A study of how specific principal behaviors affect teacher and student performance
This discussion chapter begins with an overview of the findings presented in chapter four, and is followed by an in-depth discussion of each research question: quality of teacher instructional practices; changes in student performance; and frequency and focus of teacher conversations. After the discussion of each research question, implications of the study’s unintended outcomes and recommendations for future research are discussed. The chapter concludes with a table of core questions principals may have which are affirmatively answered by this research; the audience for these questions/answers is principals interested in this work.
As a result of the four principal-teacher interactions introduced in this study, teacher instructional practices improved (according to analyses of QIR data), student performance increased (according to analysis of student grade distributions and discipline), and the frequency and focus of some teacher conversations changed (according to analysis of teacher and student surveys). While teacher instructional practices improved, the degree of improvement varied among different groups of teachers. Additionally, teachers and principals did not always agree on the quality of instructional practices exhibited in the classroom, particularly for lower performing teachers.
According to data analyses of grade distributions and discipline referrals, student performance improved more than predicted if the treatment had not occurred. Improvement in classroom grade distributions and discipline referrals may have been impacted both by changes in teacher instructional practices and by increased principal visibility. This study did not investigate the direct or mediated impacts of teacher instructional practices or principal visibility on student performance.
Results indicated that the frequency and focus of teacher conversations changed during the course of this study; however, there were essentially no indications that the frequency and focus of principal-teacher conversations or teacher-student conversations changed during the course of the study.
Teacher Instructional Practices
Teacher instructional practices formed the key construct for research question one. Because the treatment introduced was school-wide, the entire teaching staff of the school was the initial focus of data analysis for research question one. As a group, teachers in this school indicated the quality of their instructional practices improved in the two domains of Planning & Preparation and Learning Environment (see Table 15). Principals indicated the quality of teacher instructional practices improved in the two domains of Instruction and Assessment. Teacher’s perceived self-improvement in instructional practices within specific domains that differed from the specific domains principals’ rated as showing teacher improvement. Several factors potentially explain this divergent outcome in perception of improvements in different domains.
Teacher planning and preparation for lessons is generally a private endeavor that is not easily observable by principals. Rather than observing the process and intellectual effort required for quality preparation, during instruction principals observe the result of the planning and preparation (Hirsch, 1999; Yinger, 1980). Likewise, teachers would have a much deeper knowledge of their classroom learning environment than would principals who are only visitors from time to time. Thus, teachers’ self-rated improvements in the two domains of Planning & Preparation and Learning Environment suggested that their thorough self-knowledge of these two critical instructional components may not be directly observable by principals.
Principals traditionally receive extensive training in analyzing, recognizing, and characterizing a wide range of quality of assessment and instructional techniques employed by teachers. As a result, they tend to develop perspectives that are attuned to these differences in quality. Thus, principals may often characterize changes they observe in their teachers as improvements in instruction and assessment, whereas the teachers they observe may only recognize that their assessments and instructional practices have changed but not necessarily characterize that change as improvement (Fullan, 2005b).
As a group, teachers rated their instructional practices as demonstrating greater quality than did the ratings of principals (see Table 16). Because many in the teaching profession are hard working, conscientious individuals that put forth a great deal of effort in their core responsibility of planning and delivering instruction, teachers may rate the quality of their instructional practices artificially high based on that effort (Ross, 1995; Schacter & Thum, 2004). Essentially they equate effort with quality. Due to observation cycles and evaluation timelines, principals evaluate the quality of instructional practices within specific time intervals with specific standards in mind. They have not been personally immersed in the planning and delivering of the instruction they are evaluating, so they tend to bring an outside perspective to this observation task. Because principals are only generally aware of the time and effort teachers spend preparing outside the classroom, this awareness does not typically influence their interpretation of what they observe (Torff & Sessions, 2005). In many cases, teacher self-ratings may be the result of confusing planning and delivery effort with the quality of instructional practices. As discussed in chapter three this supports the stronger validity of principal ratings over teacher ratings.
The Quality of Teacher Instructional Practices
Groups of teachers identified as having different quality of instructional practices may have also differed in their self-ratings. In order to obtain as much discrimination in groups as possible, teachers were split into three groups according to the overall posttest QIR ratings completed by the principals as discussed in chapter four. Statistically, there was no difference in the QIR self-ratings completed by teachers who were identified as high, medium, or low performing teachers based on the principals’ ratings as evidenced in Table 17. According to analysis of QIR ratings completed by the principals, high performing teachers improved the most, and low performing teachers improved more than medium performing teachers who essentially did not improve (see Tables 18, 19, & 20).
High performing teachers rated their instructional practices equivalent to the principal ratings. Medium performing teachers rated their instructional practices slightly higher than principals (about 0.3 of a performance level). Low performing teachers rated their instructional practices considerably higher than principals, almost a full performance level above in every domain throughout the QIR. As a result, high performing teachers accounted for little to none of the variance between principal and teacher ratings of instructional practices. Medium performing teachers accounted for only a small amount of the variance. Low performing teachers accounted for most of the variance between principal and teacher ratings of instructional practices.
High Performing Teachers
High performing teachers primarily perceived their only improvement to instructional practices to be in the domain of Learning Environment as evidenced in Table 18. This result may be associated with teacher efficacy. Teachers with a strong sense of efficacy are open to new ideas and more willing to experiment with new methods to better meet the needs of their students (Guskey & Passaro, 1994; Stein & Wang, 1988). According to Tschannen-Moran, Woolfolk, and Hoy (1998), “teacher efficacy influences teachers’ persistence when things do not go smoothly and their resilience in the face of setbacks” (p. 222).
The belief that there is always room for improvement seems to apply to how high performing individuals in many professions view their own performance. High performing individuals in professions outside of education tend to be critical of their performance and continually search for ways to improve (Dunning, Johnson, Ehrlinger, & Kruger, 2003; Young, 2009). High performing teachers may also tend to be critical of their own instructional practices, constantly looking for ways to improve their teaching. Thus, high performing teachers may have underrated their instructional practices compared to their actual performance in this study.
Principals perceived improvement in all domains of instructional practices for high performing teachers with the exception of the domain of Planning & Preparation. This result supports the conclusion that high performing teachers actually improved their instructional practices during the course of this study. Though teachers themselves may not have perceived as much improvement, the reason that principals did not observe improvement in high performing teachers’ planning and preparation was likely due to the private nature of activities related to planning and preparation. As previously discussed, teachers’ planning and preparation often takes place in isolation away from the school and is more difficult to for a principal to observe (Hirsch, 1999; Yinger, 1980).
Medium Performing Teachers
Improvement in instructional practices was demonstrated in one domain out of eight possible for medium performing teachers; teachers who principals rated in the middle group in terms of instructional practices as evidenced in Table 19. According to principals’ ratings the quality of medium performing teachers’ instructional practices remained essentially unchanged during the year of full implementation. Like the principals, the medium performing teachers themselves perceived no improvement in their own instructional practices. The only domain in which principal-rated improvement was indicated in the QIR for the medium performing teachers was in the domain of instruction. That change, while statistically significant, was only 0.2 of a QIR performance level. These results seem to indicate that the instructional practices of medium performing teachers did not improve during the course of this study.
The lack of change to instructional practices by medium performing teachers implies that in order to improve instructional practices for medium performing teachers in this school, additional strategies might be necessary, or perhaps the treatment employed needed more repetition to be effective.
These findings for medium performing teachers contradicted some recommendations regarding the breadth of teacher improvement. According to Marzano (2003) and Downey et al. (2004) specific strategies for improving instructional practices of teachers should not be differentiated among teachers performing at different levels.
Low Performing Teachers
Low performing teachers indicated that they improved in the domain of Planning & Preparation (see Table 20). As was the case for all teachers, a possible cause for this result was that teachers are more familiar with their own planning and preparation than principals (Fullan, 2005b). This may have indicated that low performing teachers were putting forth increased effort in planning and preparation, which they perceived as improvement (Ross, 1995).
Low performing teachers in this study were rated by the principals as Developing within five levels of performance on the QIR (Unsatisfactory, Beginning, Developing, Proficient, Exemplary) on both the pretest and posttest. The low performing teachers, when rating themselves, selected the proficient range overall. Through the lens of social cognitive theory, Bandura (1986) proposed that what people perceive about themselves affects the outcome of their endeavor. Thus, if a low-performing teacher believes he or she is doing proficient work, but does not have the skill set to improve or the desire to improve, then perhaps no improvement in teaching will occur.
Although not supported by data presented in this study, principals determined through discussions and calibration meetings that the low performing teachers did not believe there was any room to improve their instructional practices. Instead, they blamed poor student effort instead of their instructional practices for poor student performance.
Across the three levels of teacher performance (high, medium, and low performing) principals and teachers perceived improvement differently, but overall improvement of instructional practices was indicated. The data also demonstrated that low performing teachers consistently overrated their instructional practices as compared to principal ratings. QIR ratings completed by the principals and teachers showed that, while the set of four principal-teacher interactions did improve the instructional practices of the low performing teachers, the principal-teacher interactions did not improve the accuracy of the low performing teachers’ perceptions of their own instructional practices. (This assumes principal ratings of instructional practices were more valid, as discussed earlier in chapters two and four). According to the findings in this study, low performing teachers either perceived their instructional practices to be of equal high quality as other teachers or low performing teachers were unwilling to admit their instructional practices were of lower quality than their peers. Both of these interpretations reflect current research in self-perceptions of low performing teachers (Dunning et al., 2003; Kruger & Dunning, 1999; Yariv, 2009).
Changes in Student Performance
For the purpose of this study, student performance was defined at the school level as school-wide student grade distributions and school-wide student discipline referrals. As established in chapter two, student grades and discipline are often approximate indicators of the overall quality of teacher instructional practices. Research question two asked if the change in teacher instructional practices would have an effect on these two measures of student performance. Analyses of data in response to question one documented a changes in instructional practices for the year of full implementation, thus a change in student performance was hypothesized.
There were some strong indicators that student performance (grades and discipline) improved more than trends would have predicted in both the pilot year and the year of full implementation. Since the changes in instructional practices clearly vary for different groups of teachers (high, medium, and low performing), the researchers expected to see differences in the student performance indicators of the students of high, medium, and low performing teachers. However, there were not significant differences in the student performance indicators for students of high, medium, or low performing teachers.
Student Grade Distributions
Although historically improvements in student grade distributions have been difficult to achieve, overall student grade distributions improved slightly (see Table 21). A ceiling effect may have influenced the grade distributions of As. The data analyses of student grade distributions suggested that some Bs moved to As and some Fs moved to Ds as shown in Figure 6. Analyses also indicated more improvement to grade distributions during the year of full implementation than during the pilot year.
As indicated earlier in a discussion of the quality of teacher instructional practices, principal ratings suggested improvements to teacher instructional practices were made in the two domains of Instruction and Assessment (see Table 15). Improvements in instruction and assessment may account for some of the improvement in student grade distributions (Haycock, 1998; Jordan, Mendro, & Weerasinghe, 1997; Rivkin et al., 2001). A behavior known as the Kohler effect, where subjects will change behavior to reflect what they perceive an observer wants, may also account for some of the improvement in student grade distributions (Lount, Park, Kerr, Messe, & Dong-Heon, 2008). That is, the nature of the data reviews, which included reviews of each teacher’s grade distributions, may have prompted teachers to alter classroom grading practices to avoid exhibiting higher failure rates than their colleagues.
It is important to note that any altering of grading policies to avoid being an outlier was a self-imposed process, since all data presented to teachers was anonymous. No one but the individual teacher and the principal knew the results of the individual classroom data collection. Although not supported by data presented in this study, principals perceived through discussions and calibration meetings that some teachers made changes in grading practices to allow failing students to improve their grades. These same changes then made it easier for other students to see an increase in their grades. According to principal discussions with teachers, many teachers extended deadlines for assignments, allowed students to retake assessments, or prove mastery of content in an alternate format. Such accommodations were often provided for all students within a class, not just those who were failing.
Another strategy that may have influenced some of the improvements in grade distributions was the use of trigger points as mentioned in chapter two. A trigger point initiated more detailed principal follow-up with a teacher who seemed to need additional assistance in order to improve grade distributions. For example, principals met with teachers whose classroom grade distributions exhibited substantially higher percentages of Fs compared to their anonymous peers (typically scores of 5% or higher than their peers teaching similar classes, grade level and subject area). Teachers who exhibited a high percentage of Fs were required to consult with the principals for help and guidance on how to improve their instructional practices so that students would be more successful academically. These teachers then began implementing various strategies, such as requiring students to stay after school to make up work and retake tests. This resulted in Fs becoming less frequent.
Overall, discipline referrals decreased dramatically during the treatment years of this study as seen in Table 23 and Table 24. The data analyses indicated that most of the reduction in discipline infractions was due to a decrease in aggressive discipline referrals as seen in Figure 7. The data analyses also indicated that essentially all of the reduction in discipline infractions was due to a decrease in male discipline referrals. For the pilot year, aggressive discipline referrals actually decreased more than overall discipline. During the year of full implementation, the drops in aggressive discipline referrals were equivalent to the drop in overall discipline referrals. Before the introduction of principal-teacher interactions, males were responsible for about two thirds of the discipline referrals; however, by the end of the year of full implementation, male discipline referrals were responsible for little more than half of the overall discipline referrals.
The drop in male discipline referrals over the course of this study may be accounted for in two ways. First, as boys mature, they learn to tame their aggressive tendencies in part through learned appropriate interactions with adults. Boys learn nurturing through conversation and example and are more aware of the presence of authority. They respond more to authority figures than do females (Gurian, 2001). Given that the quality of instructional practices did not change dramatically (see small effect size in Table 15) over the course of this study, while aggressive and male discipline were considerably less than expected, it is most likely that increased visibility of principals (as a result of snapshots) substantially contributed to the reduction of aggressive and male discipline issues. Consistent with the theory of management by walking around (Peters & Waterman, 1982), students interacting with the principals on a daily and ongoing basis may have been more likely to act appropriately. Of the four different grade levels, freshman discipline referrals were most impacted (see Figure 9). Although not supported by data presented in this study, through discussions and calibration meetings, principals concluded that freshman were more closely monitored and perhaps had more interactions with numerous adults in support of their success as they transitioned to the ninth grade. It is likely then that increased involvement with freshman by both principals and teachers impacted their behavior. Sophomore, junior, and senior discipline referrals also decreased, and saw the largest decrease during the year of full implementation.
Principal visibility increased due to the nature of snapshots which initiated a kind of positive feedback loop between interaction and discipline. That is, as principals performed snapshots, they interacted more with the students and discipline referrals decreased as a result. As discipline referrals decreased, more time was available for principals to perform more snapshots which increased the interactions between students and principals even more. This resulted in fewer discipline referrals and more time for snapshots during the second year of the study. According to Sarason (1990), eventually behaviors such as snapshots will likely become part of the school culture.
Grade Distributions for the Students of High, Medium, and Low Performing Teachers
Analyses of grade distributions indicated that there was no statistical difference between high, medium, or low performing teachers’ student grade distributions as evidenced in Table 25. Student performance data from each of these groups exhibited large variances for such a small sample size, which suggested that the variance of grade distributions within teacher groups may have been greater than variance across the three groups. Thus, one interpretation suggests that the impact of varying quality of instruction on grades may not be strong enough for this research design to measure.
Another interpretation of the grade distributions issue might conclude that improvement in classroom grade distributions was not related at all to the quality of teacher instructional practices. Although grade distributions and discipline data were reported anonymously, the data may have been affected by teachers resisting to be outliers (Lount et al., 2008). Therefore, teachers may have discussed and attempted to change and regulate performance indicators to fit statistically with their peers as a whole since the individual data was anonymous. Once again, this may be a function of the conscientious, success-driven traits of teachers (Martinez et al., 2005).
Discipline Referrals from the Students of High, Medium, and Low Performing Teachers
As discussed previously, teacher instructional practices improved for high performing and low performing teachers. Improvements in these instructional practices may have accounted for some of the decrease in student discipline referrals. However, if the number of discipline referrals were negatively correlated with quality of instruction scores, then one would expect the number of discipline referrals to also be related to the teacher score on the QIR instructional practices scale as well as to any improvement they demonstrated. Contrary to this expectation, there were no significant differences in the number of discipline referrals from high, medium, or low performing teachers as evidenced in Table 25. As with grade distributions, if the quality of teacher instructional practices were the main effect for decreasing student discipline, then students of teachers exhibiting the best instructional practices would have had fewer discipline referrals (Hall & Hord, 2000; Myers, et al., 1987).
This lack of any significant difference in the discipline referrals between high, medium, and low performing teachers may indicate that most of the reduction in discipline was due to the nature of snapshots, which increased principal visibility in the classrooms (Downey et al., 2004; Frase & Hetzel, 1990). Students actually said to the principals “Are you following me?” and “How do you know everything that is going on?” These examples illustrate student awareness of the presence of principals and the possible subsequent impact on behavior. In the future, it would be beneficial to study what effect principal visibility has on student discipline as an isolated treatment. This study used four specific principal-teacher interactions as one treatment. Principal visibility was not an intended outcome of the interactions, but it may have been responsible for much of the reduction in discipline infractions.
Changes in the Frequency and Focus of Teacher Conversations with Principals, Students and Other Teachers
Unlike the clear indication of results in student discipline referrals, results from student and teacher surveys regarding the frequency and focus of teacher conversations were mixed. According to teacher perceptions, there were changes in the frequency and focus of conversations between teachers and other teachers, but not between principals and teachers. Additionally, students did not perceive changes in the focus of teacher conversations with students.
The Frequency and Focus of Principal-Teacher Conversations
The four principal-teacher interactions implemented in this study were specifically designed to alter the frequency and focus of principal-teacher conversations, but teachers perceived little to no change in the focus of these conversations. Opportunity for principal-teacher conversations clearly increased, as indicated in the teacher survey; teachers clearly perceived a dramatic increase in the frequency of principal visits to the classroom, with at least 80% of teachers indicating more than five principal visits to the classroom during the pilot year and again during the year of full implementation. This was not surprising since a large number of snapshots did happen as documented for these years. Teachers also indicated on the survey that the average length of a classroom visit, discounting official observations, was approximately ten minutes. This result was not surprising either as, by definition, snapshots were to be about ten minutes in length. What these results do show is that teachers were aware of the frequency and the duration of the many principal visits (more than 1600 during the pilot year and more than 2400 during the year of full implementation). Teacher perceptions of what principals and teachers discussed, though, remained essentially unchanged during the course of this study as evidenced in Table 26.
Teachers did not perceive a shift in the focus of principal-teacher conversations toward curriculum, discipline, or teaching strategies. Because these topics were originally intended to be a core component of the principal-teacher interactions, this result was unexpected by the researchers. As a result of reflection and examination of the treatment in the study, the four principals who were the treatment providers were confident that all four principal-teacher interactions yielded more principal-teacher conversations focused on curriculum, discipline, and teacher strategies. However, teachers did not perceive these same changes.
There are several possible reasons for this disconnect between the perceptions of the principals and those of the teachers. First, at this school, principals and teachers had many conversations prior to the pilot year on a frequent basis. It may be that the collaborative nature of the treatment used in this study was perceived as friendly and benign as discussions about weekend plans or family. Teachers seem much less sensitive to principal-teacher interactions of a collaborative nature, not perceiving them as threatening (Peters & March, 1999). Second, it may be that the survey instrument itself was not sensitive enough to detect the actual changes in principal-teacher conversations which occurred during this study (see validity and reliability of teacher and student surveys in chapter three).
The collaborative nature of the treatment does not help explain why teachers did not perceive a shift in emphasis in principal-teacher conversation related to curriculum and instructional strategies. Each of the principal-teacher interactions implemented in this study had both curriculum and teaching strategies embedded in the treatment. According to Schunck, Aubusson, and Buchanan (2008),
“We argue that the contribution of this role [principal] is not merely in what the critical friend offers to the observed teacher, but rather, lies in the opportunity for discussion that probes the assumptions of all concerned, challenges views of what good teaching looks like, and enables analysis of the practices of all concerned. An understanding of the complexity of the factors that made this process work for us will inevitably be partial, yet important factors would seem to include: willingness to take risks; respect for one another’s expertise in teaching; and ability to reflect collaboratively on our teaching and learning. The strength of the process is difficult to quantify and lies in our acknowledgment of its complexity; it cannot be reduced to a checklist of critical factors. We are reminded by the process of how intensely personal is professional” (p. 225).
Considering the decrease in discipline which occurred over the course of this study, it is logical to think that such a dramatic drop in discipline would provide fewer occasions for principals and teachers to discuss discipline issues as evidenced in Table 23. Therefore even if principals were initiating discussion with teachers about discipline, thereby increasing discipline discussions, fewer actual discipline referrals could counteract the increased discussion and result in an overall drop in principal-teacher discipline conversations.
The Frequency and Focus of Teacher-Teacher Conversations
The set of four principal-teacher interactions implemented in this study were not designed specifically to affect changes in frequency and focus of teacher-teacher conversations but, according to teacher surveys, changes in the frequency and focus of teacher-teacher conversations were indicated in the pilot year and sustained through the year of full implementation as evidenced in Table 26. Beginning with the pilot year and continuing through the year of full implementation, teachers reported speaking with each other much more on a daily basis. This increase in teacher-teacher conversations may have been a result of teachers discussing changes in principals’ behavior or a response to the added level of attention to classroom level data, grades and discipline, emphasized during the two years of this study (Lount et al, 2008).
Results from teacher surveys also indicated that teachers more frequently discussed curriculum and discipline issues with each other in response to the treatment in this study. Each of the principal-teacher interactions in this study attempted to intentionally increase the teachers’ reflections on instructional practices. So it is logical to conclude that if teachers talked to principals about curriculum and discipline, they would talk to each other about similar concepts. According to Schacter & Thum (2004), although there are vast differences between disciplines, there are common pedagogical issues that teachers can discuss that will improve instruction. The survey results indicate that these types of teacher-teacher conversations increased as a result of the treatment from this study.
The Frequency and Focus of Teacher-Student Conversations
By definition, the quality of teacher instructional practices includes the frequency and focus of teacher-student conversations as discussed in chapter two. Yet, according to student surveys, essentially no change in the frequency of teacher-student conversations occurred during the course of this study as evidenced in Table 27. This may have been a result of resistance to change in culture as noted by Sarason (1990). The results from student surveys may indicate that students were not sensitive to changes in the frequency and focus of teacher-student conversations. Students often focus on classroom assignments and social issues in their school; and although students claim to want direct feedback from teachers, they do not care for overly personal conversations or interactions with teachers (Cushman & Delpit, 2003). Students are also accustomed to changes in classroom interaction dynamics with teachers because they have different teachers for different classes. So many periodic changes for students may result in change becoming so commonplace that students fail to recognize it. Hence, responses to a survey about changes in conversation patterns with teachers may not indicate changes in the previous patterns. It is possible that the survey instrument provided to students was not sensitive enough to detect changes in the frequency and focus of teacher-student conversations (see the discussion of reliability and validity of teacher student surveys in chapter three).
Study in the area of specific treatments with teachers, especially lower performing teachers, adds to the body of research already available on teacher treatments from Danielson (2007), Marzano (2003), Fullan (2005b) and Whitaker (2002). The amount of time required to improve instructional practices for underperforming teachers presents a significant challenge. We were disappointed by the insignificant changes noted in the quality of instructional practices for medium performing and low performing teachers after the introduction of the four principal-teacher interactions. Consistent with prior research (Dunning et al., 2003; Kruger & Dunning, 1999), low performing teachers often do not have the ability or the knowledge to accurately self-assess their own practices and may require more intense treatment to improve. Given our experience, the same treatment that demonstrates impact on high performing teachers would not induce improvements in the quality of instructional practices for medium performing and low performing teachers.
Principal Visits and Collaboration with Teachers
Teachers are a diverse group of professionals who are passionate about their students and their content (Schacter & Thum, 2004). At the high school level, the many diverse content areas and specialties make it a challenge for principals to establish themselves as instructional leaders in all areas (Ginsberg, 2001). Yet, good teaching demonstrates many of the same characteristics whether in a physics class or in a history class. If the principal establishes a clear understanding of high quality instructional practices, teachers will recognize that principals can be helpful in improving instruction in any classroom (Danielson, 2007; Marzano, 2003). Further, it is essential that teachers know that the principal is familiar with their classrooms and instructional practices in order to establish mutual trust and true collaboration. Such relationships develop from rich dialog and frequent classroom visits that create collaboration between teacher and principal (Ginsberg, 2001; Marshall, 2008).
The principal-teacher interactions used in this study were implemented as one treatment. As such it was not possible to quantify which specific interaction may have had more influence than another; however, the principals at this school suggested that snapshots had a more significant effect than other interactions. The most important component of snapshots is the structure it provided for collaborating with teachers side by side in the classroom and being involved directly with instructional practices (Hall & Hord, 2000). Although no empirical evidence is presented in this study to support it, the researchers believe that visiting classrooms daily and principal involvement in daily teacher instructional practices demonstrated more impact than the other principal-teacher interactions. It takes time and energy, but when snapshots are in place, the results are rewarding for principals and teachers (Downey & Frase, 2001).
Rubric Based Assessment of Instructional Practices
It is far easier to assess teacher improvement or growth when a principal is frequently in the classroom. The anecdotal records, conversations, and collaboration from classroom visits establish an environment for personal growth that cannot easily be documented on a more traditional checklist of teacher skills. Conversely, rubric based evaluations, as defined by Danielson (2007), provide teacher and principal an effective tool for improving instructional practices. The rich dialog that comes from discussing what is observed with the teacher and the teaching artifacts furthers the discussion and makes the resultant collaboration invaluable (Cordingley, Bell, Rundell, & Evans, 2003; Danielson, 1996; Downey & Frase, 2001; DuFour & Marzano, 2009). Traditional evaluations that rate teachers on a scale from 1-4 without a clear description of what each rating means are ambiguous and ineffective (Danielson, 2007; Halverson et al. 2004). Thus, for quality evaluations to take place the process of dialoging between the evaluator and evaluatee around a rubric-based instrument must occur; and a mutual understanding of the learning taking place in the classroom must be developed (Danielson, 2007; Downey et al., 2004, DuFour, DuFour, Eaker, & Karhanek, 2004; Halverson et al. 2004).
Working with Teachers of Differing Qualities of Instructional Practices
The various levels at which teachers perform present a challenge for principals (Villegas-Reimers, 2003). The principal must identify the strengths and weaknesses of the teachers and shift strategies when working with each individual. High performing teachers need feedback to validate their quality work. They also tend to be overly critical of their own skills; and this behavior, while contributing to improvement, often increases the likelihood of job burnout (Dunning et al., 2003; Kruger & Dunning, 1999). According to Pratt (1977), positive feedback to teachers is a powerful tool and motivator for high performing teachers.
Medium performing teachers need positive feedback as well as discussion to improve their instructional practices. These teachers may fall through the cracks and not change their instructional practices because they go about their day to day routine with little feedback, but never do anything to garner criticism. Research (e.g. Danielson, 2007; Marzano, 2003; Fullan, 2005b; Whitaker, 2002) has demonstrated that improvement occurs when principals intentionally collaborate with all teachers, not only on evaluation years, but each year. Principals must collaborate with medium performing teachers because, with coaching, these teachers can become higher performing, while others demand attention to prevent falling into the lower range of performance.
There seems to be no practical way to improve instructional practices for teachers who are unable or unwilling to accurately assess the quality of their instructional practices. Despite the treatment used in this study, low performing teachers as a whole stayed low performing while perceiving themselves as high performing teachers. According to Goldsmith & Reiter (2009), Coaching is most successful when applied to people with potential who want to improve—not when applied to people who have no interest in changing. This is true whether you are acting as a professional coach, a manager, a family member, or a friend. As discussed earlier, low performing teachers may not have the ability to self-evaluate accurately or significantly improve instructional practices even with the help of other more capable evaluators (Dunning et al., 2003; Kruger & Dunning, 1999).
When designing the treatment for this study, the researchers did not consider the possibility of unintended outcomes. We did not anticipate an initial increase in the number of weak teachers leaving the school. We did not anticipate the development of close relationships between principals and students, or that snapshots would supply valuable information to improve the effectiveness of our conversations with parents. Finally, we did not anticipate the increased level of job satisfaction for principals.
Twelve Dixie Heights teachers resigned from the start to the finish of the study; seven at the end of the pilot year and five after the year of full implementation. This level of turnover is not typical for Dixie Heights, which lost on average fewer than three teachers annually in each of the previous nine years and no more than five in a single year. Various reasons for the exodus were reported by exiting teachers. At exit interviews, nine of the 12 teachers expressed concern and dismay that principals were in their classrooms so often as a result of the snapshot interaction; 10 out of 12 cited data being distributed to the entire faculty for review and; a third factor, reported by eight out of 12, was the monitoring and discussion of discipline in data reviews. All of the exiting teachers further reported that if “left alone” they could continue to teach, but could not teach under constant supervision.
Another unintended outcome from the study was the increase in relationships that developed between principals and the students. By the end of the study, many students indicated an increased level of comfort when approaching principals with potential problems or concerns, which actually prevented more serious issues before they developed. Students often made recommendations to principals to prevent specific student altercations, and when students were asked why they chose to more readily cooperate with the principals to reduce disciplinary issues, the typical students responses were, “I see you [the principals] everywhere, so I thought you could help”. As a result of the increased principal-student communications, student fights were prevented, squabbles mediated and potentially tardy students were sent to class in a timely fashion. Principals being visible throughout the school and in the classrooms also contributed to the dramatic decrease in student discipline referrals. When asked about the classroom visits by the principals during conversations, students indicated they believed the principals were more engaged with the school than in the past. Often principals were able to prevent issues simply by being present to talk with students and redirect behavior (Downey et al., 2004; Frase & Hetzel, 1990) and these efforts helped increase the efficiency of the school as well as teacher instruction.
In addition to the unintended positive increase in principal-student discussions, the quality of principal-parent discussions also improved. In many instances, when parents came to school to discuss teachers’ instructional practices, principals were able to dispel rumors or misconceptions much more effectively than in past years. For example, a parent of a high achieving student came into a meeting with other parents and reported that a specific teacher “just sat and read books” and told the students to “figure it out for themselves.” The principal was able to access a tracking document of snapshots and provide documentation of 67 classroom visits reporting only high quality instruction.
Increased Job Satisfaction for the Principals
Perhaps the most surprising and unintended result of this treatment was the discovery that implementing the principal-teacher interactions at the core of this study may reduce problems that often lead to principal burnout. Chief among these problems is the lack of substantive principal connections to the students and teachers. Whitaker (2003) noted that in addition to this isolation, principal burn out results from job ambiguity, loss of autonomy and the increasing demands of the position. According to Duke (1988), principals wanted to be a part of the faculty and involved with students in order to increase job satisfaction, and the execution of the treatment used in this study added much enjoyment to the principals’ work by increasing the frequency of positive interactions with students and teachers. Two of the principals reported, “This is why I got into administration, to be with students, and help teachers.” Another principal said that she “would never leave the building if the principals could keep this program going in the same direction.” One principal, who was in charge of the majority of discipline for the school said, “I did not believe that being in classrooms would help keep discipline to a minimum, but the time in classrooms made my job easier.”
Recommendations for Future Research
Further research on particular treatment needed for teachers at various levels of performance would add to the body of knowledge currently being used to help teachers improve their instructional practices. The treatment used in this study must be applied to other schools to increase the generalizablitiy of this study. Since this study was conducted in one school, we cannot conclude it would work the same way in other schools.
Further research is also recommended to determine how principal interactions in the classroom could strengthen and support the walk-through model currently used by many schools and districts. This study demonstrated that a district could strengthen its walk-through with more frequent, informal visits to familiarize the principal with the teaching strengths and weaknesses of teachers. These principal-teacher interactions would also strengthen conversations between the teachers and principals and provide solid examples of quality teaching.
Additional research will be required to discover possible implications of the individual effects of each of the four principal-teacher interactions used in this study. For example, data reviews may have directly affected grade distributions in this study, but would data reviews as an isolated treatment improve grade distributions in a school?
Principals being present in classrooms on a frequent basis through snapshots seemed also to directly affect discipline, but could an increase in principal visibility in classrooms as a single treatment improve discipline in a school? Given that these effects were unintended consequences for this study, it seems appropriate to investigate them further.
Core Principal Questions Affirmatively Answered by this Research
Ultimately, there are limits to what principals can directly affect in a school, but principals do have the opportunity of changing how they interact with teachers. The overall goals of this study were to determine how a specific set of principal-teacher interactions affected instructional practices and to provide practical strategies for principals which can improve a school without accessing additional personnel or financial resources. The following table of questions and answers (Table 28) does not represent any new information beyond the discussion presented earlier in this study. These questions and answers do however address questions a principal might ask which can be answered from the results of this study.
Core Principal Questions Affirmatively Answered by this Research
In this study implementing a few specific changes to principals’ behavior enabled the principals to function more effectively as instructional leaders. A principal cannot control every aspect of a school, but they can control the way they interact with teachers. Teachers are contentious professionals who desire and benefit from quality collaboration for the purpose of improving the quality of instructional practices. Improving instructional practices is essential to increasing student performance within a school. This study concluded that the introduction of a few collaborative principal-teacher interactions incorporated into the school day, improved instructional practices, student performance, and student behavior while enhancing principal job satisfaction without the use of additional personnel or financial resources.