Randy M. Brooks, Ph.D. Millikin University

Illinois Community Networks Conference (May 7, 1997)

---

Community Information System
Development Process

The "Answer Machine" is a touch screen kiosk of community information, available in the Decatur Public Library and at Millikin University. This hypertext presents a short overview of the basic development process, purposes, and outcomes of this project. This project is an example of Millikin University's plan to develop service learning in all disciplines.

The comunity information system was developed through the collaborative efforts and support of Communities in Partnership, Millikin University and the Decatur Public Library. Communities in Partnership provided office space for development of the information system, helped organize planning focus groups and gathered information. Millikin University received a national grant to a develop service learning program including money to develop a clearinghouse of volunteer opportunities in the community. The Decatur Public Library provided funding and location for another public kiosk. The library also provided a place for public user testing and ongoing maintenance by library staff.

Focus Group Planning--goals, purposes, publicity, institutional support Gathering Information--surveys, phone follow ups, walk-abouts, editor model Building the Kiosk--electronic exhibits cabinet maker, touch screen, computer Developing Prototypes / Structural Models--visual mock-ups of screens, indexes, menus, informaton hierarchy, layers of information Information System User Testing--protocals of diverse public, videotape hands using kiosk, alternative designs, suggestions from agencies Official Launch / Publicity--public launch w/politicals, media, user giveaways Updating and Maintenance--library staff, service learning, me Converting to Web--screens to pages, buttons, logos, searches

---

Community Need?

Decatur has typical problems blue collar job loss / no job security / unemployment / underemployment / drug abuse / gangs / drop outs / criminal lifestyles

Decatur has caring, helping tradition increase giving to United Way in times of labor strife & national dishonor lots of church community action / volunteers several model social service programs (DOVE, BabyTalk, Homework Hangout) some strong, active neighborhood organizations

No Clearinghouse / Coordination of Services directory of organizations printed in 1991 (quickly out of date) United Way database lacked technical coordination & follow-through hard to find out where to volunteer / hard to know where help is available

Choose Your Community Information System Development Topic:

Focus Group Planning	Interface Design	Launch & Publicity
Gathering Information	Performance Goals	Updates & Maintenance
Building the Kiosk	User Testing	Outcomes
Hypercard Prototypes	Web Conversion	Work & Participation

back to top

Community Strategic Planning / Focus Groups

Community Information System-an initiative to develop community awareness of great social services of Decatur to encourage people to help, to volunteer to quickly refer people in need of help to reach the appropriate progra

Focus Group Planning diverse focus groups from across the community discussed & planned the name of the community information system "The Answer Machine" the type of information to be included (dynamic and static) the means of distribution or delivery of the information to the public financial support strategies

The Plan--touch screen kiosks and database of community services which could be updated as needed (without republishing the entire database).

back to top

Gathering Information is the Most Difficult Part of Project

Surveys (1) by Mail or (2) by Phone low survey completion response (too complicated/poorly designed) some organizations don't trust projects motives (cut duplicate programs) agencies claim to do everything & serve everybody classifications of services are either too technical or too general

Graphics few organizations would provide brochures or graphics for inclusion we scanned logos, took photographs and developed graphics

Walk-Abouts we finally resorted to visiting non-responsive organization for "tours" personally picked up material laying around (with permission)

back to top

Building the Kiosk

Getting Bids carpenters unaware of needs for ventilation of computers go with an electronics cabinet builder in Bloomington, Indiana

Physical Goals large and heavy enough to not be easily carried away slanted screen display (for standing users of various heights) wheel chair acessible (official guidelines) no access to keyboard, mouse shelf for bags, books, notepad while using printer out-hole for pages printed good ventilation through bottom of cabinet attractive wood with institutional sponsor logos (Millikin image)

Touch Screen-built into the cabinet but easily replaced if damaged inexpensive resistive touch window from EdMark (works like mouse)

back to top

Working With Hypercard / Prototypes

Why Hypertext? discrete units or modules (developed separately or at random) visual image, aural information and textual information easy customized interface scripting capabilities for user counts, update time stamping

Object Oriented Elements of Hypercard stack background card field button graphics/color sound bytes

back to top

Designing the Interface / Scripting Interaction

Simplicity--no extra info, minimal decorative elements, consistent screen placement of elements, consistent font use Prompt Choices--first screen welcomes and verbally prompts search method, navigational button names also prompt actions (quit, print, go to) Orientation on Two Search Screens--verbal prompt to use alphabet buttons and once there, users know to return for more searches Dual Function Headings (informational & hyperlinking)--button choices indicate type of information and link to the info Instant Feedback--choices instantly create hilited buttons, clicks or voice so user knows their touch has "controlled" the system

back to top

Hypertext Performance Goals

(1) No more than four touches to desired data. (2) No "missed" touches or inaccurate registrations of selections. (3) Multiple feedback (sound and visual response). (4) No need for help screens or instructions (all prompts and guidance needed built into the design when needed by users). (5) Total choice and control by user (including illusion of "quitting" their use). (6) User choice to print any information page. (7) Invisible use data collection / idle use detection to reset for next user.

back to top

---

User Testing

Protocal Analysis User Testing--talk aloud your thoughts while using

User Testing Facility & Equipment--library board room, videotape, two test coordinators observe (one on conceptual interaction, the other physical)

Coordinating Representative Users--invited agencies, diverse groups

User Testing Process 1. open-ended task to test general orientation and first time user prompts often checked their own organization or a known organization 2. requested specific task to find a specific organization quickly went to agency search list and found it 3. general services search task ("find help for fictional pregnant teenager") required services search helped refine public concepts of categories and key terms realized there are multiple organizations/services for most needs

back to top

---

Launch & Publicity

Launch Date provided closure and commitment (promise) to public planned celebration for official recognition of fulfillment media event / coverage (reporters, editors, radio, tv)

Primary Publicity Strategy community involvement in all phases of development user testing as "sneak previews" (requested suggestions for improvements) key "grapevine" community leaders pre-launch presentations demonstration to city council (on public tv)

Library Sign & Prominent Placement big sign over kiosk placed in main entrance to library explanation sheets and giveaway oversized paperclips given to users

back to top

---

Updates & Maintenance

Proposed Update Process dedicated phone line with actual answer machine for reporting changes advertised availability of kiosk and promotion of update phone number pro-active ongoing editorial updates (1/10 per month double checked) annual editorial review of all sites from print-out proof sheets paper reloading, basic window cleaning provided by library & Millikin technical support provided by me as needed

Natural Update Process someone looks up their organization, sees a mistake and complains we thank them profusely and change it and encourage them to inform us of new information in the future whenever changes occur

Web Version--updates could be emailed to us straight off the web site

back to top

---

Web Conversion

http://www.decaturnet.org/amachine.html

Why Convert? (duh) problems with computer platforms (database IBM / kiosk Macintosh) eventually we will have only one file to maintain more access to information from remote people

Why Not? less control of information more difficult interface for non-computer users touchwindow not discrete enough for small web links if truly online, no easy way to keep kiosk on community information only (yes, we could use html with netscape not online & enlarge navigation)

Easy Conversion cards become screens, buttons become buttons, fields become anchors search screens still conceptually work (though slow loading/scrolling)

back to top

---

Outcomes of Community Information System

Use Outcomes--over 600 use sessions per month / 1 ream of paper per week

Millikin Students & Faculty Outcomes national award for model of community education solutions several students involved gone into community service profession Millikin Service Learning program has expanded into new projects

Library Outomes applauded for provided information to community recognized as leader in the state for community information services

Community Outcomes anecdotes of people getting in contact with organizations community awareness about the extent and variety of services provided support for additional community collaboration (specifically a FreeNet)

back to top

---

Work & Community Participation

Faculty Hours Worked faculty member completed 190 from June 5, 1995 to August 14, 1995 he completed 200 hours developing the kiosk during the previous year web conversion took 140 hours of work by Dr. Brooks and his sons

Millikin University Students and Faculty Worked a total of 990 hours on this project each student took a fourth of the agencies listed in the Answer Machine each student also worked on categories of service

Participation by Area Agencies 228 area agencies participated by providing information 34 user testing sessions were held with 42 participants

back to top