WWW sites have begun to provide interactive access to spatial information. However, these sites are not supported by a documented understanding of the issues listed above. This study will examine these issues in order to create a more responsive, effective, and efficient public access tool.
The intent of this paper is to describe the results of the study with four neighborhood associations in Tucson, Arizona and use the results to define pragmatic guidelines developing a public access GIS system.
The Bauman Foundation published a report in 1995 titled Agenda for Access: Public Access to Federal Information for Sustainability through the Information Superhighway. The document was developed with input ranging from Clinton Administration officials to grassroots community organizers with the intent of "developing a blueprint of and concrete implementation steps for assuring that a broad range of federal information associated with sustainability is disseminated to the public on the National Information Infrastructure" (Bauman Foundation, 1995). The report focuses on non-profit organizations' access to federal information and how access can positively affect sustainable development. The authors of that document stress that successful access to information hinges on availability of hardware, proper training, funding to develop such systems, developing the capabilities of organizations to tap into such systems, and convincing low income users that a computerized public access system could be beneficial to them. A problem mentioned by many authors is the establishment of information and communication programs without consulting end users. They offer the solution of joint design teams including technical and non-technical representatives from all affiliated user groups.
What is not apparent in most literature dealing with public access to spatial data is a clear definition of what data the public wants access to, what they would do with that information, and suggestions as to how their needs and desires may impact the development of spatial data.
There are trends in computing and networking that support the idea of public access to data. One of those is community networks. A community network is the ideal location for a public access system. Unfortunately, there is very little published about community networks. Most information about community networks comes from where they exist; the Internet. A community network movement is one or more computers providing services to people so that users can gain access to services and to each other (Guy, 1991). These networks serve as a good model for public access systems. Organizers of community networks wish to add value to a community and make services and information more readily available to the public, however, precious few provide access to spatial information. The movement is generally championed by librarians because of their professional ties to information and increased pressure to "go digital." However, community networks are also promoted by commercial and governmental agencies.
The community network movement began with bulletin board systems (BBS). A BBS is a computer that can be accessed by another computer via a phone line and modem. BBSs are founded in order to connect people with a similar problem, concern, or interest. If someone owns a computer or has access to a computer and a modem, they can dial in to a BBS to read messages, leave messages, take part in discussions, and learn from others. The community network movement took this concept of connecting people one step further by connecting people within a geographic area to share information and learn more about what is happening in their area. Community networks, based on a BBS model, provide an opportunity to offer inexpensive, efficient access to information.
Community networks are often funded jointly by public and private organizations. They are often built with a small amount of seed money and second hand computer hardware. Once they are proved useful, equipment and services are often upgraded. Community networks offer a variety of different interfaces and designs. In many locations, citizens can access a community network through terminals in libraries, schools, government office buildings, and coffee shops, or via modem in their personal computers. Many community networks only offer access to local citizens only.
Most community networks function primarily with a text interface. Few community networks provide the ability to view graphics, query data, or analyze it. This is because expensive software and hardware is needed for both clients (end users) and servers. Click here for a list of community networks (also called freenets) that have a web site.
Services Provided by the Average Community Network
Some of the better funded community networks have branched out to the World Wide Web (WWW) to take advantage of it's graphic capabilities. This is not to say that they have eliminated their text based interfaces. Often, community networks provide both services in order to appease different user groups. The community networks on the WWW provide a more user-friendly interface than do the traditional text based networks. A graphic interface on the WWW allows hotlinking of information, display of graphics such as images and maps, and access to multimedia information (sound and video).
Even though some community networks have developed a WWW version of their network, virtually all of these lack interactive spatial capabilities. That is, users cannot query spatial data and get results. For example, users cannot view a map of landuse in their area and then ask for the land uses to be reclassified based on a different classification scheme. However, client-server technology is quickly evolving to provide these capabilities. Esri, and many of their competitors, have introduced Internet spatial data server technology that can provide interactive analysis of spatial information via the web.
In the mean time, it is important to understand and examine how community networks are currently organized and determine their strengths and weaknesses. By learning from past mistakes, a well designed, effective public access system can be embedded into a community network. What follows is a description and analysis of five different community networks.
Ideally, a community network should:
Five community networks are briefly described below.
The Cleveland Community Free-Net (Telnet to freenet-in-a.cwru.edu)
The Cleveland Community Free-Net within the Greater Cleveland area was one of the first community networks in existence. It began as a system dedicated to public provision of medical information. It quickly blossomed into a system serving 30,000 people in the Greater Cleveland area. It began as a BBS and grew quickly with support from the National Public Telecomputing Network (NPTN). The NPTN is a national organization aspiring to be the Public Broadcasting System of community networking. The Cleveland Community Free-Net popularized the term "free-net" and developed the system of a virtual community. A virtual community is a series of menus that begin generally and delve into a particular theme with a series of submenus. Currently, the Cleveland Community Free-Net does not offer a graphic user interface. However, it does provide access to a large audience, has a clearly organized interface, and is simplistic and useable on slower, older computers. The Cleveland Community Free-Net is an excellent model to be emulated.
The Blacksburg Electronic Village
The Blacksburg Electronic Village in Blacksburg, Virginia is a community network begun by self-motivated private parties. It was initiated by Bell Atlantic who provided telecommunication resources and expertise for establishment of the network. The Blacksburg Electronic Village is available from both text and graphic web browsers making it a flexible community network. It provides electronic mail, discussion groups, the ability to make commercial purchases on-line, and access to static graphic images. The Blacksburg Electronic Village is a blend of information services and commercialization with considerable activity revolving around advertisement.
Big Sky Telegraph
The University of Montana runs the Big Sky Telegraph out of Dillon, Montana. It was begun in 1988 with assistance from the M.J. Murdock Charitable Trust and US WEST. The original intent was to provide an educational tool for teachers in distant rural areas. After a startup lesson on telecommunications, teachers used the system for collaborative efforts and long distance classes. The system is a classic example of a low cost community network that reaches new audiences in remote areas. Big Sky Telegraph began running on a IBM-compatible 386 running Linux and BBS software. Funding for the system is primarily soft funding from donations and grants. The community network is text based but offers full Internet connectivity. No access to graphic information or GIS functionality is possible.
The Boulder Community Network (BCN)
The Boulder Community Network (BCN) in Boulder, Colorado is accessible via terminals in public schools and libraries as well as from dial-in access. BCN is accessible via the WWW. Like many sites on the WWW, BCN can be viewed with both text browsers or graphic browsers. BCN offers on-line information about itself, business opportunities in Boulder, education, health and environmental issues in the area, job training, human services, public transportation, and local and regional weather. BCN is the best example of a community network that emulates what the PAS could be on the WWW. BCN contains a link to the Boulder County Land Use Department GIS WWW site. Once the user is at this site, they can view spatial data and a data dictionary. Users can observe the comprehensive plan, zoning updates, and ongoing parcel changes. However, no interactive GIS functionality is provided. Electronic mail can be sent directly to BCN developers and staff. BCN tracks users and provides weekly and monthly statistics. By the first week of June, 1995, BCN had almost 52,000 users access the system. Local government leaders can also track public opinion by following who accesses particular pages in the system and compiling comments left by users. BCN is successful because it is easily accessible and user-friendly.
The Santa Monica Public Electronic Network (PEN)
The Public Electronic Network (PEN), of Santa Monica, California was started in 1989. PEN promoted computer mediated communication in the concept's earliest days. Computer mediated communication consists of electronic mail, public discussions, and providing an area for the public to post opinions. PEN limits user membership to members of the community and does not provide mail connectivity to the outside world. This is done to preserve a sense of place and culture for Santa Monica.
Free access terminals were installed in public libraries and other places around Santa Monica. Shortly afterward, users began to coagulate into interest groups. The plight of the homeless became one of the major issues of PEN users. PEN administrators noticed homeless people participating through library terminals demonstrating an example of a community network involving a hard to reach audience. This interest group eventually formed an organization to benefit homeless people. Fund raising began and a shelter was built. PEN organizers reported that a benefit of the system was that users could read system postings before becoming involved, thus retaining some anonymity. PEN has recently been upgraded to a WWW site. Although the site does not offer access to spatial data or GIS functionality, it does offer many of the attributes similar to the Boulder Community Network and Blacksburg Electronic Village.
Neighborhood and home owner associations were selected as the unit of study because they have localized interests, are well organized, a need for access to local spatial information, and they are often a conduit for residents to voice their concerns and objections to local municipality planning and policy decisions. After discussion with the City of Tucson's Citizen and Neighborhood Services office and experts working with neighborhood and home owner associations throughout the area, four selection criteria were identified:
Once the selection criteria were developed, four neighborhood associations (Figure 1) were selected:
The Balboa Heights Neighborhood (Figure 2)
The Balboa Heights neighborhood is defined by Glenn Street to the north, Stone Avenue to the east, Grant Avenue to the south, and Oracle Road to the west. Balboa Heights shares it eastern border with the Keeling Neighborhood. Balboa Heights has a strong presence of multi-family apartment complexes and is distinctly urban in nature.
The Balboa Heights Neighborhood Association is well known in the City of Tucson for its activism and successful programs. Balboa Heights was the first neighborhood association to successfully construct and manage a neighborhood park on donated land. The association has a President - Jane Baker, a Vice President, a Secretary, and a Treasurer. The group meets monthly at E.C. Nash Elementary School.
The Corbett Neighborhood (Figure 3)
The neighborhood is defined by 22nd Street to the north, Alvernon Avenue to the east, Golf Links Road to the south, and Campbell Avenue to the west. The Corbett neighborhood covers one square mile in the near southeastern part of Tucson. The neighborhood is comprised predominantly of small single family lots and is bordered and intersected by major arterial roads. The neighborhood was built almost entirely in the late-1950s as Tucson expanded to the southeast. This one-time growth spurt gives the neighborhood a common thread of house design, spatial layout, and street appeal.
The Corbett Neighborhood Association is an active association with a core membership of approximately 20 people. They have been active for many years and focus primarily on crime reduction, neighborhood clean up, and lobbying for the design and construction of a local neighborhood center. The association has a chairperson - Debbie Johnson, a Vice Chairperson, and a Secretary/Treasurer. They meet on a monthly basis in the Corbett Elementary School.
The Flecha Caida Neighborhood (Figure 4)
The Flecha Caida neighborhood covers approximately 2.5 square miles. The neighborhood is unlike the other three in that it is actually three disconnected pieces. The neighborhood is in Pima County between the City of Tucson's northern boundary and the southern border of the Coronado National Forest. The entire Flecha Caida neighborhood is part of phase one of the Catalina Foothills development. Residents refer to different areas of the neighborhood by the stages in which the original developers laid them out. However, the neighborhood is clearly unified by the deed restrictions initiated during its inception.
The Flecha Caida Home Owner's Association is comprised almost entirely of home owners. Deed restrictions limit land use within the development and have played a large role in the homogenous evolution of the neighborhood. The association charges a small fee for yearly dues and uses that money to upkeep a FAX machine and PC computer as well as publish a newsletter entitled 'The Arrow.' The association spends most of its time tracking violations of deed restrictions primarily pertaining to illegal home businesses, mailboxes painted the wrong color, and other aesthetic issues. The association has a President - Tonya Hladky, executive board, and board of directors. They meet on a bi-monthly basis.
The Keeling Neighborhood (Figure 5)
The Keeling neighborhood is defined by Fort Lowell to the north, 1st Avenue to the east, Grant Road to the south, and Stone Avenue to the east. The Keeling neighborhood is located directly east of the Balboa Heights neighborhood. The neighborhood strongly resembles Balboa Heights in terms of land use and character but is slightly larger.
The Keeling Neighborhood Association is an active association with a core membership of about 30 people. The association is concerned about neighborhood crime, graffiti removal, fund raising, and continuing their projects of a neighborhood circus and house repairs. The association has a President, a Vice President, a Secretary, and a Treasurer. The association meets on a monthly basis at Keeling Elementary School.
Because of space limitations, only perceived theme importance and awareness ratings for the overall study population and each neighborhood association are examined in this paper. Refer to Paul Braun's Masters Thesis for an analysis of the remaining contents of the pre-prototype questionnaire as well as how responses varied by sex and home status (owner versus renter).
72 people responded to the pre-prototype questionnaire.
The initial draft of the pre-prototype questionnaire was reverse engineered based on the categories of information needed to complete the study. A general idea of relevant topics and questions were also gathered from studies by the University of Wisconsin's Land Information and Computer Graphics Facility (Vonderohe et al. 1991).
After answering the pre-prototype questionnaire, participants were asked to use the Public Access System (PAS) prototype and answer the post-prototype questionnaire.
Based on pre-prototype questionnaire responses, particular data layers (e.g., land use and land values) and software capabilities (e.g., zooming in and out, panning, querying particular features, and presentation of metadata) were deemed necessary or desirable.
Two versions of the prototype were tested. After discussions on the advantages and shortcomings of the test prototypes, a final version was developed. Issues that surfaced during prototype development were:
The final prototype incorporated most of the comments and suggestions from the initial prototypes. By double clicking on the 'PAS' button (Figure 6), users would access the PAS prototype. This button would initiate a series of events; a menu prompting the user for their name(Figures 7) and a menu prompting the user to pick the neighborhood association they belong to (Figure 8). The program would then open a view and display regional information such as TIGER road data, river washes, highways, and locational text for the particular neighborhoods involved in the study (Figure 9). The image in Figure 9 is of the entire Tucson basin. The hillshade was made in the Grid module of ArcInfo version 7.0.4. Users would see the Tucson basin as a whole and could then do one of two things; begin turning on and off basin-wide data layers, such as US Census Block Group data, or zoom in to their neighborhood and begin turning on and off neighborhood-wide data layers, such as parcels.
For each data layer there are two buttons in the prototype. A button to add a data layer to the view and a button to delete a data layer from the view. The button to delete data layers from the view has the same icon as the button to add data layers but with a red circle and slash superimposed upon the icon . Some data layers required that the user determine which attribute to classify on before that data theme could be added to a view. For example, parcel data provided eight different attributes that the user could classify on.(Figure 10). Once a choice was made, the selected data layer would appear in the view. Also a metadata window (Figure 11) was shown prior to displaying the selected data layer. The metadata window would explain what attribute information existed for that data layer, where the data came from, when it was created, and which attribute would be used for the initial classification. The metadata window was a highly simplified version of the Federal Geographic Data Committees conception of metadata.
Users were given access to ArcView tools such as zooming in and out, panning, identifying singular features, and measuring distances. Within the limited time allotted for use of the prototype, anything more complicated would have warranted training before use of the prototype.
A handful of association members within each neighborhood association expressed interest in interacting with the PAS and providing comments in the post-prototype questionnaire. Prototype users worked on Pentium PCs running Microsoft NT 3.51. ArcView and all affiliated spatial and tabular data were provided locally to ensure system performance.
Of the 72 people that took part in the pre-prototype questionnaire, 16% (12 people) used the PAS prototype.
Because of space limitations, only perceived theme importance and awareness ratings for the overall study population and individual neighborhood associations are examined in this paper. Refer to Paul Braun's Masters Thesis for an analysis of the remaining contents of the post-prototype questionnaire as well as how responses varied by sex and home status (owner versus renter).
Of the 72 people that took part in the pre-prototype questionnaire, 16% (12 people) responded to the post-prototype questionnaire.
Neighborhood association members were asked to rate their perceived level of importance and their perceived awareness of 30 different spatial data layers. The list of 30 themes were determined from the literature review and discussions amongst project members (student and thesis committee). Means were calculated for each of the 30 data theme ratings from 1 - 5. Those rating categories were:
The themes were then ranked based on mean ratings. The 30 spatial data themes are listed below.
1. School district boundaries |
2. Water district boundaries |
3. Fire district boundaries |
4. Utility district boundaries |
5. Voting district boundaries |
6. Locations of schools |
7. Locations of churches |
8. Locations of post offices |
9. Locations of public transport |
10. Locations of bike routes |
11. Locations of landfills |
12. Locations of parks |
13. Locations of floodplains |
14. Variations in the terrain (Contours) |
15. Types of soils |
16. Locations of trail and trail heads |
17. Location of the City of Tucson boundary |
18. Location of the Pima County boundary |
19. Location of subdivision boundaries |
20. Locations of crimes |
21. Locations of frequent automobile accidents |
22. Locations of commercial, industrial, and residential properties within your neighborhood |
23. Census tract boundaries |
24. Demographic data about your neighborhood (e.g., age, race, and sex of households) |
25. How ethnically diverse your neighborhood is in comparison to the City of Tucson |
26. Land use of individual parcels |
27. Zoning of individual parcels |
28. Assessed values of structures within your neighborhood |
29. Owner-occupied housing in your neighborhood |
30. Housing owned by residents of your neighborhood |
Respondents were asked to rate their perceived importance of the 30 spatial data themes mentioned in the previous section. The intent of doing this was twofold. First, ranked mean ratings of theme importance would be measured before and after exposure to the PAS prototype to see if the system affected their perceived importance of the 30 themes. Second, when perceived ranked mean ratings of theme importance were found to be low and ranked mean ratings of theme awareness were found to be high, a gap in desired knowledge was noted. These gaps in knowledge could assist PAS developers in knowing where forthcoming efforts should be focused.
The following table lists the rank of mean theme importance ratings for the 30 spatial data themes by overall study population and individual neighborhood association. Ranks shown in boldface are where a significant difference occurred between the mean ratings of neighborhood associations.
Spatial Data Themes | Corbett | Flecha Caida | Balboa Heights | Keeling | Overall Rating |
Locations of crimes | 1 | 2 | 2 | 2 | 1 |
Locations of auto-accidents | 8 | 3 | 1 | 1 | 2 |
Assessed values of structures | 6 | 1 | 5 | 4 | 3 |
Land use of individual parcels | 15 | 4 | 4 | 3 | 4 |
Owner-Occupied Housing | 4 | 13 | 9 | 7 | 5 |
Locations of parks | 2 | 14 | 14 | 5 | 6 |
Zoning of individual parcels | 9 | 7 | 18 | 18 | 7 |
Housing Owned by neighborhood residents |
3 |
12 |
13 |
10 |
8 |
Locations of floodplains | 7 | 8 | 15 | 16 | 9 |
Fire District Boundaries | 17 | 10 | 7 | 20 | 10 |
Locations of commercial, industrial, and residential properties |
10 |
11 |
18 |
19 |
11 |
Location of the City of Tucson Boundary | 16 | 5 | 27 | 21 | 12 |
Voting District Boundaries | 14 | 16 | 10 | 8 | 13 |
Location of the Pima County Boundary | 24 | 6 | 28 | 26 | 14 |
Location of landfills | 12 | 19 | 26 | 15 | 15 |
Location of post offices | 28 | 20 | 8 | 6 | 16 |
Utility District Boundaries | 22 | 15 | 11 | 30 | 17 |
Subdivision Boundaries | 29 | 9 | 25 | 27 | 18 |
Locations of Schools | 5 | 24 | 6 | 29 | 19 |
School District Boundaries | 19 | 18 | 20 | 25 | 20 |
Census Tract Boundaries | 25 | 22 | 12 | 17 | 21 |
Neighborhood Demographic Data | 11 | 25 | 23 | 12 | 22 |
Locations of public transport | 21 | 27 | 3 | 9 | 23 |
Locations of trails and trail heads | 23 | 21 | 29 | 11 | 24 |
Locations of bike routes | 13 | 26 | 22 | 13 | 25 |
Water Service District Boundaries | 30 | 17 | 24 | 23 | 26 |
Contours | 26 | 23 | 21 | 24 | 27 |
Neighborhood ethnic diversity | 20 | 28 | 16 | 22 | 28 |
Types of soils | 18 | 29 | 30 | 14 | 29 |
Locations of churches | 27 | 30 | 17 | 28 | 30 |
Significantly Different Theme Importance Ratings by Association | Test-Statistic (t-test in SAS 6.0) | Less Important | More Important |
City of Tucson Boundary | 0.0129 | Keeling | Flecha Caida |
Pima County Boundary | 0.0024 | Keeling | Flecha Caida |
Locations of Post Offices | 0.0435 | Corbett | Balboa Heights |
Utility Districts Boundaries | 0.0182 | Keeling | Flecha Caida |
Subdivision Boundaries | 0.0004 | Corbett and Keeling | Flecha Caida |
Locations of Schools |
0.0069 |
Balboa Heights and Corbett |
Keeling |
Locations of Public Transport | 0.0222 | Flecha Caida | Balboa Heights |
Entire Study Population
Entire study population's top ten data themes by rank of mean theme importance ratings:
Corbett Neighborhood Association
Corbett's top ten data themes by rank of mean theme importance ratings:
Flecha Caida Neighborhood Association
Flecha Caida's top ten data themes by rank of mean theme importance ratings:
Balboa Heights Neighborhood Association
Balboa Heights' top ten data themes by rank of mean theme importance ratings:
Keeling Neighborhood Association
Keeling's top ten data themes by rank of mean theme importance ratings:
Crime, auto-accidents, and assessed value of structures stand apart from the remaining 27 themes because they are ranked first through third overall and are in the top ten for all four associations. Land use of individual parcels and the locations of owner-occupied housing follow closely behind because they are ranked fourth and fifth overall and are in the top ten for three of the four associations.
Once those top five themes have been set aside, the similarities between the four associations begin to break down. Associations can be grouped in pairs. Two of four associations ranked the following ten data themes in their top ten; locations of parks, zoning of individual parcels, housing owned by neighborhood residents, locations of floodplains, fire district boundaries, locations of commercial, industrial, and residential properties, voting district boundaries, locations of post offices, locations of schools, and locations of public transportation. The two associations which did not rank any of these ten themes in their top ten tended to agree on the unimportance of those themes. This demonstrates a rather clear border between associations who want those themes in their top ten and those who do not. An example of this dichotomy exists with zoning of individual parcels. Corbett and Flecha Caida both ranked zoning of individual parcels in their top ten while Balboa Heights and Keeling ranked it 18th.
Respondents were asked to rate their awareness of 30 spatial data themes mentioned earlier. The intent of doing this was twofold. First, ranked mean ratings of theme awareness would be measured before and after exposure to the PAS prototype to see if the system affected user knowledge of the 30 themes. Second, when ranked mean ratings of theme awareness were found to be low and ranked mean ratings of theme importance were found to be high, a gap in desired knowledge was noted. These gaps in knowledge could assist PAS developers in knowing where forthcoming efforts should be focused.
The following table lists the rank of mean theme awareness ratings for the 30 spatial data for the entire study population and each neighborhood association. Ranks shown in boldface are where a significant difference occurred between the mean ratings of neighborhood associations.
Data Themes | Corbett | Flecha Caida | Balboa Heights | Keeling | Overall Rating |
Locations of post offices | 4 | 1 | 2 | 1 | 1 |
Locations of schools | 1 | 3 | 4 | 2 | 2 |
Locations of parks | 2 | 2 | 3 | 7 | 3 |
Location of the City of Tucson boundary | 3 | 4 | 9 | 6 | 4 |
Locations of public transport | 7 | 10 | 1 | 3 | 5 |
Location of the Pima County boundary | 5 | 5 | 10 | 5 | 6 |
Locations of churches | 9 | 6 | 15 | 4 | 7 |
Locations of school districts | 11 | 7 | 12 | 21 | 8 |
Locations of commercial, industrial, and residential land uses |
8 |
19 |
5 |
8 |
9 |
Locations of voting districts | 17 | 9 | 16 | 10 | 10 |
Locations of subdivision boundaries | 13 | 11 | 14 | 24 | 11 |
Locations of bike routes | 18 | 14 | 11 | 9 | 12 |
Locations of trails and trail heads | 16 | 13 | 13 | 11 | 13 |
Locations of auto-accidents | 6 | 26 | 17 | 12 | 14 |
Locations of owner-occupied housing | 14 | 16 | 7 | 25 | 15 |
Locations of crimes | 12 | 24 | 8 | 14 | 16 |
Housing owned by neighborhood residents | 19 | 18 | 6 | 23 | 17 |
Locations of floodplains | 25 | 12 | 25 | 20 | 18 |
Locations of landfills | 10 | 25 | 22 | 15 | 19 |
Contours | 29 | 8 | 30 | 17 | 20 |
Neighborhood ethnic diversity | 15 | 21 | 23 | 13 | 21 |
Assessed value of structures | 22 | 20 | 24 | 19 | 22 |
Land use of individual parcels | 26 | 22 | 18 | 22 | 23 |
Zoning of individual parcels | 27 | 23 | 19 | 25 | 24 |
Locations of utility districts | 23 | 15 | 27 | 28 | 25 |
Locations of fire service districts | 21 | 17 | 26 | 29 | 26 |
Neighborhood demographics | 20 | 27 | 21 | 18 | 27 |
Locations of census tract boundaries | 24 | 30 | 20 | 27 | 28 |
Types of soils | 30 | 28 | 29 | 15 | 29 |
Locations of water districts | 28 | 29 | 28 | 30 | 30 |
Significantly Different Theme Awareness Ratings by Association | Test-Statistic (t-test in SAS 6.0) | Less Aware | More Aware |
Locations of commercial, industrial, and residential land uses | 0.0160 | Flecha Caida | Corbett |
Locations of auto accidents | 0.0019 | Flecha Caida | Corbett |
Locations of crime | 0.0272 | Flecha Caida | Corbett |
Locations of public transport | 0.0177 | Flecha Caida | Corbett |
Locations of landfills | 0.0299 | Flecha Caida | Corbett |
Entire Study Population
Entire study population's top ten data themes by rank of mean theme awareness ratings:
Corbett's top ten data themes by rank of mean theme awareness ratings:
Flecha Caida Neighborhood Association
Flecha Caida's top ten data themes by rank of mean theme awareness ratings:
Balboa Heights Neighborhood Association
Balboa Height's top ten data themes by rank of mean theme awareness ratings:
Keeling Neighborhood Association
Keeling's top ten data themes by rank of mean theme awareness ratings:
All four associations awareness ratings for post offices, schools, parks, and public transportation, and the City of Tucson and Pima County boundaries were in their top ten. Locations of churches and general land uses (i.e., commercial, residential, and industrial) are included in three of four association top ten ratings with Balboa Heights dissenting on churches and Flecha Caida dissenting on awareness of land uses. This surprising correlation denotes a common awareness of nine of the top ten themes.
There is also strong agreement amongst associations on data themes they consider themselves most unaware. From the overall theme awareness rating list, nine of the bottom ten data themes (ratings 21-30) have at least three associations rating them in their bottom ten. Those nine data themes are assessed value of structures, land use of individual parcels, zoning, locations of utility districts, neighborhood demographics, locations of census tract boundaries, types of soils, and locations of water utility district boundaries.
Corbett rated five themes significantly higher than did Flecha Caida. Three of the five themes deal with issues related to transportation. The five themes were locations of commercial, industrial, and residential land uses, locations of auto-accidents, locations of crime, locations of public transport, and locations of landfills. This is to be expected because the two neighborhoods that differ (Corbett and Flecha Caida) also differ in both physical and socio-economic characteristics. However, if those differences are to be expected, than Keeling and Balboa Heights should also differ somewhat from Flecha Caida. This was not the case and it cannot be explained with the data available.
The comparison of ranked mean theme importance and awareness ratings for the study population and by neighborhood association indicates gaps in study participant knowledge. From this analysis, four data priorities can be deciphered.
High priority theme are those ranked in the top ten for theme importance and in the bottom ten for theme awareness. Respondents considered this data important but they knew little about it. The PAS prototype would be of most benefit providing access to this data.
Medium priority themes are those ranked in the top ten for theme importance but, unlike the high priority themes, they are also ranked in the top for theme awareness. Respondents considered this data very important yet they were well aware of it. This data is also important to provide access to but because of respondent awareness, it takes a lower priority.
Low priority I themes are those ranked in the bottom ten for both theme importance and awareness. These PAS prototype would be least effective at providing access to these data themes.
Low priority II themes are those ranked in the bottom ten for theme importance and ranked in the top ten for theme awareness. Respondents considered themselves well aware of these data themes but they considered these data themes to be unimportant. These data themes could be withheld from development until time and resources are available and more immediate needs have been met.
Data Themes | Overall Ranking of Mean Data Theme Importance | Overall Ranking of Mean Data Theme Awareness |
Locations of crimes | 1 | 16 |
Auto-accidents | 2 | 14 |
Assessed values of structures | 3 | 22 |
Land use of individual parcels | 4 | 23 |
Owner-Occupied Housing | 5 | 14 |
Locations of parks | 6 | 3 |
Zoning of individual parcels | 7 | 24 |
Housing Owned by neighborhood residents | 8 | 16 |
Locations of floodplains | 9 | 18 |
Fire District Boundaries | 10 | 26 |
Commercial, industrial, and residential properties |
11 |
9 |
Location of the City of Tucson Boundary | 12 | 4 |
Voting District Boundaries | 13 | 10 |
Location of the Pima County Boundary | 14 | 6 |
Location of landfills | 15 | 19 |
Location of post offices | 16 | 1 |
Utility District Boundaries | 17 | 25 |
Subdivision Boundaries | 18 | 11 |
Locations of Schools | 19 | 2 |
School District Boundaries | 20 | 8 |
Census Tract Boundaries | 21 | 28 |
Neighborhood Demographic Data | 22 | 27 |
Public transport | 23 | 5 |
Trails and trail heads | 24 | 13 |
Locations of bike routes | 25 | 12 |
Water Service District Boundaries | 26 | 30 |
Contours | 27 | 19 |
Neighborhood ethnic diversity | 28 | 21 |
Types of soils | 29 | 29 |
Locations of churches | 30 | 7 |
High Priority: High Importance Theme Ratings with Low Awareness Theme Ratings:
Medium Priority: High Importance Theme Ratings with High Awareness Theme Ratings:
Low Priority I: Low Importance Theme Ratings with Low Awareness Theme Ratings:
Low Priority II: Low Importance Theme Ratings with High Awareness Ratings:
The PAS Prototype
The PAS prototype is a logical connection between citizen's rights to obtain government information and the maturation of GIS and desktop mapping technology. This combination results in easily accessible interactive access to spatial data for the public. Access will not only assist in creating more informed citizens, but improve the data by having more people examine it and comment on it.
The need for public access to spatial data is agreed upon by both data providers and users and is practiced in some form or another in a handful of organizations. However, the systems are somewhat haphazardly designed and are not based on a full understanding of what data the user community desires. A few studies have attempted to determine what spatial data different users desire. However, those studies are predominantly surveys of data and do not prioritize data. Based on this study, a public access system geared toward neighborhood and home owner associations should focus on the themes listed below in order to appease the most users with minimal effort.
A PAS must be designed with concerns of citizen privacy and data security in mind. A proactive program must also be put in place to promote the system and its usage to less fortunate members of society and groups that need assistance. The societal impacts of a public access system are directly related to citizen privacy, data security, and public access. Blank acceptance of computerized data and people's perceptions that technology is autonomous and driving change makes the simplistic and understandable design and functionality of an access system imperative. (see Master's Thesis for detailed discussion of these topics)
The PAS for this study was designed using existing community networks as a guideline. The PAS is an evolutionary step in the development and usage of community networks. This step is held back by the need for graphic interfaces and a knowledgeable user community. Community networks are also important to the design of a PAS because the people that use community networks are the same clientele that may be interested in using a PAS. Therefore, the design of a PAS should be in agreement with community network philosophy.
The impact of the PAS would be most felt by giving citizens access to crime data, locations of auto-accidents, parcel based data such as assessed values of structures, land use of individual parcels, owner-occupied housing, zoning, and housing owned by neighborhood residents, as well as the locations of floodplains and fire service district boundaries. The highly urban neighborhoods (Corbett, Balboa Heights, and Keeling) would also benefit from additional access to bike routes and demographic data.
The most promising aspect of the impact of the PAS prototype is that the system took many users to a new level of spatial thinking. After using the PAS prototype, many users expressed the desire to examine correlations between data themes. For example, users wanted to know the correlation between land uses and assessed value indicating that they had assimilated information on those two themes and wanted to analyze spatial data at a higher level. The PAS served as a tool to provide association members with the ability to explore issues and opened new connections between various physical, social, and economic phenomenon.
Users expressed their enjoyment with the ability to visualize spatial data. They were especially pleased with the ease they could turn on and off data, zoom in and out, print out hard copy maps, and perform simple spatial analysis. The metadata provided gave users a better understanding of what they were viewing.
The PAS was designed to enhance citizen empowerment. The prototype opened the door and demonstrated how people could access data. All respondents to the post-prototype questionnaire felt they had a strong tool in their hands.
PAS Prototype Shortcomings
The full analytical capabilities of the system were not implemented by neighborhood users. This is partially due to the limited time they had to use the PAS. For more advanced capabilities, stemming beyond basic desktop mapping functionality, it might be beneficial to train an association representative. The capabilities could include analyzing data themes to look for correlations between themes, joining databases either by attribute or by spatial location, and developing methodology that is consistent with the data at hand and its spatial and attribute limitations.
The log files of user activities demonstrated that there was no consistent method to their interaction with the data. Users activated particular data themes multiple times without removing them. No data themes were turned on or off more than any other. This probably demonstrates the need for a more simplistic user interface and possibly more training.
The post-prototype questionnaire asked respondents what other data themes they would like to have access to once they had used the PAS. Questionnaire respondents gave answers that fit into two categories
Themes such as:
Themes such as:
Implementation of a "starter" PAS would not be difficult. Many local governments now have a great deal of spatial information already available in a digital format. This data can be readily used in most desktop mapping and GIS applications. This study demonstrated that a large budget and effort are not needed to obtain data and put together a public access system (the project was completed without any budget). Public agencies are generally in favor of providing data for circumstances such as what neighborhood associations are interested in as long as the onus is put on the user to format data and pay for time and labor to download the data. That attitude might change as public institutions begin to see a growing user community.
CD-ROM Implementation
The PAS could be implemented on a CD-ROM at neighborhood association offices, county offices, or public libraries. The customized ArcView and data could be cut onto CD-ROM and distributed to the locations listed above. This could be done using Esri's ArcView Data Publisher. This would eliminate problems with Esri product costs and the fact that Esri software is proprietary. Neighborhood association members are used to searching out information on issues that concern them at the library and local governmental offices so this model of implementation would replicate what is currently going on. If the PAS was easily accessible (i.e., on countertop PCs), it would most likely get used. The main issue with stand alone systems would be data currency and developing a mechanism for reporting errors discovered in the data. A staff member from a public institution may need to oversee the system and handle user comments about data omissions and errors. This staff member should have connections to all PAS data providers in order to facilitate timely data revisions.
Tucson has created multi-disciplinary teams of governmental employees to focus on particular neighborhoods. These teams could run a CD-ROM version of the PAS on a laptop. The system would be portable so that it could be run at public meetings or when the team is working in the field.
World Wide Web Implementation
With the new wave of World Wide Web GIS mapping software releases, specifically the MapObjects Internet Map Server and the forthcoming ArcView Internet Map Server, the PAS really belongs on the WWW. Most local libraries now offer WWW connections. Neighborhood association members could tap into the PAS remotely and have full access to its data themes and GIS functionality. This model would remove some of the queries commonly heard at the public window at county planning or zoning offices.
Ideally, the PAS should be rewritten in MapObjects so that it is "lighter". Even though ArcView has been heavily customized in the PAS prototype, the computer running the application must still load all of ArcView even though only pieces of it are offered to the user. If the PAS was slated for WWW publication, it would be beneficial to invest in a MapObjects version.
Cost
Neighborhood association members who used the PAS prototype expressed that they would be willing to raise funds through traditional methods to pay for the development and access to the PAS prototype. Cost of a PAS should be small. Development of the PAS prototype was done with no funding (reminiscent of how community networks function). The system should be simple in its early days and grow in functionality as it proves its utility and the user community becomes knowledgeable and more sophisticated in its questions. Software donations and existing computer equipment could put the system in a variety of public libraries and neighborhood associations. Grants could be pursued to fund computer procurement for neighborhood associations.
Archer, Hugh and Croswell, Peter, 1989. Public Access to Geographic Information systems: An Emerging Legal Issue, Photogrammetric Engineering and Remote Sensing, Vol. 55, No. 11, pp. 1575-1581.
Bauman Foundation, 1995. Agenda for Access: Public Access to Federal Information for Sustainability through the Information Superhighway. Baumann Foundation, Washington, D.C.
Corbley, Kevin, 1995. "Citizens Enjoy GIS Access", GIS World, March 1995, pp. 66-68.
Friedley, Dale and Colbert, Larry, 1991. Reaching for Consensus in Public Information Access Policies, Proceedings of the 1991 Annual URISA Conference, vol. 4, pp. 51-64.
Guy, Neil, K., 1996. Community Networks: Building Real Communities in a Virtual Space?, Simon Fraser University, Master of Arts in Geography.
Lerner, B. Nancy, 1992. Developing a Successful Information Access Policy. Proceedings of the 1992 Annual URISA Conference, vol. 4, pp. 79-84.
Stipek, Lawrence, and Torraca, David, 1993. Implementing and Assessing the Value of Direct Public Access to the Loudoun County, Virginia Geographic Information System. Proceedings of the 1993 Annual URISA Conference, vol. 2, pp. 202-208.
Tosta, Nancy, 1991. "Public Access: Right or Privilege?", Geo Info Systems, November/December 1991. vol. 1, number 10, pp. 20-25.
Vonderohe, Alan, Gurda, Robert, Ventura, Stephen, and Thum, Peter, 1991. Introduction to Local Land Information Systems for Wisconsin's Future, Wisconsin State Cartographer's Office.
Warren, S.D., and L. Brandeis, (1890). "The Right to Privacy." Harvard Law Review 4.5: pp.193-220.
Paul Braun is the Campus GIS Coordinator for the University of South Carolina. He holds a MS in Renewable Natural Resource Studies from the School of Renewable Natural Resources, University of Arizona and a BS in Landscape Architecture from the University of Wisconsin-Madison. He can be reached at the University of South Carolina, Columbia, SC 29205. His telephone is (803) 777-4590, FAX (803) 777-7489. Paul Braun - paul@garnet.cla.sc.edu
D. Phillip Guertin is an Associate Professor within the School of Renewable Natural Resources, Advanced Resource Technology Group, University of Arizona. He can be reached at the University of Arizona, Tucson, Arizona 85721. His telephone is (520) 621-1723, FAX (520) 621-8801.
Dr. Phil Guertin -- phil@nexus.srnr.arizona.edu