Accountability for Numbers
Standardized testing in K-12 schools has a twenty-year history of being the one and only way that we hold students, teachers, principals, schools, districts and states accountable for public education. The problem with them is that they are designed to yield one number for every child, student, teacher, etc. These numbers can be easily compared and contrasted to other student’s numbers and based on that one number, a judgement for accountability purposes can be made. Performance Assessment, where students demonstrate their thinking by solving problems in real time—where students show what they know and can do—is a better assessment to determine accountability for what we value in education.
Standardization of student assessment has been very helpful in light of the heavy lifting needed to collect large amounts of data for all 50 million K-12 students. The argument about the efficiency of this data collection won out for the last twenty years but in recent years has been failing us in the types of accountability it provides. Big data and its analysis is one of the hallmarks of our educational era, and the criticism of how bad an idea it is to judge students by one number is very loud. Facebook—to use an example of how big data gets handled by large companies—looks at user information and makes advertising decisions that are well admired by analyzing large amounts of user opinion expressed to family and friends.
Efficiency vs. Thinking
The efficiency gained of standardized testing is technically understood by testing experts as valid and reliable. The number yield provides us with one number from a student on a test and that is said to be a valid way to measure what that student has learned, while the number can also be compared to other student’s numbers and that is called reliable. But the problem with this is the validity of one number for one student is limited to the quality of the test and does not reveal anything about what the student is thinking. It also tends to value knowledge reproduction, that is to say re-telling the same knowledge that was assigned, rather than new thinking and new learning about the knowledge that was assigned to be learned.
Factory-Model vs. Collaborative Model
For well over 150 years, the factory-model school, organized as cohorts that learn the same material at the same time and get assessed at the same time has been a hallmark of our education system. However, education is the process of drawing out of the student ever-more sophisticated thinking that will end up in qualifying that student for future levels of education and future jobs. Future schools depend on students who demonstrate what they know and what they can do with what they know—theory and practice. Likewise, employers depend more on thoughtful interviews of what students know and what they think about what they know. The one number report is not as useful as it once was perceived and the need for better assessments of student progress are becoming much more valuable.
Luckily for the world, international assessments of educational progress are now being measured using performance assessments around Literacy, Math, Science and Collaborative Problem Solving. These types of tests will have some multiple choice components but also include real life situations where the student will be asked to offer their thinking of how to solve problems. Rather than asking to reproduce knowledge, these performance assessments ask students to “Extrapolate information from what they know and apply it to new and unfamiliar situations” (What is Programme for International Assessment [PISA]). The performance assessment system where students solve problems in real time is a much more accurate test of what we value in schools and jobs—the quality and equity of our educational system and the readiness of our students to succeed in the job market.