Supporting End User Development in Community Computing: Requirements, Opportunities, and Challenges Lu Xiao, Umer Farooq Abstract End user development (EUD) tools in community computing are not well-developed and typically do not take into consideration the unique characteristics of community groups such as the lack of human, financial, and technological resources. Using a case study, we explore EUD in the domain of community computing. Situated in a community computing context, we identify design requirements of EUD tools, demonstrate the use of conceptual scaffolds to support EUD, and illustrate the need of new evaluation methods of EUD tools. We discuss the tension between pushing EUD tools to community computing for local autonomy on technology issues and the long time practice of seeking and relying on external technical expertise. We call for research studies that address the tension and explore ways of creating and stimulating the “pull” force from the community groups. Keywords: End user development, technology adoption and use, effective use, sustainability Introduction End user development (EUD) is about exercising greater control by non-developers and non-programmers over technology, such as enabling the design of computer-based applications without getting entangled in the nitty-gritty details of programming (Sutcliffe & Mehandjiev, 2004). EUD tools are critical in technology projects in community computing because non-profit community organizations often face financial issues in getting long-term technical support; therefore have to heavily rely on temporary volunteers. The use of EUD tools in community computing helps the organizations to empower themselves and take control of information technology in the projects. On the other hand, research in community computing enriches EUD. In the EUD community, researchers and developers often need to work in specific domain contexts because end users by themselves do not exist without such contexts. Non-profit community organizations provide a test bed for adopting, using, and evaluating EUD tools because people involved in these organizations provide a natural context, driven by civic and societal goals, for technology-based activities for which researchers in end user development aspire to build tools. Moreover, end users in non-profit community organizations as opposed to other contexts (e.g., individuals as end users) present contrasting requirements due to the nature of communitarian work. For example, non-profit community groups operate with little money and few resources; they rely on help from volunteers; there is usually a strong tendency to look to experts to help in the use of technology (Benston, 1990). These differences are viewed as grist for expanding the knowledge base of EUD. However, there has been only limited work in EUD for community computing. Early work in end user development had the goal of empowering computer users to pursue personal exploration and learning goals (e.g., Fischer & Lemke, 1988; Papert, 1993). Other work is more pragmatic, aiming to provide more accessible support for tasks that could benefit from programming techniques (e.g., spreadsheet manipulations (Nardi, 1993) and web-enabled technologies (e.g., Burnett et al., 2001). Towards a systematic understanding of end user experiences in a community context, Kase et al. (2008) developed a pattern schema that helps identify organizational and individual processes contributing to successful (and failed) informal technology learning outcomes in non-profit organizations. Yet, there is much research to do to fully understand the benefits of community computing to EUD and to develop EUD tools that leverage the rich context of non-profit community organizations and which support these organizations’ activities in their technology projects. In this paper, we present a case study concerning the usage and perception of EUD tools in a non-profit community organization. Driven by the case study, we discuss the EUD requirements in community computing, illustrate how conceptual tools can enhance EUD, and identify the challenges of adopting EUD tools in community computing. This is a perspective paper which focuses on the broad implications of research in community computing contexts. Specifically, we understand our work in community computing as an “incubator” (Carroll, 2001, pp. 310-311) for EUD tools. Integrating community computing with EUD can expand the frontiers of the latter conceptually, theoretically, and methodologically. This paper is organized as follows. We first discuss our methodology. We then present the design requirements of EUD tools and the conceptual scaffold of EUD tools, and describe the need for an alternative concept of usability to understand technology evaluation and adoption in community computing. We next identify the challenges that need to be addressed in appropriating EUD tools for local autonomy leveraging the “effective use” concept (Gurstein, 2003). Research Methodology The case study we report in this paper is part of a project called Civic Nexus (Merkel et al., 2004). In the Civic Nexus project, we worked with twelve community groups in State College, PA (USA) to increase their ability to enrich the lives of people and their activities in the local community by leveraging and enhancing their capacity to use information technology. Our goal has been to empower community groups, moving them toward greater control over their adoption, ongoing use, and management of technology. The context Our case study reports on the Spring Creek Watershed CommunitySpring Creek ). Spring Creek (http://www.springcreekwatershed.org) is organized around a commitment to showing how regional environmental and economic planning by watershed is more effective than planning by municipality. The mission of the organization is to explain basic terminology and information about watersheds, and to demonstrate the impacts of watersheds on people’s quality of life and the local economy. Spring Creek resembles most non-profits: they have few staff members; they often rely on volunteers to meet their organizational goals; they tend to make technology decisions based on limited human and financial resources; they often rely on small grants that give way to limited technology budgets for new technology initiatives such as developing a web site. Spring Creek hired a commercial vendor to develop and maintain their web site. Spring Creek was dissatisfied with the web site because it did not reflect their mission, overall goals, or the fact that they were a local group concerned with environmental and economic planning. For example, whereas the goal of Spring Creek was local economic planning, influencing decision makers, and encouraging quality of life through watersheds, the web site depicted them as a generic tree-hugger group. This incident was formative for the group. Kathy, the lead coordinator of Spring Creek, decided to avoid using commercial vendors and instead resolved that the Spring Creek group itself should direct the redesign process. The key players in this web redesign project are listed in Table 1. Table 1. Key players in Spring Creek Name Role Background Kathy Lead coordinator Limited technical background Emily Staff member for Clearwater Conservancy working on Spring Creek’s web site Limited technical background, trained as biologist Tim Technical volunteer for web design Technically proficient in web design and web technologies Dan Technical volunteer for web design Consultant for web design Ned Technical, unpaid intern for developing Spring Creek ’s online newsletter Undergraduate student in a Computer Science related program, owns a consulting company Our Involvement The assumption underlying our participatory design approach was that our community partners were not only shapers and decision makers in their organization, but also active technology users, developers of information content, and learners. For example, Kathy was the lead coordinator of the project from the organization, but also played an important role in the web design process as a user and in this role proposed suggestions about the web site from a user perspective. In our meetings with Kathy, she asked how to go on about the design of the web site. Together, we came up with a list of the expected features of the home page as follows: The home page should represent a local web site Images should reflect the local context, but places that are close to State College need to be included as well such as Boalsburg, Bellefonte, etc. Acknowledgements (to the photographers) for the pictures will be implicitly provided, e.g., they will show up as someone mouse's over the pictures The first page should not be cluttered, and the page should have as little text as possible. This list shows that Kathy was really involved in the design process as a user of the web site and she cared about the website from a user perspective. In the first meeting we had with Kathy, we were informed that the major technology issue that Spring Creek faced was to redesign their web site. The web site was expected to serve as a portal for local residents to get to know the organization and an information system that would support Spring Creek members’ daily work (e.g., upload and share data for the purpose of water resources monitoring). Spring Creek needed an external consultancy to help them become more actively involved in making technology-related decisions within their organization. We were involved in the web redesign process for about a year and a half, and played the role of bard (Carroll, 2006) in our mutual collaboration with Spring Creek with the primary goal of revamping their current web site. Our involvement is indicated by the timeline in Figure 1 that captures the major events in our collaboration with Spring Creek. Figure 1. Our involvement with SPRING CREEK in a timeline view, highlighting major events. In our work with Spring Creek, the primary concern for Kathy was the capability of manipulating web content by herself along with the non-technical volunteers in the organization. Based on our previous work in collaborative architectures (Ganoe et al., 2003) and following an evolutionary design approach (Rosson & Carroll, 2002), we provided a Wiki-like interface (Guzdial et al., 2000) for Spring Creek’s web site (Figure 2). We initiated the process of adding content on this web site and thereafter, Spring Creek had been enthusiastically updating and maintaining it. When we introduced the environment, Kathy remarked, “This (Wiki-like environment) is just motivating me… you’re putting something in front of me that I can use.” Our Wiki-like environment acted as an initial design artifact embodying hypotheses (Woods, 1998) about what is useful, what is not useful, and how it can be enhanced. This environment provoked Spring Creek to become conscious of their experience with it and help us understand the type of content management system desired by them and similar community groups. A Critical Incident At the later stage of this collaboration, Spring Creek faced challenges of deciding between two EUD tools. The first end user tool is the Wiki-like environment that we provided, which Spring Creek has been actively using during the summer (May-August), 2004 (see Figure 1). The other end user tool was provided by Ned, an intern with Spring Creek who owns an information technology consulting company. Ned was originally responsible for providing newsletter functionality to Spring Creek’s web site, where Kathy and other volunteers could automatically generate online newsletters by using templates and changing content without learning how to do programming. We had also been working with Ned to define specifications for this automatic newsletter generator as we did not want an artifact developed that would be unusable by Spring Creek after Figure 2. Screenshot of SPRING CREEK web site created by us Ned’s internship. In early August (2004), Ned requested Kathy's approval of the current web site interface so that the newsletter interface would be in compliance with it. After viewing the Wiki-like environment, Ned “surprised” Kathy after approximately a week, by sending her an email with a link not only to the newsletter generator but to a completely new web site. Ned, inspired by the Wiki-like environment, reused its content and developed the new web site as a content management system based on open-source software (Mambo, http://www.mamboserver.com). His email to Kathy contained the following excerpt: “With time running out I decided to make the whole effort worthwhile. I've configured and modified an open source content management system to suit your needs. Not just for the newsletter but for the entire website (emphasis added).” Kathy at that time was in a quandary as she needed to decide whether to go with our Wiki-like environment to which we were planning to add functionality based on the requirements from Spring Creek or Ned’s new content management system. In mid-August (2004), Kathy specifically asked us several times: “What do you think we should do?” Her inquiry was an interesting case from a research perspective for two reasons. First, being familiar with our Wiki-like environment, Kathy could have evaluated both systems from a usability standpoint—she did not. Second, she chose to ask us and no one else (she could have asked other volunteers or even Ned). In a phone conversation with Kathy (first week of September, 2004), she specifically expressed her concern that if she goes with Nick’s system, would we “have any bad feelings” in working with Spring Creek. All these showed that Kathy was welcoming of our advice, building on prior trust and rapport with us (because we were acting in the capacity of university researchers involved in technology projects). She once openly acknowledged our influence on both the conceptual and technical levels of end user development: “The assistance we’re getting is just incredible…I mean it’s really extremely impressive…that we’re an environmental organization, you’re a technology group and so, it’s not like our missions are necessarily coming together.” We indicated to Kathy that we would continue to work with Spring Creek irrespective of her decision, to which she responded, “I appreciate this”. Kathy chose the content management system over our Wiki-like environment in the end. Data Generation Our methods in working with Spring Creek built on previous work that takes a long-term participatory design approach in building information systems to address local needs (Carroll & Rosson, 2007). The field research with Spring Creek was carried out during a period of approximately 14 months. Data collection The primary method of data collection was observation recorded through field notes. We attended meetings held by Spring Creek’s web site committee, and other technology-related meetings, each lasting about an hour. However, observations were not just passive. We assumed a variety of roles within the case study context and participated in the events being studied in a variety roles (Yin, 2003, pp. 93-94). For example, we played active roles as both facilitators and technology consultants. We also adopted slightly more passive roles in that we were observing activities and their dynamics but not taking part in them. Secondary sources of data collection included documentation (e.g. meeting agendas, meeting minutes, and newsletters), archival records (e.g. emails and web sites), and physical artifacts (e.g. design mock-ups and scenarios). We conducted two semi-structured interviews with Kathy and one with Emily that lasted approximately an hour each. We focused on these individuals because they were the primary stakeholders of the organization and were non-volunteer members of the organization (paid staff members or in charge of the decision-making process). The interviews were tailored to each person and focused on their perception of what happened and why in relation to Spring Creek’s Web site; on how decisions and actions were influenced and made; what conflicts arose and how they were resolved; and on our particular role in the design process. The interviews were tape-recorded and subsequently transcribed. Additionally, data was collected through both face-to-face interactions and phone conversations with Kathy and Emily. Data analysis The analysis of the data collected was done using the general analytic strategy of developing a case description (Yin, 2003). We took the long term participatory design approach in working with our community groups in the Civic Nexus project (Merkel et al., 2004). Guided by our perspective on participatory design as a learning process, we took a descriptive approach to help identify the complex stages of designing a Web site and how we as researchers facilitated this process. For example, in our analysis, the email dated March 22nd 2004 that was sent to Kathy by one of the Civic Nexus researchers (see below) showed how we introduced the concept of designing with scenarios into the organizational practice to help Spring Creek move towards being autonomous entity on technology matters: I have attached an example of a scenario for a new user who happened to arrive at the Spring Creek Web site by searching on Google. Please read it and perhaps refine it further. As we discussed last time, scenarios such as the one I have attached are evocative tools to design a Web site. As you will read through it, you will soon begin to realize what kind of things you need to put up on the front page of Spring Creek’s Web site. We will discuss this further on Saturday. If you get a chance, take a shot at making some initial scenarios for decision makers and stakeholders… Data evaluation We had at least two Civic Nexus researchers attend the meetings with Spring Creek to provide independent assessments of the interpretations that were being made about Spring Creek’s Web site design process. Additionally, all Civic Nexus researchers met bi-weekly to reflect on the collected data to generate collaborative interpretations. This collaborative process of data analysis helped to remove the individual researcher’s subjective bias, thus increasing the reliability of the data analysis. Accounting for our bias in interpretation of the community context and the process of Web site design, we undertook a process of member checking with Spring Creek. We presented Spring Creek with our analysis and asked for the organization’s feedback. A research issue we encountered was anonymity. We have used the actual name of our case study organization, because we feel that knowing information about their Web site, specifically the URL and back end system, is critical to understanding the issues that arise in the course of designing a Web site. However, we have anonymized the names of Spring Creek’s key players to protect their real identities. Based on our case study with Spring Creek, we now turn to the focus of our paper, highlighting how community computing can enrich end user development (EUD). EUD Requirements in Community Computing Using our Wiki-like environment as a prototype and based on our interactions with Spring Creek, we have assimilated three design requirements of the EUD tools for supporting activities in community groups such as Spring Creek – support continued use of technology by different roles over time, provide compatibility or access to tools that are affordable, and accommodate daily tasks with support for an evocative learning situation. Eliciting and refining such requirements is part of our long-term strategy for building a palette of EUD tools that can be reused by other community groups. These requirements are not unique by themselves, but we express them in unique ways because of their context in community computing and their implications for EUD tools. Support continued use of technology by different roles A requirement in community computing is the ability of EUD tools to support continued use of technology by different people over time. In the case of Spring Creek, Kathy elicited requirements related to the content management of the web site. This included editing text, uploading images, automatically publishing newsletters, maintaining an online calendar, and so on. The key concern for community groups is to sustain the process of using technology. Because community groups often cannot financially secure a full-time position for a technician or technical consultants, their technology maintenance and development must rely heavily on technology volunteers. It is understood that community volunteers come and go in a dynamic fashion. Different volunteers have different roles and connections with the community although they alll come all are offering their time and help. EUD tools have to be designed to provide the flexibility of supporting different roles in the community. Kathy indicated an agreement with this point in one of the interviews we conducted: “What I would like to do eventually is once we get the site in a manageable point, I would like to have a volunteer or two volunteers who are willing to update the site regularly.” This poses the huge challenge of creating development tools suitable for end users who possess no programming background or interest in learning how to program (Repenning & Ioannidou, 2004). Furthermore, since there is a quick volunteering turnaround in organizations like Spring Creek, the content management system should be configurable for allocating different privileges to people contributing to the web site. This was directly referred to by Kathy: “Like there’s this one volunteer who I would like to be able to have him do certain things on the site, but I don’t trust him…if he only had control over that one thing so he’s not promoting his own agenda, that would be good…you don’t want somebody to be able to go in and screw up your site (emphasis added).” Provide compatibility to affordable tools that are accessible to the organization Another requirement is to make EUD tools accessible or compatible with technology tools that are affordable (Kathy once said: “I feel an immense obligation to make it (web site redesign process) worth the dollar value”). Spring Creek had previously relied on proprietary software from a non-profit technology assistance agency known as Techsoup (http://www.techsoup.com). Along with another volunteer (Tim), we have discussed the idea of using open-source tools such as Wiki (Guzdial et al., 2000). The tradeoff here is of course technical assistance, as open-source software tends to be less well-supported. In several meetings, Spring Creek expressed their concern to not end up in the same dilemma as with the previous vendor when he refused to furnish further technical support for the web site (see Merkel et al., 2004 for details). This resonates with an argument made by Gurstein (2003) in his “effective use” approach. “Effective use” is defined as “the capacity and opportunity to successfully integrate ICTs into the accomplishment of self or collaboratively identified goals” (Gurstein, 2003). The similar argument as ours is that the application needs to be made using the particular technical infrastructure accessible in the area where the application is to be implemented. Accommodate daily tasks with support for an evocative learning situation One of the most critical requirements in community computing, which tends to be overlooked in end user development, is to have end user tools be as much everyday technology as possible (not to be confused with ubiquitous computing). This implies that EUD tools for community groups should require little learning effort to get started. For instance, EUD tools should be used on common operating system platforms (e.g., MicroSoft Windows), should be integrated with existing tools, or should have their information/interaction design features consistent with the existing tools in the community group. Thus, for example, an environment such as the Envisionment and Discovery Collaboratory (EDC) (Arias et al., 2000), which uses electronic whiteboards, will not suit most community groups because it is an uncommon environment which demands a very steep learning curve. On the other hand, if community groups are familiar with using Mambo in managing their web site, then it makes sense to use a EUD tool that shares similar design features as Mambo. Another aspect of everyday technology is the creation of tools that end users are motivated to learn and use in daily work practices (Fischer et al., 2004). In a community computing context, community groups are already engaged in interesting and meaningful activities; hence, there is less of a requirement to use EUD tools as a vehicle for engaging people in evocative learning situations (e.g., Eden et al., 1996). Instead, the tool learning overhead for end users should be as little as possible, since these userss are already overwhelmed by various community activities and they are very often unpaid. We have observed with Spring Creek on multiple occasions that the learning of the tool is typically to be accomplished in a matter of a few minutes. For instance, in asking Kathy about the amount of time she could devote to learning EUD tools (such as our Wiki-like environment), she responded: “We (characterizing people like herself) don’t have time to learn this (emphasis added)”. Here, Kathy referred to the ease of learning Wiki in a matter of minutes rather than taking out hours of her volunteering time to learn to use a tool (e.g., through formal instruction). However here, lies the problem with such a requirement: end user developers (Spring Creek ) are trying to complete development tasks in which, by definition, they are not experts with limited learning time for development tools as opposed to users who are experts in their tasks in traditional HCI terms (Beringer, 2004). Thus, the central requirement for end user development tools is to compensate for the discrepancy between the limitations of the user’s expertise and the complexities of the development task to be performed (Beringer, 2004). In summary, for community groups like Spring Creek, sustainability of technology use is the key issue for technology adoption. That is, the process of using technology should be sustainable over different resources (human, financial, technical, and temporal). Based on this observation, we elicited the above requirements from community computing that may help us to have more of an impact in support of end user development in society. The importance of sustainability in technology adoption and use is also discussed in an “effective use” approach (Gurstein, 2003). That approach was used to show that the larger issue in the digital divide is in supporting the contextual usage of technology rather than merely providing access to technology. As Gurstein maintained, use is a situated behavior going beyond interaction between an individual and the software (Gurstein, 2003). The “effective use” approach called for policy makers and technology designers to give attention to the context in which technology is embedded. Similarly, the three design requirements we propose here are the design guidelines addressing the specific contextual aspects of the daily practices in community groups. EUD tools should be considered as open systems that evolve in the hands of the users (Arias, et al., 2000) and that empower end users to tailor their applications (Wolfe & Jarke, 2004)). Outsiders (to the community groups) that provide EUD tools should emphasize not only the tool as a product but also the process of using the tool. Enhancing EUD with Conceptual Scaffolds Community computing can re-conceptualize what we mean by EUD tools. EUD tools are predominantly perceived as tools that allow end users to write software programs. Wikipedia defines the EUD process: “people who are not professional developers can use EUD tools to create or modify software artifacts (descriptions of automated behavior) and complex data objects without significant knowledge of a programming language”. In community computing, this is not always the case. In other words, end user development is not only about “programming” or using technical tools, but also utilizing conceptual scaffolds that facilitate the realization of their organizational goals in terms of technology. By conceptual scaffolds, we mean the use of techniques to help community groups to make technology-related decisions. We illustrate the use of conceptual scaffolds as consultation with human resources, index cards, and design scenarios. These are only examples, which are obviously not exhaustive, that epitomize our view of adding conceptual scaffolds to EUD tools. Using a Wiki-like environment was merely one of the scaffolds (technical) that we provided to Spring Creek. In fact, more conceptual scaffolds were provided and used that eventually empowered Spring Creek to establish a strong foothold over their web site redesign process. For instance, we helped Spring Creek clarify various nomenclature related to the redesign process. One of the more complex issues that groups face when trying to implement a technology project is making sense of the design process. As Collins, Brown, and Newman note (Collins et al., 1989), part of the work that novices (end users) are doing is developing conceptual models that are needed to take on a task. In our face-to-face meetings, Spring Creek members had different perspectives on “design” that created a tension between technical requirements and the need to organize information on the web site effectively. One of the volunteers, Dan, was technically proficient in web site design and wanted to move directly into the interface design of the site. At this point, we intervened, suggesting that design is an iterative process. Further, we emphasized that the content of the web site needs to be designed before the layout. Kathy agreed but did not know how to go about it. She once remarked, “Do we mean the same thing by the term ‘design’”, reiterating her struggle with the difference between designing content and designing layout. We helped clarify these notions by using the old, vendor-developed Spring Creek web site as an example of “good” layout and “bad” content design. Our interventions with conceptual scaffolds are indicated in Figure 1. Once the new content for the web site was relatively stable, Tim suggested talking about “arranging the web site directory structure”. He advised Kathy to use index cards (see Figure 1) to create a hierarchy of the navigation scheme for the web site. Even mundane conceptual scaffolds, such as index cards, can act as cognitive artifacts (Norman, 1991) that eventually concretize the use of technology for realizing concepts tangibly in the real-world. In the current web site prototype, most of the navigation scheme resulting from Kathy’s use of index cards has been incorporated. In our effort to weave design into Spring Creek’s organizational practice, we introduced the concept of designing with scenarios (see Figure 1) (Carroll, 2000) to elucidate how the web site front page should be laid out. Kathy specifically used scenarios not only to illustrate different personae that would interact with the web site but also to convince other key players in the meeting on how the web site layout should be designed (XY, 2005). There was also an instance where Kathy used pen and paper to explain how the main content links on the navigation bar should be preserved but that clicking on one of those links should change content in the content area (separate from the navigation bar). To us, even this simple and tangible illustration represented Kathy’s inclination toward an interface exhibiting overview + detailed interface properties (e.g., Plaisant et al., 1995). Schön (1983) drew an important contrast between merely creating or identifying elements of the problem context and making them reasonable. He characterized design as inquiry—a property of the various conceptual scaffolds we have described in our work with Spring Creek. In community computing, technical tools only comprise one facet of what end users can employ in their daily activities. We have illustrated that design is manifested even in ordinary artifacts that entail technology affordances. End users in community computing are active technology users and designers. Designers are not just making things; they are making sense (Carroll, 2000, p. 66). To this end, the community computing context can enrich what we, as researchers and developers, believe EUD tools to be as not simply technical gadgets but rather as conceptual aids that entail the use of technology. Adopting and Evaluating Tools in Community Computing From their unsatisfactory experience with the previous vendor, Spring Creek has realized that technology is not an unmixed blessing (Postman, 1993). During the process of appropriating technology (Dourish, 2003) and revamping their web site, we have observed that Spring Creek adopts and evaluates EUD tools in ways that go beyond current notions of usability (e.g., Carroll, 2004), usefulness and understanding (e.g., Woods, 1998), and even group dynamics related to group performance and productivity (e.g., Grudin, 2004). The critical incident described in the Research Methodology section gave a good example in this context showing that the concept of usability alone is insufficient in understanding how people evaluate and adopt EUD tools in the community computing domain. In this incident, both EUD tools were similar in what they provided technically from the usability point of view. But Kathy chose Ned’s web design over the wiki web she co-designed with us and her web design team. Kathy did not seem to be concerned about repeating the history with the previous vendor because she perceived that Ned was providing relatively “better” support than the original vendor (who provided limited assistance). Prior to Ned’s unveiling of the content management system, Kathy had compared him to the previous vendor (May, 2004): “Ned has been extremely professional from a student standpoint…the difference between him and that company (vendor)…there’s no comparison (emphasis added)”. Through our work with Spring Creek, we have learnt that evaluating EUD tools in community computing goes beyond the tools themselves—they may transcend into the evaluation of providers of the tools as we have observed in the case of Spring Creek. Social influence around community groups plays an important role in their evaluative criteria of EUD tools. We have also learned that adopting and evaluating EUD tools is not merely a matter of use and non-use, but the use of technology viewed as a continuum (Malhotra & Galletta, 1999). This continuum defines the range from avoidance of use (abandoning the vendor-supported technology) to enthusiastic and consistent use (using our Wiki-like environment). Adopting and evaluating EUD tools in the community computing context, perhaps, eventually converges to the perceived fit of technology use to the users’ values (Malhotra & Galletta, 1999). Discussion Nonprofit community groups often lack resources such as money and technology infrastructure, as well as the organizational structures necessary for coping with rapid technology change in information systems (Benston, 1990; Mcphail et al., 1998; Trigg, 2000). Considering the context the community groups are embedded in, we had expected Spring Creek would quickly embrace and adopt EUD tools such as wiki technology because these support end users to develop the needed technology results and to become autonomous relative to technology need. However, as discussed in the critical incident, Spring Creek decided to go with a volunteer’s web design giving up their control on web design and content management. This crucial incident posited an interesting yet challenging point: end users in community groups could voluntarily come to rely on external technology experts even while acknowledging the potential problem of having the experts’ help only intermittently and even when they have the option of adopting EUD tools to empower themselves and be more independent in their technology projects. This indicates a tension between appropriation of ICT for local autonomy (e.g., through careful and considerate design of EUD tools) and the ways external expertise could work against such appropriation in the community computing context. On the one hand, our paper has been arguing for pushing technology development to consider the community computing context for sustainable technology adoption and use by considering ways of empowering community users to become autonomous in their technology projects and decisions. And as well, EUD researchers have increasingly paid attention to how EUD tools can facilitate end-user’s learning processes in how to use EUD tools (e.g., Dorn & Guzdial, 2009). But on the other hand, non-profit community groups have been living in the volunteer culture for a long time, and have developed their practices so as to carry out their daily operations while relying on external help. Nonprofit community groups make strategic plans for employment of technology resources, including building and maintaining collaborative relationships with external IT experts (Hackler and Saxton, 2007). Situated in this cultural context, external resources may be introduced to and leveraged by the community groups through various channels, which create temptations that attract the community groups towards giving up on the plan to invest time and effort in equiping and train themselves with EUD tools and to be more independent. In other words, pushing the technology for sustainable use and learning is not enough to support end user development in community computing. This is essentially what Gurstein alluded to in his paper. That is, community groups need to actively participate in the process of technology acquisition and implementation, and it is sometimes important to create or stimulate the “pull” from local communities (Gurstein, 2003). Our case study indicates that there are different levels of “pull” force from local communities. In the Spring Creek case, the bad experience with the prior vendor triggered the community to become more active in identifying the right web site product thus creating a “pull” action in the community. At this level, the group was committed to actively looking for a technology solution. The next level of “pull” action was the community group’s commitment to learn the technologies to support its autonomy in technology projects and decisions. This level of “pull” action requires the community groups to realize that technology literacy is important for community development. In the Spring Creek case, the choice between a wiki web site and a regular web site was essentially about whether to take full control of the technology or to continue relying on external expertise for significant managerial and administrative tasks. We did not see a “pull” from Spring Creek indicating its commitment to technology learning and Kathy did not eventually choose the wiki web design. We don’t know what made Kathy decide to choose Ned’s web design. But considering the sustainable use as a key decision factor, we think the different roles that Ned and ourselves took in relation to the organization may have had an impact. After all, we were researchers in the project with our own research agenda and we would fade away and not be able to provide continuous help on web site issues. Ned, on the other hand, was a technology intern with Spring Creek, and had a local consulting company in the area of web design. For Ned’s design, it would therefore be more likely possible to obtain continuous support in the future. Interestingly, when we checked out Spring Creek’s web site as of Feb. 2011, the organization had totally changed its web design again with a different content management system - Joomla! In today’s world, IT is vital to all aspects of nonprofit groups’ functions from member recruiting and management; to communication and visibility to the community; to realizing their organizational mission. We believe that it is more beneficial for the organization to become more autonomous and technology savvy itself in the long term. It is important for community computing researchers to acknowledge this tension and seek ways to push community groups towards learning and adopting technologies through their own capacity. It is also important for EUD researchers and developers to realize and address this tension in supporting end user development in community computing. Conclusion The end users in community computing develop tools and applications within the world of their organizational goals. In this paper, we have alluded to how community computing can engage end user development to have more of a socio-technical impact on society. End users in community computing are not natural Luddites (Snow, 1998). They are active end users with different characteristics and requirements that dictate the use of technology. The paradox of the active user (Carroll & Rosson, 1987) perhaps applies more to community computing than others due to their lack of time and other resources. Therefore, we can learn lessons from community computing to apply to design for end users and thus better understand how they actually react to and learn technology. In this paper, we have only begun to traverse toward this goal. We believe community computing presents a promising and cultivating context for scientifically enriching end user development, as we have partly demonstrated in this paper, and we hope to inspire more efforts in this direction within the research community. Our paper contributes the following notions that have not been previously addressed in depth by the researchers in the domain of communities and technology: 1.Community computing provides a foundation for eliciting design requirements for developing EUD tools. 2.Community computing expands the notion of EUD from mere technically oriented EUD tools toward more conceptual end user scaffolds. Our paper also identifies a tension between supporting end user development and relying external technical expertise in community computing. Adopting the terminology used in Gurstein’s paper (Gurstein, 2003), we argue for creation and stimulation of “pull” from community groups on active participation of technology projects as well as commitment to technology learning for community development. References Arias, E., Eden, H., Fischer, G., Gorman, A., and Scharff, E. Transcending the Individual Human Mind-Creating Shared Understanding through Collaborative Design. ACM Transactions on Computer-Human Interaction, 7, 1 (March, 2000), 84-113. Benston, M. Participatory design by non-profit groups. Proceedings of the Participatory Design Conference, 1990, 107-113. Beringer, J. Reducing Expertise Tension. Communications of the ACM, 47, 9 (September, 2004), 39-40. Brazleton, J. and Gorry, G.A. Creating a Knowledge-Sharing Community: If You Build It, Will They Come? Communications of the ACM, 46, 2 (February, 2003), 23-25. Burnett, M., Chekka, S.K., and Pandey, R. FAR: An end-user language to support cottage e-services. Proceedings of the IEEE Symposium on Human-Centric Computing, 2001, 288-295. Carroll, J.M. Making use: scenario-based design of human-computer interactions. The MIT Press, 2000. Carroll, J.M. Community Computing as human-computer interaction. Behaviour and Information Technology, 20, 5 (2001), 307-314. Carroll, J.M. The participant-observer in community-based learning as community bard. Journal of Community Informatics, 2-2, 2006, 125-127. Carroll, J.M. Beyond fun. Interactions, 11, 5 (September-October, 2004), 38-40. Carroll, J.M. and Rosson, M.B. The paradox of the active user. In J.M. Carroll (Ed.), Interfacing Thought: Cognitive Aspects of Human-Computer Interaction. Cambridge, MA: MIT Press, 1987. Carroll, J.M. & Rosson, M.B. 2007. Participatory design in community informatics. Design Studies, 28, Special Issue on Participatory Design, 243-261 Collins, A., Brown, J.S., and Newman, S.E. Cognitive apprenticeship: Teaching the craft of reading, writing, and mathematics. In L.B. Resnick (Ed.), Knowing, learning and instruction: Essays in honor of Robert Glaser, Lawrence Erlbaum Associates, Hillsdale, NJ, 1989. Cooper, A. The Inmates Are Running the Asylum: Why High Tech Products Drive Us Crazy and How To Restore the Sanity. Sams, 1999. Davis, F.D. A Technology Acceptance Model for Empirically Testing New End-User Information Systems: Theory and Results. In MIT Sloan School of Management, Cambridge, MA, 1986. Davis, F.D. Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology. MIS Quarterly, 13 (1989), 319-340. Davis, F.D., Bagozzi, R.P., and Warshaw, P.R. User Acceptance of Computer Technology: A Comparison of Two Theoretical Models. Management Science, 5 (1989), 982-1003. Dorn, B., Guzdial, M. Learning on the Web: A Case Study of Graphic Design End-User Programmers, CHI’09 workshop: End User Programming for the Web Dourish, P. The Appropriation of Interactive Technologies: some Lessons from Placeless Documents. Computer Supported Cooperative Work, 12 (2003), 465-490. Eden, H., Eisenberg, M., Fischer, G., and Repenning, A. Making Learning a Part of Life, Communications of the ACM, 39, 4 (April, 1996), 40-42. Ehn, P. and Kyng, M. Cardboard computers: Mocking-it-up or Hands-on the Future. In J. Greenbaum and M. Kyng (Eds.), Design at work: Cooperative design of computer systems, Hillsdale, NJ: Erlbaum, 1991. XY., 2005 (Blinded for double review) Fischer, G., Giaccardi, E., Ye, Y., Sutcliffe, A.G., and Mehandjiev, N. Meta-Design: A Manifesto for End-User Development. Communications of the ACM, 47, 9 (September, 2004), 33-37. Fischer, G. and Lemke, A.C. Construction kits and design environments: Steps toward human problem-domain communication. Human Computer Interaction, 3 (1988), 179-222. Fishbein, M. and Ajzen, I. Belief, Attitude, Intention and Behavior: An introduction to Theory and Research. Reading, MA: Addison-Wesley, 1975. Ganoe, C.H., Somervell, J.P., Neale, D.C., Isenhour, P.L., Carroll, J.M., Rosson, M.B., and McCrickard, D.S. Classroom BRIDGE: using collaborative public and desktop timelines to support activity awareness. Proceedings of the ACM Conference on User Interfaces and Software Technology, 2003, 21-30. Grudin, J. Return on Investment and Organizational Adoption. Proceedings of the ACM Conference on Computer Supported Cooperative Work, In Press (2004). Gurstein, M. (2003). Effective use: A community informatics strategy beyond the digital divide. First Monday: Peer-Reviewed Journal on the Internet, 8(12). Guzdial, M., Rick, J., and Kerimbaev, B. Recognizing and Supporting Roles in CSCW. Proceedings of the ACM Conference on Computer Supported Cooperative Work, 2000, 261-268. Hackler, D. and Saxton, G. (2007).The Strategic Use of Information Technology by Nonprofit Organizations: Increasing Capacity and Untapped Potential, Public Administration Review, Vol. 67 No. 3, pp474-487. Kase, S., Zhang, Y., Carroll, J.M. & Rosson, M.B. 2008. Sustainable Informal IT Learning in Community-based Nonprofits. Proceedings of ACM Conference on Human Factors in Computing Systems, CHI 2008 Extended Abstracts, (April 5-10, Florence, Italy) New York: ACM Press, pages 3435-3440 Kim, A.J. Community Building on the Web: Secret Strategies for Successful Online Communities. Peachpit Press, 2000. Malhotra, Y. and Galletta, D.F. Extending the Technology Acceptance Model to Account for Social Influence: Theoretical Bases and Empirical Validation. Proceedings of 32nd Annual Hawaii International Conference on System Sciences, 1999, 6-19. Mathieson, K., Peacock, E., and Chin, W.W. Extending the Technology Acceptance Model: the Influence of Perceived User Resources. The Database for Advances in Information Systems, 32, 3 (Summer, 2001), 86-112. Mcphail, B., Costantino, T., Bruckmann, D., Barclay, R., and Clement, A. CAVEAT Exemplar: Participatory Design in a Non-Profit Volunteer Organisation. CSCW, 7, 3-4 (1998), 223-243. Merkel, C.B., Clitherow, M., Carroll, J.M., and Rosson, M.B. Sustaining Computer Use and Learning in Community Computing Contexts: Making Technology Part of “Who They Are and What They Do”. The Journal of Community Informatics, (2005) Vol. 1, Issue 2, pp. 158-174. Merkel, C.B., Xiao, L., Farooq, U., Ganoe, C.H., Lee, R., Carroll, J.M., and Rosson, M.B. Participatory Design in Community Computing Contexts: Tales from the Field. Proceedings of the Participatory Design Conference, 2004, 1-10. Morgensen, P. B. and Shapiro, D. When survival is an issue: PD in support of landscape architecture. Proceedings of the ACM Conference Computer Supported Cooperative Work, 1998, 187–203. Nardi, B.A. A Small Matter of Programming: Perspectives on End User Computing. Cambridge: MA: MIT Press, 1993. Norman, D. A. Cognitive artifacts. In J. M. Carroll (Ed.), Designing Interaction: Psychology at the human-computer interface. New York: Cambridge University Press, 1991. O’Reilly, C., III and Chatman, J. Organizational Commitment and Psychological Attachment: The Effects of Compliance, Identification, and Internalization on Prosocial Behavior. Journal of Applied Psychology, 71 (1986), 492-499. Papert, S. The Children’s Machine. New York: Basic Books, 1993. Petroski, H. to Engineer is Human: The Role of Failure in Successful Design. Vintage Books, USA, 1992. Plaisant, C., Carr, D., and Shneiderman, D. Image-browser taxonomy and guidelines for designers. IEEE Software, 12, 2 (March, 1995), 21-32. Postman, N. Technopoly: The Surrender of Culture to Technology. New York, Vintage Books, 1993. Repenning, A. and Ioannidou, A. Agent-Based End User Development. Communications of the ACM, 47, 9 (September, 2004), 43-46. Rosson, M.B., Ballin, J., and Nash, H. Everyday Programming: Challenges and Opportunities for Informal Web Development. Proceedings of the Conference on Human Centric Computing, In Press (2004). Rosson, M.B. and Carroll, J.M. Usability Engineering: Scenario-Based Development of Human-Computer Interaction. Morgan Kaufmann Publishers, 2002. Schon, D.A. The reflective practitioner: How professionals think in action. New York: Basic Books, 1983. Schultze, U. Reflexive Ethnography in Information Systems Research. In E.M. Trauth (Ed.), Qualitative Research in IS: Issues and Trends, Hershey: Idea Group Publishing, 2001, 78-103. Snow, C.P. The Two Cultures. Cambridge University Press, 1998. Sutcliffe, A.G. and Mehandjiev, N. End-User Development. Communications of the ACM, 47, 9 (September, 2004), 31-32. Trauth, E.M. (Ed.). Qualitative Research in IS: Issues and Trends. Hershey, PA: Idea Group Publishing, 2001. Trigg, R. H. From sandbox to "fundbox": Weaving participatory design into the fabric of a busy nonprofit. Proceedings of the Participatory Design Conference, 2000, 174-183. Venkatesh. V. and Davis, F.D. A Theoretical Extension of the Technology Acceptance Model: Four Longitudinal Field Studies, Management Science, 46, 2 (February, 2000), 186-204. Venkatesh, V., Morris, M.G., Davis, G.B., and Davis, F.D. User Acceptance of Information Technology: Toward a Unified View. MIS Quarterly, 27, 3 (September, 2003), 425-478. Wolfe, V. and Jarke, M. The Economics of End-User Development. Communications of the ACM, 47, 9, (September, 2004), 41-42. Woods, D.D. Designs are hypotheses about how artifacts shape cognition and collaboration. Ergonomics, 41, 2 (1998), 168-173. XX., 2005 (Blinded for double review)