Texas Commissioner of Education Michael Williams entered into an agreement with the U.S. Department of Education in fall 2013 in which, in exchange for a waiver from federal AYP and other requirements of the NCLB/ESEA, he committed to instituting new statewide teacher and principal evaluation and support systems that include the use of student growth on state standardized tests as a significant factor (i.e., at least 20 percent). The agreement stated that the new systems must be used to determine Texas teacher and principal summative evaluation ratings by the 2015-16 school year.

TCTA believes this commitment is highly problematic and likely to be challenged for the following reasons:

  1. Neither the federal government’s requirements for teacher evaluation systems nor the conditions agreed to by the commissioner in the Texas NCLB waiver agreement were the subject of legislation or rule, thereby they circumvented the normal process by which major policy changes are developed.
  2. The final waiver request that included the teacher evaluation requirements was not posted for public comment, and there was no opportunity for broad-based stakeholder input into the final waiver request.
  3. The commissioner committed to requirements not allowable under current state law: requiring student growth on state tests to be included in individual teachers’ evaluations; requiring districts using locally developed systems to include student growth on state tests for individual teachers’ evaluations; and requiring annual evaluation of all teachers.
  4. These commitments may jeopardize Texas' NCLB waiver. Other states’ NCLB waivers have been put on high-risk status because they didn’t meet federal dictates related to teacher evaluation. Washington’s waiver was recently revoked because its Legislature declined to change state law to allow the use of student state standardized test scores in teacher evaluation. Although it is unclear what the Texas Legislature will do, in previous sessions, legislators have declined to pass bills that required a teacher evaluation system similar to the one outlined by the Texas NCLB waiver.
  5. The implementation timelines are unrealistic and would not provide sufficient time to train and certify evaluators for statewide implementation. The new system was originally scheduled to be piloted in 2014-15 and fully implemented in 2015-16, despite the fact that TEA acknowledged that value-added data from the pilot would not be available until September 2015. Although the system may now be piloted a second year in 2015-16 and implemented in 2016-17, the next Legislature will meet in early 2015, so legislators will still be in the position of having to consider a change to state law before they see final pilot data.
  6. The teacher evaluation requirements in Texas’ NCLB waiver agreement expose the state and local districts to potential lawsuits. By requiring districts that use local evaluation systems to base at least 20 percent of a teacher’s evaluation on student growth on state standardized tests, the state is setting up Texas districts, which will make personnel decisions based on these evaluations, to be sued.
  7. TEA plans to use a value-added model using a complex statistical formula based on proprietary algorithms to calculate a teacher’s value-added score. The model that TEA plans to use, EVAAS, is the same one used by Houston ISD, which has recently been challenged by a lawsuit. Even if TEA uses different specifications from those used in HISD, the algorithms behind the calculations are proprietary. This prevents teachers from being able to access and understand the scoring methods — a problem when they may need to defend themselves from adverse consequences of these evaluations.
  8. Most of the research on the efficacy of value-added measures concludes that value-added methods are an invalid and unfair means of teacher evaluation because results are unstable over time, subject to bias and imprecision, and rely solely on results from standardized tests that were not designed for that purpose.
  9. The requirement that individual teachers’ evaluations be based in significant part on student performance on state tests can be applied only to an estimated 24 percent of teachers (teachers of STAAR-tested subjects).
  10. It does not appear that the state plans to develop other measures of student growth needed for the remaining majority of teachers, leaving districts on their own to do the work. Since these measures have only recently begun to be used in teacher evaluation systems, few vetted and validated prototypes are available as models to districts.

See:

Details of new state teacher evaluation system released

Texas NCLB waiver to tie STAAR scores to teacher evaluations