Imagine that you are at a crucial juncture in an instructional unit. It is absolutely essential for students to understand the material covered so far; otherwise they will struggle with the upcoming topics. You would like to give students a task that identifies both what they know and can do and what they are struggling with, so that you can provide feedback and make instructional modifications. How can you design and conduct this activity to make the most positive impact on student learning? In other words, what are the characteristics of effective and efficient formal formative assessment?
As part of a five-year research project funded by the Institute of Education Sciences, our research team interviewed, observed, and videotaped 20 middle school mathematics and science teachers every day during 26 entire instructional units. (Some teachers were observed more than once.) We also interviewed a random sample of about 30 percent of students in these classes and collected 3,459 student work samples. Here we describe what we learned about formal formative assessment.
Defining Formal Formative Assessment
Formal formative assessment activities are planned in advance, designed to gather information from all students in the class at the same time, and intended to move students forward by providing feedback or instructional adjustments. Some testing companies portray formal formative assessment as only involving an embedded test in the middle of the unit. However, we use the term formal as meaning a planned opportunity for all students to share what they know and can do. These activities can take many forms, including homework assignments, exit tickets, and handouts. As an example, in one of the 6th grade math classes we observed, the teacher asked students to complete an exit ticket with the following task: Draw two shapes—one rectangle and one square—so that each has a perimeter of 12 units.
These assessments are only truly formative when they improve student learning. A teacher who carefully designs, plans, and implements a task but doesn't act on the results is not fully engaging in formative assessment. How can we, as educators, set ourselves up to collect meaningful information and then capitalize on it to move students forward?
Windows into Student Thinking
When we design questions or activities that help us understand how students arrived at an answer, we can be more precise about our next instructional steps. For example, knowing that a student has incorrectly identified which of two fractions is larger is not as informative as knowing that the student only considered the denominators in making her decision (Wiliam, 2011). Only the latter information can help us know where to focus our time and efforts. The key is to create activities and ask questions that make students' thinking explicit.
The most informative questions require students to explain their answers, elaborate on their responses, or provide information about why they think something. Sometimes teachers ask multiple questions with increasing focus to help identify the source of confusion. Some of these questions include Why does ____?, How would you ____?, Could you explain ____?, and Why is ____ an example of ____? These prompts can provide insight into not only what students know, but also how they know it (Stobart, 2014).
In our study, we learned that only a small portion of the work teachers graded was designed to make students' thinking explicit by revealing students' naïve conceptions, mistakes, misapplications, and common mathematical errors (14 percent). Instead, the majority of teachers' efforts were spent grading work that was designed for students to practice something (52 percent) or grading summative assessments like end-of-unit exams (24 percent).
A Prescription for Feedback
After we've collected student work, we need to carefully select what's worth our time to read and how we'll comment on it. What kind of feedback is both effective and efficient?
Face-value comments that only evaluate student work—for instance "good" and "not quite"—are not worth our time. If we write "good" on a student's paper, he won't know what, exactly, was good about his work. If we simply write "not quite," it may not be completely clear how the student can improve.
Instead, we should aim for comments that are both descriptive and prescriptive. A descriptive comment, such as, "Good explanation. You are providing data as evidence to support your claim," lets the student know why something was correct or incorrect. A prescriptive comment, for instance, "Do you have a claim? Where is your evidence? Provide some justification that supports your claim," helps the student know how to improve.
We also know that both writing comments and providing scores on papers minimize the impact of the comments (Butler, 1988; Wiliam, 2011). Students tend to pay attention to the scores and ignore the comments. So if we assign a formal formative task that we plan to provide comments on, it's best to forgo a score. Additionally, not all feedback should focus on content. Comments can also be helpful when they address overarching problem-solving or learning strategies, such as, "When you answer questions at the end of a lab, make sure to review your observation notes, data, and analysis before writing your answers."
Across all of the student work samples we collected that included teachers' comments, only 28 percent of student work featured comments that were either descriptive or prescriptive. The remaining 72 percent included only lower-level comments, such as those that were evaluative or corrective—the type that aren't very helpful to students. Instead, effective and powerful feedback should be concrete, specific, and useful; it should provide actionable information. Effective feedback in the form of descriptive and prescriptive comments should lead students to judge the quality of their work and to monitor themselves as they produce new work.
Moving Forward with Instruction
In our study, we found that teachers tended to repeat written comments from one student to the next. In other words, when different students made the same mistake, teachers recycled the same comment over and over again. We found repetitive comments in 72 percent of all products in which teachers left some type of comment. Although repeating the same comment may seem economical (you don't need to generate a unique comment for each student), it is often more efficient to take action with a group of students who all made the same mistake to model how to approach a problem.
When teachers model a problem and students help solve it, the process reinforces what students know or redirects what they are doing wrong. Even better, we can emphasize strategies that will help students check themselves the next time they are faced with a similar task. A teacher might say, "Before you add your numbers, remember to align the number by the decimal point."
In addition to making instructional modifications, we can share with students how they are doing as a class. One way to do this is to provide percentages of how many students responded correctly to each question. This immediately shows students what questions they struggled with the most. We can take the opportunity to discuss why and also what they can do to help one another.
Teachers can provide a general description of what students tended to miss, such as, "Most of your reports missed a description of the control variable. Why is this information important, and why should we not miss it?" Discussing results at the whole-class level helps students understand where they are now and what they need to focus on. Plus, it's more effective to discuss why an answer is correct or incorrect than to provide only correct answers or to hand students their reviewed products without further discussion.
Balancing Efficiency and Effectiveness
So what are the key ways to maximize your time and improve student learning?
- When you administer formal formative assessments to get a sense of students' thinking, choose a few well-designed questions or prompts. If you won't review all of the questions, there's no point in asking them. A few well-designed questions are better than many superficial ones. You will want to be able to quickly review students' work and identify their strengths and weakness.In our study, a middle school math teacher asked students to solve one multiplication problem: 2 1/9 × 11/16. The teacher selected this problem because it requires students to complete all the potential steps that might occur when multiplying mixed numbers, including converting a mixed number to a fraction and multiplying two-digit numbers in the numerators, as well arriving at a product that is an improper fraction and deciding what to do with an improper fraction that can only be reduced to a mixed number. With just this one question, she could very quickly see where students were getting hung up.
- Briefly review the students' work to determine whether a number of class members made the same mistakes or displayed the same misunderstanding. If all or almost all students made the same mistake, consider taking action the following day with the whole class by reteaching, modeling, or assigning a new or modified task. In our study, we found that in 92 percent of work samples in which all or almost all of the students missed the same question or questions, teachers left repetitive comments on the work. The teachers could have instead spent that time planning whole-class instruction targeted at the area of misunderstanding.
If you see patterns of errors in some students' work, on the following day meet with a small group of students who made the same mistake. Alternatively, it might make sense to place students in purposeful groups—in which at least one student made the mistake and one student did not—to facilitate peer learning. If one student's responses were very different from the rest of the class, you may want to plan some time to work individually with that student.
If there is a high degree of variability in the students' errors, it may be most beneficial to leave descriptive and prescriptive comments on their work. You can follow up with students later to make sure they have an opportunity to show what they've learned. For instance, if you give students the mixed numbers multiplication task mentioned above and find that a handful of students made a variety of errors, it may not be necessary to alter instruction for these students. Rather, it might make more sense to provide individualized written feedback on the students' work that relates directly to the error they made. Later, you can ask students to complete a similar task to check their understanding.
With formal formative assessments, you can pause at essential points in an instructional unit, check in with all students about what they understand and what they don't—and then make instructional adjustments and provide feedback. Make the most of your valuable time by investing in formal formative assessments that strike a balance between efficiency and effectiveness.
