A recent CALDER study examines the effects that earlier-grade teachers have on students’ eighth-grade math outcomes by analyzing Washington State administrative data.
The researchers followed nine cohorts of children who started fourth grade between the 2006–07 and 2014–2015 school years. The study focused on students whose data contained their third-grade standardized math test achievement, who could be tracked through eighth grade, who had background data including variables such as learning disability status and free-and-reduced-price-lunch eligibility, and who could be matched with one math teacher in each grade and school year of their education. With these constraints, the sample contained 173,858 students and 15,981 teachers.
Using a value-added model, the researchers ran two analyses. First, they measured the impact a student’s fourth-grade teacher has on their eighth-grade math achievement—and how that impact compares to that of their eighth-grade teacher. And second, they measured the impact on eighth-grade math achievement of replacing teachers in grades four through eight—so also grades five, six, and seven—whose effectiveness is above the 95th percentile or below the 5th percentile with teachers of median effectiveness. How much, in other words, does losing the best teachers reduce achievement? And how much does removing the worst teachers improve it?
There were three noteworthy findings. First, fourth-grade teachers do impact their former students’ eighth-grade math scores, but not as much as those students’ eighth-grade teachers—0.16 standard deviations versus 0.36. Second, replacing the worst-performing 5 percent of teachers in grades four through eight with median performing teachers raises overall students’ eighth-grade math scores by a whopping 2.5 standard deviations (mirroring past research on the topic). Third, both retaining good fourth- and eighth- grade teachers and replacing bad ones had positive impacts on eight-grade math achievement—but the magnitude of those impacts depended on the grade. Retaining a fourth-grade teacher above the 95th percentile of the value-added distribution, instead of replacing her with one of median effectiveness, had a larger impact on increasing later math achievement, 0.46 standard deviations, than replacing a teacher below the 5th percentile, 0.30 standard deviations. But in eighth grade, the results were opposite, with the replacement of ineffective teachers causing more impact than retaining the good—0.97 versus 0.57 standard deviations.
The policy implications are clear: Teachers matter—and their impacts, be they good or bad, last for years. Leaders should therefore strive to gauge teacher effectiveness in reliable ways, then strive to retain the best teachers and remove the worst. All of that is, of course, easier said than done. But when successful, the efforts would likely bear fruit.
SOURCE: Dan Goldhaber, Zeyu Jin, and Richard Startz “How Much Do Early Teachers Matter?,” National Center for Analysis of Longitudinal Data in Education Research (2022).