Delving Deeper Into the Teacher Effectiveness Gap
Here’s something you already may know: Mathematica Policy Research’s two, large-scale studies on the distribution of effective teachers across schools have made quite a splash in the research and policy worlds. So why yet another blog post on them? Because most of us need help translating meticulous research so we can use it in the field!
That’s why I asked some of the authors of Access to Effective Teaching for Disadvantaged Students—Eric Isenberg, Philip Gleason, and Jeffrey Max from Mathematica, and Michael Hansen from American Institutes for Research—what the results mean, and what we still don’t know.
First, the study findings. On average, in 29 mostly large school districts across the country, with poverty ranging from 34 percent to 78 percent, researchers found that (1) poor students in Grades 4–8 were taught by less effective math and English teachers than students who aren’t poor, (2) shrinking that teacher effectiveness gap would reduce the achievement gap by 2 percentile points, and (3) some districts are doing a much better job than others of maintaining equitable distribution of effective teachers. Teacher effectiveness was measured using value-added scores.
According to the research, the typical poor student has a teacher at the 47th percentile of the distribution of teachers in English/language arts, compared to the typical non-poor student who has a teacher at the 56th percentile. In math, these differences are between a teacher at the 48th percentile for a typical poor student and the 53rd percentile for the typical non-poor student.
Statistically significant, yes—but a large difference? The authors say no. “The distribution of measured effectiveness is bell-shaped,” they explained in a joint response, “so the difference between percentiles in the thick middle of the distribution is not that much.” Eliminating teacher effectiveness differences like these, Michael Hansen says, would add roughly two additional weeks of learning for poor students.
Still, these differences are for a single year only. So, the authors explained, “We do not know how the one-year difference in access to effective teaching would aggregate over time. In other words, we looked at the effect on the student achievement gap that would result from having equal access in a single year but we have not yet examined the effect of having equal access over multiple years.”
What else don’t we know? Why there was such variation in how equitably those effective teachers are distributed within different districts, what explains those differences, and how we can help make effective teacher distribution equitable in more districts. Later reports that cover two more years of data collection will examine these questions.
Although we don’t have answers yet, the authors noted that in English/language arts, the more urban the district, the more unequal its access to effective teaching. And in math, access is greater in medium-sized districts than in larger districts. By region, southern districts are the most inequitable.
Bottom line, we still need to better understand why some districts are struggling more than others to ensure that all of their students get access to the most effective teachers.
In districts with inequitable distribution, why do poor students get assigned to less effective teachers? Are effective teachers leaving? Are ineffective teachers staying? Are less experienced teachers being assigned to poor students? Closer scrutiny of these districts should inform policymaking.
Still on the edge of your seat? You can get up-close and personal with the researchers at an IES/Mathematica forum from 11 a.m. to 12:30 p.m. ET, on Tuesday, December 10, 2013, at Mathematica’s D.C. office or via live webinar.
So tell us, did this blog help unpack the study’s findings? What questions do you still have that would inform your work as a practitioner, researcher, or policymaker?