Collaborative design of online professional development: building the Milwaukee Professional Support Portal.

This case study describes an effort to design and pilot an interactive, Web-based, district-wide "professional support portal" to support new teacher retention and professional growth in the Milwaukee Public Schools. The study reviews the overarching goals of this program and focuses, in particular, on an account of the nature and kinds of collaboration among the district and its institutional partners that made possible the design and early use of this Portal. Findings identify key factors that proved either to facilitate or impede distant collaboration. These principles are essential to consider when planning and implementing district-wide initiatives for in-service professional development that embrace new technologies and include multiple partners.

**********

State and local policies designed to address the shortage of qualified teachers are increasingly mandating mentoring, coaching, and peer support to increase the induction and retention of new teachers and to renew the skills and motivation of veteran teachers (National Commission on Teaching and America's Future, 2003). Such a complicated and difficult task strains the resources and human capital of many schools, particularly urban districts, which tend to have large proportions of new and under-qualified teachers and which confront complex issues of diversity and equity. Findings from a variety of studies indicate that traditional methods of professional development are inadequate to meet these challenges (Choy, Chen, & Ross, 1998; National Commission on Teaching and America's Future, 1996).

New communication and information technologies appear to offer great potential to address this issue through flexible approaches that support professional development in the midst of work (Smith, 2003). Milwaukee Public Schools (MPS) has established a strategic initiative, the Professional Support Portal (PSP), designed to use technology in innovative ways to support new teacher induction, retention, and continued professional growth through online tools, resources, and expertise. The goal of this systemic project is to engage new teachers in mentoring and peer coaching activities online, while providing support and guidance through the district's accomplished teachers and university partners. This model consists of a sophisticated combination of face-to-face interaction complemented by online tools and services that address issues of access, diversity, equity, motivation, capacity, and quality. The Harvard Graduate School of Education, along with the Educational Development Center, worked across distance as codesigners, mentors, researchers, and evaluators for the project. Other initial partners were SRI International, Homeboyz Interactive, Teachscape Inc., and the University of Wisconsin-Milwaukee.

This particular case study explores the use of high bandwidth, interactive media (internet-based videoconferencing and groupware) to enable Boston-area experts to serve as active, day-to-day partners and mentors on this MPS initiative for new teacher induction and retention. In particular, the study portrays how the partner organizations built and sustained a virtual "Design Team." From September 2001 to June 2003, the group focused on collaborative design across distance through three distinct phases: (a) an exploratory stage of defining needs for new teacher induction and retention, (b) a preliminary design phase that defined the basic "look and feel" of the portal, and (c) a pilot demonstration phase that saw the first use of the portal during the 2002-2003 school year with a group of 150 new teachers drawn from across the district.

The primary tool for design group communication was Internet-based videoconferencing, which provided an inexpensive and reliable means of communication. This was augmented by several other methods of synchronous and asynchronous interaction and information exchange among group members, ranging from personal e-mail distribution lists to multifunctional groupware applications, such as Groove[R] (1) and TappedIn[R] (2). The team initially sought a dominant groupware interface that might not only serve as a medium of communication about the design process--talking about the process or outcomes of design--but also provide the means for enacting design--carrying out discreet design tasks collaboratively across distance. Based on previous studies described next, those on the team hypothesized that, by adopting new kinds of shared information and communication tools for collaboration, the use of a common interface would open the way for new processes of design across institutional and geographical barriers. Unfortunately, this Golden Grail of groupware--a technology solution that would seamlessly empower the technical, social, and task-oriented dimensions of the design process--proved elusive.

Over the project's initial 18 months, the design team's means of collaboration shifted from identifying a dominant, multifunctional application to living within the often messy, necessarily redundant, but nonetheless invaluable world of "instrumentalities" or orchestrations of information and communication technologies (Engestrom, Puonti, & Seppanen, 2001). Each application in the mix of interactive media was used and supported by one of the partnering organizations, but had to be gradually appropriated (DeSanctis & Poole, 1994) and brought into the work flow of the other partners--individuals as well as institutions. Thus, implementing a suite of tools unique to the partnership, common among all partners, and tailored to the task at hand was a complex task.

The first section describes MPS and how the Professional Support Portal came to be seen as a solution for the district's induction and retention of new teachers. The next section reviews prior studies relevant to technology-mediated collaborative design across distance. Following these background sections is an account of the three phases of the collaborative design process, with a particular focus on the evolution of the Design Team and the tools used to bridge distance. The case concludes with a set of principles for inter-institutional collaborative design across distance, based on the experience designing and developing the PSP.

THE MILWAUKEE PUBLIC SCHOOLS AND NEW TEACHERS

Milwaukee Public Schools (MPS) is the 27th largest school district in the country, educating over 100,000 students in 200 locations. The most recent enrollment data from 2003-2004 show MPS demographic statistics reflect the following ethnic diversity: 59.8% of the students are African-American, 15.5% White, 17.1% Hispanic, 4.3% Asian, 0.9% Native American, and 2.4% other nonwhite. Fifteen percent (15%) of the total student population has identified disabilities, and approximately 7.5% have limited English proficiency. Over 77% of the students receive free or reduced lunch.

MPS has made great strides in recruiting new teachers to an urban setting. However, over 10% of MPS' teaching force resigns annually, with many leaving for reasons other than retirement. With the baby boom generation nearing retirement age, the district anticipates that this attrition rate will increase. For example, for 2001-2002 over 25% of all teachers in MPS are age 45 years and older. The yearly departure of a substantial number of teachers directly impacts student academic success in the school district.

The number and percentage of new teachers hired has risen significantly over the years. Ten years ago (1993-1994), the Milwaukee Public Schools had to hire 173 new teachers to replace those who had retired or left for other reasons, accounting for about 3% of all teachers in the district. In 2001-2002 a total of nearly 1,000 teachers were hired to replace those who retired or left for personal reasons, as well as to fill newly created teacher slots. This represents about 15% of all 6,900 teachers in the school district.

MPS also has a large turnover of new teachers. Novice teachers consistently struggle with issues of group instruction, classroom management, technology integration, and logistical support that undercut their effectiveness and the educational outcomes of their students. As of the 2003-2004 school year about 37% of the new teachers hired by MPS leave the district within the first five years. New teacher retention is a problem shared by many schools--especially in urban districts--that dramatically undercuts students' academic performance (National Commission on Teaching and America's Future, 1996). Since students' success is largely dependent on the effectiveness of their teachers, sustainable strategies to support the professional development and retention of new teachers are essential for high academic performance.

MPS established a strategic initiative, the Professional Support Portal, to use technology in innovative ways to support teacher induction, retention, and continued professional growth at all stages of the teaching career continuum. Through the use of online tools, resources, and expertise, the Portal's goal was to help improve the new teacher retention rate by addressing three categories of needs. These needs were determined in consultation with an advisory group of district stakeholders, as well as through several rounds of focus group interviews with new and experienced teachers:

1. Access to classroom resources, lesson planning tools, and teaching and classroom management strategies;

2. Time and strategies to work with mentors and to observe classrooms led by experienced teachers; and

3. Ongoing social, emotional, and professional support from peers and experienced teachers.

These themes, frequently cited in the teacher retention literature, were shared with key leaders in the district, who endorsed seeking resources to build a MPS Professional Support Portal. Its purpose was to build the social and technical infrastructures needed for new teachers to easily find and interact with resources, experts, and peers, both face-to-face and mediated by technology.

The portal project has allowed for a convergence of several key technology initiatives underway within and outside of the district:

* The Curriculum Development Assistant (CDA), an online tool developed over several years by MPS, creates a collaborative environment in which teachers can post or find lessons that support their day-to-day work in classrooms. This tool is web-based and always accessible

* Teachscape[R] (3), a commercial professional development process based on video case studies, provides examples of standards-based lessons being taught in urban classrooms at the elementary level.

* TappedIn[R] (4), a nonprofit multi-user virtual environment for professional development developed by SRI, provides an online social context that allows educators to build and sustain communities of practice.

A partial model of the portal project was built in 2003, with a full launch set for the late spring of 2004. (5)

DESIGN PARTNERSHIPS

A number of different experts from Harvard's Graduate School of Education and the Educational Development Center (EDC) assisted MPS staff with the design and implementation of the portal. Using broadband telecommunications, Internet-based videoconferencing, and collaboration tools, Harvard's Graduate School of Education and EDC provided:

* online courses customized by EDC for MPS staff in a cross-section of leadership roles across the district. Over the past two years, these have focused on the use of new interactive media in professional development and on data-based decision making;

* guidance through videoconferencing from Harvard faculty for a cohort of "rising star" MPS principals, who completed the Harvard Principal's Institute in Cambridge, MA, in the summers of 2002 and 2003 and who continue to receive feedback on their attempts to implement leading-edge practices in local settings;

* shared development of a pilot graphical multi-user virtual environment to complement the intellectual and social interaction provided in TappedIn[c]; and

* guidance and facilitation through video-conferencing, teleconferencing, and e-mail from a Harvard evaluator for the design and implementation of feedback mechanisms for teachers and advisors involved in the project.

The remainder of the article discusses an additional component of this partnership: the groupware- and videoconference-based collaborative design of the PSP with interface and knowledge management specialists at Harvard.

COLLABORATIVE DESIGN ACROSS DISTANCE

The field of collaborative design across distance is a core interest of the area of study known as computer-supported cooperative work (CSCW). Arenas of work relying predominantly on visualization and/or on forms of practice that are predominantly computer-based (such as engineering, architecture, and software design) tend to dominate current studies of distributed design. Studies of cross-institutional efforts outside of these domains, such as collaborative website design (Hill, Tharp, & Sindt, 1998), are not common. A growing number of studies in the closely-related field of computer-supported collaborative learning (CSCL) broach collaborative course design across distance. These tend to focus on creating online environments for higher education students in domains whose work products and processes rely heavily on digital technologies, such as engineering and architecture (Hill et al.).

A recent study that overlaps both CSCW and CSCL and explores the use of groupware in distant mentoring is that of Hettinga's (2002) analysis of a telemedicine application in the Netherlands. Hettinga's study does not address the issue of collaborative design directly, but it does involve consultations by way of dedicated (ISDN) videoconferencing by university-based specialists with practitioners at remote hospitals. Hettinga and her colleagues examined the changes brought about by distant collaboration along three dimensions: social, task, and technical. Social change encompassed aspects such as the nature of relationships among participants, as well as shifts in participants' attitudes towards technology. Changes in task or work included the "nature and texture of work, patterns of interaction, distribution of work, nature of knowledge, and mechanisms of coordination" (Hettinga, 2002). Finally, technical changes pointed towards shifts in personal and institutional configurations of technology and, at a more fundamental level, adaptations in the design of particular technologies.

This study explores virtual team development over time, as technical shifts--the adoption of collaborative tools--came to leverage shifts in task (the way work was accomplished). Earlier studies of shifts in the ways small teams go about their work point to two well-known paths of development. The first of these is an evolutionary path in which change occurs gradually and incrementally over time, roughly corresponding to a passage from an early focus on social dimensions to a later focus on task. The second pathway is an episodic or "punctuated equilibrium," one in which change is abrupt at certain times and nonexistent in between these "leaps."

With respect to the portal project, the authors believed initially that an episodic shift in the technical aspect would lead to evolutionary shifts in task--that is, in the ways in which all partners approached design. Following Bardram (1998), the anticipated shift in tasks conducted across distance would evolve from communication--activities geared towards coordination and co-operation around the work to be done--towards coconstruction of the object of work. Previous research, based on the "media richness" ideas of Heeren and Lewis (1997), indicated that videoconferencing alone was unlikely to support such a shift. Kydd and Ferry (1994) emphasized that videoconferencing is functional for communication focused on reducing uncertainty around tasks (for example, clarifications and the exchange of information), but not useful for the kind of rich discussion and social interaction required for coconstruction and knowledge building.

This framed the project's initial hypothesis that a more robust computer-mediated environment was needed to move towards the actual creation or revision of work products across distance (such as collaborative mark-up and redesign of the website). Those working on the project sought a single but multifaceted groupware application that could be used by all partners on their desktops. Moreover, project partners assumed that--after this initial technological "leap of faith" (which might require some effort to master and sustain on the part of individual members of those involved with the design)--the technical dimension would remain relatively stable. It was assumed that such stability would bring gradual and incremental shifts in the social- and task-related aspects of communication and activity across distance. These interactions with and through the groupware would then be supplemented with videoconferencing, both for ad hoc check-ins between individual members of those participating in the design of the portal and for formal presentations about the work of this distributed team to stakeholders (teachers, administrators, advisors) within MPS. Finally, the authors believed occasional face-to-face visits would be essential, especially at those '"leaps" (following a punctuated equilibrium view) within the work cycle that required focused, "hands-on" coconstruction of some aspect of the work.

COLLECTING INFORMATION

To test these assumptions, researchers reviewed video and audio recordings as well as meeting notes from the regularly-scheduled meetings of what came to be called the PSP "Design Team," which was made up of representatives from each of the partnering organizations described. Harvard project team members also collected data (notes and video recordings) on other group meetings, both distant and face-to-face, in which members of the Design Team were involved. Individual interviews with some Design Team members were conducted both at the beginning of the collaboration and at one other point over the course of the collaboration. The results presented here derive from analysis of this data.

The Elements of a Virtual "Design Team"

The collaborative design of the PSP has moved through three distinct phases, depicted in Table 1.

Over the first 18 months of the project, the institutional partners remained constant. However, the individuals taking part in the activities of the team changed somewhat, with two new participants joining, two original participants stepping back, and "visitors" taking part for short periods of time to work on or discuss discreet tasks (one of whom is a former regular participant).

At no point over the 18 months covered here did representatives from all four institutions meet together face-to-face. Also, labeling the members of this distributed collaboration a "Design Team" did not happen until the second phase, when closer coordination became essential. Only under this pressure did the most immediately useful collaborative tools became apparent. The following describes in detail the team partners, the scope of their activities on the team, and their means of distant interaction with other members of the team.

MPS

Participants and scope. Within the MPS Department of Technology, the School Technology Support unit is the division charged with overseeing the design and development of the PSP. The director of this division and the PSP Project Manager, whom she supervised, worked closely together on the project; and both participated regularly and actively in virtual meetings of the Design Team. If neither could attend, the meetings were cancelled.

Working under the direction of the third author, MPS Director of Technology at the time, the Director of the School Technology Support Unit and the PSP Project Manager took primary responsibility for all day-to-day aspects of the project, from organizing teacher focus groups and a district-wide advisory group at the start of the project to managing the web development and production activities of Homeboyz to "evangelizing" for wider involvement in the PSP on the part of other MPS departments.

Collaboration tools. MPS staff were eager to experiment with online collaboration tools; however, no dedicated desktop groupware applications was in use at the time. Room-based and "roll-about" interactive video units were widely used throughout the district for a wide range of meetings and presentations. The two MPS staff involved with the Design Team used a room-based system about 100 yards from their offices. The room was arrayed for technology training, much as a traditional classroom, with a large screen and lectern at the front and 4-5 rows of tables, each seating about 8, facing the front of the room. Network connections for laptop use were available at each of the seats. Use of the videoconferencing equipment had to be scheduled with a technician in advance, but, once scheduled, could be operated by the users. The front of the room also had a conference phone for teleconferencing. When regular design team meetings were established in the second and third phases, these interactions entailed simultaneous use of teleconferencing, videoconferencing, and network connections for synchronous chat, e-mail, and web browsing.

Homeboyz

Participants and scope. Homeboyz Interactive is a not-for-profit web design and production firm that had already worked with MPS on the development of the Curriculum Development Assistant (described earlier) and that held the sole contract for a similar design and development effort with the PSP. Homeboyz designated a single project manager to oversee the efforts of an interface designer, graphic designers, and programmers on their staff. The project manager represented this Homeboyz team in Design Team meetings. The initial manager who worked with MPS at the beginning of the project was pulled away towards the end of the first phase. Another project manager came on board for the second phase; he was then succeeded by the original manager just before the third phase began.

Collaborative tools. Although Homeboyz had a sophisticated intranet for their web development efforts, the organization did not make regular use of either desktop-based groupware or interactive videoconferencing. For most of the meetings of the Design Team, the project manager would come to the videoconferencing room described above. When he could not be physically present, he would call the teleconference phone in the room.

Harvard Graduate School of Education

Participants and scope. In the first phase, the team from Harvard's Graduate School of Education consisted of three masters' and one doctoral student under the guidance of the second author, a professor at Harvard University. At the time, no desktop groupware applications or interactive videoconferencing facilities were in regular use. The masters' students, however, had studied and used a number of different groupware applications as part of their coursework; and the doctoral student, a coauthor of this article, had used and studied a variety of groupware in work settings.

At the beginning of the second phase, the masters' students completed their degrees and left Harvard. A second doctoral student who was experienced in interface design joined the team specifically to work with Homeboyz and MPS on moving the design of the PSP rapidly forward. Once the parameters of the design were set and after pilot testing began in the third phase, the interface expert took a less active role, although he remained "on call" and continued to offer advice on design and other issues related to the development of the portal.

At the beginning of the third phase, an expert in formative evaluation, a faculty member of the school, began taking part in meetings of the team and continues to work with MPS on deployment of the portal and establishing metrics for its use.

Collaborative tools. The recent inauguration of a Learning Technologies Center at the school of education made available several new videoconferencing facilities--both room-based and roll-about units. Despite the availability of these, the ISDN connections on which they depended were very expensive; and it was sometimes difficult to find technical support for their use. As part of this project, the professor at Harvard introduced two IP-based videoconferencing units at two different sites at the school. These were small units that could either be used as desktop devices or as part of a roll-about unit. Moreover, IP-based videoconferencing did not rely on dedicated telephone lines, but instead on a robust connection to the internet, which was relatively easy and inexpensive to obtain in a university environment. These did require special network configurations for optimal use, and one student spent time in the first phase working out connection difficulties with Harvard technicians. The unit that was up and running first was in a new classroom-sized space known as the "Experimental Lab," which was meant for media productions and offered ample network connections. It was essentially an empty room, as it had no equipment and was at that time little-used. As a result, the IP videoconferencing unit was largely available and ready for use with little or no advance notice. When things did go wrong, however, none of the staff technicians could support its use, nor could they train new team members in its use.

This all changed by the third phase. The school hired a manager to oversee the videoconferencing facilities, and he took on the support of the IP-based units. In doing so, the use of the unit in the Experimental Lab became more restricted because of new scheduling requirements put in place by the manager. On the other hand, he played an important liaison role with his counterpart at MPS, ensuring that the equipment was functioning and the connection stable before scheduled meetings.

SRI

Participants and scope. SRI began taking an active role in the Design Team towards the end of the second phase. This educational research nonprofit received a separate grant from the Joyce Foundation, funder of Harvard's activities, to explore the inclusion within the portal of its online teacher professional development platform, TappedIn[R]. SRI's Director of Community Technology, who had overseen the development of TappedIn, also oversaw the work with MPS. When SRI joined the project, the Director of Community Technology was launching a beta-version of a revised TappedIn, called TappedIn 2. SRI used this updated platform in its work with MPS. Others involved from SRI included a technical developer who worked for the Director of Community Technology and two researchers who were being funded through the grant to conduct a study of the district's planned and actual use of community-building technologies as components of the portal.

Collaborative tools. SRI staff was well-versed in the use of TappedIn (TI) not only as a forum for both synchronous and asynchronous discussion, but also as a more multi-functional groupware for collaborative work. In their own project work, TI was used to post notices, store files, and provide a basic tracking mechanism for revisions to collaborative products. However, for inter-organizational group work, SRI staff relied heavily on teleconferencing. SRI did have a room-based videoconferencing facility that used ISDN, but the expense of the connection was deemed prohibitive for daily or even weekly use. SRI made use of it occasionally for this project during team presentations to groups of MPS teachers or staff, but did not use it for regular meetings of the Design Team.

USING TECHNOLOGY TO COORDINATE THE WORK OF THE DESIGN TEAM

Employing information and communication technologies to help coordinate the work of the Design Team proved to be more a matter of fit among organizational structures than an ideal match of technology to task at hand. In the exploratory phase, MPS was handling its work with each partner as distinct activities. By the time the pilot phase was well underway, regularly-scheduled design team meetings that combined IP videoconferencing, teleconferencing, synchronous chat, and web browsing had come to be the forum for the joint coordination of activities related to the PSP. The transition from the unidirectional coordination that marked the first phase to the use of multiple technologies and multiple points of connection for coordination in the pilot phase was not smooth.

In hindsight, however, the problems that arose in coordination followed along patterns of activities around communication technologies within each organization that were already well established. The failure of Groove and the success of IP videoconferencing as collaborative tools offer insight into the reasons for this. As previously noted, IP videoconferencing was initially viewed as too lacking in "media richness" (Heeren & Lewis, 1997) for the kinds of design conversations partners envisioned as most productive. MPS willingly explored the possibility of using Groove with Harvard, installing the Groove application on several machines. But the technical shift required turned out to be a far higher hurdle than imagined. In retrospect, two conditions stood out as the most salient in limiting the ready adoption of a potentially promising collaborative tool: neither MPS nor Harvard was using Groove as a regular mechanism for the conduct of work prior to these experiments, nor were the machines on which Groove was installed, both at Harvard and MPS, the desktops used most consistently when team members were working on aspects of the portal. Meanwhile, MPS and Homeboyz had already established a work process that relied heavily on face-to-face meetings through their work on an earlier collaboration, the Curriculum Design Assistant. Homeboyz did not even have a computer capable of running Groove, which did not offer a version of its client for the Macintosh OS. The sole Windows PC this web development shop owned was not connected to the Internet and sat in a conference room on a separate floor from the project manager and those working on the design of the portal.

In contrast to the experience with Groove, which was not well-established in any partner organization, IP videoconferencing was an important internal communication channel for at least one partner. Hindsight shows that the convergence around IP videoconferencing as the primary medium hinged on two considerations. MPS was already heavily invested in interactive videoconferencing, with familiar equipment close at hand. And it had a staff person dedicated to videoconferencing support upon whom Harvard could rely. As coordination of tasks among partners became more of a concern due to the design activities of the second phase, regularly scheduled meetings through videoconferences appeared as the easiest, most affordable tack.

But videoconferencing alone was not sufficient to focus distributed work in constructive ways. When SRI joined the effort, TappedIn (TI) appeared as another readily available complement to the collaboration. During the regularly-scheduled design team meetings, the TI chat area served as a back channel of interaction, especially when technical difficulties occasionally caused the video or audio to fail. Between meetings, TI became a repository of links, meeting notes, and a handful of other files that were important to the work of the team.

Another important component was the ready availability of Internet connections in the various rooms used as part of the meetings. Many times, those taking part in meetings browsed the portal as they critiqued the design. Much of the time in meetings was spent with participants not looking at one another's images on the television monitor, but staring into their laptop screens. That is, the video conveyed a sense of collective presence, which was important for social interaction that did not primarily depend on eye contact. Despite the lack of eye contact, the social presence allowed for a more fluid coordination of tasks with only occasional visual check-ins to resolve miscommunications or points of confusion. Moreover, the TI group space supported this style of interaction with a repository of design documents and a "back channel" for conversations through internet chat when the primary medium failed.

Counter to expectations, the adoption of a somewhat uniform toolset for collaboration took place in an evolutionary way, not in an episodic one. The adoption of different tools by one partner came out of the institutional experience with that tool evinced by another. On the other hand, the necessity of face-to-face meetings arose episodically as expected, particularly in two situations. The first was when work could not be accomplished by other means. (For example, an extensive paper-based survey of new teachers applying to be part of the pilot group needed to be tabulated by hand. This activity offered a useful moment for those from Harvard to work with MPS on establishing selection criteria.) The second appeared when new members joined the collaboration. Several expressed the difficulty of understanding what constituted the project until they had been to Milwaukee to talk with those working on it day-to-day and to have topics arise out of the interaction with a panoramic, immersive environment, rather than through the comparatively narrow window of video monitor and computer screen.

DISCUSSION

Four principles appear vital to sustaining a productive team effort across distance. These are complementarity, flexibility, totality, and afford-ability. Each is discussed below along the three dimensions of change--social, task, and technical--presented earlier.

1. Complementarity:

Social: The social dimension of complementarity is akin to what the literature on virtual teams labels "cohesion" (Lea & Nicoll, 2002). It was important in the early formation of the team that both the professor at Harvard and the MPS Director of Technology had a long and successful relationship working on district technology issues. This provided an opening on both sides to explore the potential of collaborative work in innovative ways. This partnership was then reinforced when individual team members took an active interest in each other's areas of work, even when that fell outside of the immediate tasks at hand.

Task: Finding the right fit of expertise to move the development of the portal forward proved essential to the ongoing work of the Design Team. As evidenced by the three phases of work, tasks evolve and occasionally make radical shifts as work proceeds. Each partner, as an institution, needed to respond to gradual as well as dramatic shifts as these become apparent.

Technical: The experience of trying to find the right toolset in this instance teaches that essential conditions for adoption are not only individual, but also rely on institutional "ownership" of the tool, as well as a willingness by other institutions to support: the use of that tool. Equally necessary is the requirement that the tool be an integral aspect of the work process of at least some of the partners for it to be adopted by all.

2. Totality

Social: Videoconferencing proved to be a limited means of empowering the kind of co-construction sought, confirming our initial assumptions. However, it did provide for social connections in ways not anticipated, ones that called for a broader view of social connections than collaborative tools might support. The first dimension of this sort lies in the "social-political" realm. By way of videoconferencing, MPS called on outside consultants to reinforce the worth of the PSP to two distinct audiences: the new teachers and facilitators taking part in the pilot and the broader group of district stakeholders (such as the Content Development Group and PSP Advisory Group) who needed to be convinced of the worthiness of the effort. Second, the room-based set-ups used by both Harvard and MPS provided reflective spaces in which participants could step back from their physical desktops to reflect on the object of their work. This approached the kind of coconstruction Bardram (1998) discusses, one that goes beyond immediate tasks.

Task: The shift from the discoordination of the early phase to the emergent coordination of the pilot phase demonstrated the importance of an awareness of how tasks fit together across the project. Collaborative tools offer enormous advantages in coordinating tasks across distributed partners, but only if coordination is seen as a vital aspect of the work.

Technical: It became clear that no single tool could meet the emergent needs of the Design Team. Such collaborations required stepping back from discreet tasks to an appreciation of more subtle processes and the tools that support those processes as a matter of course in everyday work.

3. Flexibility

Social: The experience of the team reinforced what studies have shown to be true--that virtual teams are prone to greater fluctuations in membership than are colocated teams (Lea & Nicoll, 2002). This requires greater social flexibility on the part of all to maintain a cohesive sense of group identity.

Task: The distributed team had fewer cues to assess the state of the project at any particular moment of time. This required a readiness to respond as changes occur and a willingness to take on required tasks as they arise.

Technical: Like ecological diversity, the diversity of collaborative channels and the flexible use of those most appropriate to given conditions reinforced long-term stability. That the Design Team had many different ways of interacting at any given time--even though only one or two might have been used in a particular moment--reinforced the chances that meetings could focus on the work at hand rather than on making the tools work.

A fourth consideration, affordability, highlights the inevitability of evolutionary change for multi-institutional partnerships with limited resources. Episodic change through the joint adoption of compatible information and communication technology--the great leap--is expensive along all the dimensions discussed earlier. It is expensive socially, in terms of the strain it places on relationships among members of the team; it is expensive in terms of the immediate tasks, requiring all available attention. It is also expensive in terms of tools in that technological leaps necessarily require tremendous support to sustain. Yet, evolutionary change proved incapable of fulfilling expectations and maintaining momentum around the coconstruction of work across distance, a limitation that became apparent in the most recent phase of the project.

IMPACT ON TEACHERS AND THE DISTRICT

Before drawing conclusions from the experience of putting together a distributed Design Team, it is important to take note of the effect of the project on its intended audience--new teachers--as well as the broader impact it had on others in the district. Portal-related activities with teachers consisted primarily of "cadre meetings"--groups of approximately a dozen teachers and two teacher-leaders who checked in either online or face-to-face every two weeks, using the resources available through the portal as a springboard for their discussions. Membership in the cadres included teachers in over half of the MPS schools. At the end of the 2003 academic year, 13 cadres were operational, with 104 members ready to continue in 2004. By June 2004, 21 cadres were fully functional with 256 members and 52 facilitators. In the middle of the 2004-2005 school year, a total of 70 facilitators were working with 303 cadre members.

During each school year, the cadres met both online and in person. For the most part, the cadres were organized around grade levels and interests. The leaders/facilitators received support from MPS staff working on the portal, who organized many activities (both online and face-to-face) to buoy and train the facilitators and teachers. Key portal staff included three former teachers who, on a full time basis, had assumed the responsibilities of cadre development and facilitator training. MPS staff were monitoring new teacher retention with early anecdotal evidence pointing towards a positive impact on the 2003 group, especially in terms of motivation and collegiality. As of January 2005, it was still too early to tell conclusively whether involvement in the portal project has positively affected teacher retention past the crucial three-year mark.

The process of working with teacher cadres along with designing and piloting the portal had revealed, in particular, the need to attend to the "back end" of the portal. The early portal effort highlighted the lack of a common platform across the district for access to a wide range of shared content and digital services not just for new teachers but for all teachers and administrators in the district. To remedy this, the district contracted with a major vendor of commercial portal software to put together such a common platform. Over the 2003-2004 school year, focus shifted from the portal's "front end" design to back-end information architecture. The new portal, launched at the end of the 2003-2004 school year, carried several of the same basic design elements that were defined in the project's initial stages, but did so on top of a far more sophisticated "intranet" meant to bring central departments and schools into closer coordination.

CONCLUSION: EVOLUTION AND REVOLUTION

No single groupware application proved capable of meeting the needs of the Design Team through all three phases and across all institutional partners. However, an array of compatible tools began to emerge that allowed the team not only to talk about design but also to do the work of design across distance. The initial phases of the portal project provided an important proving ground for demonstrating the potential as well as limits of such collaborative design across distance. This experience helped bring into focus lessons that the district later began to apply as it escalated its efforts to provide a comprehensive, technology-enabled support structure for new teachers.

Out of this experience, those working on the project found four principles to be essential for effective, technology-mediated collaboration--complementarity, totality, flexibility, and affordability. Moreover, to understand fully any one of these required attention not just to technical know-how but also to the ways in which the array of tools in use helped accomplish the work at hand and enrich social interaction. The "bulls eye" of effective collaboration happened when the mix of these three dimensions across all four principles achieved some rough measure of equilibrium. Table 2 provides key questions to consider when taking an inventory of existing or promising tools that might help nurture collaboration across distributed institutional partners.

In sum, given the collaborative tools available, three steps appear critical for future efforts:

1. identify common collaborative tools that are either dominant or supplemental among at least two partners;

2. examine the existing ways that those tools are used to accomplish the social, technical, and task dimensions of work in the partner organizations; and

3. determine the degree to which other partners would be willing to adapt/adopt these tools, taking into account the four considerations of complementarity, totality, flexibility, and affordability.

The portal project revealed that evolutionary approaches may lead to identifying the need for more deep-seated--and disruptive--episodic change that came in the form of an effort to revamp the district's information architecture. The Design Team eventually confronted systemic constraints that could only be resolved by a more integrated initiative across the district. Evolutionary change paved the way towards a "tipping point" of sorts, one continuously reinforced by the often-times slow, sometimes bounding changes that occur as information and communication technologies become ever more integrated with the basic processes of work.

References

Bardram, J. E. (1998). Designing for the dynamics of cooperative work activities. Paper presented at the Computer Supported Collaborative Work Conference, Seattle, Washington.

Choy, S. P., Chen, X., & Ross, M. (1998). Toward better teaching: Professional development in 1993-1994 (NCES 98-230). Retrieved April 24, 2006, from http://nces.ed.gov/pubs98/98230.pdf

DeSanctis, G., & Poole, M. S. (1994). Capturing the complexity in advanced technology use: Adaptive structuration theory. Organization Science, 5(2), 121-147.

Engestrom, Y., Puonti, A., & Seppanen, L. (2001). Spatial and temporal expansion of the object as a challenge for reorganizing work. Retrieved April 24, 2006, from http://www.emp.uc3m.es/~quattron/conference/papers/Engestrom.pdf

Heeren, E., & Lewis, R. (1997). Selecting communication media for distributed communities. Journal of Computer Assisted Learning, 13(2), 85-98.

Hettinga, M. (2002). Understanding evolutionary use of groupware (Fundamental Research Series No. 007). Enschede, The Netherlands: Telematica Instituut.

Hill, J. R., Tharp, D., & Sindt, K. (1998). Collaborative web site design from a distance: Challenges and rewards. TechTrends, 42, 31-37.

Kydd, C. T., & Ferry, D. L. (1994). Managerial use of video conferencing. Information and Management, 27, 369-375.

Lea, M. R., & Nicoll, K. (Eds.). (2002). Distributed learning: Social and cultural approaches to practice. New York: RoutledgeFalmer.

National Commission on Teaching and America's Future (1996). What matters most: Teaching for America's future. Retrieved April 23, 2006, from http://www.nctaf.org/publicatios/WhatMattersMost.pdf

National Commission on Teaching and America's Future. (2003, January). No dream denied: A pledge to America's children. Retrieved April 24, 2006, from http://www.nctaf.org/dream/report.pdf

Smith, P. J. (2003). Workplace learning and flexible delivery. Review of Educational Research, 73(1), 53-88.

Footnotes

1. <http://www.groove.net/>

2. <http://www.tappedin.org/>

3. <http://www.teachscape.com/>

4. <http://www.tappedin.org/>

5. A screenshot of the homepage of the portal is available at: <http://www.gse.harvard.edu/~dedech/images/joycepsp.jpg>

DAVID EDDY SPICER AND CHRIS DEDE 

Harvard Graduate School of Education

Cambridge, MA USA

[e-mail]

[e-mail]

Table 1 Three Phases of Design Team Activities and the Use of
Collaboration Tools

                Fall 2001-Spring
                2002
                Exploratory         Summer 2002        Fall 2002-Spring
                (establishing       Preliminary        2003
                specifications,     (establishing      Pilot (initial
                parameters)         interface design)  use)

MPS             Teacher/advisor     Teacher            Increasing focus
                focus groups        selection.         on content.
                Securing funding    Focus on           Managing
                Securing            interface design:  expectations.
                partnerships        look & feel,
                                    functionality.
Homeboyz        Developing          Interface design;  Iterative design:
                specifications      back-end database  Moving from
                                    design             development
                                                       server to
                                                       production server
Harvard         Experiments with    Consultations on   Consultations on
                Groove.             interface design   formative
                internet-based      and knowledge      evaluation and
                videoconferencing   management         knowledge
                                                       management
SRI             Initial             TappedIn 2         TappedIn 2 roll-
                conversations       testing            out w/ teachers
Collaboration   Email listserv      IP                 IP
Tools           (IP                 Videoconferencing  Videoconferencing
Dominant        Videoconferencing)  Email listserv     Email listserv
Additional      (Groove)            Teleconference     TappedIn2
(Experimental)                      (TappedIn2)        Teleconference
                                                       (Timbuktu)
Frequency of    Arranged ad hoc     Weekly             Weekly (Oct-Dec)
regularly-                                             Bi-weekly
scheduled                                              (Jan-May)
meetings

TABLE 2 Key Questions to Consider When Examining New Technologies to
Promote the Work of Distributed Institutional Partners

                 SOCIAL

Complementarity  Does the array of tools enable team members to take
                 an active interest in each other's work?
Totality         Are the array of tools in use capable of meeting a
                 broad range of needs from casual exchange to formal
                 presentations?
Flexibility      As membership fluctuates, does the array of tools in
                 use readily accommodate new members?
Affordability    Does the array of tools lead to less costly ways of
                 maintaining the "social fabric" of the team?

                 TASK

Complementarity  As tasks evolve, are the right tools in use to
                 accomplish those tasks?
Totality         As tasks evolve, does the array lead to greater
                 coordination across tasks?
Flexibility      Are team members,
                 [TEXT INCOMPLETE IN ORIGINAL SOURCE]ew and old, readily
                 able
                 [TEXT INCOMPLETE IN ORIGINAL SOURCE] track the
                 evolution of
                 [TEXT INCOMPLETE IN ORIGINAL SOURCE]sks and shift among
                 [TEXT INCOMPLETE IN ORIGINAL SOURCE]odes of
                 communication
                 [TEXT INCOMPLETE IN ORIGINAL SOURCE] meet the demands
                 of
                 [TEXT INCOMPLETE IN ORIGINAL SOURCE]ose tasks?
Affordability    How much of each team member's attention does this set
                 of tools require? Is this commensurate with the
                 benefits they bring?

                 TECHNICAL

Complementarity  To what degree do the collaboration tools in use "fit"
                 within each partner's existing infrastructure?
Totality         For what kinds of collaborative design activities are
                 the existing tools of little use? What other tools
                 might complement the existing palette?
Flexibility      Does the array in use allow for multiple channels of
                 interaction?
Affordability    What are the costs of introducing and supporting the
                 existing array? Of introducing new tools to the array?