Evaluating Evaluation

Terry Grier in Ed Week on what’s happening with teacher evaluation in Houston:

These immediate, common-sense steps were a vast improvement over the evaluation system we were working to replace. The staff-review process helped principals hold their teachers to higher expectations, but, just as important, it helped us hold principals accountable for being instructional leaders, not just building managers. For the first time, we were requiring our school leaders to distinguish great teaching from good, good from fair, and fair from poor, and insisting that they help all teachers chart a path to take their professional skills to the next level.

5 Replies to “Evaluating Evaluation”

  1. As I understand it, Grier’s approach to improving teacher evaluation in the Houston schools is to require each principal to do at least one 1-hour observation of each teacher each year and, relying on the observation and student test scores, rate each teacher A, B, C, or D; the school system then tries to retain the As and to discharge the Ds.

    If this is all there is to Grier’s approach, one wonders Grier wrote an article about it and/or why Eduwonk posted the article.

    A principal might be able to reach some tentative conclusions regarding an observed teacher’s teaching skills based on a 1-hour observation, but with huge caveats — 1) that the principal know what good teaching looks like; 2) that the principal have sufficient subject matter expertise to evaluate the teaching techniques; and 3) that the observed instruction be representative of the teacher’s day-in/day-out work. Other huge potential problems with this evaluation “system”: 1) no check against principal bias (for or against the specific teacher; 2) observation rating is based on 1 hour out of perhaps 1,000 hours of classroom instruction in the school year — the teacher could be having a good/bad day; 3) a teacher’s observed performance/student test scores are probably impacted by student mix — i.e., percentage of students from low SES families, from single-parent families, from non-English-speaking families, with learning disabilities, who are emotionally disturbed, who are reading far below grade level; 4) a teacher’s observed performance/student test scores might be impacted by teacher workload — i.e., number of separate preparations/day (secondary teachers), total student load (elementary and secondary teachers).

    Also, this evaluation approach — relying heavily on a single extended observation — does little to address the problem of a teacher who is competent but lazy. Such a teacher will look good in the single extended observation (particularly if the teacher has advance notice of the observation), but would still be a goof-off during most of the rest of the teaching year. The fact that a teacher was lazy might be indicated by the teacher’s student test scores, but, as noted above, student test scores can be impacted by so many factors beyond a teachers’ control that personnel decisions based on teacher test scores are little more than arbitrary. And, the lazy teacher is a much more common problem than the incompetent teacher — it is unlikely that the system would hire an incompetent teacher in the first place, and, since an incompetent teacher will probably be pretty unhappy most of the time, many of the incompetent teachers will quit on their own. By contrast, given human nature and the virtual absence of supervision, all teachers will goof off occasionally and many will goof off regularly.

    If Houston is really retaining/discharging teachers based on this evaluation “system”, the predictable consequences are: 1) teachers will kiss up to principals; 2) teachers who do not like to kiss up to principals, and particularly teachers who are willing to challenge principals, will leave the system; 3) teachers who have a choice will avoid teaching in schools where a large percentage of the students are poor/LD/ED/ESL/etc.; 4) teacher unions (and individual teachers) will successfully challenge negative evaluations/discharges in arbitration/court; 5) teachers generally will feel insecure and less willing to help fellow teachers; and 6) the quality of instruction, particularly in the lower SES area schools, will be lower than it was under the old system.

  2. LaborLawyer, we probably would not agree on much, but I definitely agree that 1 hour per year is not nearly enough. With the easy availability of audio-visual equipment, every classroom should have cameras and mikes so that a principal can peer in at any minute of any school day. The rest of your post sounds like a bunch of excuses for our failing 1940’s model, union-dominated ed system.

  3. “Every classroom should have cameras and mikes so that a principal can peer in at any minute of any school day.”

    Now THAT really is going back to the factory (plantation?) model! It’s time to break the status quo of this old model and hand over schools to the people willing to teach in them. Teachers would then choose a head teacher who would lead at the pleasure of the faculty. This would be the beginning of full professional status for teachers (mainly women), similar to what college professors (mainly men) enjoy. The biggest advantage to this is that the profession would then be able to recruit and retain many more talented individuals than it does at this time. And that’s something that would really help students. Also, once teachers have the ability to decide who enters the profession and who stays, there will be no more “bad” teachers.

    Intelligent and well-educated people want to be decision-makers. It’s time to demand full professional status for K-12 teachers.

  4. Jason —

    I’m not defending current teacher evaluation systems — just pointing out fundamental problems with Grier’s not-very-unique approach.

    We can definitely improve the ways we monitor teacher competence and performance (note — monitoring competence does not necessarily result in monitoring diligence and, in my opinion, teacher laziness is a much bigger problem than teacher incompetence).

    My suggestions regarding howto monitor teacher competence/performance mostly involve getting more adult eyes into the classroom on a much more frequent basis + reducing the impact of principal bias (in favor or against a specific teacher) + reducing the impact of student-mix/teacher-workload variables in the monitoring process.

    Finally, not clear why you think “we probably would not agree on much”. Given your reference to a “union-dominated ed system”, perhaps you assume that I’m a union-side labor lawyer. Although I strongly support the federal Labor Act’s concept of unions providing counterweights to employers, in my 30+ year career as a labor attorney, I have represented govt and management, but never unions or employees.

Leave a Reply

Your email address will not be published.