Reforming Teacher Evaluation in Maine: New Haven, Connecticut’s Innovative Approach is a Model for Maine


Read the full report | By June 1 of this year, Maine is to submit its application for the Race to the Top, a federal education grant program that could bring the state as much as $70 million in funding for education reform. As a result of recently-enacted legislation, however, Maine is ineligible to compete for a Race to the Top grant unless a state-level stakeholder group, already appointed, approves a process for evaluating teachers and principals in Maine?s school districts that uses student achievement data in some way.

The stakeholder group has already reviewed two evaluation models. The first, the Danielson Framework, is used in a number of teacher evaluation systems already and rates teachers against a set of professional standards which are thought to be correlated with highly effective teaching. In a similar way, the TAP model, which is the other model the stakeholder group has reviewed, has an evaluation element that uses multiple classroom observations to establish whether teachers have met TAP?s “Teaching Skills, Knowledge and Responsibilities Performance Standards.” In neither model, however, is actual student achievement, or lack thereof, used to evaluate teacher effectiveness. Rather, in both models, teachers are evaluated on the extent to which they possess the knowledge and skills thought necessary for highly effective teaching.

The city of New Haven, Connecticut, however, is set to launch a new kind of teacher evaluation system. It combines profes-sional teaching standards and student achievement data in a remarkably innovative way to create a more comprehensive evalua-tion of teacher effectiveness. With the deadline for the state?s Race to the Top application rapidly approaching, it is a model that Maine?s education policymakers at both the state and local level, including the stakeholder group, should strongly consider.

The New Haven Model
The city of New Haven, Connecticut, made national news last year when the city?s school system and its teachers? union agreed to adopt a “groundbreaking” teacher evaluation system that makes significant use of student achievement data.1 On April 25, 2010, after months of work, the team of teachers and administrators working on the new evaluation model presented their plan to the city?s Board of Education.2 It was immediately heralded in the New York Times, which editorialized that with its new system, New Haven could be “at the forefront of a national effort to improve the caliber of instruction in public schools.”3

According to the designers of the New Haven system, its development was driven by an effort to advance four policy objectives: “measuring student learning, assessing teacher instructional practice and professional values, teacher development, and peer validation of administrator judgments.” The model that resulted, they claim, “re-crafts the NHPS teacher evaluation and devel-opment system to enable deep individualized development for teachers, ensure that development is aligned to student learning goals, enable embedded and professional evaluation and coaching for all teachers, and be bound by the consequential recogni-tion of both outstanding and poor performance.”4

Under the new model, New Haven?s teachers are to be evaluated in three areas. Through multiple classroom observations and conferences, teachers are first to be assessed on their instructional practices, which include such things as planning, classroom management, and the use of data. Second, teachers are to be judged on what the plan describes as “professional values,” which include such “competencies” as “collaboration and collegiality, self improvement, reliability, high expectations, respect, respon-siveness and outreach, and professionalism and judgment.” Assessment of professional values will be done through regular con-ferences with evaluators, but the model contains provisions for teacher self-assessment as well. The third component of the evaluation system is student learning, with an emphasis placed on student growth. The New Haven plan describes student growth as “the advancement of learning relative to peers with a similar academic history.” Using achievement growth as an indi-cator is important, say the plan?s designers, “because it enables some control for environmental factors, so that like students are compared to each other, and because it reflects the actual contributions of individual teachers over the course of the year.” The student learning data to be used will include both standardized test scores and alternative portfolio-type assessments, and learn-ing gains are to be reviewed throughout the year to ensure an accurate measurement of student achievement growth.5

Under the New Haven model, data is collected in each of these three evaluated areas, and “at the end of each year, all teachers will be assigned a rating that indicates their level or perform-ance for each component.” The system uses the following five-point rating scale:

One of the most innovative things about the New Haven model is how it brings together the three evaluation elements into a single “summative rating.” As illustrated in the matrix below, the “instructional practice” and “professional values” compo-nents are combined, with instructional practices making up 80 percent of the combined rating and professional values making up 20 percent. That combined score is placed on one axis of the system?s summative rating matrix, with the student learning growth score placed on the other axis. A teacher?s final rating is determined by the intersection of the two scores.6

Student Learning Growth

What is so ingenious about the summative rating matrix is how it integrates student achievement data into the overall rating in such a balanced way. Teachers who are found through confer-ences and observations to be very strong on instructional prac-tice and professional values will typically receive a final rating of effective or better, even if student learning growth is not outstanding. Teachers with low levels of student learning growth, though, will have to score exceptionally well on their instructional practice and professional values ratings in order to avoid being rated as “needs improvement” or “developing.” In this way, student learning growth is an important part of the overall evaluation, but not so important that teachers thought to be otherwise professionally skilled are rated as ineffective.

The asterisks which appear in several boxes at the upper right and lower left of the matrix indicate that teachers who score in these areas will have their evaluations subjected to further re-view, given that it is unlikely that a teacher scoring poorly on the professional aspects of the job will have high student achievement or that teachers who are strong professionally will have poor student achievement. This is just one of several oversight provisions built in to the New Haven plan.7

In fact, the designers of the New Haven plan highlight a num-ber of other components which they say are critical to the model’s success:

  • For the architects of the plan, the “centerpiece” of their approach to teacher evaluations is a new conference struc-ture. As planned, teachers will be assigned a single “instructional manager,” most commonly a building ad-ministrator, with whom they will meet for “evaluation and development conferences” at least twice a year. Non-tenured teachers and teachers found to be “needing im-provement” will meet with their managers at least four times a year. These conferences, say the plan? designers, “will be the anchor of the rest of the evaluation and devel-opment process, and the foundation of the professional relationship between teacher and instructional manager.”8
  • The New Haven plan also proposes to reinvent the obser-vation process, replacing the formal classroom observation with which all teachers are familiar with a more informal approach called “instructional rounds.” It is expected that in the weeks prior to each professional conference, instruc-tion managers will perform at least three instructional rounds per teacher, collecting data for later use in the evaluation process. More traditional formal observations will remain an option as well.9
  • A more muscular teacher support system is also integrated into the new model, so that teachers identified as “needing improvement” receive “immediate and intense develop-ment opportunities,” including a written “Intensive Plan of Improvement.” Teachers rated as “developing” would get additional support as well, while those found to be “exemplary” would be “eligible for teacher leadership positions,” including “modeling and sharing best practices and supporting other teachers.”10
  • In response to oft-heard complaints that administrators might not be fair in their evaluations, the New Haven sys-tem also includes a “peer validation process.” For teachers who appear to be “on track” for either a “needs improve-ment” or “exemplary” rating, additional evaluation confer-ences, this time with peer teachers attending for validation purposes, must be held. In this way, teachers can be as-sured that administrators cannot play favorites, nor target a teacher for poor evaluations for reasons other than those found within the system? evaluation standards.11
  • In an effort to ensure that student achievement data is as complete as possible, the New Haven system requires mul-tiple types of student assessments and, if available, achievement data from multiple years for each student. Students are to be measured both against their peers and against state and district learning standards.12
  • In order to accurately measure annual student growth, teachers in the New Haven system are to conduct “diagnostic assessments” each fall in order to “diagnose the knowledge and skill level of their students.” Working with their instructional managers, they are to then establish individualized yearly goals for student academic growth. That growth is then reviewed and analyzed during evalua-tion conferences throughout the year, with the teacher? final student learning performance rating assigned at the end of the year.13
  • In order to assess student learning growth in subject areas not typically measured with state standardized tests, such as art or music, the district will develop a series of valid and reliable alternative assessments from which teachers in those subject areas, working with their instructional man-agers, can choose.14
  • The New Haven system also includes sanctions for teach-ers who are found to be ineffective. Underperforming teachers who fail to improve despite additional training and support will face sanctions at the end of that school year. Even tenured teachers “should not remain „eveloping?for more than two consecutive years,” or they will face sanctions as well.15
A Model for Maine
Integrating student achievement data into evaluation systems for teachers and principals is no easy task, and the state-level stakeholder group that is currently at work developing such systems for Maine continues to struggle.

Teachers and school officials in New Haven, however, have developed one of the nation? most innovative teacher evalua-tion models, one that focuses on improving student outcomes by providing teachers with timely feedback, a steady stream of data and analysis, and targeted professional support. Struggling teachers are given additional resources and training, while ex-emplary teachers are enlisted to mentor others and model best practices. The system? ingenious summative rating matrix fairly and elegantly blends teacher professionalism with stu-dent learning growth in order to generate a truly comprehen-sive assessment of teacher effectiveness.

The stakeholder group studying evaluation systems for Maine would do well to study New Haven? new evaluation approach, which will almost certainly become a model for school systems across the nation in the years to come. They should then ap-prove a similar model for use in Maine? schools, and put Maine on track to better measure the effectiveness of its teach-ers and school administrators.





4. NHPS Evaluation and Development System Recommendations- Overview for the Board, April 15, 2010.




8. NHPS Evaluation and Development System Recommendations- Overview for the Board, April 15, 2010.



11. NHPS Evaluation and Development System Recommendations- Overview for the Board, April 15, 2010

12. IBID


14. IBID

15. NHPS Evaluation and Development System Recommendations- Overview for the Board, April 15, 2010.

Read the Full Publication