As a middle school math teacher, I taught at a school serving a wealthy student body and my students had incredibly high test scores. But I had also taught at a school where the majority of students were eligible for free or reduced lunch and my students had incredibly low test scores. In the low-poverty school, I was seemingly a very effective teacher—yet, in the high-poverty school, I was seemingly a very low-quality teacher.
This experience led me to a career as a sociologist focusing on inequities in education. In my most recent study, I investigated the use of test scores to assess the effectiveness of teachers.
Teachers are formally ranked based on their students’ test scores. Value-added scores are intended to be an objective scientific way to determine individual teacher quality. These scores attempt to be accurate and fair to teachers by focusing on growth in students’ test scores, under the assumption that a student’s social background only affects their initial achievement levels but not their growth in learning.
Schools have been ranked based on students’ test scores for even longer than teachers. Schools with lower average test scores are described as “low-performing,” and schools with higher average test scores as “high-performing.” The schools ranked as low-performing virtually always serve a predominantly lower socioeconomic status (SES) student body, whereas the schools that are ranked as high-performing virtually always serve a predominantly high SES student body, or are schools with selective admissions processes.
Nonetheless, many people believe that schools are low-performing because of low-quality teachers, and policy aimed at reducing educational disparities continues to center schools and teachers. A couple of studies find that teachers’ value-added scores are lower in low-performing, high-poverty schools. This is interpreted as more evidence that achievement disparities by SES and race are due to low-quality teachers.
Yet, there are many studies that question the validity of value-added scores, mostly because we don’t have the data to account for all of the things that are outside of teachers’ control but impact student learning. These are things like the characteristics of the school, the student body, and the students’ neighborhoods and homes. Despite their questionable validity, these scores are sometimes released publicly or are applied in decisions surrounding teachers’ salary and retention. In an extreme and tragic case, a teacher’s suicide in Los Angeles was linked to the public release of value-added scores.
My recent study focuses on Education Value-Added Assessment System (EVAAS) scores, the most widely used and least studied methodology. Whereas most previous studies have investigated score validity by comparing across schools and teachers, I investigate the validity of EVAAS value-added scores by examining changes in the scores of individual teachers over time. That way, the teacher qualities that are difficult to measure (like how they interact with students, or the way they explain things) are held constant. The idea for this approach comes from my own experience in the classroom.
This approach also aligns with the thinking of policymakers. There’s a notion that if we could just fill high-poverty schools with “high-quality” teachers, everything would turn around for those kids. One day, a school district superintendent confided in me that he’d actually tried this. He pulled multiple teachers from a high-performing school and put them into a low-performing school. In a whisper, he told me it didn’t work – the teachers who were perceived as high-quality teachers in the low-poverty school were not successful in the high-poverty school.
Although teachers may recognize how much student learning depends on resources outside of schools, it’s not something policymakers like to acknowledge. If value-added scores really measure teacher quality, they shouldn’t vary for individual teachers depending on changes in the teachers’ school or students. Yet, my study finds that scores went up for teachers who switched into a school with more wealthy students, and scores went down for teachers who switched into schools with more poor students. My study shows that these scores are biased by factors outside of teachers’ control.
This is consistent with findings from other studies but no previous studies have interrogated EVAAS scores — the most prevalent value-added methodology used across the country.
Why does all of this matter? Education is central for kids’ access to jobs, upward mobility, social development, and even mental health. It is important to understand what actually drives differences in student learning, and value-added scores misdirect attention to teachers. The pandemic has only made it more clear how society relies on schools not just for educating children but also for facilitating parents’ ability to work, and providing children with food and other social supports. The pandemic has also made it clear how hard teachers’ jobs actually are, as parents have scrambled to manage both their children’s learning and behavior. Ultimately, value-added scores perpetuate false understandings of the work of teachers and of the process of learning.
Sociological research has shown how powerfully kids’ social background and resources in their homes and neighborhoods shape their achievement. And then, with segregation, we cluster kids with the valued resources that will propel them through school and into powerful jobs into the same schools, compounding inequality. If we only emphasize teachers as the source of educational disparities, we will never resolve these inequalities. If we truly care about increasing access to learning we must address poverty and social inequality—pay living wages, ensure access to quality healthcare, and extend school services to explicitly address social issues even in early childhood. This also matters because teachers in the most challenging contexts are demonized and less likely to be retained at the schools where teachers are needed most, or in the teaching field more generally. Ideally, we integrate schools. Until then, teachers who work in high-poverty schools must be paid a higher salary.
Dara Shifrer.“Contextualizing Educational Disparities and the Evaluation of Teacher Quality” in Social Problems 2020. For a free, pre-publication version of the article, click here.
Image: Joe Brusky via Flickr (CC BY-NC 2.0)