Though spatial decision support systems (SDSS) to help solve locational problems are not new and are being used successfully in many fields of application, one field, school attendance area planning, has yet to tap fully the potential of such applications. This paper describes a spatial decision support system used by the Blue Valley School District in Overland Park, Kansas. The system developed, known as the Student Enrollment Decision Support System (SEDSS), allows decision makers the ability to view, analyze, and compare the current and future spatial impacts of various school attendance area boundary scenarios. An ArcInfo-based geographic information system (GIS) application written in the Arc Macro Language (AML), SEDSS is both an enrollment projection modeling system and an attendance area creation and analysis system. The combining of two of the most integral functions of school district planning into one system has proven to be extremely beneficial. This paper will focus on the attendance area creation and analysis module of SEDSS, and how using the system has benefited the school district in its school attendance area decision making needs.
There are not many decisions that a local school board of education makes that evoke more highly charged emotions on the part of district patrons than that of changing school attendance area boundaries. Board members are faced with the unenviable task of determining, based on projected future enrollment growth and current enrollment pressures, which school boundary configuration will be best for the district in the long-run. This task is made even more difficult if the district is large and is growing rapidly. The possible combinations of school attendance areas are seemingly endless, and each combination differs from the others in terms of what it means to the school district and its ability to provide for its patrons. For a board of education to serve its patrons with the greatest level of service possible, a tool is needed which allows access to current information on present and anticipated student enrollments, as well as allows the future student enrollment impacts of attendance boundary decisions to be viewed and analyzed. A geographic information system is well-suited for this task.
The Blue Valley School District (BVSD) in Overland Park, Kansas, is the sixth largest school district in Kansas with a 1994-95 enrollment of nearly 13,000 students. The school district is located in parts of the cities of Overland Park, Leawood, and Olathe, as well as part of unincorporated Johnson County. The BVSD has developed a spatial decision support system to deal with its enrollment projection and boundary analysis needs. While decision support systems to help solve locational problems are not new (Armstrong, et al, 1993; Harris, 1992; Densham and Rushton, 1988), the system developed by the Blue Valley Schools is unique in that it is both an enrollment projection modeling system and an attendance area creation and analysis system. These two processes are the integral parts of school district planning. The development and use of SEDSS has greatly enhanced the school district's ability to accomplish these tasks.
Several works contain descriptions and/or definitions of spatial decision support systems (SDSS). Bracken and Webber (1990) define an SDSS as having the following three distinct features: one, an SDSS should allow the asking of what-if scenario questions; two, an SDSS should have a user-friendly and powerful interface for problem formulation and expression; and three, an SDSS should use analytical tools to provide possible solutions to the problems posed.
Ryan (1992) defines an SDSS as a sequence of steps necessary to support decision making for a spatial problem. The steps are image and data capture, database management, modeling and analysis, and presentation. Anjomani and Saberi (1992) have written that SDSSs consist of four main components - data manipulation, modeling, display and mapping, and interaction with decision makers. Other researchers have written that an SDSS "integrates database management, spatial modeling and display capabilities to help decision-makers explore locational problems' (Armstrong, et al, 1993). The spatial decision support system presented in this paper fits well with the four definitions and descriptions presented above. SEDSS allows the construction of what-if boundary scenarios using a user-friendly interface, utilizes a small area population growth model for enrollment projection, and allows decision makers the ability to view, analyze, and compare the current and future spatial impacts of various school attendance area scenarios.
The Blue Valley School District has been one of the fastest growing school districts in the nation. Over the past 20 years, student enrollment in the district has increase more than 1,200 percent. If enrollment in the State of Kansas would have mirrored the growth Blue Valley has experienced since 1971, public school enrollment in the state would be more than 7.1 million students - as of 1994, it was just over 450,000.
It does not take long to realize that in a rapidly growing school district, certain pressure issues exist which must be addressed. One such issue is that of school attendance areas. Over the past nine years, Blue Valley has experienced a growth of roughly between 700 and 1,000 students annually. The need for an automated, decision-facilitating system for creating and analyzing current and future school attendance areas is paramount.
Any system designed to fill this need must also be capable of tracking developing areas in the district and monitoring student enrollment activity. At Blue Valley, the Planning and Facilities Committee, a Board of Education subcommittee, has the task of reviewing and recommending school boundaries. The committee consists of twelve citizen patrons, three board members and three staff members. In previous years, the committee would take several weeks in analyzing boundary alternatives. The length of time was required not as much for discussion purposes, but for information collection and calculation of the spatial impacts of various boundary alternatives. A method was needed which would allow the Committee to access current information on present and anticipated student enrollments, as well as allow the future student enrollment impacts of attendance area and boundary decisions to be viewed almost immediately. SEDSS was designed with that task in mind.
In 1994, the Blue Valley Schools moved its geographic information system from a PC based environment to a UNIX workstation environment. A DEC Alpha 2100 server was chosen to house the GIS. ArcInfo version 7 (Environmental Systems Research Institute, Redlands, California) was chosen as the GIS software.
An enormous amount of digital information is needed for a successful spatial decision support system for school planning. The Blue Valley School District benefits from a unique partnership with the Johnson County, Kansas, Government. The Johnson County Government initiated a GIS program in 1986. The county has an extensive database of land records and graphical information. A data licensing agreement between the BVSD and the County Government has allowed the school district to make use of important information (i.e. subdivision and parcel data, street data, address information) which would otherwise have been cost prohibitive for the school district to obtain. In turn, the School District gives back to the County digital products that are developed as a result of using the information. This type of cooperation between Johnson County and the School District is a key to the overall success of the SEDSS application.
SEDSS is a menu driven system. The system consists of five subsystems - map display, database editing and query, enrollment projection modeling, attendance area creation and attendance area analysis. SEDSS is an application written in AML, ArcInfo's Arc Macro Language. The overall aim of SEDSS is to allow decision makers the ability to view the potential impacts of various boundary scenarios immediately. While it formerly took considerable effort to calculate the impacts upon various schools of potential boundary changes, SEDSS allows such impacts to be analyzed and drawn to the screen with little delay. The ability to do this was advantageous to the Planning and Facilities Committee this year as attendance boundaries had to be redrawn with the addition of a new elementary school. In future years, several other school facilities will open, and SEDSS will undoubtedly play a major role in the boundary considerations of these future facilities.
The five subsystems of SEDSS allow users to accomplish several tasks. With the map display subsystem, users can view the school or subdivision boundaries, and change the boundaries interactively. With the enrollment projection modeling subsystem, all of the tasks involved in the projection process - from data input and maintenance to changing model assumptions - can be accomplished. The database query/editing subsystem allows users the ability to ask questions of the database (i.e. what subdivision is here, what is the future projected enrollment by grade level of the subdivision, etc.), update subdivision information, and add new or future subdivisions to the system.
The attendance area creation and attendance area analysis subsystems of SEDSS allow users to interactively assign subdivisions to certain schools and analyze the long-term projected enrollment effect of the boundary changes on the schools immediately. In boundary committee meetings, the school district and its subdivisions were displayed on a 6-foot by 8-foot overhead screen. Committee members discussed boundary issues, moved subdivisions from school area to school area, and analyzed the effects of such scenarios. Under this arrangement, committee members were able to see graphically how the new boundary configurations would look and also examined what the new configuration meant to the school district from a school enrollment perspective. This method enabled the Committee to explore and debate several boundary alternatives. Using SEDSS also helped facilitate discussion and school boundary conflict-resolution among Committee members.
The database design which SEDSS utilizes consists of several files related by a common item. Thus, SEDSS makes use of traditional relational database design. The common item links the graphical component used in committee meetings with data and information resources. The common item is a three-digit subdivision code. This item relates such files as building permit histories, student enrollment ratios, actual counts of student enrollment, available building lot information, etc., to the subdivisions which make up the Blue Valley School District.
SEDSS has several advantages over the previous, non-GIS method of school district planning. The system offers the flexibility to view various boundary scenarios and presents projected enrollment figures based on those scenarios on the fly. This proved to be more time efficient than previous methods. SEDSS has the ability to group and count students within each school attendance area, even though the basic unit of data is maintained at the subdivision level.
Several stakeholders in the school boundary process also benefited. Three stakeholders in particular - school district patrons, the Planning and Facilities Committee (the group charged with recommending new boundaries), and district planning staff - all reaped the benefits of what SEDSS offered.
School district patrons, several of whom serve on the Planning and Facilities Committee, were able to see the long-term effect of several school boundary scenarios as they were discussed. The Planning and Facilities Committee was able to perform its boundary-recommending duty in a fraction of the time than was necessary in previous years. Additionally, SEDSS allowed the committee to engage in helpful discussion by allowing them the ability to see the effects of certain boundary proposals almost immediately. District planning staff's time commitment to the calculation and re-calculation tasks of the boundary-making process was nearly eliminated, thus freeing up time for other tasks to aid the committee in its decision making process.
SEDSS has been evaluated by all who have been a part of the method as an extreme success. The system greatly reduced the time devoted to the "number-crunching" tasks which are involved in the boundary analysis and adjustment responsibilities. The system resulted in greater staff utilization as well as provided a forum for the Committee to debate various boundary proposals.
With all its successes, however, this particular version of SEDSS is still limited. For one, the system handles boundary scenarios from a macro level quite well, but needs to incorporate a micro level of analysis as well. For example, entire subdivisions are easily assigned by SEDSS to certain schools; however, individual cohorts within subdivisions are not. This precluded the committee from using SEDSS to examined certain enrollment decisions possibilities such as phasing in boundary changes over time. The future versions of SEDSS will be enhanced by the addition of this and other features.
The author would like to thank the members of the Planning and Facilities Committee of the Blue Valley School District for their willingness to be a part of the SEDSS experiment; the Johnson County Planning Department and Automated Information Mapping System (AIMS) Department for their assistance in obtaining digital information for use with the system; and Jenny Slagle for her assistance with reviewing earlier drafts of this paper.
Anjomani, A., A. Saberi. 1992. "Large Scale Land Suitability Analysis Using GIS and Optimization Models As An Spatial Decision Support System: A Multijurisdictional/Multiregion Case", Urban and Regional Information Systems Association 1992 Proceedings, Vol. 3, p. 39-48.
Armstrong, M. P., P. Lolonis, P. Honey, 1993. "A Spatial Decision Support System for School Redistricting", Journal of the Urban and Regional Information Systems Association, Vol. 5, pp. 40-52.
Bracken, I., and C. Webster. 1990. Information Technology in Geography and Planning, Including Principles of GIS. London and New York: Routledge, p. 28.
Densham, P.J. and G. Rushton. 1988. "Decision Support Systems for Locational Planning". In: R. Golledge and H. Timmermans, editors, Behavioural Modeling in Geography and Planning. London: Croom-Helm, p. 56-90.
Harris, Michael J. 1992. "Spatial Decision Support System For A Potomac River Crossing Of I-95 South Of Washington, D.C.", Urban and Regional Information Systems Association 1992 Proceedings, Vol. 3, p. 60-73.
Ryan, T. C. 1992. "Spatial Decision Support Systems", Urban and Regional Information Systems Association 1992 Proceedings, Vol. 3, p. 49-59.
Michael R. Slagle, Planning Coordinator