Just ask the kids: Student surveys and teacher evaluations
By July 1st, Ohio law will require public school districts (charter and district) to establish a teacher evaluation policy. The evaluation policy must conform to a framework that depends half on student growth on test scores and half on classroom observations. Based on these measures, teachers will earn an overall rating: accomplished, proficient, developing, or ineffective.
In our recent survey of superintendents, Ohio’s teacher evaluation policy received mixed reviews. Nearly three out of four (73 percent) said that teacher evaluations would become accepted practice five years hence. And, 42 percent said that teacher evaluations would lead to “fundamental improvement” in the state’s K-12 school system. So, there’s modest optimism toward teacher evaluation.
But there’s undeniable angst about the policy details. Nearly all superintendents (93 percent) think that there will be lawsuits when personnel decisions are based on Ohio’s evaluation framework. And nearly all (86 percent) think that the classroom observation mandate will “put too much pressure on principals.” One superintendent said
“It will over-tax the principals and render them useless. They will need to spend so much time on evaluations, they will not have time for anything else.”
When one looks at the Ohio Department of Education’s website, one can see from whence this sentiment emerges. For example, the “teacher evaluation resource packet,” which operationalizes the classroom observation portion of the policy, clocks in at 22 pages. By simple extrapolation, this suggests a small mountain of paperwork for a principal who supervises 20 teachers.
Is there a better way?
Enter the Measures of Effective Teaching (MET) study, a three-year project sponsored by the Gates Foundation involving 3,000 teachers in 7 school districts. The MET researchers probed whether student perception surveys could be an effective and reliable tool for measuring teacher effectiveness. The study used the Tripod Project survey, developed by Ronald Ferguson of Harvard, which asks students to rate their agreement on 35 or so items related to seven characteristics of an effective teacher. (For the survey items, see pages 12 and 13 of this MET report.)
The researchers found that surveys, if done well, can deliver reliable information that differentiates high- and low-performing teachers. The study finds that:
“Although an individual student may have a less sophisticated understanding of effective instruction than a trained observer, student feedback has two other advantages that contribute to reliability: students see the teacher all year (and, therefore, are less susceptible to lesson to lesson variation), and the measures are averaged over 20 to 75 students, rather than 1 or 2 observers (p. 14).”
The MET study concludes succinctly, “students know an effective classroom when they experience one.”
Despite this robust research, some may still think that surveying students in high-stakes evaluations is akin to the inmates running the asylum. One Georgia high school teacher claimed that “kids, number one, don’t have the maturity to do it and, number two, can be quite biased.” The MET study proves that this fear is unfounded. Plus, this isn’t to suggest that teacher evaluations ought to rely wholly on student surveys. In fact, according to the MET researchers, policy makers mustn’t exclude either value-added growth or classroom observation from teacher evaluation. Multiple measures are required.
Many would agree that having an effective teacher in every classroom is a fair and legitimate goal. And, identifying which teachers are effective and which are not—and tying personnel decisions to effectiveness—must happen, in order to achieve this goal.
Measuring teacher effectiveness, however, need not create procedural nightmares for school leaders, lest the bureaucratic costs eat into the benefits of great teaching. Other states, including North Carolina and New York, are implementing student surveys, and Ohio’s policymakers ought to consider mixing surveys into the evaluation framework. If done well, student surveys may be the low-cost and effective relief to the migraines that too many of our school leaders seem to be experiencing per teacher evaluations.
 The Ohio Senate passed an amendment to ORC section 3319.112, the law that governs teacher evaluations. The amdendment reads (underlined): "One factor shall be student academic growth which shall account for fifty thirty-fiveper cent of each evaluation. A school district may attribute an additional percentage to the academic growth factor, not to exceed fifteen per cent of each evaluation."