The Development of the Social Assets and Vulnerabilities Indicators (SAVI) Database

Sharon Kandris, Karen Frederickson, David Bodenhamer, Neil Devadasan, Cynthia Cunningham, and Jim Dowling

The Social Assets and Vulnerabilities Indicators Project (SAVI) was initiated in 1993 to develop a common repository of information on Indianapolis’ community assets and vulnerabilities for use by human service and community planners. Although conceived initially as a limited data set to be distributed as static maps in hardcopy and digital format, it has evolved into a comprehensive, GIS-enabled database of mapped and tabular data on the Indianapolis MSA. To keep the SAVI database flexible and robust, a sustainable data processing system has been developed for continued maintenance, processing of new data, and archiving. The principles applied and lessons learned by the design, implementation, and maintenance of the SAVI data processing system will be discussed.


Introduction


Before the SAVI system existed, a human service planner in Indianapolis would typically go to multiple agencies to collect data about his/her community. Through SAVI, planners and neighborhood leaders now have access to a robust, sustainable system of time-series data that allows them to do analysis at the regional down to the neighborhood-level.

This paper will discuss an overview of SAVI, the history of the SAVI system including the evolution of the database, the lessons learned in its development, and its potential for the future. It will demonstrate through these lessons the key elements to building a sustainable system.


What is SAVI?


The Social Assets and Vulnerabilities Indicators Project (SAVI) is a community information system that includes a comprehensive database of community assets and needs in the Indianapolis Metropolitan Statistical Area (MSA). The database was designed as a tool for human service and community planners to improve their understanding of the community. Its use is now extending to community-based organizations and residents to empower neighborhoods to make their own decisions to positively impact their communities.

The database contains data from over 40 data sources including local, state and federal agencies. The data include education, crime, welfare, human service agencies and programs, schools, and libraries to name a few. The data are standardized and distributed in a GIS-enabled system that allows users to utilize the data in concert for analysis from a regional to a neighborhood level. By giving users access to the SAVI database, users no longer have to visit several agencies to gather information about their neighborhood or area of interest because it is contained within a single, comprehensive system.

SAVI was created and is maintained in partnership between The Polis Center of Indiana University Purdue University-Indianapolis and The United Way of Central Indiana/Community Services Council.


History and Evolution of the Project


The SAVI project began in 1993 as a result of the United Way/Community Service Council seeking to update previous research reports that assessed the conditions of the Indianapolis community using vulnerability and asset indicators. It was determined that rather than merely an update to these reports, that a persistent repository of information should be created for continuous use by human service and community planners. This was during a time in which GIS was becoming more prevalent in the planning community, and the potential for GIS to make this a more dynamic tool was recognized. A steering committee was convened to discuss the concept of SAVI and its potential use and application to human service providers.

A pilot project was initiated to demonstrate the concept to funders and stakeholders. Initially, the “user community” consisted of a small number of human service and community planners. The database consisted of a series of excel spreadsheets that were linked to geographic boundaries in Atlas GIS to produce a series of static maps. A few data analysts used the prototype for limited analysis. The prototype demonstrated the potential for a much larger project, and as a result, the goals and scope of the project expanded, the database development and design evolved and the user community grew.

Once the proof of concept was accepted during the pilot phase, the need to develop a sustainable system became clear and was the goal of the second phase of the project. The database eventually evolved from flat files to a relational database driven by a high-level data processing system. The evolution of the database was incremental. The data were originally stored in Lotus 1-2-3 and then Excel as spreadsheets that could be linked to geographic boundaries to generate maps. A separate spreadsheet was developed for each data source, and the need for a single database was quickly realized. Microsoft Access was the initial solution, but as more complex data sets were introduced into the system, its limitations were discovered and the need for a more robust system became clear. Oracle was chosen because of its capacity to store large amounts of data.

In parallel to developing a database that could support the needs of the project, procedures were being developed to process data collected from various data sources in an effort to standardize data format and report the data in a manner usable by the intended planning audience. This too was an evolving process. In general, data were collected by address for each event, reclassified, and reported to the users in aggregated fashion by various geographic areas (e.g. crimes by Census Tract). Initially, the data were processed manually, but automation of the data processing soon began. Routines were created in Arc/Info to perform the geoprocessing as well as the data aggregation. While the software had the capability of performing this function, it could not support the data warehouse and storage needs of the database. Although Oracle was chosen for its capacity to store large amounts of data, it eventually became the ultimate software for designing and implementing a robust data processing system as well. Once the processing system was in place, efforts were taken to streamline and automate the process by building stored procedures, applications, and a user interface for data processing. An interface was also developed to automate the geoprocessing component in Arc/Info using AML.

As the utility of SAVI became more evident, the user community began to expand beyond the initial few planners. Human service agencies and planners saw this as a tool for making more informed and effective decisions. Access sites were established throughout the Indianapolis MSA as places where users could go to access the SAVI database. The user community continues to grow and now includes libraries, city planners, United Way representatives, other non-profit agencies, university departments, local and state government agencies, as well as several neighborhood and community organizations. There are now over 50 access sites throughout the MSA.

The products and tools for user access to the database that were developed and distributed to the users have also evolved throughout the project.As mentioned, the first product was a series of static maps produced using Atlas GIS to demonstrate how GIS could be used to visualize human service data spatially, a new approach for the human service community.A few users were later able to access Atlas GIS to generate their own maps from their desktop.The distributed data files, however, contained several additional fields that allowed users to link the tabular and geographic data in other software as well.ArcView GIS was later selected as the interface of choice to the database.Shapefiles were generated for every vulnerability indicator and asset category.ArcView GIS project files were created that contained base information such as various reference boundaries and assets, and allowed users flexibility of selecting their own themes for their particular analysis and the ability to view time-series data.An application was developed in ArcView GIS using Avenue that automated the creation of the deliverable product, including shapefile generation and project file creation and formatting.A GIS-enabled web interface (www.savi.org) was developed using Map Objects and Map Objects IMS.This was developed to be an easy-to-use tool providing immediate access to the data for all users, especially neighborhood and community leaders.


Lessons Learned


Strategic Planning

Through the evolution of the SAVI database and development of the data processing procedures, several lessons were learned that are now applied in the continued expansion of this project and to other projects that are similar in nature.The SAVI database was developed as part of a small project, but as the potential was recognized and the need for the data and its application increased, the database evolved and expanded.Initially, there was a problem associated with the continual evolution and expansion of the database design to meet the growing needs of the project.The software used for data storage and processing was changed to accommodate the increasing size and complexity of the database, and data processing procedures had to be modified based on the database design changes from flat files to a relational structure.While there should be a degree of flexibility in the design of the database, this clearly demonstrated the importance of developing a strategic plan before the project begins and conducting an up-front needs assessment.A key to successful and efficient project development and implementation is the preparation of a strategic plan that defines the purpose of the database or system that is being developed, identifies both the stakeholders and potential funders, and contains an implementation strategy.

A needs assessment is essential to determining the feasibility of developing the proposed project and to understanding the needs of the users that will be utilizing the data and how the data and/or system will be used in the end.By understanding the needs of the end-users, the system can be developed in a manner that addresses those needs with the appropriate data and technology and with an eye for the future.The needs assessment will provide information that can be used to develop an implementation strategy that addresses issues such as database design, resource requirements (human and technological), and data collection strategies.The strategic plan should be periodically reviewed and updated to reflect new audiences, changing needs and priorities, and new technology.In order to reflect new and changing user needs, needs assessments should be conducted at each significant phase of the development.

Data Assessment

The strategic planning process needs to include a data assessment.Based on the needs of the end-users and project goals, the amount of data required to meet the objectives should be identified.As the SAVI project expanded beyond a prototype, the data expansion, in part, drove the need for larger capacity systems.The complexity and variations in the data complicated the development of a data processing system for multiple reasons.By doing a general assessment of the available data in the beginning, it is easier to plan for variations in the data.During the planning stage, the availability of data required to meet the specific goals of the database should be researched and the number of indicators to be developed should be discussed.It may be that there will not be a limit to the number of indicators developed, but knowing that up front allows the developers to create innovative, long-term solutions.It is also important to have an understanding of the diversity of the data, including data formats, data quality, data infrastructure and data models.By knowing this in advance, the system can be designed to handle the variations and plan for anomalies.Data standards should be developed before the project begins and an understanding of those standards should be communicated to the data development team as well as to the user community.

Data security should be part of the data assessment.In building the SAVI system, the need for data security became clear.Data providers are more likely to contribute data to a system in which they are confident that adequate protocols can be implemented that will protect the privacy of their clients.For SAVI, this meant confidential data (e.g. welfare recipients) were classified and aggregated to various geographies to remove the address components of the data that could be associated with individuals.It also meant establishing data handling procedures for data exchange, storage, and access.The implications of releasing data should be assessed during this time.Some data, although publicly available, may have negative connotations in releasing in certain environments.It is important to protect the source provider and the intended outcomes of the end-user.

Based on the knowledge gained in the data assessment, the general processing steps can be identified that are required to translate source data to the final product.There are several key considerations in developing a data processing system.The general processing steps should be identified and generalized as much as possible to apply to multiple data sources.Each data source may require additional and more specific steps, but the general process should be the same across all sources.Anomalies can be identified and handled in the processing system structure.Documenting all processes and anomalies is very important to ensure a sustainable system that is not dependent on one person or a group of people.After the process has been tested as noted below, then it is time to streamline and automate the procedures, which is necessary in building a sustainable system.The initial investment required for streamlining may be costly, but the long-term benefits are great, resulting in a more efficient and robust system.

Once the strategic plan is in place, it is often beneficial to conduct a pilot study.This serves multiple purposes.It allows the opportunity to test the data processing systems on a small scale before making the major investment required for full implementation.It also provides a mechanism to market the concept to gain the support from project partners and stakeholders and heighten awareness about the project.Building a system like SAVI may seem like an overwhelming project initially, but a pilot study can also build confidence and support the feasibility of accomplishing such goals.

User Support

In order for a system to be sustainable, the value and applicability of the product to the intended users must be recognized.The establishment of the SAVI user community would not have been possible without two key elements:documentation and training.In the initial stages of SAVI, little data documentation was captured.As the project moved out of the pilot phase and data was being distributed to users, several things began to occur.Without the appropriate documentation of the origin of the data, how the data were processed, and the quality of the data, users had no way of knowing how reliable the data were and therefore could not confidently use the data in analysis.A very important step in building the trust of the users and data providers was to document the data, including the source, collection methodology, processing strategies, limitations, and quality.Although users and data providers may not initially value the thoroughness of the documentation, once they use it and understand the data, they are more willing to support and contribute to data documentation.

In addition, the users had to be trained on how to use the database and tools and educated on the associated risks, limitations, and cautions in the application of the data.Training users on how to interpret the data and the potential applications and analyses in which the system can be used is also critical in developing a sustainable system.If the users and stakeholders do not see the value in the system, they will not support it.

An important step in making the data valuable is putting it in a format that is easy to use and understand.For SAVI, data are collected from multiple sources in a variety of formats.Processing steps include standardizing the data and aggregating it into meaningful categories.Data analysts who have an understanding of the content of the data are very important to this process.Indicators are developed based on their knowledge of the data and of how it applies to the users’ needs.User feedback must be continually collected for the analysts to be responsive to the type of indicators that should be developed.

Quality Assurance

As the intended use and audience for SAVI increased, so did the need for more Quality Assurance.This too ties to building confidence in the system and the necessary support for making it sustainable.Every process should have a corresponding method of quality control, and the results of those checks should be documented.Developing a general quality assurance plan that establishes quality standards and ensures they are met should be included as part of the initial strategic planning process.

Once the users and stakeholders have bought into the concept, support the system, and understand the potential, their expectations for development and responsiveness will grow.By managing expectations of both users and funders from the beginning of the project, there is less opportunity for scope creep and for the system to become unmanageable.It is natural for the project definition and implementation strategies to change as new technologies become available, and it is important to keep stakeholders and users updated on changes that will impact the expected outcomes or deliverables.There was an intense resource effort that went into building the SAVI prototype.Expectations were high because many were not aware of the effort that was necessary to create the initial product.An important process that was necessary in managing the expectations of users and funders was to educate them on the amount of time and money required to develop the system.

Relationship Building

As evidenced throughout this document, relationship building in many respects is essential for the success of any project.It is an on-going, time-intensive but paramount process in creating a sustainable system.A database has no value without an understanding of how it can be applied and without the buy-in of the stakeholders, which can only be accomplished by building the appropriate relationships.  In the development of the SAVI project, relationship-building with four major categories of people were key to its success in building a sustainable system: 1) the user community, 2) funders, 3) data providers and 4) project partners.

The relationships with the user community must start at project initiation.Their data needs, application needs and subjects of interest must be understood and captured through the needs assessment process.It is important to gain their support, provide them with appropriate training, allow them opportunities to provide feedback, and be responsive to their requests.

Funders must understand and agree with the concept and intended outcomes of the system.By establishing a relationship with and being responsive to potential funders, the life of the project is more secure.Constant communication with them allows them to see progress towards the intended outcomes.

As identified above, it is important that the data providers have confidence that their data will be secure in the database.This confidence is established in part through interaction with them.Permission to use their data must be obtained, which often occurs in the high-level offices of the agency.In order for their data to be valuable to the project, it must meet standards that have been established for the database.This often requires educating the data providers on data quality standards, formatting, and documentation.In turn, it is necessary to understand the data providers’ data format, data content, collection methods, and existing documentation.

Partnerships can be valuable to the success of a project, but the roles and responsibilities of each partner in the project must be clearly defined and understood.The relationship is built through communication, collaboration, reporting, and fulfilling responsibilities and obligations.

Future Opportunities

As the SAVI project continues to evolve and stay on the cutting edge of technology, several opportunities will emerge.The lessons learned in developing this sustainable system will be applied as new partnerships and collaborations are formed and new technology is embraced.

The nature of data sharing is changing, and SAVI will seek to improve this process by researching, testing and implementing emerging technologies and open standards.Data sharing now occurs in a centralized system where data are collected, processed and distributed to users from a single server.New technology, such as ArcIMS, ArcSDE, and Oracle Spatial, is being developed that supports data sharing through a distributed approach that allows data providers to serve their data on their server for access by applications like SAVI.Rather than collecting, processing and distributing data in a centralized location, the SAVI web application can act as a portal to the data providers.SAVI users will be accessing data from multiple servers distributed on the internet rather than a single server; they will be accessing real-time data through a distributed network of data providers.In order to make this feasible, the data access through the SAVI web application will adopt the use of open standards to access distributed stored data servers.

As the future of SAVI is envisioned, it is important to assess the current status of the system and other databases that may become part of the system.Before a distributed network of providers can be in place, there must be an assessment of the source provider and their technological capabilities, data, applications, database structure and design.The lessons learned from the assessment must then be applied to educate the providers on technology, data and metadata standards to ensure they are successful in this effort.This will also allow them to share data with other organizations and agencies in the same manner, giving them more flexibility and compatibility.

SAVI is a system that is made up of a database, data processing system, neighborhood and community leaders, human service and community planners, many other users, data providers, partners, funders, GIS technology, applications, and data analysts.In order to maintain its sustainability, it is important to use innovative and cutting-edge technology to ensure responsiveness to the needs of all parts of the system.


Acknowledgements

The authors would like to acknowledge Kevin Mickey for his dedicaiton and contribution to the SAVI project.

Author Information

Sharon Kandris
The Polis Center, Indiana University Purdue University - Indianapolis
1200 Waterway Blvd., Suite 100
Indianapolis, IN 46202
Phone: (317) 278-2944
Fax: (317) 278-1830
skandris@iupui.edu