Former D.C. elementary school teacher Stephanie Black inspired a deluge of comments when I posted her account of how her school’s focus on testing and the new IMPACT teacher evaluation system had soured her enthusiasm for her job.

One commenter suggested she do another post “outlining what a fair/effective teacher evaluation system would look like.” I asked if she would, and she agreed. Here she offers some of the suggestions taken from the comments and from her conversations with fellow educators, as well as some of her own ideas, for how IMPACT can be adjusted to make it a better evaluation tool.

By Stephanie Black

Suggestion One:

Current Problems: 1. There is a disconnect between the goals of administrators, which are the missions and visions of their schools, as well as District goals, and the goals of teachers and staff members, which have increasingly become whatever they are evaluated on. 2. Schools have varying populations, with varying needs.

Suggested Solution: Create a new IMPACT component that includes the former Commitment to School Community and School Value-Added components, and call it Commitment to School Plan. Administrators then work with their Leadership Teams to create School Plans. These plans should be based on the missions and visions of the schools and include high academic and social/emotional expectations for students. The plans should also include rubrics for evaluating how well teachers follow the mission and vision of the school plan. (In the case of schools that are required to submit comprehensive school plans, this plan will be in lieu of, or in addition to, that plan.)

Reasoning: This will enable all schools to continue to hold educators to high expectations, while also accounting for the specific needs of their student populations. It will also help get teachers and administrators on the same page. Teachers will not feel as torn between trying to accomplish only the things they are evaluated on and working towards their schools’ missions. Finally, this type of system will encourage innovation within our public schools. In creating their plans, schools will have a greater ability to create and execute programs (e.g. science-technology-engineering-math, project-based learning, dual language, magnet programs, etc.) that draw families into DCPS, as well as allow families to seek out programs that interest them.

Suggestion Two:

Current Problems: In cases where the IMPACT evaluation overemphasizes certain components (e.g. test scores), teachers are narrowing the curricula, or teaching to the test. This is leading teachers to overemphasize certain concepts and skills, and underemphasize, and in some cases not teach, other concepts and skills. Furthermore, in the case where standardized tests are believed to be overemphasized, this can lead to teachers avoiding or leaving these tested grades and/or subjects, and/or avoiding schools that already have low test scores or environments that are perceived to be more challenging.

Suggested Solution: Change the evaluation system so that all groups of teachers are using the same breakdown, and no component is overemphasized. The suggested new breakdown is: 35 percent - commitment to school plan, 35 percent - observations, 15 percent - student growth data, 15 percent - personal growth and professional development goals.

Reasoning: The new breakdown does not overemphasize any component. Commitment to school plan and observations are weighted the most in an attempt to encourage teachers to support the missions and visions of their schools (which are also meant to improve learning) and to focus on great classroom instruction. Student growth data, which can include standardized test results as well as other measurable outcomes, makes up 15 percent of the score so as to continue to emphasize the importance of making gains, while not fostering an environment where teachers narrow the curricula and/or teach to the test. Last, 15 percent of the score is also determined by a teacher’s personal growth and professional development goals; this will encourage teachers of all levels to continue to reflect and improve upon their practice.

Suggestion Three:

Current Problem: Classroom evaluations are too much like checklist evaluations. This might lead to teachers trying to put on a show during evaluations rather than continuing to teach naturally; because of this, evaluators might not be able to determine a teacher’s true strengths and weaknesses. This type of evaluation might also lead to unreasonably high or low scores depending on the method or style of teaching, and/or the point of the lesson being observed.

Suggested Solution: Adopt an observation process more like the one in the North Carolina Teacher Evaluation Process, which uses rubrics to assess teachers on a variety of performance indicators and to rate them on a scale from developing to distinguished (developing, proficient, accomplished, distinguished). The evaluation system is set up to focus heavily on teacher development. Here is the description of the categories from the North Carolina Evaluation book:

These levels are cumulative across the rows of the rubric. The Developing teacher may exemplify the skills expected of a teacher who is new to the profession or an experienced teacher who is working in a new content area or grade level, or who needs a new skill in order to meet the standard. A Proficient teacher must exhibit the skills and knowledge described under the Developing header, as well as those under Proficient. Likewise, a Distinguished teacher exhibits all of the skills and knowledge described for that element across the row. The Not Demonstrated rating should be used when the teacher is performing below expectations and is not making adequate growth toward becoming proficient on the element. This rating is also used when the principal is not able to check any of the descriptors for the element being rated. If a teacher is rated as Not Demonstrated, then a comment must be made as to why.

While the rubrics themselves are more of a checklist in this evaluation, the skills and knowledge to be checked are also broader and leave more room for flexibility. DCPS would be able to keep Teach 1-9 from IMPACT, but simply create new rubrics for each “Teach.” Another option is that DCPS adopt teaching standards and elements similar to those that North Carolina uses in lieu of Teach 1-9.

A full description of the North Carolina Teacher Evaluation Process can be read online.

Reasoning: By focusing on skills and knowledge, this type of evaluation system will better account for the differences in teaching styles and methods and allow teachers more flexibility. This system also focuses on the importance of continuous improvement.

Finally, like many people, I do not think we can meet the high expectations we have set for schools and students simply (or not so simply) by improving teacher quality. Below are some suggestions for systemic changes, intended to increase student achievement, which could accompany these changes to IMPACT.

Suggestion Four:

1. Make classes more manageable either in terms of size or range of levels per room, and put more aides in the classrooms of schools with high levels of student misconduct.

2. Change from the standard academic calendar to a year-round calendar with a one-month intercession during the summer and shorter intercessions during other parts of the year.

All of the suggestions above are just that — suggestions. I did not bring up my previously published suggestion — eliminating the “highly effective” and “minimally effective” categories. I hope DCPS can change over to the language used in the North Carolina Teacher Evaluation system — developing, proficient, accomplished, distinguished. (There will, of course, have to be policies in place for teachers who continuously fail to score beyond “developing.”) This language is less judgmental and also recognizes the fact that teaching is a craft that improves over time.

As for merit pay, I believe that it is not necessary. The enthusiasm and dedication seen in great teachers is not rooted in a desire to score bonuses. The money would be better used for making sure that all educators have livable salaries — something that DCPS already leads many districts in ensuring, as well as making sure that all schools, in every ward, have the resources they need. That said, given the demands of the federal government, if merit pay cannot be eliminated at this time, it should become more of an “incentive pay” system similar to what is in place in Denver under their evaluation system, ProComp. Under ProComp, all teachers have the opportunity to earn bonuses for doing such things as: working in a high poverty school; taking graduate courses; and getting high test scores. This system is different than the one in DCPS as it is not an all-or-nothing payout system.

I want to emphasize that neither I nor any of the people whom I talked to about IMPACT are looking to get rid of accountability systems for teachers, nor are we looking to lower expectations for teachers or students. What we are attempting to show is that in order for an evaluation system to work, and for it to accomplish its goals in terms of teacher quality and student achievement, there must be more buy-in from the educators it most affects. Schools cannot function properly when educators are frequently anxious, when curricula are narrowed to detrimental degrees, or when teachers and administrators are working towards different goals. The suggestions above attempt to mitigate some of the problems related to those issues that have arisen from the policies in IMPACT.

Hopefully, this is a conversation that can broaden, continue and be heard by the policymakers working with IMPACT, the council members and Mayor Gray, and the people participating in negotiations on the next contract.