C hat as a Follow-up to Timed-Conversations : An Integrated Approach

This article introduces a technique of using online chat to focus students’ attention on problems of grammar and word choice within the context of meaningful linguistic exchanges. It outlines a process of having in-class student-to-student conversations with a specific target expression. That conversation is then repeated as a homework assignment via online chat. Using the transcript of the chat the teacher gives feedback to the students, who in the following lesson incorporate that feedback into the next conversation practice. After the description of the teaching process, this article introduces the voices of some of the participants in this study to illustrate the effectiveness of the exercise.

H ow can we use communicative activities, like conversations to improve students' grammatical understanding and accuracy? This article outlines a teaching technique that offers students feedback on grammar usage and word choice in conversation activities by utilizing online chat as a homework assignment. The author developed this technique in response to the large number of errors heard during in-class conversation activities. The author hoped to address these problems, not through a separate activity, but within the context of the conversation itself. This is where online chat proved to be an effective tool.
The use of online chat-rooms as a means for language learning is not new. Chat has been of interest to language teachers since it appeared in the 1990s. As exemplified by Beatty (2003), Butler-Pascoe and Wiburg (2003), it would be hard to find a general book on CALL that does not treat chat-rooms. However, while the existing research introduces the idea that students can "type-talk" to others over the world-wide-web, it does not give many suggestions on how to integrate that communicative activity into a lesson, in a way that will lead to a richer SLA experience. That is the goal of this article.
The technique discussed below was developed with two English classes at a university in Nagoya, Japan, that met in the spring and fall of 2006. The students were mathematics and information technology majors who met twice weekly for ninety minutes each. Both classes consisted of approximately 25 moderately motivated students, with average TOEIC scores of around 380. The author used Moodle to moderate chats among other activities. The chat system on Moodle stores the chat transcript so that the teacher can examine it later. Therefore this article is intended for foreign language teachers whose students have access to an online chat system that allows the instructor to view and printout records of chat conversations, like the system available on Moodle.

Improving accuracy through error management and error correction
The in-class conversation exercise that the students in this study were doing was an activity known as 'timed-conversation' as it is introduced in Kenny (2003) and featured prominently in Kenny and Woo's (1999) textbook, Nice Talking with You. In 'timed-conversations' students speak in pairs, exchanging information on teacher-provided topics for between two and four minutes. Many of the errors heard in those conversations would hinder communication or lead to miscommunication. On hearing such problems, the author made efforts to manage the errors, in order to boost students' awareness of them, and encourage students to address them.
At the center of error management is error correction. Error correction itself is a highly debated topic, one of the major disagreements being whether or not teacher corrections actually lead to improvements in accuracy. Freiermuth (1997) suggests that there are two camps in the error correction debate: those who claim errors should be addressed on the spot and those who believe that errors should be ignored. On the one hand, there is research indicating that spot-corrections could hinder the learning experience of many students. Chastain (1976) argues that constant correction could cause students to lose the desire to communicate in the target language. Conversely, Tomasello and Herron (1989) have shown that in-class corrections can lead to improvements in grammatical accuracy. Lightbrown and Spada (1990) also report students in an intensive language program in Quebec attained a higher level of production accuracy if they received formal correction of errors from the teacher in class. However, Bartram and Walton (1991) throw the value of corrections into doubt again, by noting that students who are corrected in class will place a premium on avoiding mistakes at the expense of communication. This, they argue, may lead students to avoid taking chances with more difficult expressions in front of the teacher, thus making their language appear accurate, but limiting what they can say, and slowing language acquisition.
Although the question of error correction is too complex to discuss in detail here, the brief overview above indicates that feedback concerning errors made when speaking is helpful in improving accuracy and communication skills. The goal then is to find a way to make students aware of their errors while avoiding the negative psychological aspects of error correction. In other words, how can language teachers make error correction into a positive learning experience, and not an exposition of students' weak points?
The author attempted a number of techniques of error correction in these two classes, before the introduction of chat explained below. One technique was error prevention. Problematic expressions were introduced before the conversation. This technique helped to avoid some errors, but did not satisfy the author, particularly because many errors were unpredictable. Moreover, a large number of students still made the very errors that they had been warned against moments before.
Another technique used was interrupting conversation to inform students that an error had been made, and giving spot-corrections when students could not correct themselves, but this kind of interruption can create a face-threatening situation for the student who has been singled out. Moreover, when conversations were interrupted by the instructor, they lost momentum as students concentrated on correcting form errors.
Another approach that the author used was listening for common problems in a number of the conversations that were going on simultaneously, and then later introducing those problems on the board. However, there were few errors that a majority of students were making. Therefore, the whole class was forced to review the errors of a minority or even the error of one student.
Finally, a forth technique used was encouraging peer-correction. Giving students the task of listening out for errors in their partner's speaking was intended to heighten the awareness of accuracy for both speaker and error corrector. This final approach resolved the problem of face-threats, however, it and the previous approach were both lacking in personal attention from the teacher. Furthermore, many students may prefer to have direct teacher-corrections as noted by Cathcart and Olsen (1976).
With each of the above techniques having its particular disadvantage, the author searched for a method of error management that would help to systematically analyze a large number of students' language while allowing in-class communication to continue, spare students from embarrassment and be time efficient. Possible techniques included having students reflect on their language by writing conversations out in class after executing them verbally. However due to the short time between teacher explanation and student writing, success in this kind of activity may reflect short-term memory rather than real learning. As Sökmen (1997, p. 242) has noted, a longer interval between meetings of particular expressions increases the likelihood of storage in long-term memory. Therefore, a homework task appeared to be more useful in improving accuracy than in-class reflection. Another consideration was that writing exercises targeting a particular form or expression may not help improve oral communication accuracy due to the reflective nature of writing in contrast to the impromptu nature of conversation. Accordingly, the author searched for a homework assignment that would both simulate a conversation, and could be checked by the instructor and later presented to the students. The result was the online "chat as a follow-up to timed-conversation" teaching pattern discussed below.

From timed-conversation to chat
The process of the teaching technique the author used is as follows.
Introduce the target expression or form: 1.
Make sure that the students understand the target expression used.
Select a topic of conversation that will elicit the target expression. Give students some time to prepare for the conversation. They may check dictionaries for useful vocabulary words, or ask questions to the teacher or classmates.

Timed-conversation: 3.
Students speak in pairs. At the end of three minutes stop them and have them reflect on their speech.

Switch partners: 4.
Students change partners and conduct conversations again. Remind students to avoid errors regarding the target expression.
Explain to students that they are to find a new conversation partner and have the same conversation using online chat. The chat should continue for about ten to fifteen minutes, as typing takes longer than speaking. Instruct the students to do the conversation just as if they are meeting their partner face to face. Tell students to finish their chats at least 24 hours before the next class, or whatever time limit is convenient for the teacher and the students.
Check the chat: 6.
After the time limit has expired, print out the conversations and note errors directly on the printout. It is recommended that errors are not actually corrected, but the existence of errors be noted using editing symbols. In this way students will need to be more active in correcting their own errors.
In the second class, have students sit side by side with their homework partners. Hand back the printout. Ask students to talk about the errors noted with their partners. Then students should correct their own errors, and check with the instructor to see if they have made the right corrections.

Revision: 8.
have students find new conversation partners and redo the conversation. Encourage them to avoid making the errors that have been noted on the printout.

An example of chat error correction
For the purpose of this research, the author chose the 2 nd conditional (e.g., "If I won the lottery, I would buy a Rolls Royce."). This is a grammar point that many Japanese students consistently fail to use correctly in conversation and writing. After reviewing the grammar point using a textbook, the students did timed-conversations with the theme, "if the world was going to end in five days, and there was nothing any human could do to stop it, how would you spend that time?" The author followed the process outlined in the above section. When checking the homework, it was clear that students were still making many errors with this grammar point (see example below). At the next meeting, the author returned the chat printouts with editing marks indicating errors and allowed students to discuss problems with partners before correcting them on their own. After corrections were made and double checked by the author, the same conversation topic was redone in class. Finally students did another timed-conversation, with a similar topic, "what would you do if you won the lottery?" in order to have more practice with the target form in the context of a conversation.

Example of chat printout with teacher's comments
Below is an example of a printout of a chat done by two students participating in the study. Their names have been changed and the instructor's comments have been retyped to improve readability. The abbreviations used below are "W.C.": 'word choice'; "W.F.": 'word form' including all verb problems; and " ": 'add a word or phrase'. As can be seen above, the existence of errors of the target expression has been noted. Additionally, some grammar and word choice errors beyond the target expression have also been pointed out. However, the students themselves must make the actual corrections. Now, I would like to move on to a discussion of the merits of the present activity.

Justification for using chat as a follow-up to timed-conversation
The use of chat as a follow-up to timed-conversation offers a number of benefits when compared to the techniques of error management that were discussed earlier. Firstly, the chat allows the instructor to spend class time reacting to the content of conversations, ("You went to the beach, too?") rather than interrupting conversations to point out problems of form. Additionally, this technique is much less face threatening than spotcorrections, and it is not disruptive to the flow of the conversation either. It also allows the instructor to spend equal time on all students.

∨ ∨
Compared to post-conversation lecturing on commonly made errors that the instructor has heard, the above chat correction allows the instructor to make corrections personalized to each student. Also students who erroneously believe they were making no errors may disregard lectures about frequently made mistakes, as 'other students' problems'. Another positive aspect is that the interactivity of the chat forces students to react in realtime, thus the chat becomes a communicative rather than a repetitive activity. An additional benefit is that students are provided with a chance to learn by seeing peer examples, which would be more difficult with homework assignments done in isolation. Moreover, doing homework by chat fosters interpersonal skills, cooperation, collaboration and community building.
One further advantage of the chat is the quick turn around from when the students do the conversations online to when they get their feedback. However, the most important point is that the linguistic interchange persists. Unlike oral communication, which disappears instantly (unless recorded and typed out, a laborious and time consuming activity), the chat is stored and available for the instructor and student alike to analyze immediately or make use of at a later date. All of these advantages help to provide opportunities for the students to improve their accuracy and communicative skill. For these reasons, the author believes that this use of chat promotes a richer SLA experience.

Student Feedback
Finally to illustrate how this activity was perceived by the students, I would like to introduce the voices of some of the participants in this study. At the end of the second meeting, students were asked to answer three short questions about their impression of the homework.
Did you enjoy the chat exercise? Why or why not? 1.
Do you think that the chat exercise helped you to improve your English accuracy? If 2.
so, how?
Please write any comment or express your feeling when you did your homework. 3.

Analysis of responses to question #1
The motivation for the first question arose from the principle that students will learn more if they are enjoying what they are doing. As Dörnyei (2001, p. 73-76) argued, having the students do tasks that are enjoyable to them is one key of motivating students to learn. Of the 54 responses received, 45 were positive and 9 were negative. Some examples of student responses were as follows: "… I could communicate with friends by the new method." • "… I can talk English with my friend. I enjoyed talking with my friend." • "…It's very difficult for us to talk to our friends in English. So using a chat is a nice • way. However, we have to make time to do that at least two persons have to be on the net in the same time.
These three answers indicate that the students enjoyed communicating in English outside of class with their peers. Other students mentioned that it was helpful for typing skills and writing skills, which are both examples of "collateral learning" (my neologism) -learning that was not originally intended by the instructor. Among the nine negative responses, one student criticized the fact that the same topic was repeated in class and on the homework. Based on this some teachers may find it preferable to assign a similar topic eliciting the same kind of target expression rather than to repeat the classroom topic. Alternatively, it might be a good idea to offer a range of topics to choose from, one of them being the same topic as that used in class.

Analysis of responses to question #2
In response to the second question, "Do you think that the chat exercise helped you to improve your English accuracy? How?" In all, 48 students answered positively, and 6 students answered negatively. Among the students who felt that the exercise helped their accuracy, a number of students commented that the personalized corrections were helpful.
"Yes! I think typing can help us and after chat, you check our sentence and then we • can realize our mistakes.
"Yes I do. The reason is that I have to write many sentences on it and you check and • return." Another student seemed to recognize the problem that led the author to introduce the chat homework in the first place.
"Yes. Because, if we did speaking only, we would not found our mistakes." • Students' opinions on whether or not their accuracy had improved is not necessarily an indication of real improvement. However, it does indicate that students themselves are becoming aware of their errors, as well as the fact that steps can be taken to improve accuracy. Also, belief in improvement is linked to self-confidence, which in turn leads to greater interest and motivation as Dörnyei (2001, pp. 90-92) has argued.

Analysis of responses to question #3
Finally the third question of the student feedback, "Please write any comment or express your feeling when you did your homework," was intended to elicit free responses that might give some hints on improving the activity. Overall the students were quite positive about the chat homework, and some students even requested to do chat on a weekly basis. It appears that many students truly enjoyed the chat and appreciated the feedback given, as seen in the following comments: "When we do chat, we do like conversation." • "We can learn vocabulary." • "It's like a real time talking! Very interesting." • "Typing words and checking my friend's sentences were very interesting." • "Chat is very interesting. When I do this chat, my feeling is pleasure. So I want to do • this chat every homework.
Finally, I will consider some problems that were exposed in the process of the activity as reflected in the final question of the student feedback. One of the concerns raised by a few students was the difficulty in arranging schedules so that the chat could be done. Another problem a few students pointed out was that despite the fact that the participants in the study all had laptop computers provided by the university, and all of them were math and information science students; some did not have Internet access at their homes. Also the server slowed down for a few hours during the week, making the chat system unusable. These are issues that must be dealt with in the future. However, despite scheduling and technical problems, it is important to note that only one student among fifty-five failed to do the homework assignment.

Conclusion
In this short study, chat as a follow-up to timed-conversations benefited the students in a number of ways. It gave them an activity that they themselves characterized as "interesting", and "enjoyable". It placed focus on form within a context of interpersonal communication and it allowed the instructor to give advice about oral communication to each student on a one to one basis without using up class time. The process alleviated aspects of error management that might cause anxiety or pressure among the students, and allowed students to focus on improving their accuracy. After students corrected their own mistakes and executed the conversation for the final time, they were able to see their own improvement, and thus the process turned error correction into a positive learning experience. Finally, from a pedagogical viewpoint this was a productive activity, however it was also a rare example of a homework assignment in an English course, which mathematics majors requested to do again in subsequent meetings.