This study examined the perceptions of public school superintendents, central office administrators, principals, and teachers to determine the importance of central office administrator facilitating practices reported in the literature for working with decentralized change processes. Additionally, the participants were asked to provide other practices for central office administrators they believed were important which were not found in the literature.

The questionnaire, the Central Office Practices for School-Based Improvement, was the instrument used to collect data for this study. It included 74 practices which were related to the literature about organizational design, administration, and school-based change. The practices were grouped into eight areas of administrative responsibility: Goals and Planning, Policy and Procedures, Staff Development, Recognition and Rewards, Monitoring and Evaluation, Management of Resources, Organizational Communication, and School-level Involvement. The data were collected though a mailed survey that resulted in an 86% return. Frequencies, standard deviations, and means were used to report the importance ratings for the data. Analysis of Variance (ANOVA) and Tukey-Kramer α statistics were used to report differences among and between responses when grouped by role.

The population of this study consisted of superintendents, central office administrators, principals, and teachers in 35 school districts implementing the |I|D|E|A| School Improvement Process (SIP). Each of the districts had trained change facilitators for school sites and the schools had implemented the components of the |I|D|E|A| SIP.

The respondents identified sixty-eight important facilitating practices for central office administrators working to support school-based improvement. Additionally, of these practices, forty were identified as most important facilitating practices for central office administrators working to support school-based improvement. Significant differences occurred in eighteen percent of the paired comparisons.