2.3 A Methodology for Framing a Collaborative Research Process
The purpose of this section is to show how the Research Web (RW) can facilitate most aspects of the research process. To do this we must bring some order to the bewildering diversity of research as practiced. The VNS (Validity Network Schema) (Brinberg and McGrath 1985) provides a complete general description of process flow for an idealized research project. It is this ideal research project that we use to demonstrate the utility of the RW.

The VNS, a multi-perspective framework for understanding the inner nature of the research process, is one of the three foundation concepts in the conceptual framework for this dissertation. Below we will examine how the Research Web environment can facilitate the research effort from the VNS perspectives of: stages, or the temporal progress of the research; paths, or the conduct of research leading to a specific product; and domains, the basic epistemic thrust of the researchers.

2.3.1 VNS: An abstraction of an ideal research process  
Research in practice is approached from a bewildering array of models. Each discipline has its own models for reports, and its own set of preferred models, a recipe. Research by recipe takes the researcher's mind away from the meaning of the research and replaces it with a set of tasks. VNS, in its ethereal abstraction, leads the research team to results by method rather than by task performance. VNS is well suited to any size of research team since the conduct of research must suit the behavioral nature of the individual team members. Unless the team is composed of "renaissance scholars," less knowledgable persons must join hands with others to explore each of the several approaches to the research. VNS depends on knowledge building from multiple points of view, and the RW organizes that knowledge.

The ideal research team needs three kinds of knowledge specialists: theoreticians, empiricists, and methodologists. Each of the team members is expected to team with members from the other specialties in order to produce new knowledge, or products. The interdisciplinary nature of many research issues suggests that scholars from disciplines outside the primary discipline can contribute to the knowledge building (Carstensen 2000), (Younglove-Webb et.al. 1999, 426). There are a number of functional roles to be fulfilled in the RW. These roles can last for the lifetime of the RW, a facilitator is always needed to reduce the cognitive load on the researchers; or they might be of short duration, as the need for a very specialized statistician for analysis of experimental results. The team members can move between roles, or researchers may serve on the team for the duration of only a well-defined task. While the RW is not a management scheme, it is an organizational shell flexible enough to adapt to changes in leadership, retirement, loss of interest, and other perturbations.

One role is pervasive: the role of collaborator. Every researcher must be, or become, a collaborator. Colleagues from other disciplines may be asked to participate, as a contributor or collaborator, in information gathering, evaluation of products, or document review. The role of author will persist only for the time needed to produce the document or research report. Authors of research essays may be asked to manage a team of contributing authors, and then be responsible for hosting a collaborative dialog leading to a succession of refining editions of the essay. The role of scientific leader (PI) may rotate, or may be shared. The role of facilitator is tangential to the research goals, but must be filled with a person, or persons, who have an interest in the research topic, and above all a willingness to serve the research team. The facilitator should teach members of the team how to use software that they might find useful. The facilitator should also train his or her replacement well in advance of departure. The role of critic is essential to maintain the quality of the RW. Criticism of essays is the engine of refinement. No document, reference or definition is exempt from criticism.

Due to the size of the RW team and the large volume of information and knowledge brought to bear on the research issue, there is a correspondingly large body of organized knowledge produced. The body of knowledge, the RW repository, can be easily used to develop multiple hypotheses, and thus will ultimately produce a stream of research results until the issue domain is well known. The RW is thus seen as an incubator of issue specific research. From this set of facts springs the necessity of long-term research. If the issue domain is dynamic, the RW might usefully persist for more than a lifetime.

The VNS applies throughout the range of RW applications between the small-scale single paper collaborations and the very large scale confederations of research efforts such as The Cochrane Collaboration (Bero and Rennie 1995), a very large organization devoted to meta-analyses of evidence-based medical studies, and the NCGIA (National Center for Geographic Information and Analysis), an incubator for diverse research efforts. Small-scale collaborations cannot afford the overhead involved in the RW, and confederations suffer from a low level of interdependency, a quality essential to participation (Markus 1987, 494). Note that large-scale research confederations can include a set of independent RWs joined by a Web site that describes the purpose and emergent qualities of the confederation and provides an introduction and index into the Research Webs where the work is done.

2.3.2 The Research Domains  
VNS divides all the knowledge and research effort in the issue domain into three domains: conceptual, substantive, and methodological. Each of these three domains attracts researchers with a preferential point of view, either theoretician, empiricist or methodologist, respectively. Of course no scholar can avoid taking part in all three points of view in the work toward an advanced degree. Still the preference remains, part of the personality that frequently becomes a hallmark of the scholar's reputation.

2.3.2.1 The Substantive Domain  
The substantive domain contains a body of existing knowledge and new research designed to capture the observed nature of the objects and processes of the issue domain, the focal topic of a research project. It is the reality of the issue domain that exists prior to our research (Brinberg and McGrath, 1985, 33) and is the grounding for the theory and methodology, such as EAST2 (Jankowski and Nyerges 2001, Chapter 4), that we might apply to our research. New observations made during the research will also find its way into the substantive domain. The substantive domain cannot simply be a repository of isolated information about the issue domain. It is so large that it needs to be indexed and organized. The organization is the function of the descriptive model. The descriptive model will describe all the objects in the issue domain and all the observed processes.

Empiricists work in the substantive domain. These researchers are primarily interested in discovering and describing the workings of some part of the issue domain. Their work includes the minute detail that others with a more grand view disparage as "stamp collecting," forgetting that grand visions are based on collections of observations. I think of naturalists like Linnaeus and Nikko Tinbergen, and the early astronomers such as Tycho Brahe and Charles Messier who built great catalogs as stereotypical empiricists: the observers, measurers, classifiers and catalogers. Brinberg and McGrath (Brinberg and McGrath 1985, 30) add to that the system specialists and practitioners, perhaps the people that Funtowicz and Ravetz (Funtowicz and Ravetz 1993, 740) call the "extended peer community," and those Donald Schön describes as professionals in his work on reflective practice (Donald Schön 1982). This large group of people would include engineers, planners, nurses, educators, social workers and many other experts.

2.3.2.2 The Conceptual Domain  
The conceptual domain abstracts the objects and processes of the focal topic, as described in the substantive domain. This abstraction characterizes the nature of the observed relationships and attempts to provide a causal explanation of why the issue domain works as it does. Each of those causal relationships, or mechanisms, becomes a hypothesis. The emergent explanations constitute the development of theory.

Theoreticians work in the conceptual domain. Basic science is practiced in this domain. Geographers such as Christaller and Lösch created central place theory in geography, and Wegener, a meteorologist, created the concept of "continental drift" that eventually became the theory of plate tectonics. The conceptual domain acts as a hypothesis mill, always turning up new research questions to examine by experiment. While scholarship is needed and practiced in all three domains, interest in the conceptual domain characterizes most scholars.

Theory can be appropriated as well as developed. A theory, EAST (Enhanced Adaptive Structuration Theory), was developed in 1996 to explain collaborative use of Geographic Information Systems for complex decision support processes (Nyerges and Jankowski 1997). EAST was then appropriated for a larger study in a more comprehensive study of participative GIS (Jankowski and Nyerges 2001), and evolved into EAST2. Whether theory be appropriated or developed, new patterns from the substantive domain need to be incorporated into the conceptual domain. New patterns can lead to confirmation of theory, to extension of theory, or to correction of theory.

2.3.2.3 The Methodological Domain  
The methodological domain contains all information regarding the treatment of the data gathered to support the study of a hypothesis. It will contain information regarding the operationalization of measurements, the measurement protocols, and the techniques to be employed to analyze the data to support or reject hypotheses. The nature of the variables, the operationalization chosen, and their relationships will determine the mode of treatment (Brinberg and McGrath 1985, 38). Very loosely, this can be referred to as the basis of research design (Nyerges, Jankowski and Drew 2002, 10).

Methodologists such as statisticians and experimentalists dominate the methodological domain. In the social sciences, measurement of a studied variable often requires selection of a surrogate that can be directly measured. These decisions, known as operationalization, comprise the auxiliary model (Blalock 1990), a model of the system that is actually measured. The RW documents the auxiliary model by making hypertextual links between descriptions of the operationalizations and the model of the system as described in the descriptive model and the explanatory models. The RW will contain several documents that describe the design of the experiment, the operationalization of variables and the analysis plan. The team members can place these documents in DocReview for local peer review.

The RW may, in mature sites, contain a simulation model of the issue domain. A simulation model can be started when the descriptive and explanatory models are sufficiently complete. The simulation model can be used for validation of the explanatory model. Since a simulation model is hierarchical, completed and validated submodels of the system may be published or placed on the RW's web site to be executed by anyone. A working simulation model will allow the research team to perform sensitivity analyses on the variables of the model. The sensitivity analysis will determine which variables contribute most to the variance of model output. Sensitivity analysis can also be used to evaluate determination of the relevance of mechanisms of submodels, in other words, can determine if certain elements can be removed from the model. The team can run scenarios that can examine unobserved situations and create plans for experiments to refine the model.

2.3.3 The Stages  
VNS suggests that research studies have three stages: stage 1, a prestudy, or proposal phase; stage 2, a central research effort; and stage 3, making the research findings more robust. Stage 1 has well defined milestones to mark its progress: the proposal, receipt of funding, and completion of a research plan. Stage 2 has less conclusive milestones; research papers might be released during and after the central study. Stage 3 involves further generalization, limitation, reduction of uncertainty, or corroboration; it is actually never closed, though activity may be slight and publications few.

In research studies, activity in every stage is potentially perpetual, as information will continue to accrue (stage 1 or 2); study proposals will be made (stage 1) and executed (stage 2) for each line of research that might produce results; and corroborating or generalizing studies may be undertaken at any time to increase the validity and reliability of completed research (stage 3).

As one considers how the work of research is done in the VNS stages, the advantages of the RW will become apparent. One of the greatest advantages of a RW is that due to its size, the diversity of interests of the team members will be high, including methodologists, theoreticians and empirical researchers. Staffing levels may reach the point that there is a semi-permanent cadre of trained research assistants that may serve several research projects. Since a RW is organized around a rich issue domain, there can be a stream of research projects passing through, providing research positions for the diverse team. Once operating under full power, the RW will always provide research of interest to any specialty or research orientation.

The RW concept includes several repositories that can store and organize data, information and knowledge both new and old. Research Web Essays can describe the objects and processes of the issue domain. Annotated HyperBibliographies (AHB) can store references in an interactive annotatable format; and Annotated HyperGlossaries (AHG) can do the same for the vocabulary of the issue domain. Annotatable models make theory accessible to criticism at even the finest level. DocReview is the critical apparatus for most of the documents produced by the research team.

2.3.3.1 Stage 1  
Stage 1 activity in a research project includes all activity from convening a research team to the completion of a research plan. Conducting research from within a Research Web provides enormous advantages. Convening a team is easier since members of the RW team are already interested and involved, perhaps only a few new members need be recruited. Gathering information is easier since there is already a knowledge base in place and the means to add new information is present in the RW infrastructure. Preparation of models of the project's focal topic is easier since a mature RW has models of the issue domain already in place.

The team is convened.
The initiators of the research initiative must first gather the core of the research team. What attracts researchers to a team? First, of course the problem itself, then the other team members, funding, and finally, how the team is to operate. The RW, with its infrastructure in place, can be a force in attracting members, especially members committed to collaboration. If the conveners have a paper outlining the nature of the research issue and the proposed products of the collaboration, then that paper can be presented as a Research Web Essay on RW web site. The invited scholars can mount this position paper in DocReview for annotation. This initial dialog not only informs, but also can be instrumental in setting the stage for the direction of the proposal.

The conveners can put other items on the RW that will inform and perhaps attract other scholars from their networks of scholars: personal web pages, and position papers outlining their personal connection with the proposed research issue. This material will show that the conveners are legitimate, have the power to attract funding and can provide an effective environment for collaboration (Wood and Gray 1991). The personal web pages should include the CV, but should also go beyond that to show how the research issue fits with their ongoing research interests.

Presenting the earliest efforts as a RW can show prospective collaborators that the conveners already have a collaborative environment in place. With a modest effort, all the web site and collaborative software can be put in place and a facilitator recruited. The content will naturally accumulate as conveners and new members present their thoughts in position papers. An Annotated HyperBibliography can be started with the works cited in the convener's position papers and the initial description of the research issue. Similarly, an Annotated HyperGlossary can be populated with the basic vocabulary of the issue domain. The facilitator and graduate assistants can perform these services.

Nothing succeeds like success, and a RW that is up and running is an impressive start. A section of the RW can be devoted to the description of the RW concept and how it will serve the research team. The RW, by its collaborative nature, is a powerful and demonstrable tool and may favorably influence the decision of the granting agency. In the event that the proposal fails, the work remains and can be revised and extended to make a better argument in the future. The RW can also function as a recruiting device to draw in new talent.

Data, information and knowledge are accumulated.
The research team will need to gather all data, information and knowledge relevant to the project's focal topic. While much basic material will naturally accumulate in the RW, what is most important is a plan for acquiring all the needed material. Part of that plan is building repositories for references (AHB), vocabulary (AHG) and organizing the knowledge (RW Essays). The descriptive model accumulates materials in VNS stage 1. Explanatory models may be initiated in stage 1, but will be finally realized late in stage 2. One of the greatest benefits of the RW is that the models, descriptive, explanatory and simulation, apply not only to the research project, but also to the entire issue domain.

Preliminary research hypotheses are developed.
The proposal must present some initial hypothesis in order to argue for funding. The research plan or proposal that the stage 1 team produces can suggest some hypotheses which are certain to be investigated.

The proposal or study plan is written.
The document is the RW's fundamental unit of knowledge representation. A proposal is a formalized document, a genre whose format is specified by the audience. A working copy of the proposal may be mounted on the RW site and made annotatable with DocReview. In that format every member of the team, and administrators from the department or college, may review and annotate the proposal at any time.

Putting the proposal in electronic format suitable for display on the WWW has a number of advantages. If the proposal states that it is available on the WWW, some referees might be inclined to access it on the web, and thus be tempted to explore links that effectively extend the page length restrictions of the proposal. Of course a web presentation can be much more attractive than a black and white laser print. Links to an Annotated HyperBibliography or Annotated HyperGlossary might serve to demonstrate the effective use of modern technology by the team's conveners.

The RW web site is organized.
The RW web site's working area needs to be organized during the stage one. The organization should be patterned after one or more organizing principle. For research issues that have a strong locational principle, maps can be the organizing principle. Research issues that are dominated by a process should be organized around a diagram of the process. Some research issues are strongly organized by time, such as a history; for these organization by timeline is useful. Some research issues are strongly hierarchical as are some organizations, or classifications of organisms. These organizing principles will eventually become the foundations of the models describing and explaining the issue domain. The standardized infrastructure of the RW web site is discussed below in §3.5.2

It is very important to avoid structuring the RW site around artificial organization. The organizing principles must come from the issue domain, not around the team members or their institutions. If the team is multidisciplinary, the purpose of the RW is to bring the disciplinary material together; the disciplinary material should not divide the issue domain. Artificial boundaries divide and encourage parochialism.

2.3.3.2 Stage 2  
Stage 2 is the central stage of the research effort. Knowledge of the issue domain is built by work directed toward three products. These three empirical products are: a set of hypotheses (Brinberg and McGrath 1985, 64), a set of observations (ibid. p66) and a study design (ibid. p62). Hypotheses will, along with the descriptive model, form the basis for the explanatory model. Observations will contribute directly to the descriptive model. Study designs will result in research experiments that will be used to support the research papers, or perhaps to become research papers on their own.

VNS provides several methods to help the research team insure that the research process produces complete coverage of the project's focal topic. In this stage, the principal method is called the matching process. This process is a collaborative process that compares patterns that emerge from three complementary points of view within the team. The patterns come from studying the issue while progressing along three different paths, discussed below. The congruence of patterns between these paths shows validity of the research by correspondence.

Essays and research papers are written.
Essays may be produced on any topic by the authoring teams or by individuals. Elements of the models, either objects or processes, will be described in empirical essays. Elements of the descriptive model may be fully described in essays that will then become targets for criticism by the research team and assembly points for further knowledge about the element. In a similar manner, hypotheses will become part of essays about the elements of the explanatory model. And perhaps when the research process is fairly mature, essays may, by operationalizing the descriptive and explanatory models, display the knowledge about the elements of a simulation model.

As knowledge grows, a set of synthetic essays will emerge. These essays will draw their knowledge base from the empirical essays on the elements of the models. They are the basis of published research papers. It is important to retain a working copy of the essay because it remains the definitive research document. If a research study is developed that shows promise for becoming an experimental protocol that might be repeated with other data, then a methodological essay will be a very useful, and publishable, product.

Modeling continues.
In stage 2 the descriptive model is largely completed. The explanatory model becomes the focus of theory-building for the research team. The simulation model may be started, though it will come into use only in stage three.

The information repositories are populated.
As the research progresses in stage 2, additional research literature will be identified. The vocabulary of the issue domain will emerge with alternative meanings and glosses. Glosses and alternative meanings will be quite common in interdisciplinary research projects. Experimental protocols, data (both raw data and reduced data), and metadata from the experiments need to be stored. Increasingly, granting agencies are insisting on the sharing of data. The National Science Foundation "expects PIs to share with other researchers, at no more than incremental cost and within a reasonable time, the data, samples, physical collections and other supporting materials created or gathered in the course of the work (NSF 1999, Chapter VII.H)."

2.3.3.3 Stage 3  

Rather than spend[ing] valuable (and limited) resources generating new (and equally uncertain) information concerning the same focal problems, we urge researchers to spend more of their future time and effort pursuing robustness analysis as a means for reducing the uncertainty associated with the findings they already have.

Stage 3 is a mature state where the objectives of the team are to extend, generalize, and explore the limits of the issue domain. Reduction of uncertainty of the stage 2 results is the goal of stage 3 research. Uncertainty is reduced by three mechanisms: replication, convergence, and boundary search. Replication demonstrates the reliability of the results using the same methods. Convergence, or triangulation, uses maximally different experimental methods to demonstrate the robustness of the findings. Boundary search, or differentiation attempts to show some of the conditions that will produce results not explained by the theory.

Brinberg and McGrath make a strong case for Stage 3 research (Brinberg and McGrath 1985, Chapter 5), pointing out that most research reports describe stage 2 studies and leave the difficult work of generalization to "future research." This future research is seldom done because one cannot build a reputation on replication of research. Certainly the value of determining the boundaries of theory and the building of robustness must be examined 1 . Boundaries and nuances can be explored indefinitely, but certainly there are practical limits to the exploration. Senior researchers should design stage 3 studies, but can delegate the more routine work of execution. Given the importance of the socialization of students and junior researchers, some stage 3 studies might be assigned as exercises in legitimate peripheral participation. Certainly a well-conceived and executed series of studies that contribute to the robustness of the original published stage 2 findings would be publishable.

The stage 3 research is likely to add valuable detail to the descriptive and explanatory models. This sharpening of the empirical description of the issue domain and the theory of its operation is very important feedback that will be incorporated in the RW Essays as they are refined beyond the snapshot bounds of the published reports.

The simulation model can incorporate the stage 3 findings of boundaries. This information will improve the behavior of the model, perhaps to the point where the model can be used to develop hypotheses to chart the boundaries of the theory. Since studies are expensive, finding the best places to test will save resources. A well-developed simulation model can support uncertainty analysis and sensitivity analysis. As the goal of stage 3 research is the reduction of uncertainty, the simulation model is an essential prerequisite for the analyses that can accomplish uncertainty reduction.

2.3.4 The Study Paths  
A Research Web will have a team that is capable of combining in many research projects, each potentially leading to research papers, as well as the larger collaboration to study the entire issue domain. VNS assumes that scholars prefer to work in one of the three broad areas of empirical description (empiricists), theory construction (theoreticians), or methodological examination (methodologists). VNS suggests that, in order to create the empirical products of the stage, the team engage in collaborations that combine empiricists, theoreticians and methodologists by pairing the interests to proceed along one of three paths, the experimental, the theoretical or the empirical, each leading to a different research product: a study plan, a set of hypotheses, and a set of observations.

As with all facets of the VNS, the flexibility of the schema neither demands nor precludes any form of team configuration, from one member to many. When I speak of a collaborator, I mean a person, or persons, acting in a role. In the social sciences, it is not unusual to have a single person acting in the roles of empiricist, theoretician and methodologist. More commonly, two people will share the three roles, and, of course, several people may share a single role.

2.4 Technology to Support Communication  
Communication technologies fall into two pairs of categories: synchronous or asynchronous, and digital or analog (hardcopy). The research team is free to choose the technologies that suit its members. When choices are made, the communicator needs to consider the ability of the communication to be shared and the ability to be indexed and searched. Most of the older analog technologies are difficult to share (telephone calls, letters, conversations) and generally cannot be easily indexed and searched. Digital communication media are easy to share, copy, and search.

Most computer-mediated communication is asynchronous and digital, hence easy to share and search. Synchronous communication media have a special problem: recording. While analog copies may be made (tapes, CDs, etc.), the recordings are difficult to use and almost impossible to search. Conversations are an important and critical media; those that have content important to the team need to be transcribed to a digital format and summarized for asynchronous use.

2.4.1 Communication Modes  
Communication is the glue that holds together all human social activity. In order to understand communication's importance to collaboration, we must investigate the communication process. The object of communication is to transfer a set of symbols, information, from one human mind to another. The elements of a model of communication contain five elements: the source, the transmitter, the channel (with its noise), the receiver, and the destination (Osgood 1954, 1).

Problems in communication begin with the source. Human sources have limitations in their ability to formulate the communication. Does the person who is the source have knowledge of the symbols necessary to formulate the message? Often not. But is communication is necessary, and the source does the best it can. The human source also may consider its target and compose the message in symbols the target is likely to understand.

Transmission of the formulated symbolic message has a set of problems as well. The set of transmitters available to the source may be limited. If the source is mute, it cannot speak. If a person speaks, the symbols may not be well articulated or heavily accented. If the source is not equipped with suitable technology, it cannot write, or keyboard.

The communication channel may not be available, or it may degrade the message with noise. The Research Web depends on the Internet for its communication channel. The Internet has essentially no noise, but is subject to variable rates of delivery and infrequent interruption of service. For speech, the air in the immediate environment is the communication channel. It is severely degraded with noise in some situations, such as large crowded lecture halls, or at social gatherings. Telephone channels are usually noise free, but are incapable of reproducing high frequency sound, thus persons with voices in higher ranges may be at a disadvantage.

An adequate receiver for the communication channel may not be available at the destination. The RW requires the availability of a WWW-capable workstation, with a display screen capable of rendering high-resolution graphics. FAX transmits only unacceptably low-resolution graphics and requires a special machine. Low-resolution cameras often used for synchronous teleconferencing cannot transmit either text of graphics, only faces or scenes.

The destination has difficulties just as does the source. Disabilities in hearing or in vision limit the ability of the destination from receiving information. Fortunately, the web browsers have the capability of displaying text in a user-selected size consistent with the reader's visual acuity. Language is often a problem, especially with speech. Some people who do not use English as their spoken language much prefer text for communication in English (Sanderson 1996, 107), " ... foreign scientists occasionally had difficulties understanding oral English and considered text to be preferable, ...". They can reflect on the context of the text message, or consult translating dictionaries.

All these communication problems, as well as many psychological and sociological problems, are of great importance when designing and operating a Research Web collaboration. The systemic effects of synchronicity of asynchronicity dominate the utility of methods of communication. This section examines, with respect to the RW, the systemic advantages and disadvantages of both synchronous and asynchronous communication. Technology also has effects on utility, especially capacity and the ability of existing software to realize the potential of the communications methods. There are detailed examinations of the application of the major methods of communication within the RW environment.

Gaines has provided a model of the temporal structure of Internet communication (Gaines, et.al.1997, 1000). There are four times involved in this model: the origination time (t1)-- the time from conception to becoming available to the recipient; the discovery time (t2) -- the time it takes the recipient to discover the message; the response time (t3) -- time taken to read, understand, prepare and send a response; and finally the response discovery time (t4) -- the time it takes the originator to discover the response. Many of the differences between communication modes are made clear by an analysis in Gaines' terms. This model shows the continuum of temporal behavior between synchronous and asynchronous communication and publication. Clearly reduction of the four temporal technological barriers will improve the efficiency of any communication. Such reductions should be implemented in any design or design revision to the tools of discourse.

Figure III Punctuated Communication

T1 is the time that elapses between the time that the originator perceives the need to communicate and the time it takes for a message conveying the originator's thoughts to a place where the recipient(s) may receive it. I decompose t1 into three microcomponents: mental processes of the originator (t1a), including formulation of thoughts, reflection on those formulations, and consideration of alternatives; composition of the message (t1b), including selection of words, assembling the argument, and entering the message into the means of communication; and finally reflection, correction and transmission (t1c). T1a is inaccessible to RW technology, except that the RW makes the entire dialog about the issue domain accessible, perhaps assisting in message formulation. T1b can be improved by the use of alternative message entry. Some members may find voice recognition software useful. Use of a modern mailer program with spell checking will improve the quality of the message. T1c can be improved by selecting the fastest means of message transmission to the pickup point. E-mail can be quite slow compared to WWW page loading, so if the pickup point can be moved to a file that can be accessed by the WWW, T1c can be markedly improved. Many of the RW tools place communications directly into WWW files.

T2 is the time between delivery to the pickup point and the time that the recipient finds it. Bringing the recipient to the pickup point when the message arrives is the problem. Simply waiting for the recipient to call is the slowest method, and may actually prevent receipt of the message. Remember that the e-mail queue is not the only pickup point! Newsgroups are pickup points; the RW has many pickup points for annotations, essentially the web page of the document that is annotated. In the RW, as currently implemented, What's New (see §4.7) is a query facility that points the user to all pickup points for messages that have been issued in a certain time period. DocReview, the Annotated HyperBibliography and the Annotated HyperGlossary attempt to shorten the time to pickup by reminding the recipient of activity by e-mail notification. While this is an improvement over simply casting out the messages to the pickup point, things could be better. We are all familiar with "You Have Mail" and beeps when new mail is put in our inboxes; such instant (and hopefully unobtrusive) notification for all RW activity is a future design goal.

T3 and t4 are simply reiterations of t1 and t2. These temporal components are included in the model simply to illustrate "round trip" dialog. The RW does not often have this sort of conversational dialog; in Gaines' terms the RW is World Wide Web publication. t3 and t4 come into play in synchronous dialog.

2.4.1.1 The Time/Place Collaboration Matrix  
Communication between humans may be classified in a two-by-two matrix based on time and place. The time values are "same time" and "different time", or synchronous and asynchronous. The place values are "same place" and "different place", or collocated and distributed. Collocated collaboration is by far the most common form of collaboration in business and classroom, due to the longstanding emphasis on the group in business and educational organizations. Distributed collaboration is a very common form of scientific collaboration, especially in academia.

The asynchronous distributed category contains the communication media most suited to the Research Web concept. While synchronous methods are deprecated, they certainly are not forbidden -- conversation is far too valuable to be discouraged. Collocation exists on several scales: in this work, collocation is considered face-to-face, not just at the same institution or city. Collocation is seldom universal in large teams, especially in academia.

Table III below is a compilation of classifications of communication media from several sources (Ellis, Gibbs and Rein 1991), (Johansen 1992), (Nyerges 1995, 269), (Hansen, et.al. 1999)

Table III Collaboration Modes in Place and Time

This space/time typology is rather primitive in its treatment of time. Most of the works cited above are discussed in a more technical treatise (Antillanca and Fuller 1999) that considers time in much more detail, using additional terms such as real-time, and concurrent. In their work, they define seven different meanings of synchronous. What is temporally important for our purposes is whether people must communicate information in scheduled session, or may communicate any time they wish.

Asynchronous distributed collaboration is collaboration among persons in different locations and/or at different times (see Table III above). This mode of collaboration is not popular for several reasons: first much of the technology that enables this sort of collaboration is new, hence not well understood or universally trusted; second, the software necessary to facilitate asynchronous collaboration is still primitive because, unsurprisingly, the intensely market-driven software industry is serving the perceived needs of conventional management by automating current methods rather than innovating; and finally, the current accepted model for collaboration is same time, same place - a well known comfortable, conventional, social activity. Each of the four spatio-temporal modes of collaboration have advantages and disadvantages (Ellis, Gibbs and Rein 1991), (Johansen 1992), (Nyerges 1995, 269), these need to be compared and contrasted in order to understand how they might all be brought to support collaboration in research within the issue domain.

This research is directed to the study and development of asynchronous distributed collaboration. My concentration on this mode of collaboration has been influenced by the general neglect of research in asynchronous distributed collaboration in favor of synchronous collocated collaboration, the most familiar mode of collaboration. Business and the academy have honed the tools supporting conventional synchronous collaboration to such a fine point that this form of collaboration has become culturally and economically accepted as the default collaborative behavior.

Realistically, asynchronous distributed collaboration must be facilitated by integration with the Internet, especially the WWW. While the tools employed must be designed to operate on the Internet, the real challenge is to not only facilitate collaboration, but to fully engage the collaborators to such a degree that the tools become genres of communication and information dissemination (Bazerman 1988), (Orlikowsi and Yates 1994), (Agre 1998), (Oravec 1996, Chapter 1), (Orlikowsi and Yates 1998).

Those wishing to categorize the myriad software programs offered to support collaborative applications can utilize a typology that extends the time-place matrix along another dimension: modality (Nickerson 1997). Modality is the combination of communication channels used -- document, audio, and video. This taxonomy is extensible. Nickerson's example includes nine types of collaborative application, but it cannot be universal since software necessarily is targeted toward the functional intentions of the collaborators. DocReview, for instance would be an application type called Document Annotation, a type not found in Nickerson's published typology.

Web-based tools are based on several simple core protocols that have collectively enabled the Internet revolution. These protocols are reliable software commodities that have two major advantages: they are free, and they are stable. Commercial companies are at a competitive disadvantage because they must charge for their products. To create competitive advantages they load minor enhancements on top of the capabilities of the commodities (Glass 2001). They are then able to market the products as "improvements" and at the same time undermine the strength of the protocols. Unfortunately, this marketing strategy works. The results are that there are many relatively inexperienced managers and users that accept the "more is better" philosophy. If these people consider how much training time is required to learn how to use these bloated products, perhaps they would see the advantages of simplicity.

Developers can resist these strategies by keeping their products simple and easy to learn. Complexity may be hidden by linking to more advanced features. The facile may become curious and investigate capabilities beyond the basic set. The tools developed for the RW adhere to the concept of hiding complexity and to the solid simple core of existing protocols.

2.4.1.2 Synchronous modes  
The most amazing thing about synchronous communication is how limiting it is! Every synchronous method revolves about the concept of meeting person to person. Technology has freed us to some extent by allowing us to communicate over distance by telephone and then by television and ultimately, over the Internet. Yet the synchronous methods are, for some purposes, vastly superior to asynchronous means. Fortunately, we understand those purposes well, having used them personally from childhood and, as a society, since speech developed. What are the bases of the frequent superiority of synchronous communication? Rapid negotiation of meaning allows us to plan and learn. Speed of communication allows us to interact rapidly. Ease of communication in face-to-face encounters is undoubtedly the greatest advantage of synchronous communication. For purposes of social bonding, the maintenance of strong ties (Wellman 2001), and ceremony (Trice and Beyer 1984), synchronicity and collocation are essential. Temporary person-to-person contact, such as conferences or even phone calls, is often sufficient to maintain ties. Transfer of tacit or poorly expressed knowledge seems to be best accomplished with the strong social ties (Augier and Vendelø 1999, 255).

Given these advantages, what are the disadvantages? Time, space, power, and permanence. The principal problem is the very practical problem of coming together in time. Scheduling meeting times requires negotiation, unless done by decree. Disruption of personal planning schedules introduces psychological and economic costs. Though being together in time has been made less an onerous task due to the telephone and the Internet, travel is still often necessary to be together in space, be it a walk down the hallway or an intercontinental plane trip. Power effects are present, though often subtle. Synchronous communication always involves turn taking, and turns are controlled by power relationships. Certainly some power effects are positive, especially when decisions must be made. Yet when contributions are sought, power can stifle participation by triggering evaluation anxiety. The recording of synchronous communication is an immense problem. Most synchronous communication is never recorded. Transcriptions of meetings are extremely expensive and often almost impossible to read. Minutes are seldom accurate and never complete. Audio recordings are difficult to understand without personal knowledge of the speakers. Graphic images from whiteboards are almost never recorded in minutes.

The social effects of synchronous communication on team integrity are very important. Team research depends on unrestricted sharing of information. If the team has a geographically isolated individual, it is important for the better-connected members to keep the isolated member informed about any synchronous communications that have occurred within the synchronously connected group. Failure to recognize the vulnerability of the isolated individual is very likely to result in the withdrawal of that member. In a distributed team with multiperson nodes, meetings at each node seem to be inevitable. Failure to share the proceedings of such meetings destroys the integrity of the team. Meetings are the natural enemy of distributed teams -- they should be avoided whenever possible, and recorded and shared whenever they do occur. The socio-psychological phenomena of the outsider and us versus them exist to some extent in all human organizations.

2.4.1.3 Asynchronous modes  
While synchronous communication is based on speech acts, asynchronous communication is based on message passing. The message is recorded with a device on some medium, is transported by some mechanism, and is then displayed by another device, or in the case of physical messages, often read directly. In most communication, the devices are computer workstations, the medium is electronic and the transport mechanism is the Internet. The address is usually a computer with e-mail software.

Messages can be sent at any time to anyone or in e-mail, to a group known at the destination address. The sender must compose the message carefully since there is no possibility of quickly negotiating the meaning of the message (though the workstation might complain about spelling and grammar). The time it takes to transmit the message on the Internet is almost negligible. With postal messaging the transmission time is measured in days to weeks. The recipient may pick up the message in seconds, or leave the message idle for weeks. In the RW some messages are posted on the RW's web site and the recipients are notified by e-mail that a message waits. Notification of "messages waiting" and changes in the intellectual content of the RW is an important issue.

Only information can be transmitted on the Internet. Uses remain for asynchronous methods such as postal services to transmit physical items. Voice mail and phone answering machines are useful for very time-sensitive messages. Legal documents requiring signatures are still bound to physical documents, though that is changing rapidly. Large datasets are still transmitted by CD-ROM and tape, but both those applications will soon be shifted to electronic means as technology improves.

The change from atoms to bits is irrevocable and unstoppable.

2.4.2 Computer-Mediated Communication  
Computer-mediated communication (CMC) is communication that has, regardless of source, been transformed into a potentially permanent digital record. All communication in a RW is CMC. Because the contents of the RW are all CMC recorded in digital format, they are permanent and may be recalled at any time. In addition to being permanent, they may also be modified. Modification must, of course, be managed responsibly.

CMC can be divided into two major categories with enormously different properties: synchronous, which takes place nearly simultaneously; and asynchronous, which is stored on receipt by the computer, and served up to the client later, or on request. Gaines has provided a model of the temporal structure of Internet communication (Gaines, et.al. 1997, 1000). In practice almost all CMC is asynchronous because the inevitable delays caused by transmission, storage and forwarding are disruptive to the smooth flow of communication seen in synchronous communication. At best, synchronous CMC approaches the jerky character of a long-distance phone call that uses satellite links.

The important differences between synchronous and asynchronous CMC lie not so much in their behavior, but in their psychological, sociological and institutional effects. The RW deprecates synchronous communication because such communication usually goes unrecorded. Similarly it has little use for synchronous CMC because of scheduling difficulties and a decidedly anti-democratic power differential.

It is important to note that the RW does not preclude the use of conventional communication media, synchronous or asynchronous -- it simply insists that the members have the discipline to record those communications in digital format. Meeting minutes and memoranda of communication summarizing important conversations have a very important role to play in the RW. If an important finding is shared between two individuals, the proper way to manage such a message is to send a message to the entire team 2 , thus not only informing every member, but also placing the finding in the permanent searchable message archives.

The principal advantages of asynchronous communication are the temporal decoupling of communicants and the permanent recording of every message. The principal disadvantage is the need to carefully compose the message, an act that takes considerable time when compared to speech. Careful composition is essential because the recipient does not have recourse to rapid verification of his understanding of the terms used and the thrust of the message. The advantage lies with the recipients who are not interrupted by the arrival of synchronous communication, and have time to reflect on the message before replying. The "democratic" nature of e-mail burdens the powerful that now may find it necessary to compose a message on the keyboard instead of using the telephone.

2.4.2.1 Dialog  
Dialog is the heart of collaboration. Dialog within a group assumes that there is a common ground, in our context a shared interest in the issue domain of the Research Web. The group attempts to accumulate knowledge by adding to the common ground through contributions. The principal means of adding to the common ground is unilateral action; contributing the right utterance at the right time (Clark and Schaefer 1992).

Since the beginning of human communication, the building of shared understanding has been enabled by dialog. Before writing, conversation and story telling were the means of sharing knowledge. These forms of dialog are synchronous, and continue in use today. The introduction of writing, and its industrial counterpart, printing, introduced asynchronous dialog. When writing began to record knowledge, the first recorded dialogues were marginal notes to religious texts, and recorded dialogs and debates. In modern times, technology has introduced new media for both synchronous and asynchronous dialog.

Modern synchronous methods basically augment speech while asynchronous methods augment writing. Dialog in either of these families of media depends on interactivity. While interactivity in speech is as old as the human race, interactivity with asynchronous methods is relatively new, beginning with message passing and industrialized with the initiation of modern postal services. Synchronous methods of communication have been a constant target of technology and have recently been epitomized by broadband teleconferencing. The introduction of the Internet and services such as the WWW has produced a vast improvement in asynchronous methods, now providing speed that rivals synchronous methods at near-vanishing monetary cost.

We examine now the different natures of dialog in synchronous and asynchronous methods. Natural dialog, or face-to-face speech in pairs or small groups, is characterized by several social conventions: presentation, acceptance, evidence of understanding, and turn taking (Clark and Schaefer 1992, 164). We can differentiate synchronous and asynchronous dialog by taking a close look at each of these conventions.

Presentation is the initial contribution by one of the team members. In speech, this is usually a sentence and seldom more than a paragraph, a set of a few related sentences. In the RW, presentation is usually longer, from an e-mail message of a few sentences to a complete essay of several thousand words. The reason for such a vast range of sizes is that the contributions of the team are submitted en Toto as a complete argument for the consideration of the team members. This mode of contribution is necessary because piecewise acceptance is impossible in asynchronous communication. Also speech is generally conducted in pairs or small groups while asynchronous communication in the RW is always a one or few-to-many transaction. Dialog in very large cognitive chunks can best be effective when the presentation is permanently recorded so working memory is not overwhelmed.

Presentation in speech is impaired by clarity of speech, ambient noise, and poor hearing (Shannon and Weaver 1949). From a theoretical communication standpoint much of speech and written communication share equally in problems of symbol selection and interpretation (Osgood 1954) (encoding, decoding) since words are symbols common to both. Written presentations have a much lower susceptibility to errors of transmission. While speech is aided by inflection and gesture, writing may also be inflected with alternative character fonts or underlining. Since speech is ephemeral, understanding diminishes rapidly with time; writing does not have that problem as the reader can return to reflect on the words. The issue of size of the contribution presented is more difficult. Speech is a serial process, while presentation of a large written contribution is a parallel process that presents several propositions at the same time. Readers can select from these propositions in a parallel manner by screening and then evaluate the selected proposition in a serial manner by reading.

Acceptance is the process of ensuring understanding between communicants. In speech, the turnaround time is very short, so a constant stream of presentations and acceptances can be interchanged efficiently. Among communicants known to each other, gestures often serve effectively as acceptance. Should acceptance be refused, repair transactions will be initiated immediately. In asynchronous communication, gestures are absent and turnaround time is highly variable and often long, so alternation of presentation and acceptance is impossible. Asynchronous collaboration in the RW depends on three mechanisms to monitor acceptance of a contribution: first, implied acceptance; second, the existence of a common language; and finally, responsibility of members to review and annotate contributions.

Acceptance in asynchronous presentation is implied by silence. Note that there is no conditional acceptance by silence; any response, other than praise, is a request to modify or clarify the presentation. Implicit acceptance is, however, not unequivocal because people are usually reluctant to challenge statements unless they are easily refuted. When it is important to be certain of approval, poll the team (Sheard 1995).

The existence of a common language is assumed in both speech and writing; but the RW has the advantage of having an integral glossary of terms. In a series of speech dialogs the meaning of a term can drift, because the context of each meeting is different. The asynchronous dialog in the RW essay has an interactive glossary so the team can have a recorded dialog about any term.

Reviewing presentations is required both in speech and in asynchronous presentation, but this process is subject to many more damaging power effects in speech. Review (as acceptance) requires evaluation, criticism, and sometimes the admission of ignorance. Face-to-face criticism, especially in a group situation is difficult for many people. It is widely held that power effects are diminished in asynchronous communication (Dubrovski, Kiesler, and Sethna 1991,139), (Spears and Lea 1994, 448), (Kurland and Egan 1996), (Postmes, Spears and Lea 1998, 693). In an asynchronous situation, the reviewer has the opportunity to communicate off the record with peers or with the presenter, thus avoiding embarrassment. Review responsibility is implied in speech, but is more explicit in review of asynchronous contributions. Hiding disinterest behind nods of acceptance is common in speech; but its equivalent in asynchronous dialog, silence, is almost an admission of inattention. This is so because in large asynchronous contributions, there is obviously little likelihood of total concurrence.

In speech the commonest form of contribution is turn taking. Each turn is a collective act consisting of two participatory phases: presentation and acceptance. The median length of these contributions is 9 to 13 words (Clark and Schaefer 1992,174). Each turn adds to the common ground. Obviously, turns must be taken in a synchronous mode with very short pauses between phases. Turn taking in this sense is clearly impossible in an asynchronous environment.

So, what is the equivalent to a turn in the asynchronous environment of the Research Web? Another speech act is the assertion, which is an autonomous act, not a participatory act. Assertions, unless rejected, become part of the common ground. Thus in the asynchronous environment, turns consist of assertions (autonomous contributions) in written form, and the frequently lengthy exchanges of commentary (participatory contributions) required to repair the original assertion. In the Research Web, most contributions are understood to be plastic, and each contribution, or work object, has an owner who is responsible for making repairs based on commentary. Contributions in the RW are all autonomous, and vary in size from a complete RW Essay down to a brief comment attached to an assertion.

The apparently different natures of synchronous and asynchronous dialog are resolved by having a common end: adding to the common ground. Synchronous dialog quickly adds to common ground in very small units. Asynchronous dialog adds to the common ground either rapidly or slowly depending on the size of the assertion. Adding a reference to the Annotated HyperBibliography is unlikely to require repair, so adds to the common ground quickly; on the other hand, asynchronous dialog on a large RW Essay may last the life of the RW effort.

2.4.2.2 Criticism  

"... Critiques pose problems and puzzles that have the potential to stimulate pathbreaking research. Critiques suggest novel directions and let newcomers know that there is a place for them in the solution of current problems. Critiques keep researchers from becoming complacent. ... The best way to advance knowledge, it follows, is to foment a constant stream of criticism and response. ..."

Critical annotation of documents is a practice fundamental to scholarly activity. Popper has argued that criticism is the basis of rationality in science (Weimar 1979, 81). Miller, in his book on critical rationalism, suggests that all science be accepted, if falsifiable, until it is proven false (Miller 1994, 6). Justificationism, on the other hand, suggests that the better approach is to justify the hypotheses before the science is accepted. This is the approach taken by the RW, that criticism is a positive force for construction of science from the very beginning of work.

Critical social theory [of Horkheimer, Adorno, Fromm, and Marcuse] states that reason and critique are inseparable, and that researchers using this approach must collaborate with those affected by the research and must open the research to public debate and critical reflection (Ngwenyama 1991, 269). The RW fulfills these requirements and also opens the way to action science through participative design and execution. Action research has been practiced with the RW's tools through the collaborative development of a metadata collection system (Hendricksen 1998d).

Criticism can take substantial forms such as disagreement, clarifications, alternative explanations, and addition of information; or trivial forms such as grammatical or orthographical corrections. In the RW, all knowledge presented is in the form of a document, and the major documents all have a critical apparatus, or means of criticizing. RW Essays and other lengthy documents such as position papers and informative introductions may be presented in DocReview (see §4.3). The Annotated HyperBibliography and Annotated HyperGlossary have built in annotation capabilities for each entry.

The environment that encompasses the Research Web and the team must support the concepts of fairness and competence in order to ensure that criticism can operate properly. The Ideal Speech Situation (see Table IV, below) described by Habermas as modified by Webler fulfills the conditions of fairness and competence required for the RW (Webler 1995, 59). Webler's rules are designed for public participation and are therefore designed for a less elite membership. See the author's notes for application to the RW environment.

Table IV
Criticism Insured by the Competent Ideal Speech Situation

 --- from Webler 1995 (Webler 1995, 57-58)

Author's notes applying the situation to the Research Web environment
(a) Anyone on the research team plus invited guests.
(b) Standards must be relaxed to allow participation of junior members.
(c) A major responsibility. If research is viewed as argumentation, then validity must be established.
(d) Every part of the team RW must be open (exception is authoring team working areas). Tacit knowledge should be made explicit.
(e) See (c).
(f) Members may query any other member to clarify a document.
(g) This condition is met by leaving all documents open to criticism.
(h) Support software should be improved by participatory design.

Each member of the research team has the duty to review and criticize every document that lies within the member's area of interest. The member is expected to be responsible for the accuracy of the commentary, and should provide references to literature where known. Criticism should be attributed to the critic as a matter of responsibility. In asynchronous dialog, presentations are assertions; and criticism of assertions is the participatory mechanism that provides information to repair the presentation. The owner of the presentation is charged with editing the presentation from time to time and representing it as a new assertion. This cycle of presentation, criticism, editing and representation is the engine of refinement of the assertions that constitute the common ground of the research team.

2.5 A Conceptual Framework for Research Collaboration  
The conceptual framework (see Figure IV, below) has three mutually supporting legs; the first is the philosophy of scientific realism, which provides the backing for the use of models in theory building. The second leg is the methodology of the VNS that, with its tripartite domain organization, provides a practical locus for the models proposed by realism. The third leg of the conceptual framework is the Research Web, the interactive environment that provides a home for models that realism proposes, and the support for the activities of the three domains of the VNS methodology. The Research Web, examined in detail in Chapter 4, is the repository for all team documents and the knowledgebase that the team assembles. The Research Web enables several fundamental functions in document management including criticism and successive refinement.

Figure IV Conceptual Framework

Agre, Phil. 1998. Designing genres for new media: Social, economic and political contexts. In CyberSociety 2.0: Revisiting CMC and Community. Editor Steve Jones. Sage.  

Antillanca, Hector B., and David A. Fuller. 1999. Refining temporal criteria to classify collaborative systems. International Journal of Human-Computer Studies vol. 50, no. 1-40.  

Augier, Mie, and M. T. Vendelø. 1999. Networks, cognition and management of tacit knowledge. Journal of Knowledge Management vol. 3, no. 4: 252-61.  

Bazerman, Charles. 1988. Shaping Written Knowledge: The Genre and Activity of the Experimental Article in Science. Madison, WI: University of Wisconsin Press.  

Bero, Lisa, and M. Rennie. 1995. The Cochrane Collaboration: Preparing, Maintaining, and Disseminating Systematic Reviews of the Effects of Health Care. Journal of the American Medical Association vol. 274, no. 24: 1935-38.  

Blalock, H. M. 1990. Auxiliary Measurement Theory Revisited. in Operationalization and Research Strategy. Editors J. J. De Jong-Gierveld J. Hox. 33-48. Amsterdam: Swets and Zeitlinger.  

Brinberg, David, and Joseph E. McGrath. 1985. Validity and the Research Process. Beverly Hills: Sage.                  

Carstensen, Laura L. 2000. Keeping Aging Minds Sharp. The Scientist vol. 14, no. 22: 6.  

Clark, Herbert H., and E. F. Schaefer. 1992. Contributing to Discourse. in Arenas of Language Use. Herbert H. Clark. 144-75. Chicago: University of Chicago Press.      

Dubrovsky, Vitaly J., S. Kiesler, and B.N. Sethna. 1991. The Equalization Phenomenon: Status Effects in Computer-Mediated and Face-to-Face Decision-Making Groups. Human-Computer Interaction vol. 6, no. 119-46.  

Ellis, C. A., S. J. Gibbs, and G. L. Rein. 1991. Groupware: Some Issues and Experience. Communications of the ACM vol. 34, no. 1: 38-59.    

Funtowicz, Silvio O., and Jerome R. Ravetz. 1993. Science for the Post-Normal Age. Futures vol. September: 739-55.  

Gaines, Brian R., L. L-J Chen, and M. L. G. Shaw. 1997. Modeling the Human Factors of Scholarly Communities Supported through the Internet and World Wide Web. Journal of the American Society for Information Science vol. 48, no. 11: 987-1003.    

Glass, Robert L. 2001. Of Model Changeovers, Style, and Fatware. Communication of the ACM vol. 44, no. 9: 17-18.  

Hansen, Tia, L. Dirckinck-Holmfeld, R. Lewis, and J. Rugelj. 1999. Using Telematics for Collaborative Knowledge Construction. In Collaborative Learning: Cognitive and Computational Approaches. Editor Pierre Dillenbourg. 169-96. Oxford: Pergamon.  

Hendricksen, C. S. 1998d. An Exercise in Participatory Design Using Asynchronous Collaborative Methods.  

Jankowski, Piotr, and T. L. Nyerges. 2001. Geographic Information Systems for Group Decision Making. New York: Taylor and Francis.    

Johansen, Robert. 1992. An Introduction to Computer-Augmented Teamwork. In Computer augmented Teamwork: A Guided Tour. Eds. Robert P. Bostrom, Richard T. Watson, and Susan T. Kinney. 5-15. New York: Van Nostrand Reinhold.    

Kurland, Nancy B., and T.D. Egan. 1996. Engendering Democratic Participation via the Net: Access, Voice and Dialogue. The Information Society vol. 12, no. 387-406.  

Markus, M. Lynne. 1987. Toward a "Critical Mass" Theory of Interactive Media: Universal Access, Interdependence and Diffusion. Communication Research vol. 14, no. 5: 491-511.  

Miller, David W. 1994. Critical Rationalism: A Restatement and Defence. Chicago: Open Court.  

NSF. 1999. Grant Proposal Guide (NSF - 00-2). Washington DC: National Science Foundation.  

Negroponte, Nicholas. 1995. Being Digital. NewYork: Knopf.  

Ngwenyama, Ojelanki K. 1991. The Critical Social Theory Approach to Information Systems: Problems and Challenges. In Information Systems Research: Contemporary Approaches and Emergent Traditions. Eds. H. E. Nissen, H. K. Klien, and R. Hirscheim. 267-81. Elsivier.  

Nickerson, Robert C. 1997. A Taxonomy of Collaborative Applications. in Proceedings of the American Information Society 1997 Americas Conference on Information systems. 560-562.  

Nyerges, T. L., P. Jankowski, and C. Drew. 2002. Data-gathering strategies for social-behavioural research about participatory geographical information system use . International Journal of Geographical Information Science vol. 16, no. 1: 1-22.  

Nyerges, Timothy L. 1995. Design Considerations for Group-based GIS: Transportation Improvement Project Decision Making As An Example. in Proceedings of the Geographic Information Systems for Transportation (GIS-T) Symposium., Editors D. David Moyer, and Tom Ries 261-82.    

Nyerges, Timothy L., and Piotr Jankowski. 1997. Enhanced Adaptive Structuration Theory: A Theory of GIS-Supported Collaborative Decision Making. Geographical Systems vol. 4, no. 3: 225-59.  

Oravec, Jo Ann. 1996. Virtual Individuals, Virtual Groups: Human Dimensions of Groupware and Computer Networking. New York: Cambridge University Press.  

Orlikowski, Wanda J., and JoAnne Yates. 1994. Genre Repertoire: The Structuring of Communicative Practices in Organizations. Administrative Science Quarterly vol. 39, no. 541-74.  

Orlikowski, Wanda, and JoAnne Yates. 1998. Genre Systems: Structuring Interaction through Communicative Norms. Center for Coordination Science Working Paper #205, Sloan Working Paper #4030. http://ccs.mit.edu/papers/CCSWP205/.  

Osgood, C. E. 1954. Psycholinguistics: A survey of theory and research problems. Journal of Abnormal and Social Psychology vol. XLIX, no. Morton Prince Memorial Supplement:    

Poole, Marshall Scott. 1994. Breaking the Isolation of Small Group Communication Studies. Communication Studies vol. 45, no. 1: 20-28.  

Postmes, Tom, Russell Spears, and Martin Lea. 1998. Breaching or Building Social Boundaries: SIDE-Effects of Computer-Mediated Communication. Communication Research vol. 25, no. 6: 689-715.  

Sanderson, Duncan. 1996. Cooperative and Collaborative Mediated Research. In Computer Networking and Scholarly Communication in the Twenty-first-century University. Editors T. M. Harrison, and T. Stephen. 95-114. Albany, NY: SUNY Press.  

Schön, Donald A. 1983. The Reflective Practitioner : How professionals Think in Action. New York: Basic Books.  

Shannon, C., and W. Weaver. 1949. The mathematical theory of communication. Urbana, IL: University of Illinois Press.  

Sheard, Sarah A. 1995. Is "silence means agreement" a good ground rule? The Facilitator vol. June: unpaged.  

Spears, Russel, and Martin Lea. 1994. Panacea or Panopticon? The Hidden Power in Computer-Mediated Communication. Communication Research vol. 21, no. 4: 427-59.  

Trice, Harrison M, and Janice M. Beyer. 1984. Studying Organizational Cultures Through Rites and Ceremonials. Academy of Management Review vol. 9, no. 4: 653-69.  

Webler, Thomas. 1995. "Right" Discourse in Citizen Participation: An Evaluative Yardstick. in Fairness and Competence in Citizen Participation. Ortwin Renn, T. Webler, and P. Wiedemann. 35-86. Boston: Kluwer Academic Publishers.    

Weimer, Walter B. 1979. Notes on the Methodology of Scientific Research. Hillsdale, NJ: Lawrence Erlbaum.  

Wellman, Barry. 2001. Physical Place and CyberPlace: The Rise of Networked Individualism. International Journal of Urban and Regional Research vol. 25, no. 2:  

Wood, Donna J., and Barbara Gray. 1991. Toward a Comprehensive Theory of Collaboration. The Journal of Applied Behavioral Science vol. 27, no. 2: 139-62.  

Younglove-Webb, J., B. Gray, C. W. Abdalla, and A. P. Thurow. 1999. The Dynamics of Multidisciplinary Research Teams in Academia. Review of Higher Education vol. 22, no. 4: 425-40.  

Action research -- Research into social processes that involves not only researchers, but also "the researched" and "the researched for", the last in two senses: those with the problem, and those responsible for caring for those with the problem. A cyclical research process deciding on actions to take to solve the problem, implementing the actions, observing the effects, evaluating the effectiveness of the actions, and finally suggest new actions. --- (Charlie Hendricksen)

Asynchronous -- Not coinciding in time. cf. synchronous. In the context of computer-mediated communication, communication by message rather than face-to-face or telephone. --- (Oxford English Dictionary)

Authoring team -- A team charged with the responsibility of researching a subtopic of the issue domain. This researching includes assembling literature, building models, and writing RW Essays. Each member of the authoring team is an author of any publication the team produces. --- (Charlie Hendricksen)

Auxiliary model -- A model designed to test a theoretical model. The assumptions of the auxiliary model operationalize the theoretical model. The incommensurability of the two models are bridged by explanation of the assumptions made in operationalization. --- (Hendricksen - from Blalock 1968 The Measurement Problem: A Gap between the Languages of Theory and Research )

Collaboration -- v. intr. To work together, especially in a joint intellectual effort. In the context of the Research Web, the creation of new shared knowledge. --- (American Heritage Dictionary, Michael Schrage)

Convenor -- A founding member of the Research Web team. Generally convenors are all principal investigators. --- (Charlie Hendricksen)

Critical apparatus -- The means by which scholars criticize the works of others. In the context of Research Webs, DocReview provides the critical apparatus. --- (C. Hendricksen)

Criticism -- The action, process, or result of passing judgement, evaluating or analysing documents. In the Research Web, criticism is taken to be constructive and is considered new scholarship, itself subject to criticism. --- (Charlie Hendricksen)

Descriptive Model -- A model that organizes selected elements of information from the source model. The selected elements are filtered for an adequate truth value, that is, errors are removed and myths are investigated before inclusion, and the elements possess validity in the context of the issue domain. --- (Charlie Hendricksen)

Document -- Any permanent recorded file of information, text, graphic, or sound, usually electronic, that teaches, warns, or serves as an example. --- (C. Hendricksen)

Explanatory model -- A description of a hypothetical generative mechanism that produces the phenomenon seen in the corresponding descriptive model. --- (Hendricksen, from Aronson, Harre and Way)

Facilitator -- In a Research Web, a person who is charged with the duties of creating and maintaining the RW's web site, training team members in the operation of software tools, and actively searching for ways to take unproductive cognitive load from the team members. --- (Charlie Hendricksen)

Issue domain -- The topic for a research effort. The issue domain is comprehensive and defines the scope of the research effort. --- (unattributed)

Legitimate peripheral participation -- A process that provides access to the research activity for students, stakeholders, practitioners and other sub-professionals. LPP provides enculturation, learning, and bonding for the team, and mentoring opportunities for the research scientists. --- (Charlie Hendricksen)

Model -- Real or imagined representations and analogues of naturally occurring entities, structures and processes. --- (Aronson, Harré and Way, 1995, Realism Rescued, p3.)

Operationalization -- The process of creating procedures to measure real properties based on abstractions from a theoretical model. For example, how does one measure argumentativeness? --- (Hendricksen - See Blalock 1968 The Measurement Problem: A Gap between the Languages of Theory and Research )

RW -- An acronym for Research Web --- (C. Hendricksen)

Research Web Essay -- A highly augmented web page that are annotatable and may contain links to new windows containing several kinds of information: bibliographic information, glossary terms, sidebars, marginal notes, ancillary information. These essays are the primary documentation for knowledge synthesized by the team: canonical documents. --- (Charlie Hendricksen)

Research Web -- A WWW site which is the electronic embodiment of the intellectual capital of the network of excellence assembled to investigate a phenomenon. It disseminates information, provides communication facilities, and an infrastructure for collaborative interaction. --- (C. Hendricksen)

Scientific Realism -- ... the common sense (or common science) conception that, subject to a recognition that scientific methods are fallible and that most scientific knowledge is approximate, we are justified in accepting the most secure findings of scientists "at face value." --- (Stanford Encyclopedia of Philosophy)

Scientific coordinator -- A researcher who is delegated to oversee the management of the content of the Research Web. Those duties include responsibility for defining the boundaries of the RW's issue domain, so as to maintain interdependence of the research carried out by the authoring teams. --- (Hendricksen)

Simulation model -- A computer program, usually very complex, that when given a set of initial conditions and a script of actions (scenario), will produce an outcome that a real system would produce given the same scenario. --- (Hendricksen)

Synchronous -- Existing or happening at the same time; coincident in time; belonging to the same period, or occuring at the same moment, of time; contemporary; simultaneous. cf. asynchronous. In the context of computer-mediated communication, communication by personal contact (face-to-face) or telephone. --- (Oxford English Dictionary)

VNS -- An acronym for the Validity Network Schema. See Brinberg, David and McGrath, Joseph E. 1985. Validity and the Research Process. --- (C. Hendricksen)

2. This action is facilitated by use of an interactive tool called the Memorandum of Communication, which is filed and then distributed to all team members.