2. Conceptual Framework

2.1 Introduction

I began the process of building a suitable conceptual framework for this study by gathering the empirical factors from research efforts conducted by the UCD Internet Task Force and my colleagues at RMC Research Corporation, starting in 1994. Next, I summarized important ideas from three disciplines: adoption theory, systems and organizational change theory, and computer-mediated environments, and linked them with parallel constructs from cognitive science, collaborative learning, situated cognition, and activity theory. Then, I compared the empirical with the theoretical factors to test for inclusiveness. This process is illustrated in Figure 2.1.

Finally, I grouped all of these factors into six naturally emerging clusters that influence the diffusion of the Internet within the School. These clusters will act as a starting point for formulating a new model of adoption of telecommunications by educational institutions. Like its origins, the adoption theories of Rogers and Hall, this model is principally descriptive in nature. Thus, a good deal of data could be collected, organized, and reported by using it as a basic framework.

A thorough investigation of the adoption process, however, encompasses more than simply listing and labeling factors. It requires a deep understanding of the complex internal, social, and cultural processes that occur over time as new users learn about the properties and affordances of the Internet; discover how to use it in ways they find useful; are (or are not) rewarded in new ways for their efforts; and become able to communicate promptly and clearly with the faculty, administration, and university network support people about the problems they are encountering.

I found it necessary to start with a fairly objectivist framework that lists factors and supports quantitative methods such as the 1995 and 1997 e-mail usage surveys among members of the SOE. This is what I will present in this chapter. However, in the data collection activities, I moved beyond surveys to a case study approach with qualitative methods and converging lines of inquiry, just as I did with my evaluation of the Boulder Valley Internet Project (BVIP). My aim in this chapter is to articulate the individual conceptual changes and group processes of members of the SOE as they learn the basics of mediated communication, deal with their concerns and learning anxieties, develop expertise, adopt, and eventually reaffirm or reject the use of the Internet for teaching and learning.

Figure 2.1

2.2 Adoption: Empirical Base

2.2.1 Preliminary Investigations

The spring of 1995 marked the first phase of my research concerning the needs, support tools, adoption processes, cultural changes, and formation of collaborative learning communities as members of the SOE began to be introduced to the Internet as a tool for teaching and learning. I worked as a collaborative member of the UCD Internet Task Force, whose activities I have already described in Chapter 1.

For my part, I designed and distributed a survey to a sample of 73 faculty and students in the Technology and Special Services (TSS) Division. Responses were voluntary and anonymous. Participants were asked about their status in the division, access to technology that would support the use of Internet tools, patterns of use, reasons for use, and challenges to use of the Internet. Respondents were also asked to rank order eight proposed supports for training and performance using e-mail and the Web. An analysis of the collected data is on-line (Sherry, 1997c). I reported the 1995 survey results in the TSS Division Newsletter (Sherry, 1995).

Despite the reported high levels of access to the technology at home and at work, and the availability of free e-mail accounts through the university, the relative frequency of use in 1995 was low. A factor analysis of the reasons for use, with Varimax rotation, resulted in four factors: (a) share and disseminate information and communicate (41.5% of variance), (b) find and organize information (11.7% of variance), (c) collaborate (8.7% of variance), and (d) consult with your advisor (7.8% of variance).

When the responses to the eleven "challenges to Internet use" items were polarized, they were re-interpreted as success facilitators. These grouped into three factors: (a) clear benefit and value (32.5% of variance), (b) self-efficacy (17.2% of variance), and (c) finding a voice and having something to say (10.4% of variance). These were three of the five factors that emerged from Wilson and Ryder’s group interviews with students from the same population and from McCahan’s in-depth interviews with faculty members. The five factors reported in Wilson, Ryder, McCahan, and Sherry (1996) were: (a) clear benefit and value, (b) self-efficacy, (c) finding a voice and having something to say, (d) personal/cultural compatibility, and (e) proper scaffolding.

The first two factors, namely, (a) clear benefit and value and (b) self-efficacy, were identified by Bandura (1982) under his theory of self-efficacy as a mediator of performance and achievement. Driscoll (1994) notes that "learners can be sure that certain activities will produce a particular set of outcomes...but, if learners harbor serious doubts as to whether they can perform those required activities, they will not put forth the effort" (p. 301).

I renamed the third factor, finding a voice and having something to say, as "mediated writing proficiency". Berge (1997) notes that students with good mediated writing proficiency are able to deal with the lack of social context cues associated with both a text based interface and the inherent time lag in asynchronous computer conferencing. E-mail conferencing is well suited for thoughtful, shy, or hesitant conversationalists who prefer to consider and carefully frame their answers and responses before presentation. Berge notes that "the very act of assembling one’s thoughts and articulating them in writing for a conference audience appears to involve deeper cognitive processing" (1997, p. 10).

This third factor was also identified and explored by Fishman (1997) as written communication apprehension. Not only must people be confident writers; they must also feel comfortable writing to a public audience. Fishman found an interesting gradation in the correlation between written communication apprehension and the amount of privacy afforded by three CMC tools, namely: e-mail (low correlation); the Collaboratory Notebook, a semi-private software platform in which students could share their project work asynchronously (moderate correlation); and Usenet newsgroups, the most public of the CMC tools used by his students (high correlation). Thus, mediated writing proficiency becomes an important factor that enables a person to find a voice and express his or her thoughts to others using a text-based interface.

The fourth and fifth factors that emerged from Wilson and his colleagues’ interviews were (d) cultural/personal compatibility between technology and people’s learning styles, self-concepts, and lifestyles; and (e) proper scaffolding. Cultural compatibility includes school policies and norms of use. Proper scaffolding refers to a support structure that includes a non-judgmental, social support system, one-on-one mentoring relationships, and removal of technical hurdles to the use of the innovation.

Cultural/personal compatibility was identified by Rogers (1995) as "the degree to which an innovation is perceived as being consistent with the existing values, past experiences, and needs of potential adopters" (p. 15), and represents an important perception on the part of new adopters. This is important because an innovation that is compatible with the norms and values of a social system will be adopted more rapidly than one that is not. The following response by a student to the 1995 survey is an example of a cultural/personal compatibility concern: "Updating of your password is too cumbersome. Mine expired and I’ve been told I need to go down to campus to renew. Doesn’t that go against a major philosophy of e-mail? Seems to me it does." Although personal concerns are highly influenced by cultural concerns, it makes sense to consider each of these separately for the purposes of the present analysis.

Hall and Hord note that as individuals evolve from non-users to expert users of a new innovation, they go through a series of six stages of concern (1987, p. 60). Once they have gathered sufficient information about the general characteristics, effects, and requirements of use, they move on to the stages of personal concern and management concern. Whereas personal concerns are characterized by uncertainty about the demands of the innovation and their adequacy to meet them, management concerns are characterized by concentration on the tasks of organizing, managing, scheduling, and finding time to use the innovation.

Proper scaffolding is crucial in the personal-concerns stage of adoption when facilitators should visit more often with potential adopters on a one-to-one basis to offer assistance and encouragement. Scaffolding is also important at the management-concerns stage of adoption when it is important to provide "how to do it" workshops that address the constantly changing topics of management concerns as they arise (Hall & Hord, 1987, p. 72).

The following suggestions by survey respondents demonstrate the spectrum of courses and workshops that might be considered as future interventions:

The Internet Task Force originally considered implementing or recommending several types of supports and aids. These comprised (a) paper-based job aids (brochures, booklets, tutorials); (b) on-line support (on-line tutorials specific to the UCD network); (c) interactive computer demonstrations; (d) formal classes; (e) one- to two-hour workshops; and (f) individual, face-to-face mentoring by graduate assistants. Since these supports comprise both impersonal and personal scaffolding, I broke "proper scaffolding" into two separate factors, namely, impersonal scaffolding (job aids) and personal scaffolding (modeling and coaching).

Based on the responses of our participants, the Internet Task Force decided to drop the idea of impersonal job aids and concentrate on individual mentoring, workshops, in-class demonstrations, and responses to individual students’ questions via e-mail. We also made a presentation to the Dean and the TSS Division Chairperson to encourage the development of supportive organizational arrangements that could underwrite the efforts of our team, including support for a prototype SOE Web Page, better access to e-mail, and more instruction and assistance for new users.

The following year, the Dean funded a graduate assistant position for individual mentoring of faculty and staff members to be shared on an hourly basis by our team members. Concurrently, CINS started offering 1- to 2- hour free, non-credit workshops in Internet basics, and the SOE created new courses in telecommunications and Web authoring.

In summary, the adoption factors that emerged from the initial research efforts by Internet Task Force members are presented in Table 2.1.

Table 2.1

2.2.2 Developmental Research

At about the same time as we were implementing our initial supportive interventions, the Internet Task Force was also actively involved in building a Web page for the SOE. This gave one of my colleagues and me a good opportunity to conduct a developmental research study (Seels & Richey, 1994) on the process of collaboratively designing a mediated learning environment. Our investigations enabled us to study variables associated with group dynamics as well as individual, organizational, and technological factors. We developed a process model that is appropriate for groups of designers working in either academic or corporate environments (Sherry & Myers, 1998). Critical features of our model were (a) reflection-in-action and metacognition, (b) knowledge development, (c) working on an authentic task, and (d) generating research questions as a result of our design efforts.

We studied, reflected upon, and documented the collaborative process as we participated in both face-to-face meetings and e-mail discussions in an attempt to work on an authentic task. We achieved triangulation by saving meeting notes, e-mail messages, and responses to a closure questionnaire by group members.

The conceptual framework for our research was based on the work of many authors, primarily Ann Brown (1994). Brown’s process of building a community of learners consists of five important constructs, all of which were present in our own activities. These five constructs were (a) metacognition, strengthened by intentional learning and reflection; (b) multiple zones of proximal development and irregularly distributed expertise; (c) legitimization and negotiation of differences; (d) a community of discourse; and (e) a community of practice.

Throughout the design process, our team’s goal of completing the design, development, and implementation of the SOE Home Page was enhanced by our commitment to reflective practice and metacognition. For our purposes, we defined metacognition as conscious reflection upon, and monitoring of, our own cognitive states and processes.

Because expertise was irregularly distributed throughout the team, we worked in pairs or small groups, modeling and sharing our knowledge and skills in uploading/downloading files to the server, mastering UNIX commands, designing and authoring Web pages in HTML, and so forth. One of the team members also created an on-line job aid called Makepage, which enabled team members to create Web pages automatically. A major advantage of this distributed expertise, shared in jigsaw fashion, was that the group as a whole benefited from the increased range of expertise and the combined knowledge of the team members who gradually acquired representational proficiency as they simultaneously learned and worked in a mediated learning environment.

Our continuing conversations, both face-to-face and via e-mail, enabled us to identify individual strengths and weaknesses and to create a good match between individual skills and necessary tasks. Then, individual team members proceeded to develop the requisite skills and common knowledge base to take ownership of their respective tasks and to participate in the group’s creative efforts. Lave & Wenger (1996) refer to this as legitimate peripheral participation.

Although we were novice designers, we functioned more as an expert team because of the distributed expertise and multiple perspectives of individual team members. Where skills were lacking, team members with greater expertise coached novices, built close relationships with them, provided scaffolding, and mentored them through the learning process, using the strategies of cognitive apprenticeship (Collins, Brown, & Newman, 1989). Where tasks were intertwined, team members worked as pairs, either sharing or linking related elements. Novices began to take on more independent activities. Fading occurred as the experts had to answer fewer and fewer elementary questions and as the novices began to assume more expert roles.

We found that it was important for team members to co-construct a common conceptual ground (Pea, 1994). We needed to establish a common understanding and vision of the final product and to share visual representations of the partially developed structure of the product as it evolved, using a "class" account on the Carbon server that supported Web authoring activities. By using the PINE e-mail system, individual team members could e-mail the group or individuals with questions, responses, suggestions for revision, new perspectives on design, ideas for changing the types of scaffolding available to group members as our expertise continued to grow and develop.

As with any design effort, this was an iterative process, subject to many false starts and revisions. Here, Bereiter’s (1994) concept of progressive discourse became important. To continue to participate in a dialogue, we often needed to suspend our original assumptions and preconceived notions and to subject our ideas to examination by the entire group, whether via e-mail or in face-to-face meetings. At times, we arrived at a mutually agreeable solution by re-thinking our original premises, negotiating meaning among group members, appropriating meaning from the group, restructuring our ideas, shifting our perspectives, and developing a unique synthesis of two alternate points of view (Brown & Palincsar, 1989; Pea, 1993). Our design team had an evolving conceptual model or shared meaning scheme of what we wanted the final product to look like, and we continually updated it. If an individual product failed to meet the agreed-upon criteria, individuals began to make adjustments without having to be told to do so by the group.

By capitalizing on varieties of talent, the group setting facilitated learning, adoption and use of the requisite Internet tools, and conceptual change. Our team self-organized in a way that capitalized on the strengths of individuals with dispersed knowledge. Members began to share leadership and discourse activities that are normally reserved for a mentor or manager, while the mentor became a full participant in the design process. Our developmental research enabled us to move from a purely descriptive model to a more process-oriented model that combined learning, adoption, collaborative design, and conceptual change by participatory observation of our interactions with other members of the design team. In summary, the new adoption factors that emerged from our developmental research are presented in Table 2.2.

Table 2.2

2.2.3 Theory-Based Evaluation

Because of my interest in research, I was hired as a research assistant by RMC Research Corporation to evaluate the BVIP (Sherry, Lawyer-Brook & Black, 1997; Sherry, 1997a). Throughout the evaluation I found that there were major differences between Rogers’ theory and the reality of the BVIP, and that diffusing the Internet throughout a large school district could not be fully described by the Rogers model.

First, the Boulder Valley School District (BVSD) was a decentralized institution, characterized by site-based management—a very different organizational structure from Rogers’ centralized management concept. Second, an innovation is generally considered to be a relatively stable technology, whereas the dynamic, evolving character of the Internet is anything but stable. And third, although both Rogers (1995) and Hall and Hord (1987) note that client-change agent empathy contributes to horizontal diffusion of the innovation throughout an organization, we observed that the use of the peer trainer-of-trainers model limited diffusion horizontally to teachers, media specialists, and technology coordinators. In the BVIP, the use of the Internet did not diffuse vertically to the administration or the policy-making bodies.

To build a new analytical framework for the BVIP evaluation, we started with the traditional Rogers (1995) model. The first module of our new model emphasized individual user perceptions. Next, we added an organizational factor module. This incorporated factors derived from several other investigations, namely: (a) Farquhar and Surry’s (1994) Adoption Analysis Tool; (b) some of the organizational factors from Gross, Giacquinta, and Bernstein’s (1970) and Teasley’s (1996) studies on the failure to implement a major organizational innovation; and (c) the explanatory framework for the varying rates of adoption of educational innovations by sociometric pairs developed by Carlson (1970).

To these individual and organizational factors, we added a third module. This comprised a set of empirically-derived technological factors: (a) access and availability; (b) network response time; (c) human-computer interface (HCI), including design factors; (d) reliability; and (e) system capacity. Important technological factors were identified by the original BVIP evaluators (Wolf & Black, 1993), by several large-scale studies on the adoption of telecommunications in school settings (Honey & Henriquez, 1993; Levin, 1995; U.S. Department of Education, 1996), and by our own data collection activities.

Throughout the duration of the BVIP evaluation, we found ourselves addressing many of Rogers’ concerns and reflecting on his findings. We were also able to investigate some new aspects of the adoption and diffusion process such as its spread within a decentralized system and the re-invention process that was taking place among new users as the innovation (in our case, the Internet) continued to evolve.

We collected data from surveys, in-depth interviews, focus groups, a work group, examination of system logs and artifacts, observations, and an embedded case study of a technologically advanced elementary school. As a result, we dropped factors that did not appear in our model and added others that emerged from our data collection activities. For example, since the BVIP dealt specifically with training classroom teachers to use the Internet and integrate it into their curriculum, we added a fourth module that comprised teaching and learning factors specific to the BVSD. Many of the factors that appeared in the preliminary investigations by Wilson, Ryder, McCahan, and Sherry (1996) and the developmental research by Sherry and Myers (1998) also appeared in the BVIP evaluation, so I will not repeat them here.

First of all, BVIP teachers who were learning how to use the district network were primarily in Hall and Hord’s informational, personal, or management concern stages. They were actively seeking information about the general characteristics, effects, and requirements for use of the network, as well as the context in which they would be using it. Their activities took place either at home, searching the Web for suitable activities and resources for their classes, or at school, working with their students who were engaged in either their individual research projects or designing Web pages for whole-class projects.

Besides the formal training offered by the BVIP, teachers reported that two types of support were particularly useful: (a) impersonal scaffolding that included job aids such as documentation, checklists, and troubleshooting guides; and (b) personal scaffolding or mentoring, either individually or in open labs. Open labs were particularly useful, since teachers could have their individual questions answered by trained technical support people.

Those who had been trained on gopher, Archie, Veronica, and ftp, and now had Netscape installed on their classroom computers, perceived the Internet as easy to use. Many teachers reported that e-mail communications enabled them to contact their colleagues at convenient times and places, primarily from home, to share promising practices and new strategies for using the Internet in their classrooms. Innovative teachers considered the system trialable and were constantly exploring, testing, and trying out new ideas. However, the affordances of the Internet—its opportunities for action—were less clear. We observed examples of incompatibility with teachers’ needs and wants, diminution of observable benefits, and lack of relative advantage as compared with traditional classroom practices, especially for those teachers who had given up searching the Web for resources after finding no information that they could actually incorporate into their curriculum.

BVIP teachers were also concerned because the key policy makers were espousing a "back to the basics" approach to curriculum at the very time that they, as teachers, were using constructivist strategies to promote a more student-centered vision of learning. This clash of visions led to a great deal of turmoil and indecision by teachers who were just beginning to learn how to use the Internet, to connect to the BVIP network, and to consider ways to use the Internet to enrich and enhance their classroom activities (Sherry, in press). There were no incentives in place, nor was there any clear payoff for using the network. Planning time was being cut due to decreased district funding. BVIP peer training sessions were not considered mainstream professional development. Concurrently, the district policy makers were trying to persuade the individual schools to re-integrate the in-building technology support personnel back to the traditional classroom.

What we observed was similar to Schein’s (1996b) lack of alignment of sub-cultures within a learning organization, where the visions of the corporate folk were not aligned with those of the ordinary folk. Unfortunately for many teachers, the cognitive dissonance between their constructivist visions of learning and the policy makers’ vision of standards-based instruction often led to rejection of the innovation, rather than renewal, reaffirmation, or reinvention. This is consonant with the view of Steven Hodas (1993). Writing in the Education Policy Analysis Archives, Hodas notes that failures of technology generally result from a mismatch between the values of the school organization and the payoff from using the technology (Hodas, 1993, p. 1).

In an insightful article in the Harvard Educational Review, Elmore (1996) reinforced our perceptions of the difference in visions between the change agents and teachers within school districts that were in the process of implementing educational innovations and its effect on the diffusion process. If the teachers considered the innovation to be a tool that made their work easier and more efficient, then they were likely to adopt it. However, if adoption of the innovation meant that they had to change the core of their instructional processes—especially without an observable incentive structure to reward their efforts—then in all likelihood, the innovation would not be adopted. In a personal communication (October 27, 1997), Dr. Peggy Bailey, Faculty Chair of the Department of Instructional Technology at Northern Illinois University, reiterated Elmore’s sentiments: "We need to find as many ways as possible to reward and endorse on-line research, learning, and collaborations".

I found all of the empirical factors in the BVIP evaluation that emerged from my two previous research endeavors, plus eight more. The new factors are presented in Table 2.3.

Table 2.3

2.2.4. Summary of Empirical Factors

Overall, a total of 26 empirical factors were gleaned from these three major research endeavors. Table 2.4 summarizes all of these empirical factors.

Table 2.4

There are many schools of thought that can be used to construct a theoretical framework for the adoption and diffusion of the Internet throughout the SOE. Among the most important are: (a) diffusion theory; (b) human performance technology; (c) transformational learning; (d) collaborative learning; (e) situated cognition; (f) computer-mediated communication (CMC); and (g) organizational learning and change.

Upon reviewing the literature in each of these disciplines, I found that all 26 of the empirical factors revealed by my three previous investigations were best covered by a combination of just three schools of thought, namely: (a) diffusion theory; (b) CMC; and (c) organizational learning and change theory. To add richness to the diffusion theory base, I will add insights from the situated cognitionists and the learning theorists before launching into the basic tenets of the CMC experts and the organizational learning and change theorists.

2.3.2 The Foundations of Diffusion Theory

Rogers (1995) and Hall and Hord (1987) were the pioneers who mapped out the problem space of adoption and diffusion theory. Both of their models were developed in a fairly objectivist manner using quantitative methods for analyzing and presenting their data, in contrast to the constructivist paradigms of learning and change that are prevalent today and the qualitative methodologies that they support.

Neither Rogers nor Hall and Hord take a situated view of the change process such as that espoused by Suchman (1987). On the one hand, Suchman claims that "the coherence of situated actions is tied in essential ways not to individual predispositions or conventional rules but to local interactions contingent on the actor’s particular circumstances" (Suchman, 1987, pp. 27-28). On the other hand, the diffusion theorists see the adoption process as a fairly predictable, linear, rational sequence with information-seeking preceding intentional action, followed by evaluation, followed by reaffirmation, rejection, or readjustment.

Overall, these diffusion theories are good examples of the information processing model of human learning—rational, minimally situated, mechanical, and planful—whereas in Suchman’s "life-world" perspective, learning is phenomenological, contextual, and experiential; and interpretations of events determine the meanings of actions. It is important to be fully aware of the underlying assumptions of the diffusion theorists while laying the foundation for this study. It is also important to be aware of gaps in their theory bases that need to be fleshed out by incorporating insights from some of the more recent constructivist theorists.

2.3.2.1 Everett Rogers

Everett Rogers (1995) states that a study undertaken by Ryan and Gross at Iowa State University in 1943, using interviews, marked the beginning of diffusion research. Since then, other researchers have studied the diffusion of innovations, primarily in rural areas, such as water purification, birth control, genetically engineered seeds, and the like. In his book, Diffusion of Innovations, first published in 1960 and now in its fourth edition, Rogers summarizes four important aspects of his diffusion framework.

First, potential adopters judge an innovation based on their perceptions regarding five attributes of the innovation (Rogers, 1995, 15-16). These perceptions are:

1. Relative advantage—the degree to which an innovation is perceived as better than the idea that it supersedes.

2. Compatibility—the degree to which an innovation is perceived as being consistent with the existing values, past experiences, and needs of potential adopters.

3. Complexity—the degree to which an innovation is perceived as difficult to understand and use.

4. Trialability—the degree to which an innovation may be experimented with on a limited basis.

5. Observability—the degree to which the results of an innovation are visible to others.

Second, the innovation-decision process comprises a series of stages through which potential adopters pass as they move from seeking information about the innovation, to making a decision to adopt or reject, and finally, to confirmation of their adoption decision (Rogers, 1995, 20). These stages are:

1. Knowledge—an individual learns of the innovation’s existence and gains some understanding of how it functions.

2. Persuasion—an individual forms a favorable or unfavorable attitude toward the innovation.

3. Decision—an individual engages in activities that lead to a choice to adopt and make full use of the innovation or to reject the innovation.

4. Implementation—an individual puts an innovation to use.

5. Confirmation—an individual seeks reinforcement of an innovation-decision that has already been made.

Third, diffusion of innovations is inherently a social process. Rogers defines a social system as "a set of interrelated units [or individuals] that are engaged in joint problem-solving to accomplish a common goal" (p. 23). Some members of a social system are relatively earlier in adopting new ideas than others, leading to five adopter categories, namely: innovators, early adopters, early majority, later majority, and laggards (Rogers, 1995, 22).

Fourth, the rate of adoption (the relative speed with which an innovation is adopted by members of a social system) follows an S-shaped curve. At the beginning of the process, the innovators who represent a small minority of the population adopt the innovation. Later, as change agents become active, the diffusion curve begins to climb. Finally, the curve reaches an asymptote as fewer and fewer later adopters remain.

Personally, I take issue with both the term "laggards" and Rogers’ simplistic labeling of people along a single dimension characterized by the speed with which they adopt an innovation. Those who are reluctant to adopt a new technology (whom I prefer to call "resisters") may have very valid reasons that they can generally explicate, given the chance to do so.

Rogers also describes three different types of innovation decisions. These are: (a) an optional innovation decision which represents an intentional decision by an individual, independent of the decisions of the other members of the system; (b) a collective innovation decision, in which the choice to adopt or reject is made by consensus among members of the social system; and (c) an authority innovation decision, which is a choice made by relatively few individuals in the system who have the power, status, or technical expertise to persuade members of the system to adopt the innovation.

This is an important distinction. Rogers observes that the fastest rate of adoption of innovations results from authority decisions, though these decisions may later be circumvented during the implementation process (Rogers, 1995, p. 29).

In this brief summary of Rogers’ four important theories, several empirical adoption factors emerge, which include but are not limited to the users’ perceptions of the innovation. This marks the starting point in building a theoretical framework. These factors are presented in Table 2.5.

Table 2.5

2.3.2.2 Gene Hall

Another prominent figure in diffusion theory is Gene Hall. Hall and Hord’s Concerns-Based Adoption Model (CBAM) is actually a theory of systemic change rather than a theory of diffusion of innovations. In their classic work, Hall and Hord (1987) described three empirically derived aspects of adoption which they named stages of concern, levels of use, and innovation configurations. Facilitators and change agents could be more effective if they considered the concerns of the new adopters.

Concern theory emerged in the late 1960s from the work of Frances Fuller and her colleagues at the University of Texas at Austin. Fuller studied the problems and satisfactions of teachers at various points in their careers and identified four clusters of concerns: unrelated, self, task, and impact. These clusters of concerns changed in a predictable fashion as teachers became increasingly experienced in their work. Hall and Hord extended Fuller’s work to encompass six stages of concern about an innovation, starting from "ground zero"—little concern or involvement with the innovation (Hall & Hord, 1987, p. 60). These six stages are defined as follows:

0. Little concern or awareness.

1. Informational—seeking more details about the innovation, its effects, and its requirements for use.

2. Personal—uncertainty about one’s adequacy to meet the demands of the innovation.

3. Management—focusing on the processes, tasks, and activities involved in using the innovation.

4. Consequence—focusing on the impact and relevance of the innovation on peers and students.

5. Collaboration—coordinating and cooperating with others regarding the use of the innovation.

6. Refocusing—focusing on the possibilities of major changes or alternatives to the existing form of the innovation.

Second, there is a pattern to users’ behaviors as they develop expertise in the use of the innovation. There are eight levels of use that can be identified and distinguished (Hall & Hord, 1987, p. 84). These are defined as follows:

0. Nonuse—no involvement with the innovation.

1. Orientation—exploring its value and its demands upon the user and the social system.

2. Preparation—preparing to use the innovation for the first time.

3. Mechanical use—engaging in a stepwise, disjointed attempt to master the tasks and activities necessary to use the innovation.

4a. Routine use—stable use with little thought given to improving the innovation or its consequences.

4b. Refinement—varying the use of the innovation to increase its impact on students and colleagues.

5. Integration—combining one’s efforts with related efforts of colleagues to achieve a collective impact on students and colleagues.

6. Renewal—re-evaluating the quality of use of the innovation, seeking modifications and new developments to increase its impact.

Finally, Hall and Hord deal directly with the characteristics of the innovation as perceived by the user, drawing upon the work of Everett Rogers. In essence, any innovation has variations and components, not all of which are necessarily utilized. The type of use depends on the type of activities that the user would like it to support. (See Engestrom, 1996.) The emphasis on the use of a tool to support intentional activity is consonant with Allen and Otto’s (1996) treatment of the affordances (action potentials) of a mediated environment, which lead to its effectivities (modes of effective use) as perceived by the potential adopter.

Hall and Hord, however, take a different approach. In their view, the innovation can be configured and used in several ways (Hall & Hord, 1987, 121, 138). These configurations span a spectrum from (a) ideally, just as the developer intended; (b) varied within limits corresponding to legitimate or acceptable use; (c) drastically mutated in a manner that is radically different from the original concept; to (d) unacceptable. Though the innovation configuration aspect of the CBAM was based on empirical observations, it presupposes an intentional design decision on the part of the developer that reflects the philosophical basis and underlying assumptions of the innovation’s types of use (Hall & Hord, 1987, p. 112).

For the Internet, nothing could be further from the truth! Nobody designed the Internet, and nobody controls it. Each individual user is free to use the Internet as he/she deems appropriate to support his/her intended actions.

Hall and Hord give a thorough treatment to the change facilitator’s style and the types of interventions that may be appropriate at various stages in the adoption process, under various circumstances. For example, a change agent can engage in any of the following activities: (a) take the lead and involve others in identifying future goals and priorities for the school; (b) establish, clarify, and model norms for the school; (c) give teachers specific expectations and steps regarding use of the innovation; (d) seek ideas, feedback, and reactions from the adopting teachers; (e) anticipate the need for assistance and resources and provide them as needed; (f) maintain direct contact and communication with individual teachers and students; and (g) respond to them with care and concern, coaching and helping them as appropriate. (Hall & Hord, 1987, 234-242).

In this profile of the change initiator, the social aspects of adoption begin to emerge, in contrast with the purely individual aspects that are emphasized in the Rogers model. Hall and Hord also deal extensively with developing supportive organizational arrangements and incentives, tactics and strategies such as working with each individual user and holding ongoing training sessions throughout the course of the implementation effort, and providing "comfort and caring" sessions.

From the work of Hall and Hord, I added thirteen new adoption factors to my list. These are presented in Table 2.6.

Table 2.6

2.3.3 Enhancements by Other Researchers

2.3.3.1 Adoption Analysis

Farquhar and Surry (1997) point out that there is no single, unified theory of diffusion. Rather, there are two schools of thought about technology: determinism and instrumentalism. The determinists see technology as either an uplifting or a destructive, autonomous force that drives social change. The determinists’ basic logic is that technological superiority of the innovation over the status quo is the primary condition necessary for diffusion. The instrumentalists, on the other hand, view technology as a tool, largely under human control, that can be used for either positive or negative purposes. Whereas the determinists see technology as a powerful force for change, the instrumentalists see social conditions as the primary drivers of change. Surry notes that, although Rogers is the founder of general diffusion theory, Hall is one of the primary proponents of adopter-based, instrumentalist theory. This is because Hall seeks to understand the social context in which the innovation will be used and the social function that it will serve.

Farquhar and Surry (1994) developed an Adoption Analysis Tool that can be used to examine the factors that can inhibit or enhance the adoption of a technological innovation. These factors fall into two major categories: (a) individual factors, which comprise all of the skills, attitudes, perceptions, and knowledge possessed by the individual potential adopters; and (b) organizational factors, which are all of the hardware, knowledge, attitudes, and skills that exist within the adopting organization.

Individual factors can be further subdivided into individual user characteristics and individual perceived attributes. Individual user characteristics include (a) motivation, (b) anxiety, (c) knowledge base, (d) prior experience, and (e) skill level. Individual perceived attributes are the same five characteristics of an innovation as described by Rogers, namely, (a) compatibility, (b) complexity, (c) observability, (d) relative advantage, and (e) trialability (Farquhar & Surry, 1994, p. 21).

The breakthrough that Farquhar and Surry made was to consider organizational factors as separate variables that affect the adoption of the innovation within the social system, namely: physical environment factors and support environment factors. In contrast, Hall and Rogers tended to view the adoption process through the eyes of the individual adopters—the people who will be affected by the change.

Like individual factors, Farquhar and Surry subdivided organizational factors into two groups: physical organizational factors and support environment factors. The physical organizational factors at the point of dissemination of the innovation were (a) patterns of use, (b) reasons for use, (c) classroom facilities, (d) student-user characteristics, and (e) administrator characteristics. Factors of particular interest in the SOE are reasons for use and patterns of use. The support environment factors comprised all of the resources and services needed to install and maintain an instructional product, namely, (a) production services, (b) storage and delivery services, (c) dissemination resources, and (d) support resources (Farquhar & Surry, 1994, p. 21). In the context of the SOE, support environment factors include scaffolding, school policies and norms of use (which are associated with cultural compatibility concerns), administrative vision or overarching aim, persuasion and incentive structure, and communication channels.

Thus, Farquhar and Surry shifted the perspective from that of the individual user within an adopting organization to a broader approach that includes both user and organizational perspectives—perspectives that are intimately linked within a comprehensive social system. In so doing, they began to merge the older diffusion theories of Hall and Rogers with those of the organizational learning and systemic change proponents, such as Schein (1995, 1996a, 1997b) and Senge (1990).

2.3.3.2 Bandura: Efficacy x Value

In contrast with Farquhar and Surry’s treatment of both individual and organizational factors, Bandura limited his investigations to individual factors. His research addressed his subjects’ perceptions of self-efficacy, which, in turn, influenced their thought patterns, actions, and emotional arousal. Thus, Bandura linked cognitive, social, and behavioral skills into an efficacy-value theory that built a bridge between the behaviorist and the cognitivist philosophies of learning.

Self-efficacy is defined as "people’s judgments of their capabilities to organize and execute courses of action required to attain designated types of performances" (Bandura, 1986, p. 391). To Bandura, "perceived self-efficacy is concerned with judgments of how well one can execute courses of action required to deal with prospective situations" (Bandura, 1982, p. 122). This is an important component of motivation—one of the individual user characteristics identified by Farquhar and Surry. To Bandura, self-efficacy is a state, not a trait: "self-referent thought is indexed in terms of particularized self-percepts of efficacy that can vary across activities and situational circumstances rather than as a global disposition" (Bandura, 1982, p. 124).

Self-efficacy can be compromised when the individuals feel that they fail to gain expected rewarding outcomes, especially when those outcomes are highly valued (Bandura, 1982, p. 141). It can also be compromised when the individuals feel that they lack personal control over their situation. "Rather than seeking personal control, they seek their security in proxy control—wherein they can exert some influence over those who wield influence and power" (Bandura, 1982, p. 142). The price of proxy control is the restriction of one’s own efficacy, because the individuals then place their security on the competency, availability, and favors of others.

Pintrich and Schunk (1996) describe a similar linkage between individual motives, probability of success, and incentive value in the expectancy-value theory that was previously developed by Atkinson, Tolman, and Lewin. If the individual’s motive for success were high, then he/she would most likely engage in achievement tasks; but if the individual’s motive to avoid failure were high, then he/she would tend to avoid achievement tasks. The expectancy (or probability) of success was measured empirically, using tasks of varying difficulty. The incentive value of success was initially defined as an affective variable, namely pride in accomplishment. It was then generalized to a belief about the attainment value, importance, or interest in a task (Pintrich & Schunk, 1996, pp. 70-74). Achievement behavior was predicted by the product of expectancy and task value, where the expectancy construct could be thought of as "am I able to do this task?" and the value construct could be worded as "why should I do this task?"

In my own research, I was particularly interested in determining participants’ perceived level of self-efficacy because I considered self-efficacy to be the opposite of computer anxiety, as described by Berge (1997). My aim was to find suitable interventions to decrease computer anxiety. One student who responded to the 1995 survey suggested an intervention that could potentially deal with this effect:

Though the motivation theorists, and Bandura in particular, add no new factors to those found in the framework begun by Rogers and Hall, they do underscore the importance of two empirical factors found in our previous research: (a) clear benefit and value and (b) self-efficacy.

2.3.3.3 Dialogue and progressive discourse

Adoption of an innovation takes place within a social system, and social systems share cultural norms and mores. Thus, it is important to explore the role of conversation among members of that social system—not only in the early, information-seeking and personal concerns stages, but also in the later, collaborative phases. With the current emphasis on cognitive apprenticeships in school (Brown, Collins, & Duguid, 1989; Collins, Brown, & Newman, 1989) and on communities of learners (Brown & Campione, 1996; in press), researchers are now expanding the study of cognition and conceptual change beyond the individual mind. This includes learning that is built up by electronic conversations (i.e., CMC) among members of peer groups, local learning communities, communities of practice, and broader cultural systems.

CMC enables individuals not only to share distributed representations (Allen & Otto, 1996; Crook, 1994a) but also to use distributed cognition (Norman, 1993; Fischer, 1995) to overcome the limitations of the individual, unaided, human mind. Pea describes two types of conversation that take place among members of a learning community. These are (a) ritual communication, with its emphasis on participation, sharing, taking part, fellowship, and continuous interaction among members that maintain the social order; and (b) transmission of messages to the learner that takes place orally, via written text, and now, via the Internet. When learners participate in inquiries at the frontiers of knowledge in a field, with mature communities of practice, "they endorse a view of communication for learning that I describe as transformative" (Pea, 1994, p. 298). This results in generative learning and expansion of the ways of knowing.

Scardamalia and Bereiter (1996) studied generative learning students who were electronically linked through Canada’s Knowledge Society network and who shared Computer Supported Intentional Learning Environment (CSILE) databases. These students became actively involved in building and richly linking electronic databases, pointing out discrepant information, contributing new information or ideas, considering ideas from different perspectives, and forming important new working relationships and study groups. Likewise, Newman and his colleagues (1989) observed that Local Area Networking (LAN) technology could be used successfully to coordinate small group investigations by science students. Information shared via the LAN provided a framework that could be utilized in class discussions to synthesize the contributions of these small groups.

Yakimovicz and Murphy (1995) studied the group dynamics of CMC among adult learners who collaborated on projects and participated in discussions as part of a distance education course. Two themes emerged: (a) process management and (b) meaning making. The learning group developed processes that enhanced their ability to work in a new, technology-based medium. They also engaged in meaning making concerning their experiences with problems they had encountered, coupled with reflection on individual and collaborative work toward solving those problems.

Herrmann (1995) carried out a three-year ethnographic study of a 400-member, international group of academics who communicated with each other on listservs. She found that three recurrent patterns of communicative activity emerged: (a) academic, or knowledge-sharing conversations; (b) administrative, or process management conversations; and (c) community-building conversations that included encouragement, warm and playful remarks, and expressions of gratitude. Herrman’s academic and community-building conversations are very similar to Pea’s information transmission and ritual views of communication.

Regarding meaning making conversations, Pea (1993) explained how people use conversational space to collaboratively construct their common ground of experiences, meanings, and understandings. Norms arise from these shared beliefs, which structure the joint activities that are carried out within a sociocultural group. Pea built his theoretical foundation on the work of Vygotsky and the activity theorists such as Leont’ev (1981), Tikhomirov (1981), Engestrom (1996), and Wells (1996). These activity theorists believe that individuals internalize symbolic representations that first take place externally in social relations among the individuals’ peers and colleagues.

Meaning making occurs through successive turns of talk and action. In this two-way transformative communication process, members of the group progressively create, share, negotiate, interpret, and appropriate one another’s symbolic actions. By internalizing these social interactions and processes, they transform their own meaning schemes. When this conversational space is mediated via electronic messaging or conferencing, the communication tool or network becomes an integral part of the system in which the dialogue takes place. A key point in Soviet psychology, attributed to Vygotsky, is the emphasis on the use of tools in the development of human mental processes. "The tool is not simply added on to human activity: rather, it transforms it" (Tikhomirov, 1981, p 270).

Engestrom (1990) expands Vygotsky’s notion to conceptualize human activity as an interdependent system that ties the individual to the larger cultural context: "Collective activity is realized through individual actions, but it is not reducible to a sum total of those actions" (Engestrom, 1996, p. 262). In Engestrom’s conceptual framework, (a) the individual or subject, together with (b) the problem to be solved or object of activity, and (c) the tools or artifacts that mediate the activity, represent three components of an activity system. These three elements cannot be viewed in isolation. Nor is the outcome of the group’s activities simply the sum of individual actions. Rather, these elements are intimately connected with the social system of which the individual is a part, namely: (d) the community of people who are concerned with the same problem; (e) the division of labor (or legitimate activities) among community members; and (f) the norms, rules, or conventions that govern legitimate, sanctioned, or appropriate activities.

In his treatment of activity theory and constructivist learning environments, Jonassen (1998) notes that, if one changes any element in an activity system, the other elements are affected as well, in a "ripple effect". Hewitt and Scardamalia (1997) also state:

Pea considers communication, learning, and activity to be intimately linked, in the same way as Lave and Wenger (1996) consider learning to take place through legitimate peripheral participation in a community of practice. As a result, "the learner’s appropriation of culturally devised ‘tools’ comes about through involvement in culturally organized activities in which the tool plays a role" (Pea, 1993, p. 269). Expertise is developed dynamically through continuing participation in communal discourse, not just through the individual’s possession of a knowledge base and a set of problem-solving skills. This is very similar to Crook’s (1994a) concept of longitudinal continuity—a shared, mediated resource of knowledge, experiences, understandings, beliefs, values, and assumptions—dispersed in time and place, that forms the "glue" that holds a community of learners together.

Likewise, Bereiter (1994) believes that meaning-making and new conceptual structures arise through a dialectic process in which members of a learning community negotiate contradictions and begin to synthesize opposing viewpoints into a more encompassing scheme. This process, called progressive discourse, only works if the members have four commitments. These commitments are: (a) to work toward common understanding satisfactory to all, (b) to frame questions and propositions in ways that allow evidence to be brought to bear on them, (c) to expand the body of collectively valid propositions, and (d) to allow any member’s beliefs to be subjected to criticism if it will advance the discourse.

What I observed within our design team as we collaboratively built the SOE Web Page was exactly what these researchers have described here. Our team used both e-mail and face-to-face meetings for several classes of activities: (a) to carry out a process of progressive discourse, (b) to negotiate meaning, (c) to develop shared meaning schemes, (d) to appropriate meaning from the group using a commonly agreed-upon language and set of definitions, and (e) to build a sense of community and shared vision. We furthered the development each team member’s expertise by sharing a common base of knowledge, tools, and symbolic representations that were explicitly mediated by our collaborative design environment—our class account on the university’s server. Moreover, our team’s efforts to learn new skills, to adopt the new technology, and to participate in a legitimate design activity were fundamentally inseparable activities. It became clear to us that adoption was not a one-time event undertaken by an individual within a social system. Rather, it was an ongoing process that was intimately linked with the processes of dialogue, learning, and polishing skills within a learning community.

Pea, Crook, Bereiter, and the activity theorists not only emphasize the importance of communication and dialogue in the process of learning, developing skills, and adopting a new technology but also add another factor to the growing conceptual framework initially established by Hall and Rogers: namely, shared tools and mediated representations. The use of shared electronic tools and mediated representations supports distributed learning within a learning group. This is a critical aspect of the use of the Internet in the SOE.

2.3.3.4 Situated Cognition

Like the activity theorists, Brown, Collins, and Duguid (1989) emphasize the link between knowing and doing, stating that knowledge cannot be treated as an integral, self-sufficient substance, theoretically independent of the situations in which it is learned and used. Important work in developing a cognitive and situated approach to learning was carried out by Collins, Brown, and Newman (1989) in their seminal paper on cognitive apprenticeship. Cognitive apprenticeship goes beyond the traditional apprenticeship model and situates both learning and activity within a social and functional context. The ideal learning environment consists of four dimensions: (a) the content taught, (b) the teaching methods employed, (c) the sequencing of the learning activities, and (d) the sociology of learning.

There are six teaching methods that characterize cognitive apprenticeship (Collins Brown, & Newman, 1989, pp. 481-483): (a) modeling, (b) coaching, (c) scaffolding, (d) articulation of knowledge, (e) reflection, and (f) independent exploration. These six methods incorporate some of the important facets of cognitive science, especially shared metacognitive or reflective activity, the development of mental models, problem-solving within a domain, and the active cultivation of intermediate states of competence as novices progressively develop expertise.

The Internet Task force concentrated on the first three of these teaching methods when we dealt with novice students, but we began to emphasize the last three methods within the design team as our expertise increased. Hall and Rogers, too, concentrate on the first three methods for scaffolding new users, but they do not treat articulation, reflection and metacognition, or exploration in any detail.

Like Collins, Brown, and Newman, and like the activity theorists, Lave and Wenger place their emphasis on the whole person and view the agent, the activity, and the community of learners as mutually constitutive. Mastery does not strictly reside in the expert, but rather, in the organization of the community of practice of which the master is part (Lave & Wenger, 1996, p. 94). They consider learning to be the development of portable, interactive skills of acting, performing tasks, and solving problems within a given domain. Moreover, like Pea, they do not distinguish between talking about the world and acting in the world—verbal communication supports both. A key concept in their model of legitimate peripheral participation (LPP) is that

Mezirow, an expert on adult learning, shares with Pea the understanding that social reality is shared, sustained, and continuously negotiated through communication. As adults learn and grow, they discover a need to acquire new perspectives so they may gain a more complete understanding of changing events and a higher degree of control over their lives. They learn to negotiate meanings, purposes, and values critically, reflectively, and rationally, rather than passively accepting the social realities that are defined by others (Mezirow, 1991, p. 3).

To engage in transformative learning, individuals must suspend their beliefs and assumptions in order to assess an experience from outside their usual frame of reference. Though this is vital for progressive discourse, it can be a disorienting process. The individual begins by perceiving a cognitive dissonance or dilemma. The process advances toward constructing a more inclusive, integrated way of interpreting reality in which the individual scans and perceives the situation more broadly, and tries to construct new meaning out of it. This is followed by an intuitive insight that leads to new perceptions, followed by release of anxiety, and finally, by the integration of the new solution into the individual’s own worldview.

Whereas Pea simply refers to "transformation", Mezirow distinguishes between two fundamentally different levels of activity, both of which take place in intentional learning communities, namely: (a) transformation of meaning schemes—transforming old assumptions, old interpretations of experience; and (b) perspective transformation—a paradigm shift; a total transformation of world-view that provides greater adaptability to one’s environment and the situation in which one lives and acts. This is also consonant with Lieberman’s (1996) view.

By couching the concepts of learning, developing expertise, and solving problems within a community of practice, the activity theorists, situated cognitionists, and experts in dialogue add three more factors to my growing theoretical framework of learning and adoption. These are presented in Table 2.7.

Table 2.7

2.3.4 Addressing the Gaps in Diffusion Theory

At this point, let us assess the development of my conceptual framework.

1. From my own empirical research, I identified 26 disparate factors that either facilitated or served as barriers to the adoption process.

2. From the initial work of Rogers, Hall, and other researchers, I identified 23 of those same empirical factors in their theory bases and added two more theoretical factors: (a) intentional decision and (b) intentional action, making a total of 25.

3. The enhanced model that I am developing emphasizes three important aspects of diffusion: (a) social and cultural conditions as mediators of activity, leading to adoption and change; (b) dialogue and communication among members of a learning community as a way of negotiating and appropriating meaning; and (c) the transformation of individuals’ meaning schemes and meaning perspectives that are crucial to the decision to adopt an innovation and, eventually, to reaffirm that adoption decision.

4. Three empirical factors are not accounted for in the theoretical framework so far, namely: (a) human-computer interface (HCI); (b) general representational proficiency; and (c) mediated writing proficiency, or "finding a voice and having something to say" in an electronic forum. The adoption theorists or cognitive psychologists do not address these; rather, the CMC experts cover them.

5. The cultural aspects of organizational learning and the systemic aspects of organizational change—two important areas of current research—have not been covered at all. This is a change in perspective, not a search for new variables.

I will now address these gaps in order to come up with a broader framework.

2.3.5 Computer-Mediated Communication (CMC)

Rogers (1995) touches on the adoption of interactive technologies as a specific type of innovation with special characteristics, but does not analyze this issue in detail. To Rogers, an innovation is a single construct, a "thing", a technology to be adopted, that never changes except through re-invention by groups of adopters to suit their particular context. In contrast, the Internet is the epitome of a rapidly evolving, dynamic, interactive technology.

Rogers does mention that the rate of adoption of interactive media such as telecommunications tends to display a certain distinctive quality called the critical mass—the point at which a sufficient number of individuals have adopted an innovation so that the innovation’s further rate of adoption becomes self-sustaining. The presence of an interactive innovation leads to a certain degree of interdependence among the adoption decisions of the members of the social system (Rogers, 1995, p. 313) since it enhances the efficiency of their communication channels. At the same time, computer-related innovations create uncertainty among members of the social system, which, in turn, often leads to resistance to the technology (Rogers, 1995, p. 397). This issue deserves further exploration.

2.3.5.1 Interactive Media

To deal with the various qualities of interactive media, we must return to the definition and nuances of the word "media". Simply put, a medium is an environment that can support an external, symbolic representation of an idea or concept and can, therefore, support communication between individuals. Horton (1994), a corporate training consultant, describes the process thusly. An individual first encodes an idea or internal representation into an image and then represents it in an external medium—in words, pictures, or sounds; on paper, film, or screen; in a static or dynamic medium; in a passive or interactive medium. The external representation—situated with other text, sounds, graphics, and stimuli in the environment, and mixed with various memories, associations, emotions, and inferences—provokes an idea in the recipient’s mind. Since interpretation depends on context, individuals who wish to make good use of mediated environments must develop a skill that I will refer to as general representational proficiency—strategies and techniques for presenting, encoding, and decoding information efficiently. To Horton, this means using good design principles and combining multiple media to reduce the effort required by the recipient to decode and interpret the information being communicated.

Shneiderman remarks, "harnessing the computer’s power is a task for designers who understand the technology and are sensitive to human capacities and needs" (1992, p. 2). Kozma, too, encourages us to forge a relationship between media and learning, between information and processes in the mind and those in the environment. This, in turn, requires us to define media in ways that are compatible with the cognitive and social process by which individuals construct knowledge. "We must design interventions in ways that embed media in these processes" (Kozma, 1994, p. 8). On the one hand, this can lead us into the fields of media selection and message design, which I shall not explore here. On the other hand, it opens the door to an examination of the affordances and constraints of interactive media—especially the Internet.

2.3.5.2 Affordances

An affordance is the action potential of a particular object or item in an individual’s environment—the range of possible uses that the person sees for that item. Researchers (Norman, 1993; Allen & Otto, 1996) have used the idea of affordances to explain situated interactions with a computer interface. Citing Gibson (1979), Allen and Otto extend Gibson’s work on visual perception to their own research on mediated perception:

To Allen, a mediated environment is more than an intervening substance or channel through which signals flow, as in the traditional information-processing theory of learning. It permits an individual to carry out specific activities, namely: (a) to engage in intentional, exploratory action and active perception; (b) to construct external representations of the world that can be manipulated in ways that complement internal representation; and (c) to make an intentional decision to use this information for influencing or controlling that environment. This tight coupling between the individual and the environment has two advantages. First, the individual’s internal cognitive loading is decreased if he/she can offload the work of information storage and processing to the environment itself. Second, it permits mediated representations to be shared among a community of learners.

Allen makes a clear connection between higher order learning and general representational proficiency. When people invest their cognitive resources in offloading representations to the environment instead of storing them as internal representations, this is an efficient process. It allows individuals to store information externally at no direct biological cost. Later, they can expend small amounts of energy to retrieve large amounts of information.

With affordances come constraints—they are inseparable. The biological cost to perceive, acquire, and retrieve information from the environment must be less than the time and energy that the individual might spend in memorizing that same information (Allen & Otto, 1996, p. 202). Thus, one would expect people who consider this biological cost as minimal to leverage relatively small amounts of energy to exploit large amounts of information electronically. In contrast, those who consider the cost to be too high would tend not to spend a great deal of time familiarizing themselves with the operation of a new medium such as hypertext documents or electronic conferences. Berge (1997) and Fishman (1997) have also explored this constraint.

2.3.5.3 Shared, Mediated Environments

Externalizing representations in a mediated environment not only permits individuals to store and retrieve huge amounts of information; it also allows them to share cognitive tools and representations. Moreover, it reduces the transactional distance between members of a community of learners. This opens up a whole new realm of possibilities for participation among participants in learning communities who are engaging in computer-supported collaborative learning (CSCL) or computer-supported collaborative work (CSCW).

CSCL, for example, uses collaborative learning as its model of instruction. Through collaborative learning, students become co-researchers, co-actors, and co-learners in a knowledge community that is characterized by intentional learning, multiple zones of proximal development, legitimization of differences among members, a community of discourse, and a community of practice (Brown, 1994). Alternatively, it can be described as "the mutual engagement of participants in a coordinated effort to solve a problem together" (Koschmann, 1996, p. 13).

Thus, CSCL encompasses many of the current constructivist models of active learning such as problem-based learning (Savery & Duffy, 1995), cognitive flexibility theory (Spiro et al., 1991a; 1991b), intentional learning (Scardamalia & Bereiter, 1989), cognitive apprenticeship (Collins, Brown, & Newman, 1989), legitimate peripheral participation (Lave & Wenger, 1996), and collaborative design (Sherry & Myers, 1998).

CMC tools facilitate mediated communication and distributed learning across time and distance with other networked computer users who have similar tools. Internet connectivity extends this to a potentially global audience. Thus, it creates a form of continuity for participants in an on-line learning community.

Crook (1994a) describes two types of continuity: (a) lateral and (b) longitudinal. Lateral continuity addresses the problem of transfer of learning and refers to generalizing students’ understandings in important ways to new situations. To Crook, CMC supports latitudinal continuity because it enables dispersed students to share representations irrespective of distance, time, or context.

Longitudinal continuity refers to the shared understandings and ongoing dialogue among members of the learning community. This, too, is facilitated electronically by sharing a common network and representational tools that can support a narrative state such as Bereiter’s (1994) concept of progressive discourse. Using a mediated communication system enables the dispersed students to share beliefs, values, and assumptions that arise from their various actions, observations, and learning experiences (Crook, 1994a, p. 107) and to actually create a self-sustaining, on-line culture. The process of creating this longitudinal continuity is an area worthy of further investigation.

2.3.5.4 Constraints

A constraint is a limitation of a system or innovation. As I have mentioned before, it can be technological in nature, or it can depend on the user’s perceptions, expertise, and skills. For example, the PINE e-mail system is technologically constrained in that it does not support the full e-mail capabilities of the Internet, as Eudora and CEO do. Similarly, until March 1998, CEO did not support the creation of Web pages as Carbon and Ouray do.

Another technological constraint of all of the CINS telecommunications systems was that, until March 1998, they did not support the multimedia capabilities of the Internet outside the UCD laboratories unless the individual student, staff member, or faculty member paid a monthly fee to subscribe to a commercial Internet service provider.

Kozma points out the fact that a medium’s technology both enables and constrains the symbol systems it can employ and the processes that can be performed with it. Computers with graphics boards or sound cards can use visual or oral symbols in addition to written text. Computers with sufficient memory to run expert systems can process information in different ways than those without such memory (Kozma, 1991, p. 181). Sophisticated systems like Koschmann’s Collaborative Learning Laboratory (Collaboratory)—a local area network linked to a schoolwide network—permits high school science students to share graphic representations on a common screen as well as accessing the electronic resources of the library and the Internet (Koschmann, Kelson, Feltovich, & Barrows, 1996, p. 106). Interactive multimedia programs, however, make class sessions dependent upon time and place. Their effectiveness, too, depends on the level of general representational proficiency of the users.

User constraints may take various forms such as written communication apprehension and lack of mediated writing proficiency. Text-based communication (as in e-mail messaging or computer conferencing) is perceived by some learners to be more reflective than spoken interaction. "The very act of assembling one’s thoughts and articulating them in writing for a [computer] conference audience appears to involve deeper cognitive processing" (Berge, 1997, p. 10). This, in turn, may lead to written communication apprehension.

Fishman (1997, p. 15) found a significant relationship between written communication apprehension and the use of Usenet newsgroups among students who were using a combination of CMC tools in the Collaboratory. If the network’s human-computer interface (HCI) is not considered to be user-friendly, and if students have concerns about their general representational proficiency or their mediated writing proficiency, these factors could potentially affect the level of use of the network.

When CMC is used in conjunction with distributed learning, "users must practice and become familiar with their system’s hardware, software, and network capabilities, to the point where the technology becomes relatively transparent, before they can focus on course content rather than use of the delivery system" (Berge, 1997, p. 12). Hall and Hord, too, state that in the early, mechanical use stage, the user is primarily attempting to master the tasks necessary to use an innovation, often resulting in disjointed and superficial use (Hall & Hord, p. 84).

There is a delicate balance between affordances and constraints. For example, learning can be more self-paced with asynchronous communication, with students working as much as they want at whatever time is convenient for them. On the other hand, students can procrastinate and eventually drop out. Thus, along with the self-paced learning that is afforded by computer conferencing, "comes the responsibility of the students to be self-motivated in their work habits" (Berge, 1997, p. 8). In summary, the field of CMC provides us with the remaining three factors that I identified empirically. These are presented in Table 2.8.

Table 2.8

2.3.6 Organizational Learning and Change

As I shifted my focus from the individual to the organization, I began to develop a clearer picture of what the "organization" is. Rogers himself stated that "In recent decades I gradually became aware of diffusion systems that did not operate at all like centralized diffusion systems" (Rogers, 1995, p. 364).

The centralized organization, which forms the basis of the Rogers model, is often replaced in actual practice by a decentralized organization in which innovations bubble up from various levels within the system. In these decentralized organizations, "the inventions are done by certain leaders", and "the new ideas spread horizontally via peer networks, with a high degree of re-invention" (Rogers, 1995, p. 365). Information about an innovation in a centralized organization spreads from the originators to the change agents, to the opinion leaders, and finally, to the adopting population. In contrast, innovations in a decentralized diffusion system spread by horizontal networks among near-peers, comprising both local innovators and adopters, in a relatively spontaneous fashion (Rogers, 1995, p. 367).

Interestingly, the SOE appears to be a hybrid of these two organizational types. Moreover, the networked learning communities that are beginning to emerge within the various programs, doctoral laboratories, and cohorts of students who share common research interests can be considered as emerging cultures with norms and conventions that may not be fully institutionalized. Thus, the reasons and types of use of Internet tools may be very different from one group to another

2.3.6.1 Social and Cultural Issues

In dealing with learning and transformation within organizations, Schein (1996a) clearly distinguished between organizational learning (learning by individuals and groups within the organization) and the learning organization (learning by the organization as a total system). To Schein, learning is ultimately a social process that occurs within a community of practice. In direct contrast with Rogers’ linear approach, Schein cautions us to focus on the system dynamics rather than to think in terms of simplistic causal models. Schein also distinguished between adaptive learning, which is the organization’s capacity to maintain itself and grow, and generative learning or transformation, which represents a change in the organization’s sense of identity, core values, and primary ways of working.

Schein treated an organization as a system. The nature of a system is to hold itself together in the face of disintegrative forces operating within its component subsystems. If a fundamental change occurs in one subsystem, this will threaten the equilibrium in all of the neighboring subsystems. As a result, those neighboring subsystems will resist the change in order to defend themselves and preserve their equilibrium (Schein, 1996a, p. 5).

Schein then extended this idea of autonomous subsystems to internal subcultures within the organization. Three main cultures exist within every organization and often work at cross-purposes with one another. These subcultures are: (a) the executive management or administration, (b) the designers and technocrats who drive the core technologies of the organization, and (c) the operator culture or the rest of the population. The basic differences between each of these subsystems are manifested in three levels: (a) the deep tacit assumptions and beliefs that represent the essence of the culture, (b) the level of espoused values to which the group aspires, and (c) the day to day behavior which is a complex compromise between the two. When the subcultures are out of alignment with one another, organizational innovations fail to occur in the first place; or, if they are initiated, they fail to diffuse (Schein, 1996b, p. 1).

The difficulty of communicating across these various subsystems arises from two distinct factors. First, each subculture of individuals shares different goals and assumptions. Second, and more fundamentally—the very meaning of the words and representations that individuals in different groups use will differ (Schein, 1996b, p. 4). The administrative culture tends to manage at a distance and to depend on rules and policies. The designers are interested in abstract solutions and products that have utility, elegance, and efficiency. The members of the operator culture realize that, no matter how clearly the rules and policies of the organization are explicated, they must work together as a team and use their innovative skills, especially when dealing with unanticipated events under varying conditions and contexts.

This misalignment of subcultures explains the lack of diffusion of telecommunications between the teaching staff and the key administrators within the BVIP. Schein emphasized that, for organizations to learn, they have to create some "slack" to allow people to learn new skills and customize the technology for their particular situation (1996a, p. 7). In the BVIP, however, cutbacks in funding and resources forced the school administration to become "lean and mean" and to centralize their support structures. In a similar vein, allocation of funding and resources within the SOE—especially the availability of graduate students who can mentor and coach faculty, staff, and students—is an issue that will become increasingly important as we consider developing interventions to support an ever-growing cadre of Internet users.

Both Rogers and Carlson (1970) recognized that the cohesion among members of various subcultures within a larger organization is strengthened through the use of personal communication channels. Rogers stressed the need for client/change agent empathy, which he defined as "the degree to which an individual can put himself or herself into the role of another person" (Rogers, 1995, p. 342). Change agents who can empathize with their clients tend to be more successful in fostering the spread of innovations among their clients.

Likewise, Fishman (1997) found that there was a strong social influence effect on CMC tool use, especially when teachers were included in the model. The effect was stronger early in the school year, perhaps because students were less familiar with CMC tools at that time, and therefore, might have been more easily influenced by how their teachers and peers used them. This is one of the reasons why Hall and Hord (1987) emphasized people-to-people communication and the building of strong relationships between new users and mentors in the early stages of adoption.

Carlson (1970) found that the rate of adoption of educational innovations within a social group was highly dependent on three factors. These factors were: (a) the personal characteristics of the adopters (amount of education, professionalism, prestige, and opinion leadership), (b) the way in which they were joined to communication channels and sources of information, and (c) their position or status within the social structure. An innovation tended to diffuse faster among individuals who were found to be sociometrically closer to one another in than among sociometric isolates. This may be one reason why the use of the Internet tended to diffuse horizontally rather than vertically within a decentralized organization such as the BVIP.

2.3.6.2 Systems theory and systemic change

Systems theorists and activity theorists do not make a clear distinction between individual and group/cultural factors. As Senge (1990) explained, systems thinking enables us to see the patterns underlying organizational change, particularly (a) reinforcing feedback loops that bring about growth or decline in a system and (b) balancing feedback loops that act as self-correction mechanisms to maintain system equilibrium.

Schein (1995) based his systemic change model on Lewin’s theoretical foundation of unfreezing, changing, and refreezing. Schein found that human change, whether at the individual level (like Mezirow, 1991) or at the group level, involved painful unlearning and relearning while individuals attempted to restructure their thoughts, perceptions, feelings, and attitudes. Unfreezing refers to removing the restraining or balancing loops that are often associated with group norms embedded within the organizational culture. Unfreezing leads to cognitive dissonance or conflict that can be very disorienting to group members as they begin to change. In dealing with such disorientation or disequilibrium, group members must learn how to reframe their thought processes, redefine the words and representations they use to make meaning out of situations, and interpret new concepts more broadly than before. This perspective shift is what Schein refers to as refreezing.

There is a strong connection between learning and comprehension at the individual level and learning and change at the organizational level. To Mezirow (1991) and to Brown and Palincsar (1989), individual learning and transformation begin with conflict—a dilemma concerning a person’s self-concept that then becomes a catalyst for change. Likewise, to Schein, change begins with a dissatisfaction or frustration generated by data that disconfirm an individual’s expectations or hopes.

Such cognitive dissonance gives rise to two kinds of anxiety: (a) survival anxiety—the feeling that if a person does not change he/she will fail to meet his/her goals; and (b) learning anxiety—the feeling that if the person admits to him/herself and others that something is wrong, this may result in a loss of self-esteem. The key to effective change management is "to balance the amount of threat produced by the disconfirming data with enough psychological safety to allow the change target to accept the information, feel the survival anxiety, and become motivated to change" (Schein, 1995, p. 4). Here is where mentoring, coaching, and scaffolding come in—they build confidence, help reduce learning anxiety, and thus create genuine motivation to learn and change.

Havelock and Zlotolow (1996) noted that the bigger the change an organization must undergo, the bigger the forces against it. This is consonant with Senge’s "limits of growth" systems archetype. Havelock and Zlotolow asked such questions as "Is the system capable of change?", "Will the organization allow me to be a change agent?", "For whom are we making these changes?", "Is there a clear benefit and value?", "Are there multiple channels of diffusion?", and "Is there a shared vision?" In their systemic change model, the change agent is not necessarily an agent of the administration. Havelock and Zlotolow did not use a rationalist sequence with information preceding action, followed by evaluation. Instead, they described a reinforcing cycle consisting of seven ideas that change agents need to consider when attempting to change a system:

1. Care. Are the rewards worth it? For whom are they worth it?

2. Relate. Build relationships. Who is the client? How do we relate to them?

3. Examine. Define the problem. Is it solvable?

4. Acquire resources. How do we get help?

5. Try a best-fit solution. How do we put the elements together?

6. Extend the solution. Translate it into action, diffuse it through the system.

7. Renew. How do we keep it going and sustain commitment?

David (1994) found a best-fit solution to the problem of implementing and institutionalizing systemic change. In her model, the system’s administrators (rather than external change agents) explicate the vision and set the goals. Then they provide the flexibility, time, know-how, and assistance to educational organizations to achieve them. For a technological innovation to be used as a powerful learning tool and as a support for systemic reform, she states that the technology must be readily accessible for use as needed and functionally suited to support the tasks for which it was intended. Moreover, the users must have the necessary training, knowledge, and technical support to use the innovation appropriately (David, 1994, p. 142).

Through RMC Research Corporation, I have been involved in the Excellence and Equity through Technology Network (EETNET) technology planning workshops through the Texas STAR Center (EETNET, 1997). These workshops are targeted toward schools that intend to apply for one of the Technology Challenge Grants. They attempt to apply David’s systemic change theory to difficult problems of practice. Workshop facilitators help teachers, administrators, and technology coordinators to develop succinct but comprehensive mission statements and objectives that are customized for each school or district. Subsequent activities emphasize several aspects of a good technology integration plan that draw on David’s ideas. The outcomes of these activities are: (a) effective local planning so that the technology will meet student needs and reflect exemplary practice; (b) a suitable infrastructure for technology with ongoing maintenance and technical support components; (c) availability of data to support effective decision making; and (d) ongoing professional development.

Like Hall and Hord, and like the EETNET research team, David remarked that "how to" workshops have an important role, but professional development and supports must also be available on an as-needed basis. "These range from opportunities to grow professionally through collaborative work with colleagues, participation in previewing and selecting hardware and software, and observation of others’ use of technology to support teaching and learning" (p. 144). David’s process emphasizes iterative, participatory planning and implementation, paying careful attention to situational details rather than working from general principles.

Steigelbauer (1994) and Fullan (1996) shifted the emphasis of change from the management and implementation of a single innovation to developing the system’s capacity for change. Steigelbauer emphasized collaboration among people who are working together to cope with problems and who are furnishing support to one another to make the risk taking more rewarding. "Thinking about change as a learning process opens the door to opportunities to reframe, look at results differently" (1994, p. 38).

Systemic change requires resources in the form of people, money, supplies, facilities, and time to learn and experiment. Change must be effective at the local level, or it will not work at the system level, no matter how good the innovation may be. To Steigelbauer, changing the culture of the organization is the real issue. "Effective change no longer affects one teacher in one classroom, but the very culture of schools" (Steigelbauer, 1994, p. 26). This requires collaborative work to institutionalize the interactions and communication channels that link the different levels and participants in the system, to address both local and global concerns, and to respect all elements of the system for what they can contribute to the change process. As the change process moves from initiation to implementation to institutionalization, "administrative support is vital to change, and policy decisions make and break change efforts" (Steigelbauer, 1994, p. 35).

2.3.6.3 Failure to Adopt Innovations

We can learn a lot about organizational change by examining innovations that failed to be adopted. Teasley (1996) studied the diffusion of educational computing within a school. The availability of hardware and software and the differing levels of computing skills among teachers were important factors in their decisions about computer use. However, she stated that the biggest single factor that contributed to the demise of computer use in the school where she carried out her case study was the lack of administrative vision and support. Despite the fact that the change was mandated from above, it was given neither the support structure nor the organizational arrangements required for its successful implementation. Thus, the conceptions of computers as learning tools by participating teachers changed radically in four years from an integral part of instruction to an expendable add-on (Teasley, 1996, p. 8).

Gross, Giacquinta, and Bernstein (1970) focused on organizational factors from a management point of view. They observed that, even though an educational innovation was initially adopted by the school’s administration, problems arose during the implementation phase. They identified a number of factors that may facilitate successful diffusion of an innovation:

1. A clear vision of the innovation provided by administrators to all teachers, stakeholders, and users;

2. The staff’s skills and capacity to implement the innovation;

3. The availability of required tools and resources;

4. The compatibility of the school’s organizational arrangements (such as the grading system and scheduling of classes) with the new innovation;

5. A set of strategies to deal with the difficulties to which teachers may be exposed as they begin to implement the innovation;

6. A set of mechanisms to identify and cope with unanticipated problems that may emerge during the period of implementation; and

7. The wholehearted support of the administration for the innovation and for the teachers who are implementing it.

Gross and his colleagues’ emphasis on the clear vision of the innovation provided by the administration and their support of it is echoed by Deming, who stated unequivocally that "A system must have an aim. Without an aim, there is no system. The aim of the system must be clear to everyone in the system. The aim must include plans for the future. The aim is a value judgment" (Deming, 1994, p. 50). This is important, because often the vision includes an implicit belief that the innovation is fundamentally good—a "pro-innovation bias", to use Rogers’ (1995) terminology.

Rogers described the consequences of innovations as desirable or undesirable, direct or indirect, and anticipated or unanticipated. Often, change agents are unaware of the ways in which they shape the consequences of an innovation when they introduce it into a culture, especially when they concentrate their efforts on the early adopters and opinion leaders (Rogers, 1995, p. 429). All too often, the consequences of innovations are not studied in detail because of the pro-innovation bias among the change agents, the inappropriateness of using survey research methods rather than case studies for investigating consequences, and the difficulty of measuring certain consequences—especially consequences that bring about deep cultural changes.

Another trap to be aware of is Rogers’ simplistic labeling of the adopting population as innovators, early adopters, early/late majority, and laggards. The term "laggard" has very negative implications. As I have already mentioned, I prefer the term "resisters" to "laggards". Such individuals often have defensible reasons for not jumping on the bandwagon and adopting an innovation.

Though change agents may feel that the resisters are not dutifully following the experts’ recommendations to use an innovation and are considered to be resistant to change, "a more careful analysis may show that the innovation was not as appropriate for later adopters" (Rogers, 1995, p. 117). This insight will be especially important for the SOE, because several of the students and faculty members that Ryder and Wilson (1995) interviewed had defensible reasons for opting not to use e-mail and voiced them quite eloquently. The same may very well be true of certain members of the faculty, staff, and student body who will be interviewed in my current investigation.

2.3.7 Summary of Theoretical Factors

Overall, 28 theoretical factors were gleaned from the review of relevant literature. Table 2.9 summarizes all of these theoretical factors.

Table 2.9

2.4 Relating the Theoretical to the Empirical Base

Now that all of the important factors from both the empirical and theoretical bases have been identified, I can compare Table 2.4 with Table 2.9. This allows me to cluster the lists of factors into several overarching themes or threads, thereby simplifying the overall conceptual framework. This grouping process should also suggest the types of questions to ask and the methodologies to use in carrying out the study, thus serving as a precursor to Chapter Three.

Six clusters of factors emerged from these two lists:

1. User characteristics and perceptions;

2. Cultural and organizational issues, norms of use, legitimate activities;

3. Tools, design, and impersonal supports;

4. Social issues: scaffolding, mentoring, and communication;

5. Individual learning, adoption, and conceptual change; and

6. Group learning, adoption, and conceptual change.

The comparison of the empirical with the theoretical factors, together with the clustering process, is presented in Table 2.10. Though individual conceptual change and group conceptual change are tightly intertwined, for the purposes of this study I will treat them separately.

Figure 2.2 presents these same six themes as a process diagram. This figure depicts the way in which I might expect these clusters of factors to be interrelated within the SOE. One can also think of the process in Figure 2.2 as a loop in which both individual and group conceptual change bring about changes in the cultural norms of the larger system. As Hanks notes in his foreward to Lave and Wenger (1991), "structure is more the variable outcome of action than its invariant precondition" (p. 17).

Figure 2.2

It is interesting to note the striking parallel between Engestrom’s (1996) depiction of an activity system and the process that may be occurring in the SOE regarding the adoption and use of the Internet as the activity that is currently under investigation. In an activity system, the activities in which an individual engages tend to connect six elements, namely: (a) the individual actor, (b) the object of action together with an expected outcome, (c) the tools used to carry out the activity, (d) the community of which the actor is a part, (e) the norms and conventions of use of those tools, and (f) the division of labor that characterizes individual actions within local collective activities. These elements are all interrelated; changing one will invariably affect the rest of system.

The clusters of factors in Figure 2.2, which have been determined both empirically and theoretically, can be loosely identified with the six elements of an activity system. This comparison is presented in Table 2.11.

Table 2.11

The parallel in Table 2.11 is simply too important to ignore. Though I gleaned my six clusters from an exhaustive review of relevant literature and from previous empirical studies that my colleagues and I carried out, these themes roughly replicate the six elements of an activity system. The primary difference between my emergent conceptual framework and activity theory is not in the six elements themselves, but in their arrangement.

A very recent paper by Jonassen and Murphy (1998) presents a simple description of an activity system that can be applied to the design of a constructivist learning environment. Throughout the rest of this study, I will refer to Jonassen and Murphy’s paper as an interpretive framework for subsequent data collection and analysis.