Herding Cats: A Descriptive Case Study Of a Virtual Language Learning Community
John H. Steele
Indiana University of Pennsylvania
May 2002
CHAPTER 5: FINDINGS AND RECOMMENDATIONS
Questions that Guided my Study
Possibilities for Future Research
As I mentioned earlier in both my literature review and my discussion of the class, research on virtual ESL/EFL classes is almost non-existent. Most of this lack of research is because there are few schools that make use of the Internet as the sole means of offering ESL/EFL classes. As Coghlan (personal communication) mentions, “most ESL students prefer a dual mode class.” Dual mode classes make use of both the Internet and face-to-face meetings. Part of this lack is also due to online ESL classes being a relatively recent phenomenon. There are a growing number of online ESL/EFL classes and research is being carried out in many of them, but little of this research has been published yet. Further, these studies generally cover a specific medium in which these classes are offered (e.g. asynchronous classes, MOO-based classes, web-based classes, chat-based classes). As yet, there is no published research on what happens in a class that uses more than one medium. To carry out this research, I decided to focus on the following question. “What happens in an online ESL/EFL class?”
With this question in mind, I decided to carry out a naturalistic study of online classes using virtual interviews as my primary means of data-generation. Since I had previous experience with the Vandar Online Language School (VOES), a school that offers free ESL/EFL classes using volunteer teachers, I decided to study classes in this school. At the time I made this decision, most of VOES classes were using an asynchronous medium (discussion lists) augmented by the use of World Wide Web sites. Writing Through the Internet (WTI) was one of the few VOES classes that had incorporated a synchronous portion into their class.
While there is a wide range of students in WTI, most of the students have upper-intermediate to advanced English skills. They are generally adults, and come from every continent and many walks of life. Some students have been in WTI since the class was first organized over three years ago; others have just recently joined the class. This mixture of old and new students, as well as the presence of three teachers and the concurrent use of two and three chat rooms for each weekly class session, leads to an interesting class which has been characterized by one member of the community as intuitive controlled chaos.
Since this is an online class, I used what Hine (1998, 2000a, 2000b) and B. Mason (1999) call a virtual ethnography. As mentioned earlier, this is an ethnography in which the culture being studied is completely on-line. Most of my data generation was carried out through virtual interviews, but I also made use of participant observation as well as publicly available documents and logs of the weekly chat sessions. Due to the nature of the medium, interviews that are carried out in the virtual world of the Internet have characteristics that are seldom found in interviews carried out in the physical world. These differences are described in detail in the section titled “Details of the Findings” later in this chapter.
Questions that Guided my Study
The basic question that guided my study was:
What happens in an online ESL/EFL class?
At the same time, it should be remembered that naturalistic studies change as they progress. Naturalistic research, according to Lincoln and Guba (1985), is emergent (See also Glesne and Peshkin, 1992). Therefore, this question was just the starting point in my research. Other questions developed through the course of the study. The following additional questions that emerged in the course of my study are especially worthy of note.
Given that most VOES classes had problems retaining students for the entire three-month term while WTI not only retained its students but even increased in size and has existed continuously for more than three years, the following question became of interest early in the study.
1. Why has WTI been so successful?
Arising from this question are the following subsidiary questions:
a. What elements of the class attract students?
b. How would the students describe the class?
After joining the WTI class, I soon learned that it was not a class as traditionally defined. There was no set term of studies with a beginning and ending date. There were no lessons, with evaluation activities to ensure that the students had successfully mastered the skill. There were no assignments that the students had to turn in and have evaluated. In short, what I found was a community in which the members all considered themselves friends instead of teachers and students. Given that the class had become a language learning community, I asked the following question:
2. How does an online ESL/EFL class become a language learning community?
This question, in turn, gave rise to the following:
a. What are the elements of a virtual community?
b. How can these elements be fostered so that the community continues to grow?
c. What changes occur in the community as it grows?
Finally, since my interviews were conducted completely on-line, a question arose related to what elements of the virtual world most affected how interviews are carried out. I, therefore, asked the following questions.
3. What differences exist between virtual and “real-world” (face-to-face) interviewing?
This gave rise to the following questions:
a. What elements of the virtual world affect interviews?
b. Are synchronous interviews adequate for data generation?
c. Can asynchronous interviews aid in increasing data generation?
d. How can other means of data generation (e.g. participant observation) aid in developing a study?
I joined the WTI class approximately four months before I began my data generation. During this time, I noticed that there was much posting on the asynchronous e-class portion of the class as well as a group of students who participated in almost all of the chat sessions. In the e-class, students would post everything from jokes to problems they had with English, while in the chat sessions they would talk about any topic of interest. These topics varied from discussions about books the students had read and were interested in through discussions on topics related to technological issues, to discussions on how to date in different cultures. Both the asynchronous e-class and the synchronous (chat) sessions were very active. As mentioned earlier, during the course of my data generation, the two most used chat clients disappeared from active use in the Internet. With the loss of these two clients, and the subsequent search for a replacement, the community began to become less active. While nobody in the community has an answer as to why the sessions (as well as the e-class) became less active, several suspect it was due to the loss of stability within the community. This suspicion is supported by the fact that lately, after the community selected a new home at Tapped In, both the amount of activity in the chat sessions as well as the number of postings to the e-class appear to be increasing again.
The student activity within the class translated into an interest in participating in my study. When I sent out my call for participants to be interviewed, I received seven responses (six positive and one negative) from the participants in WTI, and only two responses from students in other VOES classes. The one negative response I received was from a “student” (EFL teacher) who was researching the use of chat rooms in teaching EFL. In her response, she indicated she was willing to participate, but questioned whether she was the type of student I wanted to interview since she was not in the class to learn English but to learn how to use the technology in her teaching.
As with everything else, there are negative aspects to virtual classes such as WTI and with virtual research. On the downside, virtual classes, unless they are simple e-mail and/or text-based chat sessions generally require high-end equipment. Also, most classes that are simple asynchronous discussion lists, as is often the case, seldom last for more than two or three months. When I was actively teaching for VOES, the teachers constantly complained about the lack of participation by most students and the high dropout rate. Most students in these classes appeared to be interested only in receiving help on specific problems related to assignments they had in their regular classes or in receiving help in writing reports and letters for their jobs.
Another downside is related to the type of chat rooms that are available. Unless the school is willing to charge students for the classes and pay for the use of password protected chat clients, decent private rooms no longer exist in the quantity previously available. In the case of free classes such as WTI, the disappearance of “private” rooms at sites such as The Palace means that the class must make use of sites that are less than private. Whereas, before, visitors had to be told specifically how to access the sites being used, in WTI’s new home, visitors and guests can drop in at any time, and often do so quite by accident. While this can be positive since many of these visitors later join the class and contribute to the weekly sessions, they have led to a change in the community make-up. There are now more community members, including those who post in the e-class, who are more interested in online education than in language learning. While some of the original students still participate, a shift seems to be occurring in the emphasis from language learning to technology and online education.
While much of what I found in my study appears on the surface to be limited to free ESL/EFL classes, there are six general insights that can be drawn in relation to online ESL/EFL classes, building language learning communities, and in conducting virtual interviews.
Findings Related to Virtual ESL/EFL Classes
1. Multiple means of interaction are important in virtual classes, but each student should be allowed to select which means of interaction he/she will (or will not) use.
Although the use of asynchronous communication (email discussion lists and bulletin boards) may be a vital part of having a successful online language class, so also is synchronous communication (chat rooms, MOOs, IM clients, etc.). In carrying out virtual studies, Frank and Davie (2001) indicate that chat is not very useful for in-depth discussions because the participants do not have time for considering their answers. At the same time, they also suggest it is an excellent medium for the social interaction that is a requirement of community formation. Further, according to several researchers (see, for example, B. Mason, 1996, 1999; Warschauer, 1996a) students tend to produce more in asynchronous communication such as email than in a chat session. This is due to the amount of time the student is able to dedicate to his/her answer in each medium. For this reason, a method of asynchronous communication, either through an email discussion list or through a bulletin board, is a necessary adjunct to any successful online class.
In WTI, the asynchronous portion of the class was the e-class discussion list. Although the postings on this list would vary in quantity, (sometimes being as active as five or six messages a day and at other times as slow as two or three messages a week), the list was always an important part of the class. Towards the end of my study, the community also incorporated a voice bulletin board for those students who wanted to leave an oral message and receive feedback on it. This client is not used as much as either the e-class or the chat sessions, probably due to the nature of the bulletin board. As mentioned earlier (Chapter 1), bulletin boards differ from email in that the participant must visit the bulletin board to receive any messages while the email message is sent directly to him/her.
The use of asynchronous communication, however, is not sufficient. My informants were unanimous in their opinion that students need to know that they are communicating with a real person, and not just a computer program. For this reason, chat sessions or other means of synchronous interaction are important. The teachers at WTI found that the use of chat aided in developing a sense of community. While text-based chat clients are helpful, they are not as effective in building the necessary sense of community as are voice-based clients. Voice-based clients allow the group members to hear what the others sound like and add an extra dimension to the chat session. In a voice-based chat room, the data is no longer just text. With voice, the participants can hear the tone of voice and other extra-verbal cues that add so much to meaning. As in the importance of chat clients in general, all of my informants were in agreement on the importance of hearing the other person’s voice in developing a sense of trust and community.
The weekly chat sessions are only part of the synchronous portion of the class. Another important part of the class is found in the Instant Messaging (IM) clients. Since many members live in time zones that do not permit them to attend the weekly chat sessions, the IM clients allow them to participate synchronously with members of the community who they find online and available for chatting.
This multiplicity of media of interaction is one of the main reasons
that a virtual class becomes successful. Of course, merely allowing for different
means of interaction and communication is not enough. There must be a conscious
effort made to integrate them into a whole. Each medium should support the
others.
2. Change is an inevitable and necessary part of virtual communities if they are to continue for more than a few months.
As mentioned earlier, change is a necessary adjunct to growth. Without it, the class is in danger of becoming static, in which case the members often lose interest in participating and eventually drop out. In the case of virtual language learning communities, change usually occurs in the programs used, in the make-up of the community, or in the community goals.
With the disappearance of two primary meeting places (The Palace and HearMe), WTI had to search for a new place to meet. While this took time, it allowed the community to practice with different types of chat clients and decide if they were of use to them or not. While the programs the community practiced with were seldom of acceptable quality, a home was finally found in Tapped In with Yahoo! Messenger as an acceptable voice conferencing space. But the change was more than in a move from The Palace to Tapped In and from HearMe to Yahoo! Messenger, it involved, also, a move from a private meeting place (The Palace) to a less private meeting place. Even when the group meets in one of the offices, visitors can, and do, drop by to see what is going on.
There has also been a change in the make-up of the community. Originally, WTI was a group of teachers and EFL students who met for the purpose of improving their language ability. While this is still the core of the WTI community, its make-up has recently been changing to include more native speakers of English who are interested in online education. While most of these new community members are language teachers, many of them are also interested in other fields. This, in turn, is leading toward a change is some of the goals of the group. While language learning/improvement is still the basic goal of the group, a secondary goal of learning the technology of Tapped In is also developing. Further, there appears to be an attempt on the part of a few of the new members to limit the “intuitive controlled chaos” that has characterized the community almost from its inception so as to allow for a more structured class-like environment during the second half of the weekly chat session.
Further, students find interacting with people from other parts of
the world to be a valuable learning experience. For example, at a conference
where several WTI students were present, somebody asked what it was like to
talk to students from another part of the world, online. Some of the students
who were present gave the following answers. “Wonderful.” “We can transcend
time and space.” “The world is smaller.” “Miraculous.” “I have friends
all over the world.” But they also see it as an excellent occasion to develop
their English in a socially interactive group while also learning about new
technologies for language learning.
3. Language learning communities take a conscious effort to develop and maintain.
Although virtual language learning communities are possible, they take a conscious effort to create and maintain them. As mentioned earlier, Blanchard (2000), Frank and Davie (2001), and Müller (1999) identify four components that communities usually share. These are a) a sharing of some common characteristic (i.e., location or interest), b) interaction between/among the members of the community, c) defined boundaries between inside and outside which leads to the development of a common identity, and d) members’ shared identification with and attachment to this community (i.e., a sense of community). Further, there should be some equality among the members of the community. While there may be experts in language learning, and these experts may be called teachers, there should be some means by which both the teachers and the students communicate as equals. These are not characteristics that normally occur spontaneously in a virtual ESL/EFL class as taught by VOES. Since most of these classes are still little more than email discussion lists or classes that offer tutoring, there is little chance for these characteristics to develop. Even in those classes that are offered in synchronous mode, unless a conscious effort is put into developing these characteristics, they will seldom occur of themselves.
As was explained and described in the previous chapter, community was developed in WTI through a four-step process. First was the incorporation of a focus based on exploring different technologies for language learning. All members of the community were expected to be interested in technology based language learning (otherwise they would probably not be taking a virtual language class). The next step was to encourage interaction. This interaction was fostered by requiring all community members to participate in either the synchronous (chat) sessions or the asynchronous (email discussion list) e-class. The third step was to develop a sense of trust. To build this sense of trust everybody who joined the class was expected to send an introduction and, if possible, a picture of themselves and their community. These were then placed on a class web site where each member had his/her own page. Community members were encouraged to add to this page so that other members could learn about them. The fourth step involved sending announcements to the e-class whenever a community member had a birthday or other good news. This last step also included announcements and questions about natural disasters that occurred in the country of a group member. In that way, members of the community were encouraged to think about their classmates and consider them as part of a community.
Another means of community building used by the WTI community is seen when group members show a willingness to contribute to each other’s projects and goals. One way in which that has shown itself is through community members being present at and actively contributing to online workshops and other presentations that sometimes occur. While it might be expected that these workshops and presentations were offered by the community teachers, students in the group have also presented conferences. In these cases, not only did other students appear and show their support, but the teachers also participated in the presentations.
Findings Related to Virtual Interviews
4. The context-poor medium of virtual interviews requires that the interviewer compensate by paying more attention to the text he/she receives.
The main media difference between virtual and face-to-face interviews is directly related to the amount of data that the interview carries. In face-to-face interviewing, the researcher gathers much of his information through non-verbal cues such as body language (gestures, posture, etc.) and tone of voice. The researcher and the informant see each other. Each can tell if the other is really interested in what is being said. Each knows how the other is dressed or sitting/standing. Further, each can see the expression on the other’s face, and can hear the tone of voice that the other uses.
None of these cues are available in the virtual interview. The virtual interview is completely text based. Text only data can be seen in both a negative and a positive light. For example, the lack of non-verbal cues means that the interviewer does not have access to information such as the informant’s gestures or tone of voice. At the same time, since virtual studies are text-only, background data on the community being studied is generally available. For example, in this study, I was able to take advantage of publicly available documents regarding the history of WTI as well as publicly available chat logs from the weekly chat sessions. This allowed me to make use of information I received from these other sources in preparing my interview questions. I was also able to make use of these sources both in interpreting what my informants told me and in describing the community itself.
Because of the text-only quality of the virtual interview, certain questions and conditions arise as described below.
a. Has the informant finished answering the question?
This question specifically relates to the physical aspect of answering the researcher’s question. In the face-to-face interview, once the researcher asks a question the informant either begins answering it immediately, or indicates in some way (verbally or non-verbally) that he/she is thinking over the appropriate response. Even when the informant pauses in the middle of an answer, there is usually some type of cue (e.g., tone of voice or a gesture) to indicate that he/she has not yet finished. In the virtual interview, on the other hand, the researcher has to wait while the informant types in his/her response. This wait can cause the interviewer to start asking him/herself several questions. Is the informant is answering the question? Is he/she thinking about the answer? Did the informant understand the question? Is the informant a slow typist? Is he/she still online? In the case of a slow typist, this may be a long wait. But even after the respondent replies, there is still the question of whether or not this is the entire reply. In many chat and IM clients (e.g., Homestead, Yahoo! Messenger), simply hitting the <ENTER> key (by accident or on purpose) causes the message to be sent to the chat room or IM contact. In other programs (e.g., ICQ), the typist must actually press a send key. Since most of my interviews took place using ICQ, this particular problem did not occur.
There are, however, other reasons an informant might post part of an answer and then continue to answer the question. One is to let the interviewer know that the informant is answering the question. If the informant is a slow typist, for example, he/she might send an answer in two, three, or even four parts. Another reason is the length of the answer. When the informant is sending a long answer, he/she may break it into more than one section either so the interviewer can start reading the answer, or because it is too long for the chat client’s buffer. While ICQ, for example, has a fairly large buffer (in relation to other IM clients), a long answer may still overflow the buffer, thereby erasing the first part of the answer. Further, an informant may send part of an answer while considering what else he/she wants to say about the question.
An interviewer who does not take this into consideration may interrupt the informant without realizing it. On more than one occasion, for example, I received the continuation of an answer after asking a follow-up question. This can lead to the informant either feeling rushed or deciding to send only partial answers. One of my informants began to include phrases such as “more” and “wait” at the end of partial answers so I would know she was still answering. At times, she also included cues such as “thinking” before answering the question.
b. What else is the informant doing besides answering the interview?
In the face-to-face interview, the researcher can see what the informant is doing. Also, the researcher has usually made arrangements for the informant to be able to pay attention to the interview by scheduling it at a time or place in which interruptions should not occur. Seldom do face-to-face interviews take place in conditions where the informant is paying attention to other activities at the same time as answering the interview. The interview seldom takes place, for example, at a time and place where the informant is also taking part in another conversation or watching television. This cannot be said about virtual interviews. Since there is no way to see what the other person is doing (unless the interview takes place with a web cam), the informant may actually be listening to people around him/her, watching television, eating, or working. What is more, the informant may actually be taking part in more than one chat session at a time. Although most of my informants appeared to be paying attention to the interview, one of them usually participated in interviews while preparing to go home from her place of employment. There were also cases where an informant received telephone calls while participating in the interview. In one case, however, it was not the informant who was otherwise occupied. I had accidentally scheduled two interviews at the same time and did not realize it until both informants paged me and told me they were ready to have the interview. Fortunately, IM clients generally allow chats to take place in two or more windows at the same time without any of the participants realizing that one of them is involved in more than one session at the same time. In the previous chapter (Controlled Chaos), I described how this type of activity occurs in almost every weekly chat session. If I had not previously participated in the weekly chat sessions, I am not sure I would have been able to successfully complete this particular interview.
c. Since it is text only, there is a complete record of everything that was said.
In a face-to-face interview, the researcher/ interviewer has to keep a record of what is said, how it is said, and what other (non-verbal) cues were also involved. While part of this (the verbal part) may be tape recorded, the interviewer must still pay attention to whether or not the tape needs to be changed, or if the batteries are good. He/she must also pay attention to keeping notes on the non-verbal parts of the interview. In a virtual interview, on the other hand, there is a complete record of what was said. Since there are no non-verbal cues to be considered, the interviewer can pay closer attention to what is said. He/she also needs to make sure that his/her questions are stated clearly, since the informant will not have non-verbal cues to help decipher the question. At the same time, while the face-to-face interviewer will have to transcribe his/her notes and tapes, the virtual interviewer should format the interview into an easily readable form. Anyone who has seen a raw chat log will realize that it is difficult to read without prior formatting in a word processing program.
5. Virtual interviews require attention to time, including the time it takes to conduct the interview, the time it takes to gain the trust of community members, and a consideration of what time zones are involved in any interview.
In a study that takes place in a virtual space, the researcher needs to consider the time that needs to be dedicated not only to learning the rules of the community, but also to learning how to use the particular programs and chat clients used by the class. No two chat clients or programs are identical. While some clients may be more similar than others, it still takes at least a little time to learn to use them. A person who is used to Yahoo! Messenger, for example, will find ICQ quite different. Even MSN Messenger will be somewhat different. While the researcher who knows one of these programs will not need to spend much time learning to use any of the others, if the client used in the community is vastly different, he/she may have to spend much more time learning how to use it. For example, a researcher who knows how to use ICQ may need to spend a large amount of time learning to use a MOO. Further, in participating in a virtual class, the researcher needs to know the different programs used. WTI used The Palace, Homestead, and HearMe. Each of these was different from the others, and the commands used in one would often interfere in one of the others. For example the function key F4 in HearMe was used to talk. In The Palace, that same key was used to change the person’s avatar. Since WTI used both of these environments in most weekly sessions, and usually at the same time, this interference was something the participants had to learn to cope with. Once these two clients disappeared, the community had to learn to use new clients. A researcher who decides to study a class in which this condition exists must expect to spend a little time learning to use the programs.
Of course, besides learning to use the specific clients, if it is the first time the researcher has participated in a chat room, he/she may be scared off the first few times. At the same time, since the researcher also needs to become familiar with the community norms and rules, the time spent learning the programs (and getting used to chat in general, if necessary) is usually also spent in getting to know the community.
As mentioned above, the informant in a virtual interview has to type his/her answer into the chat client before the interviewer can receive it. This can take a little time (if the informant is a fast typist or the answer is short) or a longer time (if the informant is a slow typist or the answer is long). But it is time that the interviewer must take into consideration in planning his/her interviews. An interview that would last one hour in a face-to-face environment can easily last ninety minutes or more in a virtual environment.
Of course, one aspect of time that may appear self-evident is also that aspect that is often overlooked. The world is divided into 24 time zones. If a researcher and informant in a face-to-face environment in New York City make an appointment for 3:00 P.M. January 5, for example, they both understand the time to be 3:00 P.M. Eastern Standard Time. In a virtual environment, however, they cannot take the time for granted. If this same researcher is in New York and the informant is in Caracas, Venezuela, for example, the following questions need to be answered.
3:00 PM where? In New York? In Caracas? What time zone is Caracas in? Is either of the sites using Summer (Daylight Savings) Time? If the researcher does not answer these questions correctly, the interview might never take place.
A similar problem did occur in my study. It took three attempts before we finally clarified the time zone problem. In WTI it is common for the participants to use Universal Time (UT or GMT). In that way, every member in the community can make an appointment with any other member without worrying about specifying “My time or yours.” In virtual interviews the researcher needs to keep this time zone difference in mind.
6. Since the members of the community will have a wide range of technological equipment and experience, the teacher (or researcher) needs to adjust his/her plans to what is available to a specific student. interviews.
Other than audio and video recorders, there are no technical considerations in face-to-face interviews. Virtual interviews, on the other hand, are generally high-tech. They require the equipment, connections, and knowledge, to use the Internet. At a minimum, this means a computer with a modem, an account with an Internet Service Provider (ISP), and some experience both in browsing the Internet and in using chat rooms. Virtual researchers should keep the following online conditions in mind.
a. Different media use different amounts of bandwidth.
Bandwidth is related to the amount of information that can be sent/received in a particular situation. Stated simply, the closer the environment comes to replicating a face-to-face environment, the more bandwidth is needed to carry the information. In the case of virtual interviews, a text-only interview uses the least amount of bandwidth. After that, in ascending order, comes voice, video with text and video with voice. The main condition that should be considered in relation to bandwidth is the necessity to match the medium the interviewer wishes to use with the bandwidth that the informant has available. If the interviewer wishes to use video with voice, but the informant has the bandwidth for only text, the interview will need to be carried out in text.
b. Connection speed varies with the time and the amount of bandwidth being consumed.
There is a saying that the Internet slows down everyplace else after 8:00 on a California morning. This is simply a way of saying that as more people connect, the available bandwidth must be divided among more users. But a computer connection can also seem slow because computer resources are low. If a researcher has scheduled an interview during peak hours, the speed with which he/she and his/her informant communicate may be quite slow. This also happens if the person is trying to run too many programs as one time. In one example that took place during a regular WTI chat session, two of the members decided to use their web cams while also communicating in voice chat and taking part in a text chat at Tapped In. When I tried to view the two web cam videos, the voices became choppy, one of the videos froze and the other became jerky, and text in the Tapped In session began to take several seconds to appear in the screen. Finally, my computer froze and had to be restarted. This appeared to be a case of system resources being used up. This is a condition that should be taken into consideration when planning any virtual interview.
c. Internet connections may disappear for what appears to be for no reason at all.
A third problem that needs to be taken into consideration is the possibility of either the interviewer or the informant being disconnected from the Internet. During two interviews, I lost my connection several times during the course of the interview. In many chat clients, a loss of connection will empty the buffer, thereby causing a loss of data. One reason I preferred using ICQ for most interviews was that all sessions with a particular informant was saved in chronological order. All I had to do was open the history folder and the data was there.
Related to this problem is the inability for either the interviewer or the informant to connect to the Internet. The few times this occurred during my study, it meant using email to reschedule the interview.
d. Sometimes a particular chat client is unavailable.
Related to the above problems is the inability of either the researcher or the informant to connect to a specific chat client. This inability to use a particular client may be due to equipment problems, due to the chat client’s server being shut down for maintenance, or due to the client being blocked by the researcher’s or informant’s server. The one time this occurred during my study, I was fortunate to have made plans to use an alternate chat client, so my informant and I were still able to connect and have our interview session. This supports the old rule to always have a back-up plan.
The virtual ESL/EFL classroom appears to be gaining ground. Although Coghlan (personal communication) mentioned that the researchers he has read all indicate that ESL/EFL students prefer dual mode (face-to-face classes with an online portion), virtual ESL/EFL classes are appearing not only in free language schools and classes such as VOES and WTI, but also in universities around the world. Admittedly, most of the university sponsored virtual ESL/EFL classes require physical contact at some point (an orientation meeting, for example), enough virtual classes are appearing to make research into this type of community a more common experience than it was previously.
At the same time, more researchers are carrying out virtual research studies, especially in the area of virtual communities. For this reason, it is important to see what problems these researchers might face and to give recommendations on how they might be met.
The following are specific recommendations for virtual ESL/EFL teachers and students as well as for virtual researchers. I have divided these recommendations into three groups depending on whether they are specific to the virtual classroom, or virtual methodology, or if they apply to both the virtual classroom and virtual methodology.
General Recommendations for Virtual ESL/EFL Classes and Virtual Interviews
Have a Back-up Plan
Teachers must not assume that the virtual classroom will always work. Numerous technical problems can, and occasionally will, occur without warning, thereby causing the class to hang up for the day. The same is true of virtual interviews. Technical problems can cause an interview to either be cut short or cancelled altogether. For this reason, back-up plans are indispensable in both teaching and research.
Imagine planning for a class in which you will use an overhead transparency projector, but when you arrive in the classroom, you find that the projector is not working. In such a case, you can use the chalkboard or some other lower-tech means of giving your class. If something similar happens in a virtual ESL/EFL class (you’ve prepared PowerPoint slides and placed them on a WWW page, but the students cannot access that page) however, it takes much more time to change to another form of presentation. This should be considered and planned for from the beginning. One way might be to put the presentation on more than one site; another might be to have a text-only presentation also available. But, without a back-up plan, the class may have to be cancelled for the day.
Imagine, also, that you get to your class only to find your room locked, and nobody knows where the key is. A similar occurrence often occurs in a virtual class when a specific site where the class normally meets is unavailable. This can occur for a variety of reasons including (but not limited to) the server being down for maintenance, bandwidth problems (too many people trying to access the site at the same time), or the site may have been blocked for some reason. It is important, therefore, for the teacher and students to have (and know about) at least one additional site they can access for the class.
In virtual interviews, it is often common to find that the site or program being used is unavailable (or that one or the other person cannot connect to the internet). In this case, the researcher should have made arrangements with the informant on additional sites to meet, or to meet at another time. In some cases, it may even be useful to plan for an asynchronous (email) session to take the place of the session that had to be cancelled due to technical problems.
Establish Ground Rules at the Start
One problem in both classes and research is when the teacher (researcher) and the student (informant) make up the rules as they go along. By establishing ground rules before beginning either the class or the research project, both parties can dedicate more time to the task at hand without worrying about what rule would cover it. For example, in a class, is the student expected to remain on-topic, or is it permissible to change topics at will? How long will each class session last? In an interview there are similar questions that need to be addressed. How many interviews will there be? How long will each interview last? What mode of interviewing will be used (voice, text, video, a mixture of all three)? Will the interview be formal (question and answer format) or informal (conversation format)? Answering these questions at the beginning will allow both participants to know what to expect.
Allow Sufficient Time
As mentioned above, time is a very important part of any virtual class or research study. Students and teachers (as well as informants and researchers) need sufficient time to get to know each other. But they also need time to get to know the programs being used in the class or study. While many people may feel that a virtual class or study is easier or takes less time, they are really much more time consuming than might be expected.
By allowing sufficient time for everybody to be comfortable with the class or study, both the class and the research project can be more useful and interesting to all involved.
Take Time Zone Differences into Consideration
While this may appear to be self-evident, many teachers and researchers seem to forget that it is a different time at different places in the world. By keeping this in mind, both the teacher and the researcher can make plans to use this in their favor. In the case of WTI, this problem is handled by using GMT as the basis for time, making active participation in the chat sessions optional, and incorporating the use of IM software for one-to-one sessions among students (and teachers) who are on-line at the same time. Something similar can often be included in many classes. Further, the time difference needs also to be taken into consideration in planning virtual interviews. It makes little sense to plan for a 3:30 p.m. interview, for example, if it means one of the two participants will either be working or asleep.
Recommendations for Offering Online ESL/EFL Classes
Provide for Multiple Means of Participation, but Make Specific Means of Participation Optional
Time zones will often result in some students always being unable to participate at a particular hour. This problem can be countered By planning for both synchronous and asynchronous participation. However, if students are required to participate in both modes, the problem can actually be compounded, and students may find themselves dreading the class. Unless there is a large amount of money involved, those students who find participation onerous will tend to drop out of the class. By allowing students to participate in either mode (although preferably in both), however, they class can become an enjoyable experience in which students are eager to participate.
WTI arranged for this time zone problem by making the e-class (asynchronous) portion of the class the mode that decided if a students was really in the class or not. Since the synchronous portion is so important, however, the teachers also encourage those students who are unable to attend the chat sessions to communicate using IM programs and by reading (and commenting on) the weekly chat logs.
Provide a Means for Students (and Teachers) to Get to Know Each Other
When a course is offered completely on-line to people in all parts of the world, face-to-face meetings become difficult, or even impossible. Allow some means for the students to get to know each other. One way to allow for them to get to know each other is by encouraging them to hold individual chat sessions outside of class time. It is also possible to design writing assignments (to be posted to the class discussion area) in which the students describe something about themselves or the culture they live in. The use of voice chat and video is another way to allow students to get to know each other. Hearing a voice or seeing a video of a classmate can go far in allowing students to realize that the person he/she is talking to is really a person and not just the computer. Finally, consider placing pictures on a web page. Like participation in the chat sessions, however, this should be done on a voluntary basis, where students who want to can send in their pictures for the teacher to place on the page.
Recommendations for Conducting Virtual Interviews
Provide Extra Time for the Interviews
As mentioned earlier, the informant in a virtual interview has to type his/her answer into the chat client before the interviewer can receive it. This can take a little time (if the informant is a fast typist or the answer is short) or a longer time (if the informant is a slow typist or the answer is long). But it is time that the interviewer must take into consideration in planning his/her interviews. An interview that would last one hour in a face-to-face environment can easily last ninety minutes or more in a virtual environment.
Consider Using Email as an Interview Technique
As mentioned earlier, B. Mason (1996, 1999), Slater (message on Virtual Interviews posted to the Virtual Methods discussion list, 27 Feb 2000), Waern (message on Virtual Interviews posted to the Virtual Methods discussion list, 27 Feb 2000), and Warschauer (1996a) all point out that people tend to write more in asynchronous communications than in chat rooms. Further, as Waern points out, chat rooms are not the most interview friendly places that exist. Judicious use of e-mail, at least for follow-up questions and as a back-up plan in case the chat room becomes unavailable, can go a long way to allowing an interviewer to gather the data he/she requires. At the same time, these should be used judiciously since, as B. Mason (1996, 1999) points out, email interviews have the danger of leading to information overload due to the number of email messages being received.
Possibilities for Future Research
Among the many possibilities for future research in this area, I point out four studies that I find of particular interest.
Although this study is based specifically on the development of a community from an ESL/EFL class, it does not consider either how the student learns the language within the community nor how well a student improves his/her English. Therefore, one study I find of particular interest would focus on how students learn in a virtual language learning community. This study could easily be carried out through interviews with students in which they describe how they participate in the community, especially in the relation to how they answer assignments and prepare for participation. It might also be helpful to interview the teachers about how the students in general participate and answer assignments. Related to that is a second possible study that evaluates the learning outcomes of ESL/EFL students in virtual language learning communities. This could be carried out through an analysis of the student’s work (both assignments and chat room participation) and improvement over a period of time.
Two other possible studies would analyze these same area in both virtual language classes that are offered in a more traditional manner. My study specifically describes a community in which students do not have a series of lessons to be learned with evaluation activities that are carried out. How would students participate in a class that included these elements? Related to this would be a study in a virtual language class as offered by an institution of higher (tertiary) education.
Throughout my participation in WTI, the coordinating teacher, Stonners, kept saying that online classes, especially WTI, are a lot like herding cats. Anybody who has ever owned cats knows how independent they are. Cats have a mind of their own. Try to make two or more cats go where you want them, and you will find they each go their separate way, no matter what you do. I found this to be the perfect metaphor for understanding what goes on in this class/community. Students (and teachers) insisted on opening their own topics, changing topics, switching chat clients, and in general doing their own thing. I would like to express my appreciation of Stonners for allowing me to use his term as my metaphor.
I also wish to thank Genevieve, another researcher who takes part in the TappedIn MOO environment for permission to use “controlled chaos” as my title for chapter 4. She is studying online interaction and uses this term to describe how people interact in chat and MOO environments.
Both controlled chaos and herding cats are very descriptive of the WTI community. The classes are definitely chaotic (although the members are able to navigate through this chaos as though it made plain sense). But just like cats, the students (and teachers) are independent, self-confident people, who cannot be herded. They are not sheep or cows that will follow the directions of a leader. This freedom of thought and action is, in my opinion, the community’s greatest strength.