Cross-School Teacher Evaluation

Thomas Van Soelen
Associate Superintendent, City Schools of Decatur




 Thomas Van Soelen

Since 2007 the City Schools of Decatur (CSD), Georgia has gradually implemented the Georgia CLASS Keys, a teacher evaluation system in which teachers and leaders collect performance evidence in 26 instructional areas using rubrics that detail quality. The carefully phased approach to implementation has resulted in an effective evaluation system that supports teacher development, capability and performance, and serves as a reliable measure of teaching quality to guide district and school leaders and teachers’ decisions and actions. The quality of the implementation process has produced more effective instructional leaders as the professional learning supporting the implementation of the new teacher evaluation system focused on leadership first and the teaching ranks second. Critical to this process has been the implementation of cross-school evaluation visits.

City Schools of Decatur is a unique independent Georgia school system, touching the city of Atlanta. CSD was approved as one of the nation’s first conversion charter systems by the Georgia state Board of Education July, 2008. The system serves approximately 3,250 students (58% white, 30% black, 12% other) in seven K-12 schools and 355 learners in an infant-preKindergarten center. The district employs 550 employees including 290 teachers.

In a multi-year process of changing teacher evaluation systems, CSD has modeled best practices both with its leadership and teaching staff. Strong collaborative processes have accompanied high-quality professional learning that has produced higher student achievement, stronger teaching practices, and a higher engagement level among teachers.

CSD prioritized teacher evaluation by placing it at the center of their work. In some school districts, school reform is at the center surrounded by such items as professional learning, peer observation, common instructional expectations. Teacher evaluation is often a necessary evil, placed at specific times during the year, decontextualized to anything else meaningful occurring at the school. By placing teacher evaluation at the middle and making it matter, all the work improves because every moment with students truly matters.

In transitioning from a check-mark based evaluation (either Satisfactory or Needs Improvement) to a rubric-based system, feedback became the operative word. Producing high-quality feedback for teachers to read, reflect, and actually use is the number one goal.

Toward that end, building and central office instructional leaders started out one year before implementation, monthly engaging in Teacher Evaluation Professional Learning for two hours, solely focused on instruction. These meetings were characterized by the following:

Collaborative processes. These meetings were a change from other district meetings, responsive in nature, using participant’s reflections from previous meetings to create agendas for the next gathering. Meeting content tightly focused on instruction and assessment thereby creating common district language for important concepts (e.g., differentiation, flexible grouping, formative assessment). Meetings often used protocols from the School Reform Initiative facilitated by a trained Critical Friends Group coach.


sep-nsj-sep2012-2Localizing our work. After using videos created by other entities, leaders requested local videos. With a FlipVideo in hand, both central office instructional personnel and building instructional coaches began to collect video footage that was used immediately in professional learning. Many of these clips are used today with teachers as they need to see examples or non-examples of particular elements at work.

Manageable learning segments. Each meeting held clear learning targets, exactly the same as the learning expectations for teachers. At the beginning of the work, the group identified practices when they watched short video clips – the identification was the goal. In year three, leaders moved toward judging the quality of the practice, using rubrics. This phased in approach – starting first with identification, later with quality – was also phased in with teachers over multiple years. (See New Link)

Accountability for improvement. was a popular tool that continues to be used in quarterly meetings now in year four. After watching a video clip and focusing on a particular teaching element, PollEverywhere offers each person an opportunity to text or type their feedback to the teacher, if the teacher were to receive it. The feedback appears on the screen in real time and all participants read each other’s feedback. Following each session, goal-setting occurs where participants claim how they want their feedback to improve (look like and sound like some of the feedback they read). We contend that this is the hidden part of instructional leadership: no one (except the teacher) ever reads evaluation feedback, and certainly doesn’t give feedback on the feedback! This continued practice has made our feedback proficiencies public and very open for improvement.


This intense professional learning is critical to a unique part of our teacher evaluation plan: cross-school evaluation visits called Standards-Seeking Visits (SSVs). Since CSD (and Georgia) had never defined quality teaching, we were truly doing exactly that – seeking those moments out. SSVs involve a day in the fall and a day in the spring when each school is visited by a team of colleagues from around the district (other principals, assistant principals, instructional coaches, central office instructional staff). A schedule is created for three hours where pairs of visitors visit classrooms for 15 minutes and then debrief in the hall for 15 minutes, offering ratings on 15 of the 26 elements from the teacher evaluation tool. These SSVs qualify as informal observations for the teacher for the annual evaluation cycle. This also means for the building principal that four other district leaders have observed each teacher in his/her building. This data is extremely helpful in triangulating other observational data collected within the building. To reciprocate, each principal then serves on other SSVs outside his/her building.

The group discovered at the end of year two that their feedback was better in the monthly professional learning compared to their own solo observations. To build better scaffolds toward improving feedback, all SSV observations are completed in pairs. These pairs submit only one evaluation (using an online tool called MxMobile). Many call SSVs some of the most rewarding professional learning they have ever experienced.

In Year Four of this phase-in plan, teachers have joined these meetings, now called the Teacher Quality Team. The name change is significant, as we realized that focusing on teacher evaluation has such positive outgrowths that teacher quality is changing. In addition, a Teacher Quality Academy was developed where 16 teachers next year will participate in at least two SSVs, paired with more experienced leaders. These teachers do not have their eyes on principalships; instead, this academy is focused on developing strong instructional models within the district. These teachers realize that watching others and offering feedback is a phenomenal way to improve their own teaching practice. The curriculum for this academy is the same as the Teacher Quality Team – building coherence across and within the district.

Online journals can be a boon to educators – providing opportunities to update reform efforts at a speed quicker than paper:

City Schools of Decatur’s Teacher Quality Team met July, 2012 and launched a new iteration for year five of their Teacher Quality Plan: Paired Informal Visits (PIVs). The goal hasn’t changed:

Sticky note description:

Paired Informal Visits (PIVs) increase feedback quality, accelerating improvement in teacher practice.

PIVs replace the large-scale SSVs with paired visits in the buildings where the building leaders work. Thus, one of the observers in each pair is known by the teacher being observed. In addition to that benefit, this plan reduces the number of times building leaders are out of the building, includes 16 Teacher Quality Academy graduates (all current teachers) to be partnered with leaders outside their buildings, increases the likelihood for authentic teaching since PIVs are completely unannounced, and reduces administrative workload to orchestrate a SSV schedule for multiple observers. Each building leader (the “host”) determines the schedule when their “visitor” arrives. The schedule has been predetermined for the entire year with each building leader hosting 8 different visitors.

This Double Bubble Thinking Map below was used along with a district-created Prezi for teachers to understand the changes in Teacher Quality practices for the 2012-2013 school year.


Perhaps the greatest benefit of this iteration is that building leaders have accountability for engaging in the instructional leadership that is germane to their jobs. They now host colleagues several times during the year and collaborate on the work that is the greatest lever for instructional improvement.

Since 2004, CSD has closed the achievement gap between black and white students on state assessments by over 75% in reading and 50% in language arts. CSD gives the Gallup Q12 Staff Engagement survey each year and teachers’ overall engagement grew 6 percentile points in one year, over 19% in an item about supervisors giving them feedback on their performance. We credit much of this improvement to a strong focus on high-quality teaching, all centered around feedback.