(written summer semester 1996)
Findings Communication Cybercommunity Recommendations
By Eric Wiltse University of Wyoming (Author's note: All names have been changed to ensure the anonymity of participants in the pilot project.) INTRODUCTION In the computer lab, 20 graduate students perch before their Gateway 2000s, staring intently at screens and pecking at keyboards like a flock of feeding flamingoes. Occasionally, one will cackle out loud and speak to a person in the next chair or even across the room. A few students migrate from the computers back to a large table in the center of the room, reading or talking with April, the professor. Back at the computers, the rest of the students are reading messages on their screens and typing responses. Dick turns from the center table where he's been reading and glances at the computer screen behind him. "Oh, look at all the people," he says. I look to see to whom he's speaking and to which people he's referring. But Dick is looking at and speaking to his computer screen, which is full of messages written by other students in the classroom. He swings his office chair around 180 degrees and starts typing on the computer, which he has ignored for the last 15 minutes or so. Earlier, I had asked Dick why he wasn't yet posting messages on the lab's network. "I don't want to start too soon since there isn't anybody here to converse with," he replied, even though there were a half dozen other people in the room. Dick and the other grad students were conversing with each other on the Daedalus Integrated Writing Environment's Interchange (DIWE), a program that allows classroom discussions to take place without speaking. The conversations are not verbal but text-based, and the writing is synchronous in that students can read others' messages at the same time they are writing their own. These conversations are real-time, meaning that messages appear on everyone's screens nearly as soon as they are written. I was observing an eight-week, summer class of master's and doctoral students attending a university in the Rocky Mountains to see how DIWE was being used by students to learn qualitative research. Eventually, though, I found more subtle, profound changes in the way class members interacted and communicated not only with each other, but with the computers. For instance, people who didn't know each other by name or face would converse casually on DIWE, even cracking jokes as though they were long-time friends. Other people, such as Dick, personified their machines, giving the computers human characteristics, if not a human identity. Rather than only evaluating DIWE's effectiveness in teaching qualitative research, I also approached the pilot project from a constructivistic perspective. I planned to examine how the DIWE technology could help learners build their own knowledge about qualitative research. This seemed more appropriate for qualitative research than trying to determine simply whether DIWE worked or not. The answer likely would be "That depends on how you look at it" (Bogdan & Biklen, 1992, p. 207). However, my evaluation of how students used DIWE to learn about qualitative research also revealed information about DIWE's effectiveness as an instructional technology. The learning that took place on DIWE and during subsequent verbal class discussions was constructivistic in that the construction of knowledge took place both individually and collectively. Constructivism also involves a community that provides the setting for building knowledge, poses challenges to foster the building, and offers support that will encourage the construction (Kent, 1995). For instance, on DIWE each student would build his or her thoughts individually by writing about them. Then the class, taking the role of a community, would collectively examine those writings and expand upon them, offering other ideas and experiences, and further building the knowledge base of not just the individual, but the others in the class. However, as I observed the class, talked to classmates, and read the literature, I realized that constructivism couldn't adequately explain everything that was going on in the lab. Postmodernism seemed like a more encompassing paradigm to explain the events taking place between the people and the machines in the computer lab. Students would talk at (and often swear at) their machines. They would grant the machines and the DIWE software human characteristics, such as the ability to converse. More intriguing, though, was how students engaged in DIWE conversation would switch with ease from keyboard to verbal communication with classmates, typing or reading a message one second and then turning to say something verbally the next second. Some of these characteristics of the class echo some of the postmodernist research taking place involving human-computer relationships. "Granting a psychology to computers can mean that objects in the category 'machine,' like objects in the categories 'people' and 'pets,' are fitting partners for dialog and relationship," Turkle (1996, p. 150) noted. The manner in which students smoothly shifted from textual to verbal communication underlines the postmodernist concept that the distinctions between what is real and what is virtual are being blurred, if not erased. To the qualitative research students, there was little or no difference between the virtual conversation taking place on their computer screens and the simultaneous verbal talking in the classroom. Their adaptability and flexibility, demonstrated by the textual-verbal shifts during the DIWE sessions, are among the characteristics of postmodernist behavior (McCorduck, 1996). METHODOLOGY I investigated this pilot project from June 11, 1996, until July 16, 1996. My investigation consisted of interviews, observations, photographs and documents. The literature notes that most qualitative studies use a variety of data sources because they are the "evidence and clues" that the researcher will analyze (Bodgan & Biklen, 1992, p. 106). I interviewed five people as part of the investigation: April, an associate professor who taught the class, interviewed in her office July 3 from 2:35-3:45 p.m.; Joan, a director of a university writing center, interviewed at her home by telephone July 11 from 4-4:30 p.m.; Bob, a 54-year-old secondary school educator, interviewed on campus July 12 from 1 to 1:42 p.m.; Louise, a 31- year-old graduate student, interviewed at her home by phone July 14 from 1:55-2:20 p.m.; and Paula, a 51-year-old university instructor, interviewed on campus July 16 from 9:05-9:30 p.m. I used a combination of semi-structured and informal interview techniques, and kept written notes. I observed a graduate-level qualitative research class in a computer laboratory. The class started with 24 students enrolled and ended with 23. Field notes were kept as a written record of classroom observations on June 27 and July 2, 9, 16. My role as researcher shifted back and forth from observer to participant- observer as I sometimes worked on Interchange while observing and other times didn't. Other data sources I used were photographs of students working on computers and printouts of the DIWE Interchange sessions on June 11, 18, 20 and July 2, 9, 16. The Interchange documents were collected both chronologically in the order messages were posted, and individually with all of a person's messages for the session compiled under his or her name. Four photographs were taken from the four corners of the lab on June 27 and July 3, 16 as a visual record of the DIWE sessions while they were in progress. In addition, I kept a reflexive journal that recorded my reactions to readings, problems faced during observations and interviews, mistakes I made and how I might avoid them the next time (Kleinsasser, 1991), and memos of information I needed to gather. In deciding the kinds of data to gather, I also had to consider conflicting recommendations in the literature regarding ethnographic educational evaluation. Fetterman (1987, p. 54) says that "good" ethnographic educational evaluation will use a mix of qualitative and quantitative data. However, Guba and Lincoln (1988) argue that mixing the two research paradigms is illogical, "the logical equivalent to calling for a compromise between the view that the world is flat and the view that the world is round" (p. 93). For my purposes, Bodgan and Biklen's advice for the beginning qualitative researcher seemed more pragmatic -- trying to do both types of studies simultaneously can be "a big headache" (1992, p. 43). Looking at other data to evaluate the success of Interchange as a teaching tool would be useful, such as looking at grades in classes taught with and without DIWE, and questionnaires for each person in the class to take regarding their attitudes toward the success of DIWE on learning about qualitative research. If I had the chance to repeat the study with more time, a questionnaire would be useful to gauge reaction from the entire class toward Interchange, rather than from just the three students interviewed. However, to stay in a qualitative methodology, questions would be based on my analysis of data collected from interviews, observations and documents, not on preconceived ideas I had about what I might find out about Interchange (Guba & Lincoln, 1988). The questionnaire would be one of the final steps I took in the study, rather than one of the first, as might be done in a scientific inquiry. To help me achieve trustworthiness and credibility, Louise agreed to perform a member check of my interview with her and my interpretations of what she said. Firestone and Dawson (1988) recommend this intersubjective approach to data analysis involving interaction between the researcher and participant to enhance understanding of the data and to help verify the researcher's findings. I also used triangulation to seek convergence across the various data sources (Firestone & Dawson, 1988). In addition, a peer review of my field notes from the July 2 observation was conducted by two classmates was conducted as another check on trustworthiness and credibility. To maintain a high ethical standard for the study, I developed consent forms after consulting the University of Wyoming Institutional Review Board for Projects Involving Human Subjects consent form. I was particularly concerned about protecting participants in my study from emotional or psychological harm (Eicher, 1991), such as fearing their grades could suffer if the professor discovered they said something critical about her or her class. But at the same time, I did not want to "paralyze" my fieldwork by overstating the possible risks to people (Wax, 1971, p. 274) and making them unwilling to participate in my study. After interviewing April and observing her in class, my intuition was that she would not lower a student's grade as retribution. Part of my hunch was based on the emphasis she placed on ethics in the course content. Another reason was that she is an active researcher herself and has likely encountered similar problems in her work. So my consent form stated there were minimal risks involved. One ethical issue that arose concerned my access to the Interchange printouts, which were stored on the lab's network after each DIWE session. My concern was that, because everyone was identified in the Interchange printout, I would have to get a signature on a consent form from each member of the class, which would be inconvenient in an eight-week class. If one person refused to grant consent, could I ethically look at the printouts since that person's writings would be included. I also considered the DIWE sessions to be public record, because anyone in the class could get access to them. I resolved the question by explaining my dilemma to the class and asking for a consensus allowing me to use the printouts. If anyone objected, they were asked to contact me. Nobody did and the study went forward. When it came time for data analysis, I basically followed the advice of Bodgan and Biklen (1992) about how to sort through and organize my various data sources. I did not follow their advice about cutting out pieces of my field notes and sorting them into folders. That seemed too unwieldy. I looked for repeated words and phrases, behavior patterns and events. I then assigned a code to each category with abbreviations and page numbers of where the information could be found. I originally had 10 categories: technical difficulties, interaction with the instructor, interpersonal interaction among class members, lurkers, computer as personification, more time, room problems, recommendations, and benefits of DIWE. I looked for relationships among those groups and narrowed down my data analysis to three major themes: problems, communication and cybercommunity. I decided to put recommendations in a separate section of the ethnography because my pilot project was an educational evaluation. Data source codes I used the following codes for citations from my data sources to ensure trustworthiness and credibility. A citation that reads (I-2, p. 4), for example, means second interview, page four of my notes. Dates of interviews, photos and observations are in the methodology section above. O -- Observation fieldnotes I -- Interview notes D -- DIWE records R -- Reflexive journal entries F -- Photographs from observations The lab environment The computer lab for the class was located in the basement of a business school building. There are 25 Gateway 2000 4SX-25 computers with Crystal Scan 1024 NI monitors in the lab. Six are lined up on the south wall of the room, eight on the north wall, and 10 on a U-shaped table at the east end of the room. One is located on a stand at the front or west end of the classroom for instructor use. Another U-shaped table is located at the west end for students to use when they're not on the computers. Two dot- matrix and one laser printer are located in the front of the room, as are a TV and VCR on a cart, and a LCD panel and overhead projector on another cart. There are 24 office chairs upholstered in pink fabric. The chairs swivel and roll, allowing students freedom of movement around the lab. Three pink, plastic flamingoes normally found in yards and one pink flamingo stuffed toy are in the back of the room. The pink theme continues on the north and east walls where shocking pink signs give instructions on how to use the computers. The lab was created in 1992 when the university received a $300,000 educational trust fund grant from the state, according to Joan (I-2, p.1). The computer lab was funded with $75,000 of that grant. Originally, university officials planned to buy 286 IBM- clone computers, but when computer prices fell, more powerful 486 computers were purchased instead. Until 1992, the room had been a lecture-type classroom that held 50 students. Joan said that research on computer lab design revealed that for a 25-computer lab, at least 1,000 square feet of space is needed. At first, they considered designing the lab with pods of three or four computers arranged in small groups but wiring the lab would have been difficult. So what Joan termed a W shape was used, which she said "proved workable" (I-2, p. 3). The rolling chairs "are probably the feature that students like the most," she said, and they allow for interaction as students roll around the classroom in them. One design flaw, as she termed it, was putting the instructor's computer on a higher table. "The pedagogy of a computer classroom should take the limelight off the teacher," Joan explained (I-2, p. 4). DIWE was purchased for the lab in 1994 when instructors who taught in the lab decided a synchronous writing program was needed. Most of the teachers who use the lab for their classes are in the English Department. Twenty-five copies of DIWE software were purchased for about $4,000. The software includes Mail, Interchange and a word-processing program. Technical requirements The DIWE software is produced by The Daedalus Group, Inc., located in Austin, Texas. Versions are produced for Macintosh and IBM systems. The system requires at least an AT computer installed with DOS 3.3, and with 640 kilobytes of RAM. A hard disk and mouse are recommended. Network requirements are computers with at least a 286 or 386 microprocessor, four megabytes of RAM, and 80 megabytes of hard disk. The average class will generate 10 to 20 megs of data per semester. FINDINGS My data analysis resulted in three major themes that will be explored in this section. One involved problems with DIWE, the lab environment and lack of time to work on DIWE. A second dealt with communication: interpersonal communication, discourse with the instructor, class discussion, and the use of research language. The third involved cybercommunities: personification of computer technology and lurking. Because one of my goals for the pilot project was to evaluate how DIWE helps students learn about qualitative research, I will conclude the findings with recommendations for future use of DIWE in this class and will take an advocacy position based on my findings (Fetterman, 1987). Problems with DIWE and the lab Three problematic areas were identified in the data analysis: technical difficulties using DIWE, conditions in the lab environment, and pedagogical problems, particularly lack of time. Observations, documents and interviews revealed that messages scrolled across the screen too quickly and that students had trouble stopping the scrolling. Even as late as the third class that DIWE was used, Maynard asked April, "Is there anyway you can stop it from scrolling?" (O-1, p. 2). Students came up with several ways of resolving this problem and often shared them with others. One way was to move the cursor to the top of the screen, which would freeze the messages until the cursor was moved. However, even on the final observation, at least one person was still having problems (O-4, p. 2). Paula left her computer and walked to the back of the room to show Louise how to move her cursor to the scroll bars to stop the messages. "Scrolling was a problem," said Paula, who was shown how to stop the scrolling by Bob. "It's intimidating seeing the comments flying past. It needs a programming solution" (I-3, p. 2). Bob also said the scrolling was distracting. "If you focus on one thing, two or three others scroll by," he said (I-3, p. 1). The rapid scrolling was the No. 1 problem with the software, according to observations, interviews and documents. April also said she found the scrolling to be "annoying" (I-1, p. 3). She said that initially she was concerned DIWE would be too complex to add to an eight-week semester. However, only one or two people seemed to be struggling with the technology by mid- semester, she said (I-1, p. 1.) Another problem had to do with time, both the lack of time to use DIWE and the time that it took to log on to DIWE. April said that one reason she decided to go with the DIWE writing sessions was because there was not enough time to use outside readers for the class' pilot projects in an eight-week semester. She also comes to the lab at 4:30 p.m. because DIWE requires more of her time to prepare for class. Her teaching strategy was to have the class write on DIWE for only 15-20 minutes a class so people wouldn't get tired (I-1, p. 4). Some class members would show up 30-60 minutes before class and would start writing on DIWE early. Others showed up at 6 p.m. when the class was scheduled to begin. "If you come right at 6, it's 10 or 15 minute sessions. That's not enough time," said Louise (I-4, p. 4). "We don't have enough time to really talk to people." She added that if she enjoyed working on DIWE more, she would go to class earlier to use it. If she knew that each class would begin with a DIWE session, she also would go early. In my observations, I noticed that some students would get to class about 6 p.m. and not even log on to DIWE. Some would sit at the main table and talk, while one student spent the entire DIWE session reading the assignment for that class (O-2, p. 2). Several students complained about the slowness of the software. Frustration seemed high on July 16, the night of my fourth observation. Both Sam and Candy complained that the messages were taking too long to come up on the screen (O-4, p. 1). The lack of speed could have been due to the software, the hardware or the volume of messages posted that night (85). Louise also mentioned her frustration in logging on to the lab's network before she could start working on DIWE. "You have to boot up the computer, then it takes a long time to get into DIWE. It needs to be more mouse oriented to speed it up," she said (I-4, p. 2). Cognitive time also was mentioned several times in interviews. Joan said slow thinkers and slow typers usually don't participate actively in DIWE sessions (I-2, p. 2). I asked Bob why he always came to the lab by 5 p.m. "I'm slow. That's why I start early," he said. (O-3, p. 1). In a subsequent interview, he added that he considered himself a "pretty passive" participant in DIWE sessions (I-3, p. 2). Paula also considered herself a "low user, because I can't think fast enough. If I write one thing in a session, I feel successful." When asked how much time would be enough, she said that if she had an hour on DIWE, she would write more comments (1- 5, p. 1). Some other technical problems were observed. On June 18, Sam gave instructions by Interchange to another student about how to send a message in Interchange. On June 27, another person couldn't get from a T:/ prompt to Interchange. On July 16, Candy accidentally called up the Help screen in DIWE and couldn't return to the Interchange session. That same night, Sara couldn't get from WordPerfect into Interchange (O-4, p. 1). Although most people seemed happy with the lab environment, a few criticisms were heard. April complained that the lab was too hot (I-1, p. 3). Although she liked the tables and chairs, Louise mentioned a "lack of air" because of poor ventilation. She also said the narrow aisle between the two U-shaped tables was "awkward" (I-4, p. 2). I noticed that aisle became congested whenever people were moving around the lab. While I was trying to walk through the aisle I knocked somebody's papers off their desk (O-4, p. 2). On the other hand, Bob said he liked the "coolness" of the lab and the lab layout "handles everyone remarkably well" (I-3, p. 2). Communication The main communication categories that I noticed in my fieldwork were the amount of interpersonal communication between students during DIWE sessions, the number of communications directed to April, the level of discourse about qualitative research and the discourse that digressed from research topics, and the use of subconferences and electronic mail for discussion. In interpersonal communication, I examined the DIWE documentation for messages posted directly to another person by name. For example, a message that started "Bob" would be considered interpersonal. I noticed that as the Interchange sessions evolved over the semester, the number of interpersonal messages increased significantly. On June 11, there were only four interpersonal messages, but on July 16, 51 percent of the messages were interpersonal. At the same time, while observing Interchange sessions in progress, I noticed that the number of verbal interpersonal exchanges between class members dropped from the first observation, dropping from 11 on June 27 to 8 on July 16. That indicated to me that the class might have been becoming more comfortable with the technology, as well as more comfortable conversing one-on-one using computers. April noticed a similar transition occurring, saying in our July 3 interview that she saw more "back and forth dialogue directed to specific people" (I-1, p. 3). The types of interpersonal communication varied from personal to academic topics. For example, in a highly focused DIWE session on July 16, Donald mentioned he was using stimulated recall by showing participants in his study photographs. Cassidy answered: "[Donald]: interesting idea. Are you going to include those photos in the appendix?" (D-6, p. 4). That kicked off a thread of discussion about stimulated recall. Some nights, though, the discussion could get into other threads that I called socializing online. The June 18 session began with a series of bad puns about the ValuJet crash in Florida. "You can't get a ticket on Valujet - they are 'swamped,' Bob wrote (D-2, p. 1). Many nights there was a fair amount of joking around during DIWE sessions, both on and off line. The room would be quiet for a while, and then the silence would be broken by laughter and humorous exchanges between people (O-1, pp. 1-2). One night Paula brought a treat into the lab during a DIWE session. "Did you bring chocolate cake?" asked Veronica. "You evil woman." While I observed DIWE sessions, classmates would often spend time talking with me instead of typing and reading on Interchange. These could become valuable discussions about research-related subjects, not unlike the conversation taking place on line. During my first observation, Veronica asked me if I was "watching" her. I told her I was observing for my study. That evolved into a discussion about who gives permission to a researcher in a classroom setting, the instructor or the students. I said I thought the instructor had that authority. Veronica wondered who she would ask for permission to observe a library. We speculated that it would be the librarian. I later noticed in the DIWE document that she didn't post any messages on Interchange that evening. That led me to wonder whether my presence as an observer was a distraction to classmates (R, pp. 5-6). On July 2, I was observing DIWE when I engaged in a discussion with Smitty about his pilot project. He was sitting next to Veronica at the computers on the southside of the lab. It was a revealing talk in that I learned a lot about Smitty, such as how long he had been married, and his life philosophy. But he, too, didn't file any messages on Interchange that night. I speculated that the distraction I might be causing to some people could prove useful in my pilot project. Tracking Veronica's and Smitty's Interchange participation through DIWE documents showed that they were among the least active users in terms of the number of messages posted. Did they feel more comfortable talking verbally than online? Did they simply not like DIWE or was there some other factor leading to their low participation? It would have been interesting to interview them to find out, but I also was facing time problems. Bob said he thought the verbal interaction taking place in the lab indicated problems with DIWE. "When people talk, it tells you the software is too cumbersome. It's easier and faster to talk than go on DIWE" (I-3, p. 3). Along the same line, Paula said that she walked across the room to show Louise how to stop the scrolling on July 16 because "it would take too long to write how" (I-5, p. 2). Louise said she felt comfortable verbally talking to people only if they were sitting next to her (I-4, p. 1). She also became "irritated" when the discussion on DIWE digressed from research topics. "It wastes class time" (I-4, p. 4), she said. Paula said she didn't always see a dialogue taking place on Interchange. "You're commenting to someone when you get 40 other posts at the same time" (I-5, p. 3). April expressed concern that students would write for her -- the teacher -- instead of seeing their DIWE audience as themselves and the rest of the class. Having students write for the entire class was important to her because it was a way to reach her objective of the class becoming a research community (I-1, p. 2). The number of DIWE posts directed to April by name were low. The most posts directly to her were the seven on July 16. In my five observations, I only noticed four instances in which students verbally talked to April while DIWE was in progress. However, in the interviews only Paula said she was writing "for the rest of the class." Others saw themselves writing for the instructor. "Most of the time I just answer the questions she poses," Louise said. "When she posts a question I feel obligated to answer that" (I-4, p. 1). Bob said, "I'm writing for her as much as anything because most of my writing is in response to her." He said he wasn't expecting feedback to his responses from other students. In fact, he considered the "highlight" of DIWE sessions to be "watching what (April) does. She knows exactly what your concerns are (I-3, p. 2). Another of April's reasons for employing DIWE writing sessions was that students would be performing "authentic research tasks," which include writing fieldnotes and reflexive journals (I-1, pp. 1-2). She also hoped that DIWE writing would encourage students "to think like a researcher" and to "practice the language" of research. The DIWE documents showed that the level of discussion about qualitative research intensified during the semester. I looked for posts that were digressions from qualitative research topics or that didn't use vocabulary associated with research. The June 18 DIWE session (which will be called the ValuJet session from now on) had the most digressions (13 out of 59 total posts), while by July 16 the discussion was extremely focused on research with only two digressions out of 85 posts. During my observations, I noticed that even the distractions that I caused were related to research questions. My discussions with both Veronica and Smitty were as much about researching as they were socializing. During the July 16 DIWE session, Louise verbally said to Bob, "Well [Bob], we seem to be converging," referring to their DIWE posts about credibility. Her use of the word converging indicated to me that students might be using research language even in casual conversation. In interviews, students had mixed reactions about whether DIWE helped their pilot projects or helped them learn about qualitative research. Bob, for instance, said that DIWE was of "no benefit" to his project, nor did DIWE lead to the formation of a class research community. There wasn't enough time to form a network, he said, because "it takes time to get to know people" (I- 3, p. 3). Louise said DIWE helped her "get ideas on things I wouldn't have thought of," and she learned from other people's mistakes that they disclosed on Interchange. "In that sense, it's part of that whole research community thing," she said of DIWE. She thought DIWE was successful in getting the class talking about such research issues as interviewing and journaling. But she didn't feel part of a research community in the class because most classmates were in education degree programs and she was not. Paula said that her small group had become more of a research community for her than the DIWE discussions. "DIWE is too fast and furious," she said. "I still like to have conversations with people" (I-5, p. 3). However, she found DIWE to be valuable in her research task of reflexive journaling. In class discussions, Paula had expressed her reluctance to keep a journal, saying that she didn't like to be forced to write. After class on July 2, she approached Cassidy and me about getting her DIWE posts saved on a diskette or printed out. She said she wanted to use her posts for her journal. I thought that using the DIWE posts might be an effective way for her to organize her thoughts and reflections on different aspects of research and decided I should try to interview her about that (O-2, p. 3). She told me that it "made sense" to use her DIWE posts in her journal. "I'm looking at it as being eclectic -- my reactions to books and directed writing activities like DIWE. I probably wouldn't sit down and write about data analysis, for example" (I-5, p. 1). Despite several suggestions by April and Cassidy about forming Interchange subconferences, the class seemed reluctant to break off into alternative discussions. I noted that the April 2 DIWE session, which consisted of four subconferences created by April, had no digressions out of 50 total posts. Several students also commented that they thought the April 2 session was very productive. I started one on June 20 to discuss beer and only one other person posted a message there. A subconference was set up on June 18 about covert research, but a suggestion by Cassidy to set up an ethnography subconference got no action. A June 18 subconference called Bananas had three respondents. Although April thought that the small groups in class would set up their own subconferences on DIWE, that didn't happen. April explained that her strategy in setting up subconferences on April 2 was "to get away from eight million messages scrolling past, to try to slow it up." However, she thought the result was "kind of boring." Bob said he thought the subconference DIWE session produced "almost too much information" but it did get him more actively involved than had past sessions. "I felt obligated to answer each question" (I-3, p. 2). Louise, the only person to join the Beer subconference, said the smaller discussions "give you the opportunity to talk more ... It's easier to have conversations." She joined two subconferences on April 2 and said she "checked out" all four of them (I-4, pp. 1-2). Paula also found subconferences "easier to do because there aren't as many people." She joined one subconference on April 2. "I would do more if I had more time," she said (I-5, p. 2). Electronic mail or e-mail was used with mixed results by many in the class. April speculated that some of the people who weren't active on DIWE were reading their e-mail instead, and that was causing DIWE sessions to get off to slow starts. Smitty posted several messages asking people to e-mail him with their thoughts on spirituality, research ethics, and journaling. As of the July 9 class, he hadn't received any e-mail in response. Other students, however, found e-mail was invaluable to their work in the class. Bob was using e-mail 15 years ago on the original ARPNET system when he worked in the defense industry. He said he would give up DIWE before he'd give up e-mail. He e-mailed research questions to April frequently during the semester and said he would feel "comfortable" e-mailing her in the future about his research (I-3, p. 3). Paula said she was saving her e-mail correspondence with April to use in her reflexive journal. She preferred writing e-mail to writing on DIWE because she had "more time to think ... DIWE is not the type of writing that allows for a lot of reflection" (I-5, p. 1). A list of the 20 students in class who had e-mail addresses was compiled and April made copies for everyone. Bob said he thought it was more likely that students would form "e-mail communities" than DIWE research communities. April said she intended to use DIWE as a way to get more people participating in class discussions. Joan also noted that in her experience other students like how people who dominate verbal classroom discussion "lose their voice" in DIWE conversations (I-2, p. 2). This proved to be a valid concern early on in the semester. After class members complained about a few people dominating verbal class discussion, April implemented her "DIWE" policy: When questions involved individual concerns instead of broader issues of interest to the whole class, she would say "DIWE" and further discussion of the individual topic would have to take place online. DIWE did get at least one quiet class member involved in discussion online. After the June 27 class, April mentioned to me that Marty never speaks in class discussion, but that his DIWE posts were lengthy and insightful (R, p. 11). Bob said he prefers DIWE discussions. "It allows a cross section of the whole class that I wouldn't get otherwise ... It takes the pressure off you to get recognized in class discussion," he said (I-3, p. 1). Cybercommunity As I observed the DIWE sessions and read more of the literature about computer-assisted instruction, it became apparent that some of the behavior in class was similar to that of a cybercommunity in which humans interacted not just with each other, but with their machines and with personalities that existed only through their computer networks. What first struck me as a postmodern phenomenon was when Dick referred to his monitor as "look at all the people." That started me thinking about how other students related with their machines and with their world through their machines. Candy had a love-hate affair with her machine and with DIWE. Despite complaining about the slowness of DIWE, she was active in the July 16 session, posting four messages. But she also said out loud to nobody in particular, "I'd much rather talk to a face than to a computer screen" (O-4, p. 2). April had, perhaps unconsciously, alluded to this type of behavior when she said DIWE was "a nice way to think out loud." When I asked her to explain that phrase, she changed it to "thinking made visible" in that DIWE made a person's words visible (I-1, pp. 2-3). Were DIWE and computers merely tools for learning, or were they becoming extensions of human cognitive processes? Louise said they her "thoughts come together better" when she's writing on a computer. She also saw Interchange as "a talking." Numerous descriptions of Interchange as a conversation implied that students considered the software not just as a channel for communication between people, but as a communication in itself. One feature of Interchange is that it allows people to log into discussions anonymously or with pseudonyms as a way to encourage candid discussions, much like MUDs (Multi-User Dungeons) in which people can create their own persona, or as Turkle (1996) calls them, their "multiple selves" (p. 152). Louise referred to this ability to remain anonymous online when she talked about the comfort she feels hiding behind the monitor. Although she's entered chat rooms on the Internet and was active in DIWE, Louise said she doesn't consider those to be substitutes for socializing. "I very much value the human contact aspect and wouldn't use this as a substitute," she said (I-4, p. 4). Yet some class members were able to share personal information as a result of communicating online. On June 27, Kurt broke the silence of an Interchange session by turning around to ask Candy when she changed her name. He had noticed on Interchange that she was using a different surname than he had known her by. She explained verbally that she had remarried. Even though everyone in the class probably knew each by face, at least one student wasn't sure about people's names. Bob wrote interpersonal communications on DIWE to Connie a couple times, including a funny comment that she should join Lurkers Anonymous (D-5, p. 6). When I asked him if he and Connie were friends, he said he wasn't sure who she was (I-3, p. 2), even though he once sat just a seat away from her during a DIWE session (F-3). Even though in my data analysis I considered it a digression from research discussion, lurking took up much of the July 9 DIWE conversation and showed me something else about class dynamics. About halfway through the DIWE session, Adam wrote a post that there seemed to be a lot of lurkers on Interchange today. April asked him to define lurker, which kicked off a discussion about DIWE users who "listen in" but don't participate in the discussion (D-5, p. 4). Sam wondered about the apparent contradiction between listening and reading off a monitor. "How can you be listening if I am looking at a screen. Lurcking (sic) implies 'looking' so a lurker is a virtual voyer (sic)," he wrote (D-5, p. 5). That led to an analogy between virtual voyeurs and field researchers, and made me think about something Bob had said in an earlier discussion about covert research. "I think it should be 'passive' as well as covert. I mean that one could observe a subject's actions and collect data without intruding," he wrote (D-2, p. 4). When we talked about lurking during our interview, Bob said he was "a lurker at heart" and considered himself a passive participant on DIWE. He likes to "see who's talking and what was said" during Interchange (I-3, p. 1). Rather than creating a multiple self in the cybercommunity, Bob and other lurkers in class sometimes were opting for no self, an invisible, unintrusive persona that still had a role in the conversation -- that of the eavesdropper. Recommendations Based on the problems noted in observations and interviews, and also on suggestions made by participants in the study, the following steps are recommended if Interchange is to be used again in this course: 1. Until Daedalus Inc. makes a programming correction to stop the rapid scrolling, the instructor should give a brief demonstration or make a handout on the first night of class instructing how to solve the problem using the cursor. 2. A brief demonstration or handout about setting up and accessing subconferences would encourage participation in more specific discussions and reduce the volume of posts in the main conference. 3. DIWE sessions should be scheduled in the syllabus so interested students or slow thinkers could come to class early to increase the time on Interchange. Starting the class with DIWE seemed like an effective way to seque into the rest of the evening's activities by focusing students' attention on research topics. 4. Comments made in interviews about writing for the instructor and the number of Interchange posts directed to April were disturbing and seemed to defeat one of the main attractions of Interchange -- to get students writing for themselves and the rest of the class instead of for the teacher. To bypass this tendency -- which is a learned behavior for many students -- the instructor should become a lurker on DIWE. Instead of the instructor doing it, students should generate the questions and issues to be discussed in directed writing exercises on DIWE to stimulate the creation of a research community in the class. 5. Journaling could be mentioned early in the semester as one benefit of DIWE. More students might become actively involved in Interchange sessions if they knew what they wrote could be used in the required reflexive journal. 6. If there is controversy in the classroom, an anonymous or pseudonymous Interchange session might be a way to reduce tension by letting everyone speak frankly and candidly. All other DIWE sessions probably should be done with users logged on under their own names to keep the discussion tactful and courteous, and to encourage the development of a class research community. Researcher's advocacy Having both used and observed Interchange for the semester, I think Interchange holds many benefits for a research class of this type. Primarily, it gives everyone a chance to participate in class discussions in a non-threatening environment. Writing also tends to be more introspective and thoughtful than speaking, producing a higher level of class discussion. Since most people in the class were educators, introducing them to a synchronous writing program could lead them to adopting this instructional technology in their own classrooms. Reflecting, reflexing, reacting I made mistakes. I can't even call them well-intentioned mistakes, since I didn't have any intentions when I made them. Perhaps they were honest mistakes because there was no attempt to deceive anybody. To repeat the mistakes would be dishonest. My mistakes would mainly be in the dumb category. But they were valuable mistakes because I realized what I had done soon after doing it. Then I would try to rationalize my mistakes. That would turn to reflection about why I made them and finally to reaction about how I would avoid making the same mistake again. I might make a different mistake, but not the same one. It didn't take long to make mistakes in my field work. The most embarrassing was when I interviewed the instructor and didn't give her a consent form to sign first. I had one with me, but just plain forgot to use it. Two days later, when I was writing in my journal, it struck me what I had done. I rationalized my mistake by saying I had an e-mail record of her granting me the interview and setting up an appointment. Besides, she was the instructor and should have known the purpose, methods and risks of my pilot project. Interviewing her was different than interviewing other people, I told myself, because she's THE PROFESSOR. However, to follow the ethical standards of qualitative research, I should have had her sign a form to prevent future misunderstandings when I write up my findings. I need to get in the habit of asking participants to sign consent forms. I also wondered whether verbal consent and e-mail consent would be sufficient if I ended up in court someday. Verbal contracts have been upheld in several famous palimony cases. But, I think a written contract avoids any ambiguities, misunderstandings and misrepresentations. Better get those consent forms signed in the future. I sheepishly mentioned my oversight to April at the next class, and she said it was the first thing she noticed in the interview. But why didn't she mention it then? After reading her article about the educational value of making mistakes, I understood. By failing to inform a participant of the risks of my research, even though they were minimal or non- existent, was I really any better than Milgram? That's probably being too hard on myself, but the point is well taken (R, pp. 19- 21, 25). The next dumb mistake involved the field notes I typed up for the two members of my small group to read as part of our reflexive journal grading contracts. The idea was to read each other's field notes from an observation and then comment on them as a critical friend. I was in a hurry to get the notes typed before class and typed them verbatim, including people's real names. Since my critical friends are in the same class, they would know everybody I wrote about in the notes. The next day I realized what I had done. My attempt to rationalize the mistake was that only two people would read the notes and, as critical friends, they wouldn't tell any of the participants about my fieldnote comments. But I considered it an ethical violation because I didn't protect the identities of participants in my study. I had compromised their trust in me and my research methods. In the future, I should assign pseudonyms or let participants choose pseudonyms and probably use those names even in my fieldnotes in case someone else got hold of them. Certainly, I should change people's names if I'm going to let anyone else other than a co-researcher see my notes (R, p. 25). I also thought and wrote a lot about my role as a participant- observer. Often, I was in an awkward position because I was a student in the same class I was observing. One classmate tossed out the phrase "suck-up" within my hearing range. I think it was in jest, but that might have been the perception some people had of me. Rather than feeling like a suck-up, I felt that I gets have put myself in a tougher position as a student than others in the class. April had suggested she'd like to see somebody research DIWE and I took her up on it because of my interest in instructional technology. Later I realized some of the possible repercussions of my decision. The instructor was observing the same class that I was. Would she be seeing important things that I missed? Was this a pet project of hers with expectations that I couldn't meet? Would I fall behind in the class because, instead of participating in the DIWE sessions, I was observing them? So I found myself between the proverbial rock and hard place, with classmates seeing me trying to ingratiate myself with the instructor for good grades and my self- perception of not being able to meet the instructor's expectations. My typical response to this situation is that I don't care what people think about me as long as I feel comfortable with what I'm doing. That's the attitude I ended up taking in this case. But were my observations making others in the class uncomfortable and perhaps hurting their learning experience in the class? The first time I took photos during a DIWE session, people joked about whether their hair looked OK and laughed. But they seemed uncomfortable. I guessed that they expected to be researchers in this class, not the researched. I would have liked to interview people about that: How did it feel to have the research table turned on them? I justified the photography by telling myself the photos were nearly all of the back of people's heads, which might help reinforce the anonymity of participants in my study. I had no trouble changing hats from researcher during DIWE to student when DIWE ended. But I wondered whether some people in the class may have had problems making that distinction of me (R, pp. 3, 5-6). A lot of my journal writing involved comparisons between journalistic reporting and qualitative research methods. I was surprised to see Hunter S. Thompson's "Hell's Angels" referred to in Punch's (1986) book as an example of a researcher who was threatened and beaten up during fieldwork. Thompson was a writer who got me interested in journalism, but I never thought of him as a qualitative researcher. If you overlook his exaggerations, though, his techniques of first-hand involvement with subcultures, such as biker gangs and federal drug agents, observing and interviewing them, and then writing about his field experiences in a subjective style were a lot like those of qualitative researchers (R, p. 8). Although I interviewed hundreds if not thousands of people during my career, I was uncertain about interviewing as a researcher. For one thing, I have never used a tape recorder because transcribing the tape would slow me down when writing on deadline. But it seemed like many if not most ethnographers taped their interviews. I felt confident enough in my note-taking ability, though, to stay the course. Secondly, in journalism you interview for information, for content, while in qualitative research your interview goal seemed to be to generate meaning and interpretation. In journalism, we'd try to gather facts and let the reader figure out what they mean. Research interviews went an extra step and did the interpretation for the reader. I was apprehensive about my first interview with the instructor. I figured she was an experienced, if not expert, research interviewer. I asked her to critique my interview style, which I changed to pick up on key phrases and then ask her to explain what she meant by those phrases. I used my usual unstructured interview format, going into the interview with some general ideas of what I wanted to talk about but with no prepared, written questions. Afterward, she said my interviewing was OK, but I wondered if she was just too tired to really critique it. In the course of the interview, she also gave me some ideas of things to look for in my observations. Later, when I became familiar with the term, I realized that she would fall into the key informant category because of her assistance (R, pp. 16, 20). I was also struck by Punch's (1986) concern about Great Britain's strict libel laws and how that fear could paralyze a lot of worthwhile research in that country. In the U.S., journalists are always conscious of libel when they write stories. But the main difference between here and Great Britain is that you have to prove malice to win a libel case in this country. The reporter must have deliberately set out to damage somebody's reputation to commit libel here. Proving that state of mind has proven difficult. In England, though, damaging a person's reputation is enough to prove libel, even if it was unintentional. If his participants decided to pursue a libel case against him and did need to prove malice, Punch probably wrote his own guilty decision with his admission that he didn't like the English upper-classes (R, pp. 8-10). If I had it to do over again, I would take at least six months to get more depth and breadth in my data. I would have interviewed more people in the class, used a questionnaire to gather information from everyone in class about their attitudes toward DIWE, and read more literature, especially about learning theories. In particular, I would have liked to tackle Lacan to learn more about postmodernism, and read more Turkle to see how she's applying Lacan's thinking to her cyber-studies. I'm also still fuzzy on the distinction between constructivism and constructionism, and would do more reading in that area. From what I've been able to figure out, they share a lot of qualities but constructionism is based more upon experiential learning: To learn about a hammer, you need to use a hammer. In retrospect, collaborative learning might be a more accurate theory to explain DIWE (R, p. 32).