The Classroom Teacher, winter 2013-14

Get ready, because under a new deal Texas made with the federal government, Texas must develop a new teacher evaluation system that incorporates students’ STAAR performance. The NCLB waiver the U.S. Department of Education granted Texas this fall — the one that allows Texas school districts to avoid federal Adequate Yearly Progress requirements — came with strings attached. Those strings will tie students’ standardized test performance to teacher evaluations.

To keep the waiver past 2013-14, the Texas Education Agency must submit final guidelines for teacher and principal evaluation and support systems to the U.S. Department of Education by May 2, 2014. These guidelines must include “the use of student growth, as defined in ESEA Flexibility, as a significant factor” in teacher and principal evaluations. ESEA Flexibility, the set of principles by which the USDE grants waivers, says student growth is measured by “annual, statewide, aligned, high-quality assessments.”

This is not welcome news to teachers because the use of student performance on standardized assessments to evaluate teacher effectiveness is not valid for a number of reasons. Many factors outside a teacher’s control affect how well students perform on standardized tests.

Poverty is perhaps the biggest factor in a state like Texas, where 60 percent of students come from economically disadvantaged homes. Studies show that students who come to school hungry do not perform as well as their peers academically. There is also research that shows student tests were simply not designed to measure teacher performance. And then there’s the fact that test data doesn’t even exist for many teachers who teach non-tested grade levels or subjects.

Up to now, Texas has chosen to remain among the few states that have refused to adopt teacher evaluation systems that rely on the ill-considered and unreliable linkage between student test scores and teacher performance. But the acceptance of the federal waiver is now taking us down that path.

Commissioner of Education Michael Williams — who negotiated the waiver with the USDE and, in the end, committed Texas to including student achievement growth as a factor in teacher evaluation — has said publicly that school district participation in any new state-developed system would be voluntary. However, the final waiver request specifies that if a district does not implement the state model or a system with the same components as the state model, TEA will require districts to implement the state model.

Further complicating the matter is that Texas state law (which likely couldn’t be changed until the next Legislature meets in 2015 — the same year the new teacher evaluation system is supposed to launch) does not currently require locally developed teacher appraisal systems to include the same components as those in the current state system. So there is the question of whether TEA even has the authority to require districts to adopt a new teacher evaluation system, unless state law is changed in 2015. Can Texas state legislators be persuaded to review the research and refuse to adopt this faulty metric for judging teacher quality?

How we got here

We can’t predict if the 2015 Legislature will change the law to mirror the waiver’s provision so that student performance on state tests becomes a significant measure in individual teacher evaluations. And we don’t yet know if state lawmakers will require districts to use the state system or a locally developed system with this same component. We do know that actions by state leaders in the past few years demonstrate Texas’ previous resistance to tying student test performance to teacher evaluations.

The 2013 Texas Legislature rejected a bill that would have required teacher evaluation to be based on student test scores. (The same lawmakers said “no” to a bill that would have ceded authority for development of the framework for a state teacher evaluation system from the Legislature to the commissioner.) Texas also has chosen not to pursue funds from the federal Race to the Top Program, which also requires that student performance on state tests be a significant measure in teacher evaluation systems.

Commissioner Williams expressed confidence in the state’s own system in the original waiver request he submitted to the USDE in early 2013. He not only asked for flexibility from AYP but also from federal “highly qualified” teacher requirements. The request asked that Texas instead be allowed to rely solely on the state’s rigorous teacher certification standards, supported by Texas’ own teacher evaluation system.

TEA invited public comment on this initial draft of the NCLB waiver, and TCTA submitted comments in support of the request as a whole. At the time, in keeping with Texas’ prior rejection of Race to the Top, TEA emphasized that Texas was not seeking a conditional waiver, but rather one under the secretary of education’s general waiver authority.

After months of negotiation, however, the final waiver request submitted in September 2013 was changed significantly from the original draft without opportunity for stakeholder input. One likely reason for the change is that Texas districts and superintendents may have pressured TEA to make concessions to secure the waiver.

School administrators wanted to avoid the consequences of failing to meet AYP, which start with having to offer transfer options to parents and gradually escalate to school restructuring. Without a waiver, all but a small percentage of districts would fail to meet 2014 AYP standards, which require 100 percent of students to be proficient on the state’s reading and math assessments.

With the waiver as granted, AYP is set aside, and only the lowest performing 15 percent of schools will face any consequences. But, as a result of the conditions placed on the waiver, Texas teachers will pay the price of having their evaluations tied to student test scores.

Across the nation

One reason that Texas’ waiver came with conditions (and incidentally, no relief from the “highly qualified” requirements) is that the USDE was in a position to rigidly enforce its rules for granting such waivers. After getting 40 other states and the District of Columbia to agree to similar conditions to secure NCLB waivers, the USDE simply didn’t give Texas’ waiver request special consideration.

According to Holly Eaton, TCTA director of professional development and advocacy, the conditions USDE places on NCLB waivers are the department’s way of pushing the current administration’s education reform agenda, which includes test-based teacher evaluation systems.

"Tying student test performance to teacher evaluations was one of the ‘strings attached’ requirements of many of the USDE’s previous education initiatives, including the Race to the Top Program, which Texas declined to pursue,” she says. “TCTA supported the state’s decision not to participate in that program because of the strings that were attached related to teacher evaluation.”

According to the Center for Public Education’s October 2013 report, “Trends in Teacher Evaluation,” Race to the Top and NCLB waivers have required states to redesign their teacher evaluation systems in response to the advent of new tools such as value-added models (VAMs), which allegedly isolate the teacher’s effect on student performance using statistical methods. The report claims these tools are more accurate measures of teacher and school performance than previous evaluations, which “identified nearly all teachers simply as satisfactory or not.”

According to the report, 38 states now require (and eight recommend) that the measures used to evaluate teacher performance include teachers’ impact on student achievement. A few more require or recommend “multiple measures” of teacher performance. However, Eaton says, “If those measures aren’t widely available, ‘multiple’ is an illusion, and state tests will be the default.”

Other measures often used to evaluate teachers, in addition to student performance measures, include classroom observations, student surveys, lesson plan reviews and teacher self-assessments. How the various measures are weighted varies, but according to the report, no state requires or recommends that student achievement indicators comprise more than 50 percent of a teacher’s evaluation, and no state evaluates teachers on test scores alone.

Research questions validity

Even proponents of making student achievement a measure in teacher evaluation say such a measure should not be used alone in teacher evaluation. The CPE report describes VAMs as “not perfect,” admits they “don’t provide teachers feedback on how to improve their performance,” and states that “all experts agree they should be used in concert with other measures of teacher quality.”

“Perfect? What happened to valid?” asks Eaton, who points to research that shows VAM results to be unstable over time and subject to bias and imprecision. A 2010 study by the USDE’s own National Center for Education Evaluation and Regional Assistance says that with VAMs, incidents of misclassification of teacher performance are as high as 25 percent, meaning one in four teachers is incorrectly rated.

Other evidence indicates the margin for error is much wider. For example, New York City admitted in 2012 that its complex formula had a 35 percent error rate in classifying math teachers’ performance and a 53 percent error rate for language arts teachers.

Yet, despite these types of findings, which overwhelmingly show that methods used to tie students’ standardized test scores to teacher proficiency are not reliable or valid, the idea that a substantial part of a teacher’s appraisal should depend on someone else’s test scores continues to be pushed.

A few states have experienced push back, however. Last year, for example, a coalition of researchers, educators and teacher evaluation reform advocates wrote an open letter to the Georgia governor that expressed alarm about the state’s new VAM teacher evaluation system. They cited research that supported their recommendation that the state return federal monies related to the project and opt out of Race to the Top as nine other states already had.

If we are to see any real push back in Texas at this point, it must come from state legislators when they reconvene in 2015. In the meantime, the development of Texas’ new teacher evaluation system begins immediately.

Timeline for implementation

The timeline for development of the new system, as specified in TEA’s final waiver request, includes the continued development of evaluation tools already being piloted in 2013-14. These include new observation rubrics, campus climate surveys and value-added measures. (For the past three years, TEA has worked with outside contractors to develop both a campus-wide and individual teacher value-add metric. The agency says it will continue to refine the model and gauge the appropriateness of its use in the evaluation system.)

The new system itself will be piloted in 40 school districts in 2014-15. Then, after updating state rules to reflect a new system that includes all of the components, statewide rollout of the new evaluation system begins in 2015-16. (See a detailed timeline posted by TEA.)

TEA’s final waiver request states that the agency “will engage superintendents, the Texas Association of School Boards (TASB), the Texas Association of School Administrators (TASA), and teacher organizations” as it implements the terms agreed to in the waiver.

The agency will use its stakeholder input process to determine which of the following methods is the most appropriate method for making student growth a significant measure in the new teacher evaluation system:

  • minimum percentage weighting of 20 percent based on statewide assessments in tested grades and subjects (other measures can be added on top of the 20 percent)
  • student growth matrix based on statewide assessments in tested grades and subjects (other measures may be added in addition to the matrix)
    • Possible methods for measuring the change in student achievement for an individual student between two or more points in time include: the use of performance on state assessments at the campus and individual teacher level, team or individual student-learning objectives, and performance on district-based assessments.
  • the trigger method whereby teachers and principals who do not achieve a minimum student growth amount cannot be rated as “effective” or higher and for tested grades and subjects (the minimum student growth measure must be based on statewide assessments)

TCTA will provide input on behalf of our members at every opportunity as we monitor the development of the new system. TCTA’s participation on the Teacher Standards Steering Committee, which is charged with developing new statewide, foundational teaching standards and an observation rubric, will likely give us an additional opportunity to provide input. Read future TCTA communications, especially our eUpdate newsletter, for the latest news on the new teacher evaluation guidelines.

More facts on the new evaluation system

Purpose

The new teacher evaluation system will be used for “continual improvement, providing timely and useful feedback and to inform personnel decisions.”

Components
  • Classroom observations and feedback
  • Student growth and learning
  • Professional engagement and growth
Appraisal frequency

“Districts shall perform annual or more frequent evaluations with no more than three years in between.” (Due to TCTA-initiated legislation, current state law allows districts to adopt policies allowing teachers whose most recent evaluations were proficient to be exempt from annual appraisals for up to five years.)

Source: Final NCLB waiver request submitted to USDE by TEA

See also Texas surprises nation by accepting conditional NCLB waiver