A trio of Washington researchers wondered if the mentors who helped student teachers learn the ropes made a difference in student outcomes. They found it does, especially in math.

Share story

About three years ago, Kim Harmon started overseeing the aspiring, still-in-school educators who teach in Spokane Public Schools as a part of their training.

Student teachers work alongside district teachers to run a classroom. The idea is that they can learn more from engaging in the daily work that so depends on the unpredictable nature of kids than they can from solely sitting in a theory-heavy symposium.

The status quo, as Harmon recalled it, involved splitting the management of the student-teaching program among multiple human resources workers on top of their regular duties. Neighboring colleges and universities would call principals directly, asking if they had any experienced teachers who could take a hopeful graduate under their wing.

Education Lab is a Seattle Times project that spotlights promising approaches to persistent challenges in public education. It is produced in partnership with the Solutions Journalism Network and is funded by a grant from the Bill & Melinda Gates Foundation and City University of Seattle. Learn more about Ed Lab 

“We found that was, well, a little cumbersome,” said Harmon, director of recruitment and retention for the state’s second-largest school district.

So Spokane shifted gears, and tasked a single central administrator with all placements of student teachers each school year. That administrator also works closely with the host teachers, making sure they know how to help the novice educators improve their instruction and lesson planning.

Harmon said it’s still too early to tell what impact the overhaul has had on the K-12 students taught by newbies. But a new study that looks at student teaching across Washington suggests colleges of education and districts like Spokane could partner more deliberately to improve student performance the moment a first-year teacher takes the reins of a classroom.

We spoke to two researchers with the University of Washington-based Center for Education Data & Research (CEDR) about their findings, and asked Harmon what they mean for her teachers-in-training.

(The Bill & Melinda Gates Foundation, which funds Education Lab through a grant, financially supported the CEDR study.)

What the data tell us

In a working paper released this past week, researchers Dan Goldhaber, John Krieg and Roddy Theobold examined nine years of student teaching data, focusing on whether soon-to-be teachers were paired with a highly effective, average or very ineffective mentor. They define “effective” based on a measure, known as value added, that compares how much or how little progress each teacher’s students made on standardized tests when compared to expectations for similar students.

The study also followed the student teachers once they formally entered the profession, and gauged whether training under an effective or ineffective mentor made any difference during the new educators’ first six years of teaching.

Their main conclusion: Beginning teachers who worked with a highly effective mentor start their first year with students posting test-score gains similar to that of an average, three-year teacher.

“Put another way,” said Goldhaber, director of CEDR, “if you work with a top-flight (mentor) teacher right out of the gate, you tend to be better than average.”

It’s an advantage that might make schools or parents less skeptical of rookie teachers — when they’re paired with the right coaches.

The opposite, Goldhaber said, is also true. “If you work with a real ineffective mentor, it takes you a full six years before you even hit average.”

The authors emphasized that the above-average test-score gains for teachers who worked with a highly effective mentor appeared only in math. In English language arts, some positive growth was measured but wasn’t strong enough to be considered statistically significant.

What the data don’t tell us

Goldhaber and Theobold acknowledged the possibility that, in the data, students teachers already poised for success in the classroom specifically searched for and worked with highly effective mentors.

The authors tested their findings against that concern — and still found a benefit translated to K-12 students’ math test scores.

Harmon said she plans to bring the CEDR research to the attention of a new regional network of school districts and teacher education programs in Eastern Washington.

But, she said, the numbers don’t tell the full story. “I’m hesitant to base a match,” she added, “just on test scores alone because there’s so much more involved.”

What questions remain

Curiously, the new study found that the positive impact in math decays over time, meaning the relationship between a mentor’s effectiveness and future teacher’s effectiveness is strongest in the first year of teaching and gradually flattens by the sixth.

In Spokane, the district pairs all first-year teachers with a coach who each week observes their class and offers feedback on how to improve. Other districts, Harmon noted, offer similar coaches over the second and third year of a beginning teacher’s career.

She wondered if losing that intensive support explains some of the decay in the study’s findings.

As for future research, Theobold said the team is already investigating whether student teachers who never formally entered the profession went on to work in the tech field or as baristas. Connecting the existing set of data with college transcripts also could reveal any relationship between how many math courses a teacher-to-be takes in college and their effectiveness as a math teacher.

“Part of the story is the existence of the data at all,” Theobold said. “This isn’t happening anywhere else. Teacher education programs aren’t volunteering their data for research anywhere in the country, and that’s a big deal.”