Research Articles and Papers from ASSC





A Paper Presented at the Annual Meeting of the American Educational Research Association, Chicago, April 2003


Sharing the Data along with the Responsibility: Examining an Analytic Scale-based Model for Assessing School Climate

Abstract
This study reports the findings of a pilot effort to examine the efficacy of an analytic trait scale school climate assessment instrument and democratic change system in two urban high schools. Pilot study results indicate that the instrument showed promising soundness as it exhibited high levels of both validity and reliability. In addition, the analytic trait format demonstrated the capacity to promote deeper levels of participant reflection and encourage greater participant empowerment and better process efficacy than its objective survey counterparts. Data analysis suggested a 4-quadrant model for examining "climate goal orientation" (included in the paper). Given what appeared to be an inherently socially constructed quality to each school's collective definition of what they considered "good school climate," the use of the analytic instrument and a transparent assessment process proved useful. Moreover, the participant-driven feature of the system demonstrated a high degree of efficacy in helping participants move effectively from the climate assessment phase to subsequent vision setting and action phases. Findings related to the climate and assessment process at the two schools as well as overall study conclusions and implications for practice are offered in the paper.


As schools seek ways to improve educational quality, the search often leads to an examination of their climates. While school climate improvement appears to have taken a back seat to more "direct" efforts to effect academic performance, evidence persists that initiatives for improving school climate can positively impact student learning (Bulache and Malone, 1994; Freiberg, 1999; Hoy & Hannum, 1997; Keefe and Keller, 1990; Stewart, Evans and Kaczynski 1997). Moreover, because school climate is predictive of many other indexes of school effectiveness, it is a fertile arena to explore. Many of the critical issues that face schools today are related to climate (Bobbett, 1991; Bulache, 1994; Nidiche and Nidiche, 1986). No wonder that so much has been invested into assessing school climate. One indication of this investment is the number of school climate and/or culture inventories. In a 1987 Report (ERIC Reports: Northwest Regional Educational Laboratory), 42 separate school climate inventories were identified. Since that time other school climate inventories have also emerged (Bernardo, 1997; Bobbett & French; Crawford and Irwin, 2002, Butler and Rakow, 1995; Haynes, Emmons and Comer, 1994; Jones, 1996; Olafson, Bendixen and Tirella, 2002; Worrell, 2000). Given the shear number of these instruments and the diversity of their representations of "climate," there appears to be a need to clarify the basic theoretical conceptions underlying the domain of school climate. Moreover, there are fundamental concerns with the efficacy of any school climate improvement system featuring a survey type inventory driven by university researchers, given that the school stakeholders are often not meaningfully involved in the assessment process (Fullan, 1992; Zeichner, 2002).

The Pilot Study of the ASSC Assessment System
The paper reports the results of a pilot examination conducted by the Alliance for the Study of School Climate (ASSC) of an analytic scale-based system for assessing school climate in two urban high schools. The eight-factor analytic trait instrument piloted in the study was derived from school climate factors identified by previous research into high functioning schools as well as theoretical relationships among the eight factors. The assessment instrument under investigation was part of an overall system for facilitating a participant-led democratic climate reform process. This proposed process would include an initial phase related to reflection and self-assessment, a second phase for democratic vision setting, a third phase for strategic planning, followed by a fourth phase characterized by faculty-led plan implementation. This process is consistent with those of other research-based reform frameworks (Fullan, 1992; Marriott, 2001). The foremost purposes of the pilot effort were to validate the soundness of the instrument as well the broader system for school change. The study offers findings related to the system validation process as well as the process of school climate assessment within two schools.

A Theoretical Comparison of Assessment Instruments and Systems
The prevailing model for assessing school climate involves the use of survey-type objective inventories, which are administered and/or scored by outside agents. These objective surveys are often given to parents, teaches, and/or students from which a form of numerical "school climate index" is derived. This approach to assessing school climate is typically driven by educational researchers who conceive their task within a quantitative framework. The advantage of such a model is its potential ability to make comparisons of statistical significance among groups or schools.

The alternative model examined here, and developed by the researchers who formed the Alliance for the Study of School Climate at California State University Los Angeles in 2001, utilizes an analytic trait scale instrument for use by a committee of stakeholders assisted by university partners. Analytic trait type instruments have been shown to have several advantages over surveys, have the capacity to be more sound than objective surveys (Shindler, 2002; Stiggins, 2001), and are gaining wider use, especially in the area of program assessment (Henderson and Hawthorne, 1995). One of the significant benefits of using analytic school climate instruments may be thier capacity to provide their users with a basic construct for defining good school climate, a capacity virtually absent in objective survey instruments.

The two survey items shown in Figure 1 illustrate the differences in item construction between the two types of instruments. Item A is taken from an objective survey type format instrument whereas item B is taken from the revised version of the analytic trait scale type instruments examined in the study.

Figure 1: Comparison of Assessment Instrument Item Formats

Item A: Objective type Survey Item Example1:


Item B: Analytic Trait Scale Instrument Item Example:

High	Middle	Low
Teacher-student interactions could be typically described as supportive and respectful.	Teacher-student interactions could be typically described as fair but teacher-dominated.	Teacher-student interactions are mostly teacher-dominated and reactive.

Methodology
Data were collected in four phases. First, in each school, the principal selected a convenience sample of teacher participants to act as the pilot "steering/assessment committee." (N=6 and 7). In each case the sample consisted of teachers who shared a particular planning period. This committee was interviewed regarding overall perceptions related to the state of affairs at their school and were given version 1 of the instrument to examine and make suggestions. After receiving participant feedback, revisions were made to the instrument items and design. Second, the revised analytic instrument (see figure 1 for sample items) was re-administered to the teacher participants, and two classes of student participants at each school, representatively sampled from the school as a whole (N=51 and 57). After administration of inventories, participants were interviewed using a semi-structured focus group technique. Third, research team members made independent assessments of the school in each area of the instrument. Data for this external evaluation included classroom observations, informal interviews with student and faculty participants, participant-observation from steering committee meetings, and observation of day to day school life. Fourth, teacher participants were interviewed regarding their perceptions of the potential efficacy of the instrument and the system as a whole. Data were analyzed using a constant comparative method. A grounded theoretical framework for analysis was developed (see Figure 2).

Study Findings

Soundness of the Analytic Scale-based Instrument
The pilot version of the analytic trait instrument showed promising soundness as it was administered to teacher participants at each school. Subsequent analysis of soundness was done for the revised instrument (version 2) in the areas of validity, reliability, efficacy and usability.

An examination of the validity of the instrument demonstrated four findings. First, after the initial administration of the instrument (version 1), teacher participants reported that the items were relating to "actual school reality," and "made sense." While they had minor suggestions for item changes, overall they felt that the items depicted a genuine picture of school reality. This appraisal suggested a sufficient level of what could be considered content and/or face validity.

Second, participants, given the revised instrument (version 2), reported seeing the relationship among items in each sub-scale area, showing evidence of sub-scale construct integrity. Moreover, their ratings for items within each sub-scale tended to show a great deal of consistency. This integrity could be seen further in the reliability of mean ratings of items in each sub-scale between teacher and student participants.

Third, further construct validity was observed in version 2 in the degree to which its sub-scales scores assessments inter-related. Most participants suggested that their overall ratings for any one sub-scale were similar to their ratings for the others. This consistency was confirmed by the high correlation among participant's sub-scale mean ratings. Notwithstanding the high degree of rating consistency, the intention of such an assessment instrument is for some independence of factors. An analogy might be that of a race among 8 runners that are all linked by ropes. There will be a first and last place finisher, but in the end all of the racers will finish within a limited range. In other words, a school may perform better or poorer on any particular factor, yet the performance on any single factor will always be buoyed or pulled down by the location of the mean of all factors collectively. This finding suggests that sub-scale domains may be theoretically independent, yet the practical interdependence of any teaching practice or aspect of school life to another appears to be evident.

Fourth, as participants and researchers examined the 3 "levels" in the instrument, a theoretical construct emerged for each. This was first observed when participants often noted that the "3 level" was impossible or could not be achieved in their urban school. Participants at the lower performing school suggested that with the students they had, "3 level" teaching and climate were beyond their reach. This sentiment reflected the identification of a "type" of school that was being characterized at each level. This analysis led to an attempt at conceiving these levels within a theoretical framework to explain each school "type" (Figure 3). This appearance of a relationship between item level and something that could be considered school "type" demonstrated further strength of the fundamental constructs of the instrument. While the classification of any particular school into a type is not the intention of the instrument or the assessment exercise as a whole, there seems to be some practical value to the existence of patterns among the levels.

The reliability of the pilot instrument was seen as important because it demonstrated that the instrument depicted commonly understood reality and articulated that reality accurately. However, variation in each participant's climate ratings would be assumed due to differences in perceptions.

Overall, most items demonstrated acceptable inter-rater reliability, as shown by a strong correlation among student, teacher and researcher ratings. From data collected in the focus group interviews, it appeared that most of the variance in ratings was due to subjectivity in perceptions of the reality that existed at each school, and little seemed to come from a disparity in item interpretation. This finding supports the theoretical assumption that an analytic scale assessment instrument would promote a high degree of inter-rater reliability due to the concreteness and clarity of its structure (Stiggins, 2001, Shindler, 2002). In addition, the concrete language provided by the three descriptive options (versus an objective survey item with a agree of disagree choice from a single description, see Figure 1), provided participants a better conceptual anchor, improving intra-rater reliability. Figure 2 illustrates the language used in the revised instrument.

After administering the pilot version (version 1) of the instrument, interviewing participants and analyzing the phase 1 data, the following changes were made to the revised version (version 2) to improve its technical performance and theoretical soundness. First, language was changed to make items clearer, more universal and inclusive of regional vocabulary and jargon. Second, some items were improved to better create a parallel structure and better reflect a continuum of practice/behavior. Third, items were added that were seen as missing elements of each of the eight factor constructs. Fourth, given that each item reflected a continuum of practice, it was determined that giving participants three options per descriptor provided an opportunity to give "high," "middle," or "low" ratings for each. As a result, each item could be scored on a nine-point scale. This change provided better performance level discrimination. And, while it may have weakened the theoretical reliability of the items, this was offset by improved consequential validity. That is to say that it better provided participant raters the perception that they were rating practice/performance across a range or continuum. This benefit would prove additionally useful later in the process of application as participants attempted to diagnose where changes were necessary. In this planning phase, participants, given their ratings in front of them, instead of asking "which descriptor best characterizes where we are?" were more inclined to ask, "where do we currently fall on this continuum?"

Figure 2: Sample items from scales 2 and 7 from revised analytic instrument:

2. Faculty Relations
Level - 3			Level - 2			Level - 1
High	Middle	Low	High	Middle	Low	High	Middle	Low
Faculty members commonly collaborate on matters of teaching.			Most faculty members are congenial to one another, and occasionally collaborate.			Typically faculty members view one another competitively.
Faculty members approach problems as a team/collective.			Faculty members attend to problems as related to their own interests.			Faculty members expect someone else to solve problems.

7. Attitude and Culture
Level - 3			Level - 2			Level - 1
High	Middle	Low	High	Middle	Low	High	Middle	Low
Students feel as though they are part of a community.			Students feel as though they are part of a society.			Students feel as though they are visitors in a building.
Students self-correct peers who use destructive and/or abusive language.			Students seek adult assistance to stop blatant abuse.			Students accept abuse as a regular part of their day.

The revised instrument proved to be sufficiently user friendly. Both teacher and student participants were able to complete the 73 item ratings in about 20-30 minutes. Only minimal instructions were required relate to the "high," "middle" and "low" classifications. Moreover, the analytic trait instrument demonstrated high levels of "usability" within the context of the vision-setting phase. The analytic instrument provided participants in each school not only a sound diagnostic instrument, but also a basis for initiating a discussion about prescriptions for improvement. As participants noted where they were on the continuum for each item, they could be seen taking advantage of the conceptual and practical language the items implied for what might constitute improved performance.

Functioning of the Participant-Driven Assessment Process
Considering that the "planning phase" of the process for teacher participants was undertaken as a "hypothetical" exercise, the process itself showed the promise of effectiveness. Participants agreed strongly with the contention that assessment data in the hands of the users would help promote a better and more enduring set of outcomes. And while the exercise was done within a pilot context, each set of teacher participants grew in the sense of ownership of their respective processes.

Recall that each group of teacher participants was selected for convenience. This quasi-random sampling method was useful for purposes of assessing the efficacy of the instrument and obtaining a representative sense of the faculty attitudes in general, but create limitations for a genuine change effort. As the efforts progressed, each of the school's principals questioned whether the same teachers who took part in the assessment phase of the effort would be the best ones to be involved in the planning phase.

A more purposeful selection of team members seemed necessary if the efforts were to be fully successful. While the representative composition of the participants was a benefit, there appeared to be qualities that would be requisite in those that could effectively guide the next phases of the process. Some of the necessary qualities identified by the principal of the high performing school included a commitment to the effort, a visionary orientation, and some political sensitivity and/or political capital. While, both principals concluded the members of the faculty group were in many cases not those that they would have chosen if the task had involved an implementation phase, enlisting an entirely new team for a subsequent phase of the process would prove problematic. Due to the significance of the learning curve for participants who had gone through the process of acquainting themselves with the "state of the school climate" while taking part in the assessment phase of the process, new members may not be able to duplicate this immersion into the process of determining the starting point for planning and taking action. Therefore, it appears that a thoughtful selection process of vision committee members at the beginning of the process would be desirable for the coherence and the long-term efficacy of the effort

The response of each school to the opportunities provided by the assessment pilot efforts varied somewhat. In the low performing school, the exercise was greeted with a sense of hope. These participants reported seeing a series of new programs come and go over the years with few seeming to systemically address the school's more substantive needs. They could see the validity of an effort driven by faculty and guided by a more democratically derived vision. They saw the assessment instrument providing the beginnings of a road map for their reforms. When asked how much of a role the university team should play in the implementation phase of the process, these participants suggested it should be about 70% school participant-driven and 30% university partner-assisted.

Participants in the high performing school shared the perception of the need for the process to be driven primarily by the school team, however, their perceptions of the pilot process varied because they did not feel the same degree of need for change. Like the low performing school's teachers, they saw the validity of the democratic process feature of the system, and thought that the instrument would be useful in their efforts to formulate plans for change, but were not as emotionally committed.

Academically the two sets of participants provided similar insights, largely supportive of the components of the pilot system, yet the school that perceived a greater need showed an eagerness to "do it for real," while the other school felt relatively comfortable with where they were going. These teachers seemed not to move beyond an academic engagement in the process. However at the low performing school there was evidence of emotional buy-in to the process. For example, when the principal at that school publicly made the decision to discontinue the process after the assessment phase, the teacher participants were visibly disheartened. They felt an opportunity was missed. In contrast, a vision and ethos of change appeared to be in place at the high performing school, which may explain what appeared to be a less enthusiasm for participating in the school climate assessment effort.

Assessing Climate: Common Findings from Plot-Schools
Several findings were suggested from an analysis of the data regarding the quality of school climate at each of the two schools. Many were unique to each school in particular, but four were common for each and therefore could be considered valuable in examining the nature of the process itself.

First, the school climate experienced by students in various academic "tracks" at each school seemed to vary significantly. Both teacher and students participants suggested that there were essentially three climates at each school - the advanced placement track, the regular track and the basic track. The methods used and the experiences of the students in each track were rated at different locations on the continuum, with the advanced placement track being the highest and the basic track being the lowest.

Second, an analysis of the data from the focus groups and the process of survey revision demonstrated the "socially constructed" aspect of defining any "good school climate." Participants agreed that a questionnaire style survey is more likely to mask this element. With an analytic trait type instrument there is a relative transparency to the constructs. Instead of making the assumption of an "a priori" definition, examining the notion of what makes for good school climate appears to have benefits for teachers undertaking an effort to first diagnose their school's ills and then prescribe remedies.

Third, this socially constructed nature to the concept of "good school climate" could be seen clearly as participants examined the "level 3" characterization of school climate in the instrument and found that the descriptions of practices to be in many ways unfamiliar not just relative to what they were doing, but what they were trying to do. The process of reconciling these level 3 descriptors with their own situation, needs, capabilities, and goals, proved useful but at times troubling to many participants. For example, after completing the initial version of the survey and being asked for their impressions, one participant's response was, "My thought was that now I need to change everything I do." One could conclude from this statement the goal assumptions he held were in conflict with the implicit goal logic reflected in the instrument. This conflict could be seen in many of the participants as they rated their schools 5 or 6 out of a possible 9, in areas where they had reported being rather satisfied with their own performance or the performance of the school. Each participant's and/or school's definitions of "practices they felt promoted a good school climate" could be seen as bound by individual socio-cultural values influenced by region, class, ethnicity, religion, and teacher training.

Fourth, through analyzing the data related to that which participants expressed as their "goals" related to school climate, a construct emerged to classify "goal orientation." Figure 2 outlines this two-axis matrix for climate goal orientation.

Figure 3: Four-quadrant matrix of school "climate goal" orientation:

Figure 3 depicts a useful framework for examining the "goal logic" of a particular teacher or school as it relates to school climate. The matrix shown in Figure 3 is derived in part from the attempts of previous research to classifying teacher practice (Finn, 1999; Shindler, Jones, Taylor and Cadenas, 2003), and data from the current study. It reflects the intersection of a continuum characterized by two key variables that define goal orientation. First, any teacher and/or school practice could be considered more or less effective, inasmuch as it produces desirable outcomes. This variable includes the component of intentional versus unintentional practice. Data from this study supports previous research (Shindler, Jones, Taylor and Cadenas, 2003) suggesting that in the absence of intention, practices tended to be either ineffectual or authoritarian. Second, any educational practice could be placed on a continuum from more teacher-centered to more student-centered. Applying the findings of Finn (1999) the component of liberation/facilitative practice vs. domesticating/managerial practice was overlaid upon this continuum. The resulting construct provided an addition element to the examination related to the "effect" of practice in contrast to simply examining the nature of the practice itself. The juxtaposition of these two axes provides a four-quadrant matrix with which to classify school practice and/or school goal orientation.

A relationship can be seen among the 4 quadrants of the goal orientation matrix and the 3 levels in the analytic trait assessment instrument used in the study. This relationship logic was not employed intentionally in the construction of the instrument, however it does seem to be evidenced in an empirical examination of these particular schools. Many of the 3 level descriptions for items in the instrument could be characterized by the approach defined by quadrant 1. Quadrant 1 is defined by an intentional student-centered approach. Underlying this construct is the value for the presence of community and the use of empowering pedagogy.

Similarly, many of the 2 level descriptions for items in the instrument could best be classified by the approach defined in quadrant 2. Quadrant 2 is defined by a successful and intentional effort to manage/domesticate students. Underlying this construct is a value for an effective societal structure and pedagogy characterized by direct instruction. Likewise, most level 1 items characterize practices defined by quadrants 3 and 4. These lowest levels could be characterized by a lack of intention and a corresponding dysfunctional set of outcomes. These constructs are defined respectively by either an absence of structure or an overly controlling environment, and the consequential disconnection of students to their schools and one another.

Interaction of the Assessment Process with the School's "Implied Climate Goal"

Vision team participants
Participants at each school tended to rate their climates somewhere in the level 2 range. As discussed earlier, the participants found the process of rating their schools on the climate assessment instrument relatively straightforward. However, the ease of rating belied the inherent conflict participants experienced in placing their school along the continuum of performance depicted by the instrument. While at each school, participants could readily accept that the practices and realities depicted in the "level 1" options were undesirable, the practices and realities characterized in level 3 brought a degree of consternation.

Using the paradigm offered in Figure 3 as an analytic lens, what participants seemed to be expressing was a conflicted response to the "constructed realities" characterized in the items at the 3 level. This reality could be depicted as primarily intentional, liberating, and student-centered, and in many ways uncommon in large public high schools. One the one hand they could appreciate the worth of such practice on an abstract level, but they did not see them as possible or in many cases desirable. At the low performing school, the most common reaction to the level 3 item options was to suggest that they were not possible with the students they had in their classes, with the possible exception of the advanced placement level students. Common initial responses to items at the 3-level were "No school could expect all teachers to... " or "that would not work with our students." Participants were not asked about their personal goals related their "intentional practices" and whether what they saw as "best practice" would fit more into the 1 or 2 quadrant in the goal matrix. However, it was clear that many teachers suggested that they wanted to achieve good school climate, yet as they discussed the practice they believed would help them meet that objective, they often described 2 quadrant and in some cases, 4 quadrant teacher practice.

At the high performing school, there was similar perplexity. While the teachers were relatively pleased with their climate as they saw it, they also rated their school a high 2 or upper middle level in most areas on the instrument. Satisfaction with a more intentionally managerial/domesticating agenda was more prevalent at this school. Again, these participants generally demonstrated less emotional investment, and in most cases viewed the gap between their rating and the higher 3 levels on the instrument as somewhat academic. It appeared that their goal of school climate defined by 2 quadrant practice was acceptable to them, and that what was characterized by the 3 level items/1 quadrant practice was too idealistic and/or conflicted with some of the more traditional values held by the participants.

At each high school, there seemed to be four common factors that acted to mitigate what might have been a more overt level of conflict related to the process, the instrument's implicit "goal logic" or the research team members themselves. First, the process was being undertaken as a hypothetical exercise. The stakes were low. Second, there was a great deal of effort put forth by the research team to built trust among all participants. Great care was taken to provide confidentiality, limit the perception of external judgment and assume a posture of listening. Third, while the process was undertaken with undefined end-points at each school, care was taken to let the participants drive the effort. Last, the process was undertaken with deliberate transparency. As each set of participants grappled with where they judged their schools to be operating on the climate continuum and how that picture fit with the assumptions they had used previously to judge "good school climate," the fact they had their completed instruments in front of them proved valuable. The descriptive language in each of the items provided concrete ideas for use in both assessment and planning possible next steps.

Leadership
The research team gained access to the high schools studied by securing the cooperation and participation of the principal. The data yielded by the new instrument affirmed some revealing insights about the power and influence of the principal on the climate of the school and the scope of reforms open to the faculty.

The principal of the low achieving school embraced the pilot study during its early phases and provided support for the research team to conduct its visits and interviews. However, she remained detached from nearly all interactions between her teachers and the research team. Interviews with the principal and analysis of the data now shows this principal regarded the pilot study as a routine favor she would grant the university because it provided many of her high school's student teachers. She did not consider the possibility that an analytic scale-based instrument to measure her school's climate might uncover problems that require attention. Her teachers, on the other hand, regarded the instrument as a description of what their school might become, a center of dynamic learning for their students and themselves.

Near the end of the study, when asked if she would like to work with the research team and the teachers she selected to validate the new analytic scale-based instrument to then address some of the problems identified, she expressed little interest. When later presented the formal findings of the study in a private meeting with one member of the research team, she dismissed most of the results as inaccurate. Notwithstanding the fact these findings stemmed from data yielded by the new climate instrument, talks with students and teachers and several observations of daily routines at the high school.

Applying the matrix of school "climate goal" orientation to this principal suggests she perceived herself as an Intentional Climate leader, but one who seeks to domesticate and manage her students and teachers. She endorsed the idea of a healthy school climate, but only insofar as it serves the purpose of preserving order and a diffuse sense of pride in the school.

As she glimpsed the climate survey's indicators of a liberating/facilitating intentional school climate while hearing findings that suggested passivity and isolation among her students and teachers, her dissonance grew. This may have caused her to affirm her role as a strong leader of the status quo who rejected findings that might alter this role in the future.

In contrast, the principal of the high achieving high school engaged the research team from the beginning. He quickly saw the survey's indicators of a liberating/facilitating intentional climate as a means of detailing the vision he held for his school. However, the results of the study thus far suggest the principal's vision stands far ahead of his teachers' and students'.

The teachers at the high achieving school took some interest in validating the assessment instrument, but none saw it as a means to reform their practice or to change the climate of their school.

At the high achieving school, the principal seems eager to strengthen its climate, but the research team has found little desire for improvement among the teachers and students interviewed. Their scores on the faculty relations and student interactions items were no higher than those at the low achieving school. Teachers seemed to accept their isolation as a fact of life and the students expressed little desire for a greater voice in the governance and operation of their school.

The research team will soon confer with the principal of this school to determine what he might do next. He knows if he is to succeed he must recruit a new group of teachers to anchor a "school vision team" who possess influence with their colleagues and the political savvy to navigate change.

This illustrates another dimension of the principal's influence. In the case of the high achieving school, the principal perceives his role as an intentional liberating/facilitating climate leader who desires a more student-centered culture. To achieve this goal, he knows he must use his influence with the vision team of adults (and students, perhaps) to alter their beliefs and practices in favor of students.

Study Conclusions

1. The analytic instrument demonstrated promising soundness in this limited examination. Further validation is necessary, however it appeared that the theoretical assumptions of the study related to the comparative soundness of the analytic trait based instrument relative to a survey type instrument were supported.
    
2. The predicted capacity of the analytic trait type instrument to act as both an educational tool as well as an assessment instrument was supported. The process of using the instrument as well as the instrument document itself provided the internal assessment team a useful construct to diagnose their school's current condition, while providing language for articulating both a system for thinking about change and a prescription for action.
    
3. Given the differential goals for school climate demonstrated by the participants in this study, it might be concluded that no a priori construct for the concept of "good school climate" exists, or should assume to exist. School Climate is more accurately represented by an ongoing construction undertaken by those who work in and interact within a particular school. Any effort to define climate must consider its "socially constructed" nature and how that nature might affect the implementation of an assessment of a school's climate.
    
4. The assessment instrument under investigation here assessed climate within the sociological construct of an "intentional liberating/facilitative" goal orientation. This conception is potentially inconsistent with other goals and definitions of school climate, including many of those held by participants in the study. However, the data here appears to support the idea that the kind of climate indicators that were the most desirable to the majority of participants would be more achievable with an intentional and student-centered approach. In other words, no matter how intentional their efforts, participants would not be able to reach level 3 performance using 2 quadrant practices.
    
5. We can conclude a significant role for the principal in forming new climates or reforming existing ones. No principal of a large and complex high school leads unhindered by the actions and influences of teachers, students or the community. However, the principal's role remains critical to improving a school's climate because as the findings suggest, the principal can either inhibit or enable efforts to examine the school's culture and it is chiefly the principal who holds the fate of this examination.
    
6. The participative system driven by insider ownership seemed to be workable and supports the research in the field that suggests this is a desirable approach, but as indicated by this research, there are countless pitfalls related to efforts at democratic restructuring (Fullan, 1993; Glickman, 1992; Shindler, 1998). While this study found support for this being the "best" way to approach reform, it must accept that achieving meaningful reform in any such effort has a poor track record. Nonetheless, it appears that building in remedies to these pitfalls increases the chances for success.
    
7. The need for change and a discomfort with the status quo shared by the teachers and the principal may be the most important catalysts for strengthening a school's climate.

Implications for Practice
Given the conclusions above, those endeavoring to initial a school climate improvement effort might consider the following implications:

1. The analytic trait instrument provides an accurate mirror with which a school can reflect upon their own practice and performance. The use of the instrument within a participant-driven assessment and planning process provides a mechanism for process ownership and vision building. In contrast, the use an objective inventory scored externally may not provide the transparency necessary in the assessment process, nor the ownership necessary to facilitate vision building.
    
2. The membership of the teacher leadership/assessment team is significant. The team that takes part in the assessment phase of the process will likely reach an understanding of the fundamental definitions at work and the implicit climate goals at the school. If those involved in a planning phase have not participated in this assessment phase they may not have an adequate foundation for their work. Therefore, it appears that having a team that will continue for both tasks is desirable. Likewise, selection of the initial team should consider the skills necessary for visionary leadership.
    
3. A transparent examination of the construct of "good school climate" is desirable if not essential in any effort at fruitful democratic reform. A steering or vision team entrusted with the task of first assessing and then diagnosing their school's needs, must posses the capacity to understand the broader implications of their work.
    
4. The role of the school leadership, which in most cases involves the principal, is crucial. The leader must possess a vision of what the school can become, have a clarity about the school's "goals," and the ability to articulate that vision to the rest of the staff. No other person in the process is in the same position to create the necessary shared values and vision.
    
5. The presence of vision and shared values appears to lead to an "intentional climate." The absence of vision leads to what appears to be an "accidental climate." It appears that little meaningful change can occur within a school no matter how sound the instrument or the process if a sense of vision and a collective belief in the value of change is not cultivated.
    
6. While the analytic trait assessment instrument can provide a team with a effective mirror in their efforts to reflect on the quality of their school's climate, the mirror may be unwelcome. The reflection provided by the instrument may prove unsettling. This is especially true if the goal of the school tends to fall in the "domestication" category or if there is a misperception on the part of the principal about the true state of the school's climate.
    
7. The process of school climate examination should be undertaken with a full and deliberate understanding of its delicate nature. Trust among the leadership, the staff team, and outside consultants is critical. If the process is seen as a one that is "out to get someone" or as having an agenda it will have little chance of success. There are countless ways to derail the effort, but there may be only one way to produce meaningful reform. The effort must move slowly and deliberately with a broad base of participation and a sensitivity that improvement will in many respects imply that practices must change. Without collective trust, vision building and stakeholder commitment to the common interest of the school it is likely the effort will become at best, ineffectual; at worst, a source of strife. Furthermore, research suggests (Shindler, 1997) that if a first attempt at a reform such as this does not work, there are few if any second chances.

References