supports the development of broadly applicable cognitive skills which can be applied to any situation or problem.  Working with teachers this summer has emphasized the importance of keeping this assessment work closely tied to instruction because ultimately the goal is not to develop another set of assessments, but to have all students leave us as effective critical thinkers and problem solvers.