Research has shown that student achievement is directly related to the preparation that teachers receive, particularly in the subjects they teach, and to the overall effectiveness of teachers in delivering instruction (Darling-Hammond, 1998; Elmore, 2000). How teachers acquire this effectiveness is at the heart of the challenge that schools face in raising student achievement.
The vast majority of teachers do not enter the profession as highly effective instructors; even those with strong academic backgrounds typically take several years to hone their craft and acquire the repertoire of skills necessary to meet student needs. More often than not, improvement occurs in isolation and largely through trial and error. In many schools, highly skilled teachers work side by side with teachers who are struggling. Sadly, poorly trained and poorly performing teachers disproportionately work in low-income school districts and in districts serving communities of color (Esch, Chang-Ross, Guha, Tiffany-Morales, & Shields, 2004; Lankford, Loeb, & Wykoff, 2002).
With the advent of No Child Left Behind, the pressure to raise student achievement has become greater than ever before. In response to this pressure, a growing number of school districts are adopting strategies to reduce teacher isolation. Many districts and even states have developed mentoring programs to provide greater support to new teachers (Esch et al., 2004; Fideler & Haselkorn, 1999). In addition, many districts have deployed academic coaches to classrooms to assist all teachers in working with new curriculums and to strengthen their performance in such content areas as literacy and math (Platt, Tripp, Ogden, & Fraser, 2000).
A key question remains, however: Once districts have provided such professional development and support to teachers, how do they weed out those teachers whose performance does not improve? In many school districts, the removal of ineffective teachers is a rare occurrence. Teachers unions typically receive the bulk of the blame for defending colleagues of questionable competence. It is equally true, however, that many administrators fail to carry out regular and meaningful evaluations, even though the current emphasis on instructional leadership encourages principals to increase their presence in the classroom and conduct observations and evaluations of teachers on a regular basis (Elmore, 2000; Marshall, 1996). Principals are too busy (Copland, 2001; Grubb, Flessa, Tredway, & Stern, 2003), lack the necessary expertise (Danielson & McGreal, 2000; Darling-Hammond, 1984), or seek to avoid the potential conflict (Bridges, 1986).
A more thorough and thoughtful approach to teacher evaluation exists, one that provides support to teachers who need help and promotes instructional leadership. Peer assistance and review (PAR) reduces the burden on principals, the isolation of the classroom teacher, and sometimes even the antagonism and hostility between labor and management by involving teachers in the formal evaluation of other teachers and making them responsible for employment recommendations.
How PAR Works
PAR, historically referred to aspeer review, was born in Ohio's Toledo Public Schools in 1981. The president of the teachers union, frustrated with the caliber of new teachers, suggested during contract negotiations that the district create an intern program that would enable veteran teachers to play a role in providing support to new teachers. The district agreed to support the intern program if the teachers union would agree to take responsibility for intervention with ineffective tenured teachers. The union accepted the challenge, and the “Toledo Plan” of peer review was born. Today, it is still the most well-known blueprint of the PAR approach.
In peer assistance and review, coaches who have been identified for their excellence in teaching and mentoring support new teachers as well as veterans experiencing difficulty in their teaching. The coaches are also responsible for the formal personnel evaluations of teachers in the program. Coaches are usually released full-time from their teaching duties for two to three years so they can provide this support and conduct evaluations. Typically, coaches are not based at a school site, but are matched to teachers across a district on the basis of grade level or subject area.
Some participating teachers receiving the support are new to the profession or the district; others are more experienced teachers who are perceived as needing some form of intervention related to instruction, such as veteran teachers who have received an unsatisfactory evaluation from the principal. Participating teachers typically spend one year in PAR, after which they either successfully exit the program with the principal resuming responsibility for the personnel evaluations or are released from teaching duties in the district.
At multiple times throughout the school year, coaches report on the progress of participating teachers in meeting specified performance standards to a districtwide teacher/administrator board, called the PAR panel, in which teachers typically represent the majority. The teachers union president and the director of human resources (or some other high-ranking district office administrator) serve as panel cochairs. The panel questions the coach about the support that he or she has provided and suggests other approaches that the coach should attempt. At the spring hearing, if not sooner, coaches—and sometimes the principal—make recommendations about the continued employment of each participating teacher, which the panel either upholds or questions. The panel's employment recommendation then passes to the superintendent.
Many PAR programs also include an alternative professional development and evaluation option for tenured teachers who aremeeting standards. Rather than work with a coach, satisfactory tenured teachers form pairs or small groups and identify an area for mutual professional growth for the year. In some programs, these teachers present their work to the PAR panel at the end of the year.
In most districts that have adopted peer assistance and review, the roles of coach and panel member provide powerful professional development opportunities for teachers looking for a new challenge. These roles are not intended as stepping-stones to administration. In fact, some programs require participating teachers to sign contracts stating that they intend to return to the classroom.
PAR in Action
In 1999, the California legislature initiated a statewide PAR program that essentially required all school districts to have PAR in place for veteran teachers by academic year 2000–2001. Many districts took advantage of the opportunity to include beginning teachers in their newly designed PAR models.
Findings generated from a longitudinal study (Goldstein, 2003, 2004) of an urban school district in California that uses peer assistance and review may prove helpful to districts that are considering implementing PAR. The district in question comprises approximately 100 schools and 3,000 teachers and is ethnically and economically diverse. The first year of program implementation involved 10 coaches and 91 participating teachers (88 beginning teachers and 3 veterans) across 28 schools. Participation grew during the next two years, with approximately 12 coaches and 140 teachers becoming involved in the program.
The research involved interviews and surveys of approximately 200 participants—panel members, coaches, principals, participating teachers, and key district officials—and included more than 300 hours of observation at panel and coach meetings. It examined perceptions of PAR's effect on the support and evaluation of participating teachers and on overall teaching practices. These are the study's findings.
Time for Support and Evaluation
Teachers rarely have opportunities to discuss their practice with their peers or critically reflect on their teaching practice. This is especially true for teachers in high-poverty schools, which frequently report lower collegial interaction among teachers related to instructional improvement (Shields et al., 1999).
In the district where this research was carried out, PAR led to frequent, ongoing contacts between coaches and participating teachers. Caseloads for the coaches consisted of approximately 10 participating teachers in the first year of the program, increasing to 12–15 participating teachers in subsequent years. Compared with the caseload of the typical elementary school principal, who may be responsible for 25–30 teacher evaluations annually in addition to myriad other responsibilities, the more limited workload of the coaches typically allows for more thorough and consistent teacher support.
On average, coaches were present in each participating teacher's classroom once every one to two weeks for an hour or two. Visits might be announced or unannounced. Coaches observed the participating teacher, discussed the observation afterward, assisted in planning and modeling lessons, and, in conjunction with the participating teacher, observed and discussed other teachers' teaching practices.
Coaches diagnosed the strengths and weaknesses of a given participating teacher and tailored support accordingly. Some participating teachers needed to focus on classroom management, whereas others focused on strengthening content knowledge or planning more engaging lessons.
Research shows that many new teachers do not receive enough support during their first few years on the job (Fideler & Haselkorn, 1999). This is especially true in large urban districts, where teacher turnover is typically high and large numbers of new teachers are hired each year. One principal noted thatIf you've got 35 people to evaluate, your contact is going to be limited to what's required. Those times are going to be like snapshots: I'm going to come at 10:00 a.m. on March 2 to evaluate you in your classroom on a lesson that you choose to teach. You would think everyone would be ready!
In contrast to this “snapshot” approach, participating teachers in the peer assistance and review program reported that they had regular contact with their coaches, with many describing their coaches as accessible, available, andwilling to help me when I need it. Moreover, research demonstrates that school districts that implemented peer assistance and review retained more new teachers than they had prior to PAR implementation (Brown, 1993; Hewitt, 2000).
Responding to the Needs of Veterans
Most districts have almost no formal structures in place to support struggling veteran teachers. Many educators believe that veteran teacherscannot improve, and research on teacher evaluation has claimed that teachers cannot be effectively “remediated” after their initial years in the profession (Bridges, 1986). In addition, providing support to a veteran teacher typically takes twice as much time as providing support to a novice and is a larger emotional drain on the mentor.
Peer assistance and review programs are one promising approach to providing veteran teachers with the help they may need. Principals generally give few unsatisfactory personnel evaluations of teachers (Loup, Garland, Ellett, & Rugutt, 1996; Tucker, 1997). Once a PAR program is in place, however, the number usually increases because, for the first time, a mechanism is available to provide support to veteran teachers who are experiencing instructional challenges.
Linking Professional Development and Evaluation
Practitioners and researchers alike have criticized principal evaluations of teachers because principals often lack the specific content expertise of those they evaluate. The PAR program addressed this issue by matching coaches with participating teachers by grade level and subject matter. Moreover, principals' ratings of teachers are rarely grounded in performance rubrics, whereas coaches use standards to evaluate teaching practice in their work with participating teachers.
In the district studied, the peer assistance and review program was based on the California Standards for the Teaching Profession, which were modified to address specific district priorities. Coaches received professional development training on coaching methods and standards-based evaluation. Principals and members of the PAR panel reported that they were impressed with the coaches' knowledge of the standards and the way they were able to use the standards for diagnostic purposes. Several principals asked coaches to teach them how to conduct this kind of teacher evaluation. One principal noted,We finally have specific, observable behaviors in line with rubrics. That is really good because it's much more objective. I'm confident that the difference between any two people evaluating one teacher is going to really narrow with PAR.
In addition, coaches had firsthand knowledge of how a teacher performed at the beginning of the year, and they knew what struggles and growth the teacher had experienced throughout the year. With this background knowledge, the coaches were able to provide the necessary support or locate resources in the district that would meet the teacher's needs.
Some educators have raised concerns about a potential conflict of interest in having the same person provide professional development and conduct a teacher's personnel evaluation. Only 3 of the 15 participating teachers interviewed were reluctant to discuss with their coaches areas of instructional weakness. All 3 of these teachers were in need of major remediation, however, and 2 were not renewed for employment at the end of the year. The other 12 teachers interviewed said they experienced no lack of trust toward their coach, and the survey data support this finding. By linking personnel evaluation to a substantive, ongoing, data-rich support process focused on improving instruction, PAR coaches have an edge in their work with teachers that most support providers simply do not have.
A Transparent Evaluation Process
Because principals typically evaluate teachers in isolation, their assessments are rarely challenged by public scrutiny—with the exception of the expensive legal process that often occurs when a teacher is fired. In contrast, the peer assistance and review process in the district studied was relatively transparent and open. Coaches discussed their assessments of participating teachers with other coaches at weekly meetings, sometimes visiting teachers' classrooms in pairs. Coaches also frequently conferred with principals to inform them of participating teachers' progress, and coaches and principals sometimes observed teachers together and then met to confirm the steps needed to support the participating teacher.
The formal public examination of participating teachers' practice that takes place during panel hearings is an effective way to ensure transparency in the evaluation process. The teachers union president explained the role the panel plays as a check and balance against the role of the coaches:When coaches make an employee recommendation, they have to be able to justify it to the panel. If they can't justify it, then the recommendation gets overturned or the coaches get sent back [to gather more evidence]. The panel has the responsibility of making sure that the recommendation is supportable.
The public aspect of the peer assistance and review evaluation process is in many ways the most crucial distinction between PAR and traditional principal-centered approaches. Panel members can offer input supporting a participating teacher or challenge any apparent bias that may be influencing the review.
A District/Union Partnership
Peer assistance and review creates a venue for experimenting with a new model of labor relations in education. With PAR, teachers exercise greater responsibility in assessing and monitoring their colleagues. Teaching begins to take on characteristics common to other professions, such as law and medicine, in which professional associations monitor the performance of peers (Etzioni, 1969). The confrontational labor relations that often make dismissing a teacher costly and time-consuming are replaced by a more cooperative approach, in which the district administration and the teachers union share responsibility for supporting and evaluating teachers in the program.
In California, the state legislation requires districts to submit a PAR plan for state approval, which local teachers union presidents must also approve. With the teachers union president as panel cochair, the peer assistance and review program fostered a partnership and trust between teachers and administrators not previously seen in the district. One principal reported how surprised she was when the union president supported her recommendation that a teacher be fired:It takes forever to move teachers out if they're not doing the job. All your eggs need to be in a row, and the union really comes at you. With PAR, I'm working collaboratively with the union. It's a whole different feel. There's a sense that the union and I agree that we need teachers who use best practices, and we're working together to have best practices occur. I feel that we're all on the same team and that it's about the students and the kind of teaching they receive.
Confident Evaluations
Despite principals' regular complaints about teacher quality and performance, their formal assessments find almost all teachers competent (Loup et al., 1996; Tucker, 1997). Principals ultimately doubt themselves when making evaluative decisions about teacher competence because they usually have limited time to bring to the task. With peer assistance and review, however, coaches know how much support they have provided. They know the support is grounded in standards of good practice and is part of a transparent process involving a forum for discussion and a partnership with the teachers union. This leads to higher confidence in evaluative decision making.
In the first year of PAR implementation in the district studied, 14 of 91 participating teachers were not renewed for employment: 11 of 88 new teachers (12.5 percent), and 3 of 3 veterans. In the second and third years of implementation, although the number of participating teachers in the program increased, the nonrenewal figure for new teachers dropped to 10 percent, a figure consistent with statistics from other districts with PAR programs. The change was likely the result of hiring fewer uncredentialed teachers after the first year of the PAR program.
By the third year of the program, one of the four veteran teachers who had been recommended for intervention (25 percent) successfully exited the program. This percentage placed the district below the average of other districts with established PAR programs, in which 30–60 percent of intervention cases successfully exit PAR and remain in teaching (Darling-Hammond, 1984; Hewitt, 2000; Kelly, 1998; Murray, 1999). In the district studied, however, veterans placed in the program in its initial years were thought to be notoriously below standard, which might explain the district's lower success rate.
When a district undertakes the systemic changes involved with peer assistance and review, the quality of teacher performance becomes a central concern. Losing 10 percent of a cohort of new teachers as a result of rigorous evaluation is virtually unheard of in urban districts across the United States. It is even rarer to find a district that removes underperforming veteran teachers without a high degree of conflict and rancor. Districts that have successfully implemented peer assistance and review have found it an effective means to systematically improve the quality of teaching and, in the process, to honor and recognize the best teachers.
Instructional Leadership Redefined
Despite the benefits experienced by many of the school districts that have implemented peer assistance and review, many educators will undoubtedly be reluctant to attempt the program because it represents such a significant break from past practice. Evaluation has historically been the principal's responsibility. Regardless of how ineffectively or irregularly these evaluations have been conducted, many principals may be reluctant to give up what they regard as an important source of power and authority.
In addition, teachers have not been particularly willing to take responsibility for evaluating one another. Many educators—teachers and administrators alike—hold deeply ingrained beliefs regarding long-established authority relations within the hierarchical structure of education. However, the mere fact that the removal of ineffective teachers is a rare occurrence in so many districts should be enough to encourage even the most cautious education leaders to consider a new approach.
Such an approach necessarily involves reframing how we view instructional leadership. Peer assistance and review moves supervision beyond the factory model that puts the school principal in charge of a large number of workers to a professional model in which teachers hold one another accountable for quality practice. Rather than reducing the principal's power, this approach presumes that strong principals serve as effective instructional leaders when they use the strengths of those around them.
Peer assistance and review may not help teachers figure out how to teach 40 students crowded in a classroom or how to work successfully in chaotic and out-of-control schools. Nonetheless, it is a promising approach for providing both classroom-based support and serious teacher evaluation.
Turning to teachers for help in improving the quality of teaching is just common sense. Unfortunately, common sense is often not that common.
