Student performance based on assessments can be measured in two quite different ways, both of which are important. Achievement describes the knowledge or proficiency of an individual in something that has been learned or taught. This is measured by achievement tests where students are compared to other students in their grades. Growth, in comparison, describes the increase made over the academic year. The student is their baseline. As different value-added models have become increasingly widespread and carry higher stakes, questions concerning the validity and reliability of their results have grown more important.
To determine growth in Tennessee, statistician Dr. William Sanders invented a value-added model. Sanders and his team at the agriculture school at the University of Tennessee developed a model based on theories applied in agricultural genetics. This theory was developed during a perfect economic and political storm. By 1992 the Tennessee Supreme Court ordered a more equitable funding system for public schools. This action created renewed interest in education accountability, and Sanders’ formula was included in Tennessee’s Educational Improvement Act.
In general, value-added models use sophisticated statistical algorithms and standardized test results, and other information about students to determine a “value-added score.” The running joke among Tennessee educators is that the algorithm of our model is a more closely guarded secret than the Coca-Cola formula. The Tennessee Value-Added Assessment System (TVAAS) is only one of many models used across the country.
No one will deny that educators need to be held accountable just like members of any other profession. The federal government put an increased emphasis on accountability systems in the Every Student Succeeds Act. States like Tennessee further linked teacher evaluation plans to value-added measures. A future debate must examine how states are being coerced with federal money to adopt specific education policies.
There is an expectation that there will be accountability and action to effect positive change in our lowest-performing schools, where groups of students are not making progress, and where graduation rates are low over extended periods. As value-added models have become increasingly widespread and carry higher stakes, questions concerning the validity and reliability of their results have grown more important. Year-to-year growth provides a more accurate picture than multi-year composite data. The teacher has often been deemed as the determining factor in a student’s growth (or lack thereof). The problem with this from an educator’s perspective is at least three-fold:
Are the students’ socio-economic statuses considered when growth is determined? Using a student’s original scores as a baseline appears to address this issue, but educators know that circumstances can change in an instant. What if a student’s physiological needs (food/shelter) along with the need to feel safe or loved are no longer being met? How can any teacher expect a student to learn grammar rules or multiplication facts if a parent lost a job or passed away? Valued-added does not address emotional or behavioral issues surrounding a child.
What if a school or district students consistently score in the 80-90 percentile? When student scores are consistently above average, there is little room for growth. This negatively affects teachers’ scores.
What is really measured? Most achievement tests measure English/Reading and Mathematics. Often tests will only test aspects of Science and Social Studies. What if you teach PE or music? Where will your value-added scores come from?
It has long been rumored that the Tennessee Department of Education would like to see changes in this area of accountability. We welcome that discussion. On the surface, value-added assessment appears to be a reasonable method to measure the growth of a student on more than just how they perform on achievement tests. The problem lies when the teacher is evaluated based on these value-added scores. There is too much room for subjectivity. It simply is not fair for a teacher’s livelihood or paycheck to be based on extenuating factors that he or she cannot control.
Moreover, a teacher should never be punished for the zip code in which they teach, who their students are, or which subjects they teach. Each school is as unique as the community it serves. There must be a more equitable way to determine teacher quality, success, and effectiveness. The value-added accountability model, which was created in 1992, needs to be re-examined, updated, and replaced by something possibly less contentious and less likely to be misused by administrators and policymakers. That is a debate we should have.
No decision about an educator, school, or district’s performance should be based on a single data point or a single test. A teacher’s career, salary, and livelihood should never be determined by how well their pupils perform on a single exam, one or two days of the year. If proposed changes to the TVAAS formula occur, they must be assessed and demonstrated on the front end with teachers, and used for improvement, not for continued high-stakes consequences. Educator evaluation scores and salaries should not be tied to it.
Executive Director of Professional Educators of Tennessee
Member Services Coordinator for Professional Educators of Tennessee