By Christian Schunn

IFL Research Liaison

Assessments play a powerful role in education.

A reoccurring theme in IFL’s guides for practice relate to the powerful role of formative assessment—brief, ungraded, targeted assessments that occur in the middle of instruction to surface student thinking on content of the day or week. These assessment help teachers build upon what that group of students know, addressing the specifics of what they don’t know, and orchestrating a powerful classroom discussion around the different specific ideas that various students have. Further, improvement science, a major trend in education, fundamentally calls for practical measures to shape educator’s decisions, and those practical measures usually involve in-the-moment student assessments.

Various summative assessments (those designed to see how much students have learned on a larger topic or set of topics) also have a large effect on teaching. These might take the form of benchmark testing with the district, pop-quizzes, end-of-unit assessments, or state testing. These assessments signal to students (and various levels of educators) what is valued. It is a fundamental fact of public education in the US that these summative state tests shape what is taught and how it is taught.

Just like one can eat too many carrots, one can give students too many assessments. This especially applies to summative assessments. I think eating too many carrots is relatively rare. Unfortunately, in the US today, most students are getting too many assessments in total across the year, with large negative effects on student outcomes. It is time to recognize this systems problem and reset our approach to assessments. Different educators control different parts of the system; every educator can play some role in addressing aspects of this systems problem.

When assessment load prevents instruction.

IFL Fellows regularly hear from our partner districts that little classroom instruction can happen between March and May because of state testing. Three months is a large fraction of the school year! But my observation is that actually a much larger fraction of the school year is lost to testing, once we add in all the other summative assessments being given to students and the time devoted to preparing for those summative assessments. In one recent example I have seen, half of September (benchmark), half of October (end-of-unit), half of January (universal screener and end-of-unit), and half of June (benchmark) were all also lost to summative assessments.

It is true that practicing a summative assessment will help students do better on that kind of assessment. However, research is pretty clear that this kind of learning is very limited. It only improves outcomes by a few points. Repeated practice with a given assessment format does little to further improve performance, and assessment format practice does not put students in a better position for learning next year’s content. It is also true that retention of information is improved when students test themselves on content. However, those testing effects are largely limited to simple factual information. If students do not really understand the mathematics, the science, the ELA, and the history, poor performance on the assessment will remain the outcome of extensive test practice that year and into the next many years.

To build student understanding, many real instructional days are required. Students need time to engaging in the kinds of meaning-making learning processes that produce large effects on their understanding, lead to much improved retention of information across months and years, and set them up for improved outcomes for years to come.

There is also an important equity point here. Too often I have seen classrooms, schools, and districts that have poor outcomes on state tests decide that their kids require more summative assessments and summative assessment practice; only the high performing students are given the luxury of losing only minimal instructional days for such testing. But from a learning opportunity perspective, it is those poor performing students that need more instructional days (including the use of formative assessments!) to have a chance at substantially better outcomes.

When summative assessment load turns students off of learning.

Recently, when we did empathy interviews with students in math and in ELA across middle school and high school, students frequently named assessments as a main culprit in their feelings of low self-efficacy in that content area. As another datapoint, a few years ago, I led a large study in four districts varying in size and student demographics. We were tracking student attitudes towards science from the beginning of middle school to the end of 9th grade. Attitude assessments were given each year at the beginning, middle, and end of year. We saw little overall changes in average attitudes from the beginning to middle of year. Some students went up, while other went down, related to the experiences they had in their science classrooms and in optional out-of-class experiences. But at the end of the year, average attitudes saw a massive decline in every attitude, in every district, in every classroom, regardless of student characteristics. What changed? Students had just experienced multiple months of testing hell.

When we can balance a summative assessment experience with a growth experience, students’ self-efficacy and interest in an area can grow. When the summative assessments are frequent and not balanced with growth opportunities, the negative motivational effects of summative assessments cumulate. In addition, beginning the school with a wall of diagnostic assessments does not help build student motivation or teacher-student relationships that are so fundamental to having a productive school year.

Collaboratively do an assessment audit.

Teaches and leaders, collaboratively examining the total testing load within a particular class, are likely to provide some insights to everyone, particularly the individuals who have the power to make largescale changes. For example,

    • How many instructional days in middle school math are taken up with taking state tests in math, doing practice state tests in math, going over the results of a practice test, resting because students are taking state tests in other subject areas, diagnostic testing (e.g., NEAP, NWEA’s MAP testing at BOY, MOY, EOY), English-language proficiency exams, PSAT, end-of-marking-period assessments, and preparing for end-of-marking-period assessments. Don’t forget to include makeup testing days.
    • What fraction of the instructional year in that area is consumed with assessments? If it is more than one-fourth, negative effects are very likely.
    • Which summative assessments provide new information? When two summative assessments occur in close proximity, no new information is being gathered and student motivation is negatively impacted.
    • How often is a specific testing format practiced? If it is more than a couple of times, it is unlikely to be helpful.
    • Which assessments return results that are relevant to educator decision making? Most summative assessment reports do not track student growth and therefore provide little information on how educational decisions (by central office, principals, or teachers) are influencing student learning outcomes.

By working together, we can dismantle the problem, assessment by assessment.