It’s Time to Move on from Using Student Growth in Evaluating Teachers
Ivan Silverberg teaches his American Studies class to eighth and ninth graders at the Niles North High School in Skokie, Ill., Tuesday, Aug. 20, 2013. That was the first school year that 34 Illinois school districts beyond Chicago began grading teachers on whether their students' test scores are improving, but the state's slow method of implementing the new evaluations was creating problems. (AP Photo/Scott Eisen)
Using test scores to evaluate teachers came into vogue with Race to the Top. This program drove this policy with the simple requirement of linking test scores to personnel decisions. The idea was that school systems could be held accountable for hiring and firing based on an objective measure of effectiveness. The policy also had a scientific pedigree from the research using test score-based value-added modeling to measure the teacher’s contribution to student growth.
Largely as a result of U.S. Education Department policies, there is widespread acceptance of basing personnel decisions on meaningful evaluations, and 38 states now include measures of student growth in teacher evaluations.
Unfortunately, using test scores to evaluate teachers was never practical because only a small portion of teachers instruct math and reading in grades three to eight where standardized tests are mandated. It appears that the excitement about research on VAMs led policymakers to overlook this glaring gap. It has led to unintended consequences.
Where a measure of student growth is mandated, something has to fill the gap. A widespread solution is a process called Student Learning Objectives. The problem is that SLOs are designed to vary by grade and subject and must be personalized for each teacher. This personalization prevents a systematic use for evaluation. However, since no other solution is affordable or practical, ED and state agencies have accepted SLOs as compliant.
Advocates point to SLOs as a useful process for reflection and goal setting between the principal and teacher, and for this function, they may continue to have value. Unfortunately, the process requires an enormous time investment, and the lack of evidence linking SLOs with improved student growth means that there’s no clear ROI.
In addition to the impracticality of using student growth as a measure of teacher effectiveness, there are several other weaknesses:
- A measure of the teacher’s contribution to student growth doesn’t indicate how to improve. We are seeing an increased orientation toward using evaluations to inform professional development, not just terminations and pay increases. Instead, classroom observations—conducted by a mentor or administrator—combined with feedback and guidance directly maps out where the teacher can improve.
- Test scores are associated with a narrow range of classroom practice. My colleague, Val Lazarev, and I found an interesting result from a reanalysis of the Gates Foundation’s Measures of Effective Teaching project data. A teacher’s VAM score is mostly associated with keeping order in the classroom, which is essential for teaching but does not capture other important elements.
- Student growth measures don’t capture a teacher’s longer-term influence after their students take the spring test. A teacher who gets a student to start seeing math in a new way may inspire the student to enroll in a more challenging course the next year. A teacher who makes a student feel at home in class may help prevent the student from dropping out two years later.
- The focus on student growth has strengthened opposition to testing. The introduction of the new Common Core-aligned tests could have been welcomed by the teaching profession as a stronger alignment with widely shared beliefs about what students should learn. Instead, it was seen as unfair to evaluate teachers using a test they had no experience preparing students for.
It is time for ED and states to loosen up on the student growth requirement. By allowing SLOs, which do not provide an evaluation, ED and many states have essentially conceded that an evaluation based on student growth isn’t necessary. There’s no sense in substituting make-work in the form of SLOs just to assure technical compliance.
Redirecting the effort, currently expended on SLOs, into a well-developed alternative—such as classroom observations—would allow the principal to put the effort into more productive use. Observations can evaluate the characteristics of teaching effectively for short-term growth, as well as detect warm, engaging, and inspiring teaching. A supervisor or mentor in the classroom using a framework shared with the teacher is practical and addresses the above four points. Districts and states have shown that observation scores can be calibrated to a level necessary to be useful as a system-wide evaluation. There are issues of potential bias to address, and the validity of particular observation frameworks should be tested against benchmarks of college or career success.
Now that meaningful evaluation of teachers is an established policy, it’s time to sever the link to student testing. Let testing play its role in ensuring that schools are maintaining standards, and most importantly, in considering the reauthorization of ESEA, not failing to educate the students who have the most need.