TeacherMatch CEO and co-founder Don Fraynd has a favorite exercise he likes to put principals and school human-resource teams through: He passes out the same stack of 20 teacher r茅sum茅s, then asks each person in the room to identify two candidates to bring in for an interview.
鈥淣o matter where I do this, almost every single person gets at least one vote,鈥� said Fraynd, whose Chicago-based organization uses sophisticated software applications to help some 200 school districts find the best teachers. 鈥淚t鈥檚 pretty clear we鈥檙e all over the place when it comes to this decision.鈥�
The point, of course, as research has shown, is that personal biases, however unconscious, come into play in the hiring process. Experts like Fraynd, on the other hand, say that more can both help districts attract and secure top-notch candidates and more accurately predict whether a teacher will be effective in the classroom.
In addition to minimizing teacher turnover and absenteeism, both of which negatively affect student outcomes, better use of data has helped some districts streamline their hiring processes and hire candidates earlier.
鈥淭hat鈥檚 one of the biggest factors in effective hiring, to really maximize the pool when it鈥檚 at its deepest and widest,鈥� said Joseph Hettler, a partner at , a New York City-based nonprofit organization. TNTP helps districts use data to track historical vacancy trends in specific subjects and schools to take advantage of that ideal hiring time, which Hettler said is between March and May.
With TNTP鈥檚 assistance, the Shelby County district in Memphis, Tenn., sends an annual anonymous survey to teachers to ask if they plan to return the following school year. The data from this survey, along with information from other sources, help the district predict the number of vacancies likely to open, and in turn, inform recruitment and staffing strategies.Data-driven-recruitment advocates say they鈥檝e seen the value in switching to a more objective, fair, and research-backed process. In fact, 鈥済ut instinct鈥� hiring has become a forbidden phrase in at least one district.
鈥淲e鈥檙e not even allowed to mention it during the hiring process,鈥� said Dale R. Fisher, the assistant superintendent for human resources at the 3,000-student Deerfield Public 69传媒 District 109 in Deerfield, Ill.
Beyond 鈥楪ut Instinct鈥�
Using proprietary screening tools from AppliTrack and HumaneX Ventures, the district first looks for themes in candidates鈥� application materials and screening interviews that indicate a commitment to growth, both professionally and with students. Initial high scorers participate in a series of structured interviews, and based on those scores, get invited to a site-based interview. The district then spends the first two years evaluating whether each new hire鈥檚 performance matches expectations.
Using information pulled from a variety of assessments, including the Northwest Evaluation Association鈥檚 Measure of Academic Progress, the Deerfield district has preliminary data that 85 percent of the people it hired for 2014-15 had a positive impact on students.
鈥淲e鈥檙e to the point now where we have such faith in our hiring protocols that r茅sum茅s and letters of recommendation take a backseat to the online application,鈥� Fisher said.
The Cleveland school district started analyzing data four years ago to increase the diversity of its workforce. At the time, 85 percent of its 40,000 students鈥攂ut only one-quarter of its teachers鈥攊dentified themselves as a person of color. The Ohio district set a goal to bump up that teacher ratio to 30 percent of new hires; those in charge of recruitment met every two weeks to review data and check applicant-pool statistics.
鈥淭racking that information drove our strategy,鈥� said chief talent officer Lora Cover.The district surpassed its goal. Thirty-eight percent of the teachers hired for 2014-15 identified themselves as a person of color, upping the districtwide statistic to 34 percent.
Cleveland鈥檚 data also verified that early hiring is essential: Of the teachers hired through June 2014 for the 2014-15 school year, 14 percent were rated as 鈥渁ccomplished鈥� (the highest rating) using an evaluation that includes student-achievement data; only 2 percent of those hired in July and August shared that designation.
The district is now transitioning to the Workday Human Capital Management system, a cloud-based application suite that Cover says 鈥渨ill give us a much more robust set of data and a much easier way to pull it.鈥�
With just under 14,000 instructional staff members and 212,000 students, the massive Hillsborough County district in Tampa, Fla., started working with TNTP last fall to be more strategic about where to take recruiting trips to career fairs鈥攁nd avoid places that have traditionally turned out weak applicants. Summer hires alone account for 1,200 new employees in the district.
鈥淚 knew I was going to hire 600 elementary teachers this past year based on trend data through attrition, resignation, and leave of absence,鈥� said Dena Collins, the general manager of Hillsborough鈥檚 personnel services. 鈥淧redicting those numbers well in advance is the easy part.鈥�
Finding the Right Fit
The hard part, she said, is making sure candidates are well-suited鈥攇eographically and culturally鈥攆or the school environments in which they will be assigned.
In Tennessee, Sheila Redick can relate. 鈥淪omething we鈥檝e analyzed multiple times here is to see if there鈥檚 a profile of a candidate who tends to do better in an urban setting, and unfortunately, there isn鈥檛,鈥� said the director of talent management for the Shelby County district. Using the Tennessee Higher Education Commission鈥檚 annual Report Card on the Effectiveness of TeacherTraining Programs, the 100,000-student district has, however, discovered that Teach For America candidates tend to perform as well as or better in low-performing schools and high-needs subject areas than new hires from other sources鈥攁 finding that helps guide its recruitment strategy.
While believing that better data lead to better hiring decisions, Dale S. Rose, the president of consulting firm 3D Group, based in Emeryville, Calif., cautions against putting too much emphasis on 鈥渂ig data with blunt data points.鈥� The author of Hire Better Teachers Now: Using the Science of Selection to Find the Best Teachers for Your School, Rose said screening candidates 鈥渋s a contact sport. You鈥檝e got to get them in the classroom, talk to them.鈥�
Even Fraynd from TeacherMatch acknowledges that revolutionary software programs鈥攊ncluding those from his company, which adjust weighting and other factors to make them more predictive over time for particular locales鈥攃an only go so far.
鈥淭he big data can get you that crop of people who are statistically going to move the needle,鈥� he said, 鈥渂ut you still need to know if those folks are going to be a good fit.鈥�
