More Than the Next Big Thing

Predictive analytics in teacher hiring.

GUEST COLUMN | by Donald J. Fraynd

CREDIT TeacherMatchWith hundreds of resumes for a handful of teaching positions, school district hiring managers face an immense challenge. Getting the right teacher in students’ classrooms for the first day of school is the most important job of school and district leadership. Yet teacher selection processes in schools consistently use informal or personal preferences to screen and identify the top teachers who will generate massive student learning. Objective assessments of the factors that predict teacher performance may improve the quality of teaching and reduce the burden for hiring managers.

Education and improving student achievement is a much more complicated mission than using predictive analytics to hire call-center employees at Xerox or fill positions at Google.

When many of us think of predictive analytics, we visualize data and graphs that sum up past metrics to gain insight on how an organization should perform in the future. However, predictive analytics has advanced and is much more than a generic pattern of data. It is being used to solve complicated problems, make decisions and identify opportunities. One major opportunity is the use of predictive analytics in the hiring process.

Within the hiring process, predictive analytics help identify top talent by connecting candidate data to a set of key measurable factors. As a result, organizations are able to process and manage candidates coming through the job-posting pipeline, select candidates that are a best fit, and hire individuals that will lead their organization to success. According to a Wall Street Journal article, “For more and more companies, the hiring boss is an algorithm.”

Who is using predictive analytics in hiring?

Top organizations are increasingly using data to improve hiring practices. Xerox Corporation had previously relied on candidate experience, only hiring applicants who had done the job before. Yet data showed that new hires quit Xerox before the company recovered its $5,000 investment per employee in call-center training. The company’s hiring managers realized their hiring techniques were based on untested assumptions, therefore, Xerox invested in predictive analytics to evaluate characteristics and skillsets of top, call-center employees. Today, Xerox uses talent data to hire candidates in all of its 48,000 call-center jobs.

Google also eliminated components of its hiring process based on data. The brainteaser component of the interview process had little correlation with success of the overall organization. So they removed it. Google’s interview methods have become much more data-driven and the company uses hiring tools to help identify candidates that have the ‘Googliness’ they’re looking for as well as drive business growth. Additionally, candidate data helps them hire talented individuals faster – an important factor since speed is essential for Google when hiring recent graduates. 

How does data-driven hiring relate to education?

The use of predictive analytics by corporate giants, such as Xerox and Google, has clear corollaries to education. Like Xerox, the traditional hiring process for teacher positions uses a candidate’s work history, credentials and in-person interview. However, education and improving student achievement is a much more complicated mission than using predictive analytics to hire call-center employees at Xerox or fill positions at Google. Therefore, while some high performing districts require demonstration lessons, teacher impact on student achievement cannot be directly measured during the traditional hiring process. So school districts have begun to identify indicators that are predictive of teacher performance for use in the hiring process.

TeacherMatch EPI (Educators Professional Inventory) is an instrument designed exactly for the purpose of informing hiring. Districts using it trust the EPI to predict the impact teacher candidates will have on student achievement through four core success indicators: teaching skills, cognitive ability, attitudinal factors, and qualifications.

Within the EPI, teaching skills analyze success planning attributes, ability to create a learning environment, and a candidate’s analyzing and adjusting characteristics. Cognitive ability addresses candidate’s awareness and perception by evaluating analytical reasoning and problem solving skills. Attitudinal factors looks at teacher candidate’s motivation to succeed and maintain a positive attitude, and lastly, qualifications considers candidates’ education background and professional fieldwork.

Consequently, school districts adopting predictive analytics, such as TeacherMatch EPI, are improving student achievement by identifying teachers that are the strongest candidates from day one.

Who is behind the research?

TeacherMatch EPI came from years of internal research as well as professionals that have deep experience and understanding of the industry. It was developed alongside the Northwest Evaluation Association’s research specialists and psychometricians, and researchers from The University of Chicago and the Value-Added Research Center of University of Wisconsin-Madison. Using real teachers and value-added model scores, a TeacherMatch validation study found that the EPI predicted student learning. Teachers with higher scores on the EPI were also teachers whose students learned more (controlling for demographics and initial proficiency).

Hiring any individual in any organization is an important decision, especially in the education environment. Therefore, as school districts integrate predictive analytics into their hiring processes, the validity of the analytics are critical. This particular solution has been validated and one prediction is clear: school districts’ will have the necessary data to identify a quality teacher that will improve student achievement.


Halzack, Sarah (2013, September 4). An inside look at Google’s data-driven job interview process. The Washington Post.

RAND’s Center for the Study of the Teaching Profession. (1987). Effective Teacher Selection From Recruitment to Retention. Santa Monica, California: Wise, Arthur E., Darling-Hammond, Linda, Berry, Barnett, Berliner, David, Haller, Emil, Praskac, Amy, & Schlechty, Phillip

Walker, Joseph (2012, September 20). Meet the New Boss: Big Data. The Wall Street Journal.

Donald J. Fraynd, Ph.D., is CEO of TeacherMatch, a data-driven, people-powered formula for success for K-12 education talent management. As a principal in Chicago Public Schools, his school was rated in the top 100 by US News & World Report and was the first Blue Ribbon School ever for a CPS high school. He is part of a team that spearheaded the design and implementation of a comprehensive hiring and professional development plan involving thousands of teachers and used by the US Department of Education to shape their multi-billion dollar school improvement program. Contact him through TeacherMatch.

  • Mark Gura


    I would love to read more about the actual predictors and not just about the concept of using them!

Leave a Comment

%d bloggers like this: