What we’ve learned so far:

Key 1: Clear Purpose

Summative assessments give us a snapshot of a student’s learning at the end of a sequence of instruction. Parents, students, teachers and others view summative assessments as an important indication of what students know and can do, both in relation to the content and in relation to other students.

When summative assessments cover a wide range of learning and are the results are aggregated at the classroom, school or district level, they are an indication of whether the teacher, school or district is teaching the curriculum well.

Formative assessments are short and used often, to inform teachers of whether instruction is hitting the mark with all students, and to inform students of where they are in the learning progression and where they still need to go (what work they still need to do). They are for information only.

Teachers, schools and districts should be clear about the purpose for each assessment they use.

Key 2: Clear Targets

Standards are often multi-faceted. They need to be broken apart into individual learning targets (learning goals). They should also be written in student-friendly (and parent-friendly) language.

There are different types of learning targets. Some are primarily about knowledge, some specify patterns of reasoning, others describe skills, product or dispositions. Reasoning requires knowledge; proficiency in a skill often requires higher order thinking (reasoning) as well as a base of knowledge; creating a product requires skills, reasoning and knowledge.

Reasoning and higher order skills are essential for our modern economy and our democracy. Yet there are too many standards to teach critical thinking, reasoning and problem solving well.

Assessments often address knowledge-level learning targets, but they must also address reasoning, higher-order skills and the creation of products. (The pillowcase problem and the moon phases problem are examples of assessment tasks that require higher-order thinking.)

Key 3: Sound Design

Some assessment methods are more suited to particular types of learning targets. Matching the method to the target is essential for creating a useful assessment.

An assessment plan shows the learning targets that were taught and will be tested, the spread of methods used on the assessment, and they weight given to each learning target.

Items should meet certain criteria:

1. The content should match the learning target and should be sufficient for solving the item (along with prerequisite knowledge from earlier units or grades); items that require connecting knowledge from two learning targets are useful to assessing whether students can make the connection, but knowledge of each learning target must be verified with separate items.

2. Mastery of the content of the learning target should be necessary to solve the item; the item should not be solvable by mere test-wiseness.

3. The performance in the item should match the verb of the target.

4. The context should be engaging but not distracting or biased against any group of students.

5. For multi-faceted learning targets, all of the content should be represented by items on the assessment.

6. For multiple choice items, the wrong answers should be plausible and reflect common student misconceptions or errors when possible.