A study of how specific principal behaviors affect teacher and student performance
One-on-One Summer Meetings
During the summer of 2008 (between the pilot year and the year of full implementation) each teacher individually attended a one hour summer meeting with at least one principal. Each of these summer meetings were preceded and followed by a principal team discussion to properly prepare for the meeting and then debrief. Each meeting followed the same basic structure but details of the discussions varied depending upon the needs and reactions of the teachers. The length of each meeting was planned to last about one hour, based on the number of teachers and available time of the principals during the date range.
Each summer meeting addressed past performance and goals the teacher had for the coming year; a review/discussion of the teacherís instructional practices as related to grade distributions/failure rate, discipline records, state tests scores, previous formal evaluations, the teacherís former and future individual professional growth plan, the teacherís use of technology, and the future focus for the principals regarding each teacherís needs over the next year. These were detailed in the QIR (see Appendix C for a copy of the Quality Instruction Rubric). The topics of discussion for the summer meetings were decided by the principals following a review of data collected during the previous year, grade distributions and discipline referrals, and synthesis across prior snapshots from the teacherís classroom.
Snapshots were formatted to draw from observer behaviors as associated with, but surpassing, the walkthrough processes such as Management by Walking Around (Peters & Waterman, 1982) and the Three-Minute Walkthrough (Downey et al. 2004). Specifically, principals visited several classrooms weekly to assess each teacherís progress as it pertained to items on the QIR. Each snapshot visit lasted approximately five to fifteen minutes and the principal became part of the class when possible by taking part in the educational process to aid and model proficient instructional practices. For example, if a class was involved in a discussion the principal contributed or asked questions as appropriate to improve the educational process. If students were working independently, the principal walked around the room and interacted with the students to observe and improve the quality of the independent practice. If students were taking a test, the principal walked around and aided in supervision while observing the type and quality of the assessment. If students were involved in group work, the principal circulated while observing and interacting in a way which advanced the studentsí learning. Regardless of the educational practice being used in the room where a snapshot was taking place, the principal attempted to become part of the teaching process while modeling proficient instructional practices. For detailed descriptions of proficient instruction practices, see the QIR in Appendix C.
After a snapshot or series of snapshots, not to exceed three visits in one week, the principal provided feedback to the teacher based on the QIR. Feedback took the form of email, hand written/typed notes, or verbal exchange, based upon the teacherís preference as determined in advance between the principal and the teacher. For example, some teachers requested oral communication so principals communicated with those teachers orally. Other teachers requested communications by email, so those teachers received email communications.
Once every two weeks the principals met to calibrate snapshots taken during the previous weeks as well as to make adjustments to patterns and behaviors for future snapshots. Special attention was given to snapshots and interactions two or more principals had with the same teachers. This was used as an indicator of how well each principal was calibrated. Notes, quality of instruction, frequency, and number of QIR indicators were discussed. At least once every two months district level personnel accompanied principals on at least four snapshot visits in order to fine tune the authenticity of what principals were noting during snapshots.
The principals did not keep notes or other qualitative data during snapshots for the following reasons:
∑ Maintaining the collaborative nature of principal-teacher interactions
∑ Assuring teachers of the intentional separation of the set of principal-teacher interactions and official evaluation
∑ Avoiding policy issues associated with official records and official evaluation
∑ Avoiding issues with the teacher union
Qualitative data is an essential component of the snapshot interaction as well as an important component in all treatment. As described earlier in the section Treatment Specifics, the follow up to snapshots included feedback from principals to teachers which was intended to promote collaboration and improve instructional practices. Recording qualitative data of snapshot observations may have inhibited the productivity of these principal-teacher interactions, so the principals did not take notes. Additionally, according to policy officially recording observations on teacher behavior becomes part of a teacherís official records. Such official observations could have invoked a number of rules and policies that would have a detrimental effect on the collaborative intention of the snapshots. Thus, the decision was made to intentionally NOT retain any written documentation of the snapshot observations or feedback in order to avoid undermining their core purpose. Officially recording observations would have been cumbersome and may have been perceived as negative by the teachers.
Each teacher received several sets of data throughout the year related to their studentsí performance. These data included details of how many students were assigned an A, B, C, D or F for a final grade in all the classes which were just completed (see Appendix F for examples). The numbers were compiled into one graph. At the end of each trimester each teacher received numeric and graphic representations of their classroom grade distributions as well as data indicating how these results related to the other teachersí data within the school and the school as a whole. At the end of each trimester each teacher received numeric and graphic representations of discipline infractions they reported to the principals, as well as how these results compared anonymously to the other teachers within the school and the school as a whole.
Teachers were instructed by email to examine their data and reflect upon their growth plans as they examined their data. Teachers who exhibited unusually high failure rates, abnormal grade distributions, or an unusually high number of discipline referrals were called for a one-on-one meeting with the principal to brainstorm strategies for improving teacher instructional practices.
In September, each teacher received numeric and graphic representations of their studentsí performance on the previous yearís state assessment as well as results for each department and the school as a whole.
Teacher Self-Assessments (QIR)
At the beginning (August 2008) and then again at the end of the year (May 2009), teachers completed a self-evaluation using the Quality Instruction Rubric to evaluate their perception of their own instructional practices in four domains of the QIR. The four principals independently completed assessments of each faculty memberís instructional practices using the same assessment tool at the same times as described earlier. Each set of results, teachers/principals/beginning/end of year was analyzed and compared using t-tests; see data analysis section for more detail.
For this study trigger points were defined as indicators of concern from summer meetings, data reviews, snapshots, or formal observations that prompted an increase in the number and frequency of principal-teacher interactions with specific teachers. Often these trigger points were discussed during the calibration meetings principals held once every two weeks and included items such as: teacher seemed resistant to change exhibited during summer meetings or planning period meetings, students expressed concerns about teachers instructional practices to principal, continued poor performance on the same indicators for the QIR after multiple snapshots and feedback from principals, teacher request, or poor official classroom evaluation.
Teachers with poor performance as defined by the QIR or data anomalies in the grade distributions or discipline referrals were called to the principalís office for more intense principal-teacher interactions. The focus of this meeting was to trouble shoot and design strategies to improve teacher instructional practices and increase student performance.