Graduate programs in educational administration are replete with courses focused on leadership theories and change strategies. Many of the theories include the concept of building shared vision (Cooperrider, Whitney, & Starvos, 2005; Deming, 2000; Senge, 2006.). Indeed, shared vision and effective change efforts are deeply interconnected. Senge (2006) identifies shared vision as one of the cornerstones of learning organizations, a framework for promoting organizational change. Shared vision, according to Senge, is the collective caring behind an organization. When all members of an organization convey the same vision, this acts as a positive force for change.

The concept of shared vision is well accepted; yet, as graduate students leave the classroom and enter the complex environment of a school building or district office, they are confronted with putting theory into practice. At this point new professionals discover that building and maintaining shared vision are difficult undertakings, requiring complex skills that remain elusive for many educational leaders. As a result, it is not uncommon for practicing educational leaders to implement a change effort with an assumption of a shared or common vision, only to find the effort less than effective. In turn, the leader attempts another effort, then another, then another. Fullan (2008) argues that the education profession suffers from what he calls “initiativitis” (p. 1): the implementation of change effort after change effort without regard to how these efforts interact with each other, the existing system, or the players within the system. The cycle of ever-new initiatives can create confusion, lead to exhaustion within both teacher and leadership corps, and produce few systemic improvements.

This study, which was designed to explore the mental models held by educational leaders with regard to one change effort, provided evidence that leadership cannot assume the existence of a shared vision. The study explored various aspects of walkthroughs, a single change effort implemented district-wide eight years prior to data collection. Walkthroughs within the educational setting are parallel to the business approach of “Management by Wandering Around” (Peters & Waterman, 1982, p. 122). Developed at Hewlett-Packard, the practice, whereby supervisors build relationships by physically walking throughout the plant and engaging with workers, is a strategy that many leaders in highly successful companies employ. Subsequently, Frase and Hetzel (1990) introduced the practice into the field of education as a tool for principals to positively impact schools. Because walkthroughs had been implemented for almost a decade, the study explored educational leaders’ understanding of this established change effort.

The district chosen for this study was a large urban district in the western United States that serves approximately 63,000 students over a large geographic area. The district spans approximately 6,500 square miles and encompasses rural, suburban, and urban neighborhoods. Although the district covers a vast geographical range, over 90% of the schools are located in an approximately 100 square mile metropolitan area. In 2009, the district operated 92 schools including 63 elementary schools with a student population consisting of 54.41% white, non-Hispanic and 45.59% minority. In addition, at the time of the study, the district employed approximately 7,200 people thus making it one of the largest employers in the area.

A qualitative methodology was utilized to explore two research questions: 1) What mental models (beliefs, assumptions, attitudes, and knowledge) do educational leaders responsible for school improvement hold about the role of walkthroughs as a vehicle for system-wide change; and 2) How do different contexts influence the mental models that educational leaders hold? The university Institutional Review Board (IRB) provided oversight for the design and implementation of the study.

The participants in this study were 17 practicing elementary principals and six district level administrators responsible for supervising principals. The principals were selected from a group of 63 potential participants – the total number of elementary principals in the district. The criteria used to select potential participants were social economic status (SES) of the school and the number of years since the participant received his or her first administrative credential. This purposeful sampling strategy was utilized to gain a broad level of both experience and context with practicing principals as it related to the research questions. All elementary level principals were categorized using publicly available information, (i.e., SES of school and years of experience). Socio-economic status (SES) of each school was based upon classifications utilized by the district in the study. In this case, schools classified as high SES were those in which less than 40% of students are eligible for free and reduced lunch under federal guidelines. Mid-SES schools were those in which 40% to 59.99% of students are eligible for free and reduced lunch. Finally, Title 1 schools were those schools in which 60% or more students are eligible for free and reduced lunch. Nineteen principals, representing the various categories, were initially contacted by phone. Of the 19 potential participants, 17 agreed to participate in the study. Table 1 summarizes the socio-economic status (SES) of each school, number of years since certification, and number of years as principal for each of the 17 principals in the study.

At the time of the study, six district level administrators provided oversight of the elementary principals, either in a direct supervisory role or one level above, supervising all schools. All district administrators agreed to participate in the study, thus providing a full population of district level administrators directly responsible for supporting walkthroughs. Table 2 illustrates the characteristics of district administrators based upon administrative experience.

Two data collection procedures were used: 1) semi-structured interviews and 2) survey results. A standardized, but flexible, open-ended interview guide allowed participants to share their ideas in their own voice as well as provide a basis for comparison among participants. All principals were asked the same questions with probes reflective of their responses. The questions were slightly altered for district administrators to account for their positions within the organization. Again, appropriate probing was employed to capture their complete thoughts. The interview guides are found in Tables 3 and 4.

The second data source was a survey that participants completed after the interview session. Survey responses provided supplementary data and afforded participants with an opportunity to consider questions in a different manner (Hill, Thompson, & Williams, 1997).

Because many principals had incorporated forms into the walkthrough process, a third data source, walkthrough forms, was collected from this group. Fourteen principals were using a walkthrough form with 13 principals providing a sample for inclusion in the study. The participants who did not utilize walkthrough forms were not excluded from the study. Indeed, the interview process was designed to expose the reasoning that prompted principals to either use or reject walkthrough forms, further revealing their mental models.

Data analysis occurred in several phases. First, data from the principal interviews were analyzed so responses could be explored independently from district administrators. However, the identical analytic process was used for both sets of data; NVivo8™ software was used for data management. Initially, transcripts were coded line-by-line by the principal investigator, highlighting specific details such as purpose of and feelings about walkthroughs. Sections were coded as free nodes, i.e., codes that stand independently. Once each transcript was coded in this fashion, the free nodes were clustered into tree nodes, i.e., codes that form hierarchy based upon commonalities. Memos were created for data points that either seemed to represent emerging themes or appeared to be counter to prevailing patterns.

Following initial coding, peer debriefing was conducted with the other investigators as a means of establishing validity of the three themes revealed in the data (Creswell, 2003). At this point, adjustments were made to the descriptions of the themes, sub-themes were identified and interview sections were coded to the appropriate theme or subtheme. A constant comparative method between participants was also employed (Glaser and Strauss, 1967).

The original study was designed to explore the mental models that building and district educational leaders hold regarding one change effort, walkthroughs. The data actually revealed the effects upon a particular change effort when shared vision is absent. The only shared vision held by all respondents in this study was that walkthroughs benefit the school and the principal; therefore, walkthroughs should be conducted.

That the shared mental model or vision did not develop within the context of an openly shared understanding suggests a lack of organizational learning. Senge (2006) notes that people construct mental models about everything. From an organizational perspective, however, individual mental models that remain private do not represent an organizational mental model. In fact, hidden mental models can be major barriers to organizational growth (Argyris & Schön, 1996; Karp, 2005; Kim, 2001; Senge, 2006.) Senge (2006) states,

The data from this study suggest that although participants held a collective mental model of the value of walkthroughs, the mental model did not represent one that was consciously shared. Rather, the mental model was born of individual experience and practice. Participants possessed little awareness of the mental models that other leaders held with regard to walkthroughs.

This mental model, though collective, was based upon individual assumptions, perceptions, and comprehensions versus an intentional dialog among leaders. Individuals were provided a general expectation by district leadership to conduct walkthroughs; however, the purpose of the practice was left to individuals to determine. In terms of district-wide impact, participants had difficulty envisioning impacts on the entire system because mental models were constructed individually. Many provided educated guesses about potential impacts at the district level; however, such posits were generally based upon impacts experienced at each participant’s school site.

Though all the participants adopted the practice of walkthroughs and constructed individual mental models about purpose, a collective sense of why the practice was important to the organization was missing. Kim (2001) maintains that it is critical for organizational leaders to not only provide training on how to implement initiatives but also to create a conceptual understanding of the purpose. He contends, “…conceptual learning emphasizes the why of doing things – that is, it has to do with the thinking behind why things are done in the first place” (p. 23). Unfortunately, the absence of an organizationally-shared vision of the purpose of the walkthrough process appears to have resulted in a reluctance to lead on the part of the district administrators and a sense of isolation, frustration, and confusion on the part of the building principals.

The second major finding suggests that the mental models surrounding the construct walkthroughs were highly individualized. Both principals and district leaders viewed the practice through individual lenses. The mental models constructed by participants were based upon individual experiences and how each practitioner utilized walkthroughs. Participants also held individual perceptions with regard to the district leaders’ purpose for ushering in walkthroughs as a leadership practice. No collective sense was evident with regard to the overall intent of walkthroughs from a system-wide perspective. Therefore, context had little impact upon the development of mental models. The experiences of participants were highly individualized. No apparent pattern emerged based upon school type, years of experience, or years since administrative certification.

Because mental models were highly individualized, many participants expressed a sense of isolation with regard to the initiative. Karp (2005) notes mental models and the “underlying assumptions” from which they are created are rarely discussed openly within organizations (p. 89). Little forward movement of an organization can be accomplished if the deeply held assumptions are not addressed openly (Chrispeels, Burke, Johnson, & Daly, 2008; Fullan, 2008; Kim, 2001; Senge, 2006). Kim (2001) further suggests that if organizational learning is to occur, then the knowledge of individuals must be integrated into the broader context of the whole organization. He articulates, “If we are interested in innovation and in the vitality of large institutions, then we are interested in creating learning communities that integrate knowledge instead of fragment it” (p. 16). The findings of this study clearly illustrate that the innovation of walkthroughs remained at the site level.

The awareness of isolation carried over to other educational issues including the definitions of effective instruction, instructional leadership, and change efforts. For example, opinions varied about how different change efforts such as professional learning communities and Response to Intervention interacted with one another. Of note, this individuality was both celebrated and disdained. Several participants indicated that they were proud of their individual stance on walkthroughs and their abilities to move their schools forward. District leaders similarly expressed amazement at some of the innovation within the principal corps. However, the data suggested that many participants felt uncomfortable with the perceived lack of district-wide vision for walkthroughs. A desire existed to bring order and a collective meaning to the practice. This was illustrated by a desire to have common standards with regard to effective instruction and additional accountability for principals.

The third major finding highlights the effect these collective, unshared mental models had on district-wide leadership. Because leaders held the collective, but unshared mental model that walkthroughs were an individual endeavor, district administrators were reluctant to lead the effort. Leadership did not arise from the principal corps, either. With the absence of system-wide leadership, individuals experienced isolation and uncertainty with regard to the change effort. Perceived lack of system-wide leadership reinforced the mental model that development of walkthrough practices resided with individual principals.

Holistically, the three major findings appear interrelated, associated, and somewhat reinforcing. The Individual Endeavors are associated with Reluctance to Lead, which is associated with Isolation and Uncertainty. In turn, Isolation and Uncertainty seems to reinforce Individual Endeavors. This process can be predicted to continue because a shared vision is lacking. Figure 3 illustrates this interaction of these mental models. The unvoiced, collective mental model that walkthroughs reside in the realm of the individual principal actually inhibited the rise of system-wide leadership with regard to this effort.

Figure 3: The interaction of mental models illustrates that though walkthroughs are highly valued, they are perceived to be an individual endeavor that inhibits district administrators or principals from taking a system-wide leadership role, thus creating a sense of isolation and uncertainty among principals that reinforces the mental model that the effort resides with individuals.

As faculties of education prepare their students to enter the complex and dynamic world of school leadership, the role of building shared vision cannot be overemphasized. Building shared vision requires a willingness to openly discuss mental models held by all within leadership positions. Complex skills such as team building, discussion and discourse strategies, and establishing effective feedback loops are essential to creating a culture in which the vision of leaders is articulated throughout an organization. As this study illustrates, when leaders are left to construct their own visions in isolation, results are varied and impact to the system is weakened. In this case, when principals were left to construct their own meaning of the change effort, isolation and a sense of independence emerged, and an uncertainty with the change efforts resulted.

As Senge (2006) notes, the antidote to isolation and independent islands is creating structures by which leaders can communicate and build the vision. This particular change initiative is compelling in that principals and district leaders universally accepted walkthroughs as a leadership practice. The reasons for this are beyond the scope of this article and will be reviewed in a subsequent article. However, the manner in which this practice evolved and the resulting sense of isolation provides those who prepare leaders for their role in building shared vision an opportunity to explore possible outcomes when leaders resort to implementing initiatives rather than using initiatives to support the vision of the organization. Faculty of educational leaders can use the results from this research as a case study as they guide new leaders into their roles to engage future leaders in strategies to build shared vision through a complex system. Additionally, further research on shared vision and change efforts should be explored more fully, especially in school systems, as leaders adopt programs and strategies to impact student achievement.

Argyris, C., & Schön, D. A. (1996). Organizational learning II: Theory, method, and practice. Reading, MA: Addison-Wesley.

Chrispeels, J. H., Burke, P. H., Johnson, P., & Daly, A. J. (2008). Aligning mental models of district and school leadership teams for reform coherence. Education and Urban Society, 40(6), 730-750.

Cooperrider, D. L., Whitney, D., & Stavros, J. M. (2005). Appreciative inquiry handbook: The first in a series of AI workbooks for leaders of change. Brunswick, OH: Crown.

Creswell, J.W. (2003). Research design: Qualitative, quantitative, and mixed methods approaches. Thousand Oaks, CA: Sage.

Deming, W. E. (2000). Out of the crisis. Cambridge, MA: The MIT Press.

Frase, L., & Hetzel, R. (1990). School management by wandering around. Lancaster, PA: Technomic.

Fullan, M. (2008). What’s worth fighting for in the principalship (2^nd ed.). New York: Teachers College Press

Glaser, B.G. & Strauss, A.L. (1967). The discovery of grounded theory: Strategies for qualitative research. New York:Aldine de Gruyter.

Hill, C. E., Thompson, B. J., & Williams, E. N. (1997). A guide to conducting consensual qualitative research. The Counseling Psychologist, 24(4), 517-572.

Karp, T. (2005). Unpacking the mysteries of change: Mental modeling. Journal of Change Management 5(1), 87-96.

Kim, D. H. (2001). Organizing for learning: Strategies for knowledge creation and enduring change. Williston, VT: Pegasus.

Peters, T. J., & Waterman, R. H. (1982). In search of excellence: Lessons from America’s best-run companies. New York: Warner Books.

Senge, P. M. (2006). The fifth discipline: The art & practice of the learning organization. New York: Doubleday.

WestEd (2009). Teach for Success®. Accessed on December 19, 2009, from http://www.wested.org/cs/we/view/pj/551.