Quick trip to the doctor office. After explaining that my arm was sore, my doctor told me my arm was broken. They were going to put in a cast.
“Why?” I asked.
“Other patients with pain like you described had broken arms,”
“No X-ray? No testing?” The doctor was busy–didn’t have time. Too many patients to deal with; they would rather just head to treatment.
Wait a minute. Does something seem wrong? Absolutely. Yet, in classrooms every day teachers are prescribing instruction for students at times without the proper formative tests. Without the test to tell them what the student needs instructionally.
1. Using the wrong assessments–tests that don’t tell us what to do next
2. Not collecting data in a usable format (not easy to interpret or analyze)
3. Failing to assess enough for fear of “over assessing” (which is a legitimate fear)
Ultimately, educators sometimes move to the instructional “treatment” without the “right” tests.
So… how do we make sure we have the “right” tests? Student assessment is the single most important thing that happens in any teacher’s classroom. I’m not talking about the required state assessments or even district diagnostics (although there is value to those tools in their own right). I am talking about formative assessments created by teachers within a district to measure and monitor learning. The assessments that check student understanding, and tell teachers what to do next.
Here are some best practices to help teachers create quality assessments:
1. Use Careful Alignment
Assessments should measure what we ultimately want students to know and be able to do. Therefore every single item should be aligned to a specific target. This might be a standard or a district learning target. Every question should link back to the objectives aligned to what we want students to know and be able to do.
Writing questions targeting specific standards will allow us to measure which standards or learning targets students are mastering, which present some difficulties and which require reteaching. I typically suggest writing the standard directly on the answer key or even the student version so we are absolutely clear about what we are attempting to measure with each and every question.
2. Align More than One Question Per Target
If you think you can ask one question and know whether a student “gets it” or not…you’re just wrong. Kids get questions correct and incorrect for a variety of reasons that might not be aligned to what you are trying to measure (even a lucky guess!). Additionally each standard and learning target likely has more than one level of difficulty embedded within the target. It is best to vary the level of difficulty and present questions at more than one level of difficulty: maybe a hard one and an easy one.
Here is some food for thought:
AT LEAST Triangulate your data: 3 questions per target (one at each of 3 different levels of complexity)
3. Use a Variety of Formats
There are a few different formats for delivering the assessment. Any assessment that relies exclusively on one format or another does not allow teachers to see students who may excel in one type of communication or another (Ex: writing their thoughts versus discussion or multiple choice questions)
Typically Assessment Formats:
Selected Response Assessments: Ask students to select the correct answer from a provided set of answers. This includes multiple choice, matching and true false. Multiple choice questions are powerful when the distractors (wrong answers) are well written and provide teachers information about student misconceptions. True-false items, while they can assess a variety of cognitive demand levels, often fall into low level categories and may not provide good data when students have a 50-50 chance of getting the correct answer by guessing.
Constructed Response Assessments: Ask students to construct their own answer to a question. Open ended short answer or essay questions provide an opportunity to learn about student thinking as teachers watch them explain or work through a problem. All constructed response questions require a rubric for measuring student responses in a reliable and repeatable way.
Performance Assessments: Ask students to demonstrate understanding by performing or creating a product. This could include performing a piece of music or giving a speech. These performances typically reach into the highest levels of cognitive demand and also give a great window into student thinking. Performance assessments require a rubric for measuring student responses in a reliable and repeatable way.
Personal Communication & Discussion: Ask students to demonstrate understanding through discussion, oral questioning, teacher-student conferences. Socratic seminars or one on one student discussions are great ways to learn about student misconceptions. These assessments also require a rubric for measuring student responses in a reliable and repeatable way.
4. Use a Variety of Complexity Levels
Perhaps one of the most overlooked areas of assessment is the level of complexity of a single assessment item. Teachers should be as intentional about the difficulty level of each question item as they are about its alignment to a specific learning target. Using a guide such as Web’s Depth of Knowledge (DOK) or Bloom’s Taxonomy when writing questions will allow teachers to be intentional about the complexity within each cognitive presses they are assessing.
It is best to code each question with not only the standard or target assessed, but the level of intended complexity as well.
Lets be honest: nobody wants to give up instructional time for an assessment if doesn’t tell us anything valuable. Good assessments need to tell us what to do next as instructors. Using these best practices can help us create valuable tools that will help us help us monitor learning and help our students all grow and learn each.
Anne is an assessment and curriculum specialist best known for her work in assessment design, data analysis and instructional effectiveness. Anne is a sought after speaker in the area of assessment design, curriculum and instruction.