How is research conducted in a social situation? There are many tensions and conflicting goals to be found in contemporary research. Disharmony exists between the project of traditional research and its priority for understanding and creating generalizations in a phenomenon, and more narrative inquiry into the distinctive and individual aspects of the phenomenon (Eisner, 1993). There have also been two general schools of thought about human nature; one that all is original and native in the individual, and the other that depends upon the influence of the social environment (Dewey, 1930/1922). These are but two examples of the challenges facing a researcher today.
The social sciences originally were received by the university on the premise that they would empirically prove something about human behavior (Novick, 1988). History, anthropology, political science and most other "soft sciences" set about discovering truths about their disciplines with the same spirit and priority of physics. Research was driven by the notion that anything that exists is in some way measurable, to paraphrase Edward Thorndike (Jonçich, 1968). This urge for measurement echoes the more general priorities of human study, all of which are jokingly said to suffer from "physics envy" (Gould, 1981). Following the practice of borrowing the tools of "hard" science has given education a body of research that seeks generalities and proofs over all else, a search that has dominated the field until only recently (Borg & Gall, 1989; Eisner & Peshkin, 1990).
I do not wish to continue in this empirical tradition. Rather, this inquiry sought to study phenomenon in the context of its action, taking inspiration from both early work in understanding the social nature of human action (Dewey, 1930/1922; Vygotsky, 1962), and more recent inquiry into the importance of narrative in understanding human learning and action (Bruner, 1990; Bruner, 1996; Eisner, 1993; Eisner & Peshkin, 1990; Rorty, 1979). I agreed that research into human interaction can quickly become "a hall of mirrors" for those seeking to reduce human behavior to laws (Cronbach, 1975), and rather entered that "messy" arena of human action.
Philosophically, I have been driven by the tenets of pragmatism. As post—modern thought begins to take hold in educational research, the separation of research from practice is increasingly called into question (Doll, 1993), just as the separation of theory and practice has long been critiqued (Dewey, 1938). Evidence used in this inquiry was primarily the message content of the participants, a narrative stream which I believe had most value in seeking the existence of democratic action (Bruner, 1990). I acted as a biased, involved participant in this work, realizing that to act in any other manner would only be self-delusion.
As a result of this work, I wish to provide a report that other educators can access as "intellectual instrumentalities" (Dewey, 1929) and use creatively in their own worlds (Goodman, 1978). In telling the story of the subject matter, I sought to explore an environment to ascertain whether computer conferencing could be a democratic educational environment. This description was based on observation and participation; as a researcher I was involved with the students’ discussions, projects, and difficulties. The result of this study will not be a recipe to follow, but rather a highly biased accounting of what one participant observed in a specific educational setting.
Data were drawn from two sources; student questionnaires filled out prior to the course and transcripts of the participants’ postings to the FirstClass bulletin board. In all cases the data were accurate and literal representations of the subjects’ writing taken directly from the context of class interactions. In the case of the questionnaires, students’ answers were transcribed and recorded in a database for analysis. Posted FirstClass messages were entered in a second database and were also exact copies of the participants’ online contributions.
In selecting representative evidence for this study, all of the class transcripts were studied. From this review, certain subsets of the data were chosen for the purpose of answering the research questions. Early postings and questionnaires were most relevant to questions of students’ attitudes about or toward CMC, while later message postings were used to gauge the mastery of the medium as well as the presence of democratic action. Messages were tagged and categorized according to their relevance to the research questions (see below). The design of this study (see Figure 3.1) is triangulated using these three focus questions. It is the conclusion of this study that evidence of democratic practices was indicated by the participants’ posted messages. These actions were found within the questionnaires and postings contributed to CAVE 641.
The students’ approach to computer conferencing was gauged primarily through their responses to four inquiries on the questionnaires (Appendix A) that were relevant to the study:
Also included in measuring this first question were the actions of students in the initial class meeting, especially those posts made to the practice conference.
Figure 3.1 - Triangulated Design of the Study
This study’s second focus question concerned itself with participants’ ability to master the medium of computer conferencing. The H E L P ! ! ! ! conference was an early source of data regarding the participants’ frustrations with the medium. It was designed as a place for all to ask questions and share problems. Supporting these data were comments made by participants in other conferences regarding difficulties they were having using the software. Evidences of mastery were found in the manner participants used the medium late in the semester. Also, comments regarding the use of computer conferencing posted in the Reflections conference were examined.
Evidence of democratic action is key to this study. I chose to concentrate upon two time periods within the semester in order to find evidences of democracy. The first time period spanned the third and fourth weeks of the semester, a time when small discussion groups were first being formed. Also, participants by this point in the semester had gained enough familiarity with the medium to allow them to concentrate on discussion. The second time period included the eleventh and twelfth weeks. This period was late in the semester and was chosen to capture any differences in the participants’ democratic actions. Week eleven stretched over a two week period, as it fell during the Thanksgiving holiday and students were asked not to post messages during that time.
There were two sources of data for this study. First, all students enrolling in the course completed a two page questionnaire (see Appendix A). This document, created by Drs. Gibson, Hayes and me, was based upon past experiences teaching on-line courses. It sought to measure a number of items, including hopes and goals for the course itself, measurement of computer experience, determination of computer access and availability, and finally any thoughts about taking a course on-line. Several of the questions were open—ended and provided an opportunity for students to reveal the emotions with which they entered this class. While this document was not originally designed to act as a research instrument, it did provide valuable insight into students’ abilities and attitudes.
The second and main source of data for this study were the interactions created by participants in the FirstClass environment. Each message sent by a participant was posted as an individual text message on a centralized computer server at DoIT. These messages were downloaded and then entered into a database that contained fields for the name of the sender of the message, the conference to which it was sent, the time and date of the posting, and the subject and body of the message. Threads were preserved in the database, as FirstClass uses the ordered duplication of message subjects within responses to signify the progression of these asynchronous interactions between participants (see Appendices I and J for examples of threads). Data were analyzed and sorted using the search and tagging capabilities of the database.
A more detailed description of my sources of data and research instrument will follow later in this chapter.
Educational computer conferencing is a relatively new field of in education, and therefore has not developed general methods that may be used to guide research. Most studies that have been conducted in this field have been quantitative in nature, typically using the number and lengths of messages exchanged as a basis for their findings (Burge, 1994; Henri, 1992). This manner of research is perhaps inevitable given the ease with which numerical data may be gathered and analyzed in a computer mediated environment. Numerical data was of some secondary use to this study, but did not guide its main findings. I felt that finding answers to questions regarding the creation of a democratic CMC—based educational environment required analysis of message content and context in a manner that went beyond mere numbers.
Qualitative researchers in the area of CMC have approached often massive amounts of data by taking two or more readings through the content to identify useful categorizations for the purpose of their studies (Burge, 1994; Henri, 1992). I began my analysis by creating three general catagories at the beginning of the study as reflections of the research questions; barriers, equality and responsibility. Within these main catagories patterns of action emerged from the data itself, and were used to create sub—categories (Woods, 1985). From these data I endeavored to create a rich descriptive account of CAVE 641 that will serve to guide further inquiry into the creation of democratic CMC learning environments. Given that CMC is a relatively new field, I have relied upon research methods that allow for the sort of improvisation that maintains an open eye in viewing the occurrences of the class (Oldfather & West, 1994). This flexibility allows the data itself to guide the researcher and allows me to tease out areas of interest and evidence (Borg & Gall, 1989).
I relied primarily upon responses to the initial questionnaire (Appendix A) to determine the attitudes towards computer conferencing held by the students prior to the beginning of the course. Emphasis has been placed upon questions relating to the expectations and fears connected with taking a course using CMC. I sought to measure if students approach this medium with enthusiasm, or with fear and frustration based upon previous experience. Also of interest were questions regarding the amount of computer experience the participant has had, and whether that individual had ever taken part in a course offered through CMC before. Technical proficiency and experience had a great deal to do with the participants early experience in this course. In addition, the postings and experiences at the initiation face—to—face meeting of the course were reported.
Closely related to frustrated expectations was the degree and manner in which the medium itself dominated the interaction. Did the content of the course get lost in the CMC context in which it is held? The transcripts of the on-line communications were the primary source of data for this concern. Here, direct comments upon the medium were tagged in the database as to their general nature and intent. As expected, early in the study there were a great number of comments upon the medium, coupled with a frustration at the clumsiness of using a new technology. Also, the number and types of postings in the H E L P ! ! ! conference directly reflected the experience participants are having with the CMC technology. The importance of participants developing a fluency in CMC has been noted (Gunawardena, 1992); did a time come when the CMC environment shifts from a focal to subsidiary role in the conduct of the course (Polanyi, 1974/1958)? This was determined by the absence of direct references to CMC and the dominance of course—content in participants’ discussions.
Since the primary focus of this study was to find evidence of democratic action in the CAVE 641 CMC environment, I was seeking to determine if and how the re—allocation of the classroom teacher’s authority took place (Apple, 1993; Dewey, 1916; Gutmann, 1987; Warehime, 1993). Did students wait to be told what to do, or did the conduct of the class become a shared responsibility among all participants? Were the teachers’ postings more numerous than the students’, and did their words seem to carry more importance? A democratic learning environment also should have been marked by student initiative in the selection and interpretation of what was to be learned, and this quality was also gauged. Overall, I sought clues to the creation of a community that exists on-line; self—regulating, inviting, and with equal opportunities for all members.
Some work has begun in developing a theory for interactions in CMC. Henri (Henri, 1992) proposed that to fully understand the workings of a CMC class one must go beyond analysis of quantitative data and examine the content of the messages. Her general categories for interacting with message data are broken into five general areas; participative, social, interactive, cognitive and metacognitive. I used Henri’s first three categories as models from which to generate codings that may reflect content qualities that are in some way democratic.
I developed three general categories in order to answer questions regarding mastery of the medium and democratic action (see Figure 3.2). Barriers concerned those messages that indicated some difficulty with the medium itself, ranging from technical difficulties to problems of isolation and message overload. Overcoming these barriers would be key to a participants’ ability to interact within CAVE 641.
Equality and Responsibility were concerned with the main purpose of this study, democratic action. Equality captured instances of actions that indicated a balance of power within the computer conferenced classroom. Democracy in America is broadly conceived as consensual rule by equals (Wiebe, 1995). In a democratic educational environment, there exists a like need for equality, especially in the areas of access and power (Gutmann, 1987). Responsibility recorded occurances of participant initiative in controling the learning environment. Equality is not sufficient ; responsible actions were needed by all in CAVE 641 in order to create a democratic environment (Hickman, 1990).
|
Dimension |
Indicators |
|
|
Barriers |
Concerns with problems of hardware/software. |
Difficulties using conversational features of the software. Difficulty logging onto the discussion system. Answering questions by starting new threads. |
|
Equity |
Action that shows the traditional cues and power relations of the classroom are not present. |
Social interactions Solicitations of responses. Discussion of topic |
|
Responsibility |
Action that may be construed as taking responsibility for one’s learning. |
"I would like to discuss..." Raising questions. Reporting absences |
Figure 3.2 — Categories for Analysis
Sub—categories of Analysis
Within each of the three categories there became evident a series of examples that relate to that category. These sub—categories served as instances of the major category.
Dangling thread This was marked when a message was intended as an answer to a previous post but the threading feature was not used. Dangling threads may break up interactions and make them more difficult for other readers to follow. These postings may be evidence of a lack of understanding or skill in the medium.
Report of difficulty This identified comments of software/hardware problems that have made participation more difficult.
Login problems Specifically looked for reports of difficulty or inability for the participant to connect to FirstClass through modem connections with the DoIT server.
Overload of Messages This identified those instances where students felt the volume of communication on-line had become too great.
Addiction Some students reacted to the twenty four hour classroom by logging on an excessive number of times per week.
Miss Face—to—face Reports the instances when students complained about the lack of human contact in the class. In CMC the shorthand for this personal contact is F2F.
Topic discussion Most messages in a discussion conference had this quality. This showed that a post contained some part that was concerned with the topic at hand.
Collaboration This denoted exchanges that showed evidence of participant action to change or solve some part of the process or discussion under way. Cooperation that solved the problem must be evidenced for this category to be used.
Process reflection Here students reflected upon their experiences with CMC as a medium of discussion. Reports of strategies for interacting with on-line information were found in this category.
Participation by teacher The teachers took part in the discussions as peers. Here they merely added to points being made by other participants rather than giving directions or answering procedural questions.
Community Building Acting in a manner that built a sense of shared action and responsibility.
Raising a question The participant went beyond answering questions and began to expand the discussion by raising questions to the group.
Reporting an absence Realizing one's value to a group led to explaining one's absence from discussion. This informed other participants to not expect the absent person's comments for a period of time.
Personal Exchange A personal exchange is discussion that was directed specifically at another person in the conference. Generally information was being solicited or the posting was an answer to the named person's post. It should be noted that there was a difference here between stating that "Phil said..." and "Phil, you said..." The former was not personal exchange, the latter was.
Social Interaction This was an area where students interacted informally with each other, and began to show the development of community. Informal comments, jokes, and light restatements of points were be examples of social interaction.
Solicitation of Response Awareness and craving of interaction were found in comments that solicited response. These were specifically calling for other group members to comment upon some point the speaker has just made rather than raising more general questions, which are measured by the "Raising a Question" category. It also showed some insecurity with the isolated quality of the medium.
Self Direction The participant branched from the material under discussion and began to explore areas of personal interest.
Initiating a thread Here a person has identified a discussion topic and taken the initiative to begin conversation about it. Only messages that actually began threads were counted here; those who did not properly respond to previous discussions are dealt with in the Barriers section.
Summarizing Discussion This is generally a teacher's function in face-to-face learning environments. Students performing this function were showing a willingness to take on that responsibility (or that the teacher has ceded that responsibility to them).
Asking teacher for Help This would tend to underscore the authority of the teacher in the classroom, and may have served to provide instances where students were failing to act independently.
Directions from teacher These were postings that were placed in the discussion by the instructors as a guide to the participants work. Discussion questions, directions on how to participate and other such "housekeeping" were a part of this category.
These categories were used as a way to organize and tease out the data in this study, rather than as numerical proofs in themselves. Again, it is the words of the participants themselves that were of most interest to me.
Data for this study were collected in two ways. The initial questionnaire was developed by Dr. Gibson, Dr. Hayes and myself to assess the profiles of students entering the class (Appendix A). These questionnaires were either mailed to each pre—enrolled student prior to the class orientation meeting, or handed out at that meeting. I then entered each form was into a Filemaker Pro database of my own design. Care was given to entering the responses as faithfully as possible.
The main body of data was found in the content of the messages posted by students. Though all messages were stored on the DoIT server throughout the semester, they were not going to be permanently available beyond the end of class. The challenge then was to find a way in which to download and format the class’ messages in a manner that preserved all the information associated with each piece of data. Each message contained a ‘header’ that held the sender’s name, conference to which the message was sent, date and time of the posting, and the subject, where it would be possible to follow the various threads of the discussion. The greatest problem in accomplishing this was caused by FirstClass itself; this software is designed to have messages read while on-line, not downloaded for later reading. While it is possible to save and print out individual messages, the process is too cumbersome to be suited for the large amount of data generated by CAVE 641. I decided a Filemaker Pro database would best facilitate the organization and analysis of the messages, but the challenge of getting messages to this environment remained.
Inquiry in various Internet sites led me to purchase and use the shareware program Bulkrate, which was designed specifically to act as an off-line reader for FirstClass. This software logs directly onto the FirstClass server and, once set up properly, identifies and downloads all unread messages. The one difficulty with this process is that Bulkrate only works with a direct modem connection to the FirstClass server, and our access was through a TCP/IP Internet connection. Without going into technical details that I frankly do not understand, let it basically be said that Bulkrate did not know the ‘language’ of the Internet, and therefore cannot work with it. Conversations with the programmer of Bulkrate did not resolve this problem, as he reported frankly that he did not have the skill or time to modify the program to work with an Internet connection. Discussions with DoIT led to the installation of a single modem and phone number onto the FirstClass server to facilitate my collection of data. Through this connection I was able to establish the type of connection that enabled Bulkrate to operate.
The result of a Bulkrate session was a single text file containing, typically, over a hundred messages from various conferences. In order to enter these data into the database each individual message needed to be separated into an individual file and the data formatted in a way that would allow it to be imported into a database file. Further conversation with Bulkrate’s programmer led me to understand how the text file was formatted, and allowed me to develop a HyperCard stack that prepared each message for database use. Once formatted, the stack was programmed to use Applescript, an inter-application language, to directly transfer the data to the Filemaker Pro database (see Figure 3.3).
Figure 3.3 — Flow of Data Capture
It is something of a joke among computer programmers that it takes all day to create a procedure that will save five minutes. Certainly the process above took a great deal of time and effort to accomplish. However, once in place it also provided a seamless manner in which to reliably collect data with complete integrity. It was greatly preferred to the use of a memo—transcriber typical for most ethnographic studies.
This study was conducted in conjunction with a course offered by the Department of Continuing and Vocational Education (CAVE) at the University of Wisconsin—Madison the Fall semester of 1995. "The Adult Learner," CAVE 641, is part of the required curriculum for graduate students in this school. The class syllabus describes CAVE 641 in this manner:
This course is designed as an introductory overview of topics important for professionals concerned with facilitating adult learning. It deals with concepts and information that provide a foundation for understanding the adult as a learner in American society. As time, interests, and talent permit, we will also examine related issues from international perspectives. The focus of the course will be on the application of theory and research to adult learning situations (Appendix B).
As is indicated by this description, CAVE 641 is a class that is not directly concerned with instructional technology or computer conferencing. As such, the content of the course in no way interacted with my research questions. Rather, I was interested in the actions of the participants within the FirstClass bulletin board. The only effects that the course content may have had was in regards to those who decided to take the class.
The use of computer conferencing as a delivery medium had more to do with the mission of the CAVE itself. Students in this discipline typically enter fields of education that go beyond traditional classroom teaching, with a large number being involved with some form of distance education. Thus, computer conferencing was selected as an added experience for the class, a chance to work with an educational methodology many students would later use in practice.
Participants in this study consisted of nineteen students who completed the entire semester enrolled in CAVE 641, two professors, and myself. The students in the course consisted of eight Ph.D students, ten Masters students and one undergraduate. Sixteen students reported CAVE as their major area of emphsis, with the balance consisting of one Ph.D student in Nursing, one Ph.D student in English, and one undergraduate in Family and Consumer Education. There were eleven females and eight males amongst the students. Ages were not recorded.
CAVE 641 was co—taught by two professors, Dr. Elisabeth Hayes and Dr. Chere Gibson. Dr. Hayes has taught this course in a traditional classroom format for a number of years and was primarily responsible for the content of the course. This was the first course she taught using CMC. Dr. Gibson brought experience in CMC to the team. She taught the first course offered entirely on-line at UW—Madison, CAVE 643, The Adult Independent Learner, a class I participated in and from which I developed my enthusiasm for educational environments using CMC. This class used a LISTSERV as the sole method of class meeting and discussion, and was linked to classes at a number of other campuses across the country. Since this time, Dr. Gibson has offered several other CMC courses, including participation in the initial trial of the FirstClass software that was used for this study.
My duties with this course were to both act as a technical support resource for the students and faculty of CAVE 641 and to facilitate the collection of data from this course for various research projects. Prior to the semester I assisted Dr. Hayes and Dr. Gibson in creating the various conferences and folders that were a part of the on-line environment. At the initial class meeting I assisted personnel from DoIT in training the students in the installation and operation of FirstClass. After this first meeting I spent the majority of my time assisting students with the inevitable problems associated with installing the software on their home computers and connecting via modem to the UW-Madison computer server. Also, I monitored the H E L P ! ! ! Conference closely for questions from students having difficulties with specific actions in FirstClass, as well as those wishing to improve their skills. As the course progressed and students become more competent with the software, this support role diminished in importance. At that time I was able to collect data from the class and participate in discussions. Each student’s posting was entered individually in a database for later analysis both for the purpose of this study and for other researchers who are also involved with this work.
The computer conferencing software that was used for this class is part of an ever—expanding collection of Internet access software that is made available to the students and staff at UW—Wisconsin by its Department of Information Technology (DoIT). FirstClass, by SoftArc, Inc., had been chosen by DoIT as a response to an increasing need for conferencing capabilities in support of education. In the Spring semester of 1995 four classes used FirstClass as part of a pilot study to assess its viability as a teaching aid. The success of this pilot let DoIT to offer access and support for FirstClass to the entire university community.
FirstClass contains a number of features that made it useful for the purpose of facilitating a computer—mediated class. The class environment is entered through each student’s regular electronic mail connection using a second user ID and password. This allowed students access from any Internet connection in the country, a feature of importance to the seven members of the class who lived a significant distance outside of the city of Madison. Once logged in, students entered a conference area, or folder, that was dedicated to CAVE 641 students. Only students enrolled in this class could enter this conference; it was invisible to other users of the system. FirstClass operated by storing all messages and files upon the host computer, which were then accessed by personal computers either through modem or network connections. Participants needed not download any messages to read them; indeed, FirstClass was designed to have each user remain connected to the host server throughout a session of reading and writing contributions to the class. This had the advantage of saving storage disk space on each users’ computer, as each message did not need to be stored by each participant. A possible disadvantage may have arisen if a student living a distance from Madison had to make a long distance telephone call into the host computer and maintain a lengthy connection to participate in the class. While having a local Internet connection would address this problem, such resources were not available to all students in this study.
Figure 3.4 — Main Conference Area in FirstClass
Perhaps the greatest attraction of FirstClass lies in its ability to group discussions into topics of interest. Prior experiments with other software packages resulted in all discussion being mixed together, resulting in difficulties with keeping the flow of conversation intact. FirstClass allowed the creation of conferences, which acted in much the same way as folders on a Macintosh or Windows desktop. A number of conferences were created and given titles to signify what sort of activities are to take place within them. Each conference window was divided into two sections; the top half was an area for further sub—conferences and folders while the bottom half was a sequential list of messages that had been posted (see Figure 3.4). Conversation took place through the posting of these messages; each message consisted of a subject or title, the name of the sender, the date and time of its sending, and the body of the message. Participants could engage in direct interaction with a posted message through the use of the "Reply" menu command. Replies were coded with a subject line that contains "re:" and the subject line of the original message (see figure 3.5). These sequences of responses, or threads, then grew as other participants add to them, which numbers added to longer thread of messages, i.e. "re(2):<message title>,re(3):<message title>, etc." FirstClass provided commands that allowed a participant to read these threads in sequence by starting with the original message and then opening each response in the order of its posting.
Another feature of FirstClass was the availability of synchronous conversation. These "public chat" areas were signified by a special icon that was placed on—screen next to the conference folders. Participants logged on to the system could join these chats at anytime. Upon entering the chat area a participant communicated to others by typing a message on screen. This message was then sent by the writer and immediately posted on the chat screen for all participants to see; responses could be made in a similar manner. FirstClass also provided the ability to begin "private chats" through the selection of a menu item. These were initiated by a member who then invited others to join; the only way to take part or
Figure 3.5 — Conference Message List with Threads
even be aware of private chats was through the invitation of one those involved in the chat.
FirstClass also allowed the posting of computer files. These could be of any format, from word processor files to spreadsheets. Given the compatibility problems between different computer brands and software packages, it was decided for the purpose of this class that only files saved in ASCII text format would be used. It was hoped this would enable all participants to read each others files easily. Students uploaded various assignment papers as text files. Also, FirstClass allowed for the editing of created text documents on-line. The class experimented with this feature to create summaries of discussions and perhaps collaborative documents.
There were several other features provided by FirstClass that were mostly ignored in the conduct of this class. Internal electronic mail was available, but access was limited to those logged into the FirstClass environment and its utility was limited. FirstClass allowed for the formatting of text within messages, giving such standard features as bolding and underlining. Also, text color could be altered. Formatting was ignored, as past experience had shown that the formatting did not display across computer platforms. Finally, users were provided with the possibility of creating a resumé that could be attached to their user name. We saw little application for such a feature, though some participants created one.
Given the complexity of FirstClass the availability of computer support was viewed as vital. DoIT provided twenty—four hour technical support via telephone, staffed with trained personnel to assist participants in the installation and operation of FirstClass.
After an initial classroom meeting, this study was conducted solely within FirstClass’ computer conferencing environment. Students met Saturday, September 9, 1995 for a five and one half hour introductory session to the course. The first part of the class meeting was concerned with the typical paperwork and introductions that take place at the outset of any graduate course. The final two hours consisted of a training session on the software that would be used to facilitate the class. After this first meeting, CAVE 641 did not meet face-to-face again as a class. All work took place on-line using Internet connections to access the FirstClass server at DoIT.
The organization of the CAVE 641 conference has been put in place by the instructors and me prior to the first class meeting. The conference was organized with the following major sub—conferences: