You are here

News Update

Can analytics predict the best teachers?

Programs designed to gauge impact teacher candidates will have on student test scores
Predictive analytics is becoming more common in both public- and private-sector hiring.
Predictive analytics is becoming more common in both public- and private-sector hiring.

Big data and analytics now offer districts some clues about which teacher candidates will be the most effective in the classroom.

These programs are designed to accurately gauge the impact teacher candidates will have on student test scores. Analytics companies such as TeacherMatch and Hanover Research are working with hundreds of districts nationwide to aid in the hiring process.

Though still a relatively new practice, predictive analytics is becoming more common in both public- and private-sector hiring, says Jonah Rockoff, a Columbia University associate professor of finance and economics who studies teacher hiring systems.

“People have realized that subjective opinions from personal interviews and other common hiring practices tend to lead to less accurate judgments of who is going to be effective on the job,” Rockoff says. “With the resurgence on the focus on teacher effectiveness and evaluation, hiring is a really important lever that school principals have for improving the quality of the staff.”

Research needed

There is not yet much independent research on the effectiveness of analytics programs, Rockoff says.

But many districts report anecdotal success. San Marcos Consolidated ISD, a suburban district of 7,800 students located between Austin and San Antonio, usually receives 200 to 400 applicants for every one teaching position posted, says Lolly Guerra, assistant superintendent of human resources.

Administrators began using TeacherMatch in March 2014. A 100-question survey that candidates must complete with their job application generates the analytics. The questions cover broad areas that impact teacher effectiveness, including:

• Qualifications, such as the selectivity of the candidate’s teacher prep program • Attitude, such as how the candidate handles challenges • Basic subject knowledge • Teaching strategies, and how the candidate would respond to specific classroom situations

The analytics tools then devise a score that estimates how effective the teacher will be.

Applicants who do not have the adequate skills can be quickly eliminated. As for new hires, administrators are pleased with what they are seeing on evaluations and in walk-throughs, Guerra says. The new teachers are succeeding in the district’s push for project-based learning, and students seem engaged, she adds.

The service is priced per student, and San Marcos pays about $24,000 per year.

“It’s another tool to make hiring decisions,” Guerra says. “An interview is important to make sure they are going to be a fit for your district. But this is quantitative data that is predictive of a teacher’s ability to teach.”

Improving hiring

In the end, teacher hiring should be based on a combination of factors, including analytics, an interview and a mock-lesson, Rockoff says. Most principals do not require that candidates teach a lesson in the hiring process, but it is a good way to tell how the candidate will perform in the classroom, he adds.

“It’s much easier to improve the quality of the teaching staff on the hiring side than it is once you already have somebody in the building,” Rockoff says. “The worst mistake a principal can make is hiring an ineffective teacher and exposing a classroom of students to someone who does a bad job.”