«Item type text; Electronic Dissertation Authors Goertler, Senta Publisher The University of Arizona. Rights Copyright © is held by the author. ...»
However, teachers have several different participation styles at their disposal. As was discussed in chapter four, while there were differences between the two teachers’ participation styles, there were fewer differences among the students’ experiences and language use. Hence, even though choosing one participation style over another may not significantly affect the students’ experiences or language learning, a teacher still has to make multiple decisions about details of implementation. While one choice over another with the data from this study cannot be shown to be more or less effective, every teaching moment still requires a decision from the teacher, whether it bears consequences or not.
In addition to reporting that they like chatting, students also expressed their clear preference for specific kinds of feedback. However, such preferences were not matched with actual practices. Consequently, a teacher has to make a decision as to whether to accommodate these preferences or not. While corrective feedback was generally infrequent in this study, it came from all sources, i.e., teacher, students, and peers. There was a relationship between the source of feedback and the feedback type. A teacher may want to consider adopting either the peer-feedback or the self-correction feedback type to match the students’ own use of corrective feedback. Teachers in training should receive example feedback moves to use in their teaching. These sample moves could be provided to them in their pre-service or in-service training and have their sources in the body of CMC research or in the experiences of other teachers at the institution. Tables such as provided throughout chapter four summarizing feedback and teacher moves could be used in teacher training to (a) raise teacher trainees’ awareness and (b) have a class or departmental discussion about the moves that best fit the group’s desired teaching philosophy and current SLA theories.
In this study, the amount of teacher input did not have an effect on student output.
Furthermore, in this study more input came from the peers, i.e., as non-target like input, than from the teacher. A teacher may accommodate a situation like this by either participating more, or by using transcript work to identify non-target like language.
Teacher training should discuss the advantages and disadvantages of higher levels of teacher input.
One of the most crucial differences between chatting and other classroom activities is that there are two instructional spaces: the physical and the virtual space. In this study the physical space was used by the teacher mostly for logistical and procedural interactions. Allowing the teachers to experience chatting from a student and from a teacher perspective may help them to raise awareness about these two spaces. Another beneficial activity may be to show teachers in training a video tape of a chat session and have them record the teacher’s behavior in the physical space, so that they can collect ideas. In general, while classroom observations are often part of the requirements of a teaching methodology course, pre-service teachers (in my experience) usually choose and may even be encouraged to only observe regular class-time activities. Since teaching in the lab and implementing CMC activities require different teaching strategies than the regular classroom, such sessions should be part of the classroom observations required in pre-service or in-service methods courses. Furthermore, in-service teaching evaluations should also provide feedback for both types of class sessions.
Another factor related to the medium that has been discussed in previous research, yet was not observed in the transcripts of the case study subject’s in this study, is flaming (Abrams, 2003b). However, there were some indications that rude behavior was present at least in the ESC where the teacher was using more topic-expanding and fewer topicfocusing teacher moves. For example, during one classroom observation, one student repeatedly tried to get EveningTeacher’s attention and finally called me (the researcher) over and told me that her partner was “being mean” to her; in the survey one student reported that some students were “dumm,” and in the transcripts (not from the case study subjects) some potentially offensive language was used such as calling another student “Idiot.” Teachers in training need to be aware of these behaviors and be presented with techniques to prevent or redirect discussions that make one or several students feel uncomfortable, offended, or even disrespected. One such recommendation is to provide clear guidelines at the beginning of the semester that explain acceptable and unacceptable comments. Such guidelines may differ from teacher to teacher, and institution to institution and will have to be developed locally in the same way a student code of conduct is developed locally. Writing such guidelines will require discussions among teachers with experience in CMC.
The last issue that I want to address in respect to teacher training is the “bells and whistles” of chat software. Different chat programs have different features. In the case of the chat server used for this study, the features included the “invisible” function used by MorningTeacher and the “to all” function used by the EveningTeacher. During teacher training, such features and their purposes need to be discussed. From my observations in this study and my own experience, students appear to be accepting of the teacher’s use of the invisible function once they know that the teacher can be invisible, i.e., observing their session without appearing on the list of participants. However, informing students that the teacher has this “lurking” option may be helpful. Some students in the ESC appeared to be experiencing EveningTeacher’s use of the “to all” function as distracting or not relevant for their conversation. From my experience the “to all” function has been very useful, when a teacher tries to do most of the interaction in the virtual space and has announcements for the entire class, such as “class is over now” or an additional explanation of the task in cases when most students appear to be completing the activity incorrectly. However, using the “to all” function to converse with the students is challenging, for it is impossible that the posting matches the conversation in all chat rooms. Presenting teachers in training with transcripts of different implementations of the various features, and discussing them, will provide them with ideas of how and when to use which feature.
In summary, in order to implement chatting effectively in the classroom, teachers need to be specifically trained and prepared for the SCMC context of teaching. Such preparation needs to occur on two levels: (1) technological training, and (2) pedagogical training.
5.9. Connections to SLA Research While some connections to research have already been made in the presentation of research results in this chapter, this section will reiterate these connections more specifically with second language acquisition (SLA) research. The order of discussion items will be identical to the presentation of the previous research in chapter two, thus providing a parallel structure.
As mentioned before, input, output, noticing, and the negotiation of meaning occurring during interaction are the foundations of the interactionist framework and have been claimed to be beneficial for learning (Beauvois, 1992; Chun, 1994; Kelm, 1992;
Kern, 1995; Warchauer, 1996). During the chat sessions, students had the opportunity to receive input both from their peers and to a lesser extent from their more proficient teacher. Furthermore, they had the opportunity to practice their language through output considered an essential step in SLA by Swain & Lapkin (1995). In addition, the written format gave them the opportunity to compare their own sentences with those of others.
Most students reported noticing errors in the other students’ writings, which may have been actual errors, or wrong hypotheses of their own about the target language. The caution that needs to be expressed here is that the students were exposed to more input from their peers, which was not always target like, than from their teacher, which was target-like. In addition negotiation was observed in the transcripts. However, a direct relationship between the opportunities for input, output, noticing, and the negotiation of meaning with a positive effect on language development cannot be established in this study due to the limited amount of time spent chatting in comparison to the time spent with other classroom activities. As Gass (1997) pointed out, some of the effects of interaction may be delayed, and no long-term effects were measured. However, in connection with other research, this study’s results may suggest that the chat acticities completed by the students in this study may be beneficial for language learning. These benefits resulted from activities that provided opportunities for language input, and output, and negotiation of meaning. During the chatting, students had the opportunity to produce L2 output, which as Gass and Selinker (2001) reported, plays the following roles: (1) creating knowledge from semantic to syntactic processing; (2) practicing or applying existing knowledge; (3) creating automaticity; (4) eliciting further input; and (5) testing hypotheses formed and receiving feedback about them in regard to the target language. Several students also stated a paraphrase of (2) above as one of the reasons why they thought that chatting improved their language ability. However, while students may have been able to test their hypotheses, they received only little feedback in response to such testing. Since Long (1996) considers feedback received during conversations beneficial, the low frequency of feedback may decrease the benefits of negotiation in chat.
As has been argued by researchers within an interactionist framework, the negotiation of meaning is considered beneficial for language learning (Doughty & Pica, 1986; Ellis, Tanaka, & Yamazaki, 1994; Gass & Varonis, 1994; Long, 1985; Pica, 1991;
Pica & Doughty, 1985; Pica, Young, & Doughty, 1987) though such effects may not always be immediate (Gass & Varonis, 1994). Some of the strategies identified as components of the negotiation of meaning were also found in the interactions in the case study subjects’ transcripts sets, such as clarification requests, self-repetitions, and otherrepetitions. Furthermore, several researchers claim that negotiation of meaning is the highest in dyads between non-native speakers such as the interactions during the chat session in this study (Gass & Varonis, 1985; Oliver, 2002; Pica & Doughty, 1985;
Varonis & Gass, 1985; Yule & MacDonald, 1990).
Some researchers have argued that not just negotiations experienced directly by learners, but also observed negotiations of meaning may benefit language development (Ellis, Tanaka, & Yamazaki, 1994) though the benefit according to Mackey (1999) is greater if negotiations are directly experienced. Nonetheless, in this study, both observed and experienced errors and feedback moves were discussed as potential contributors to changes of learner language through uptake, error uptake, and hypothesis testing as observed in the detailed analysis of the pre- and post-tests. However, this conclusion is based on only a sample of the data. Firmer conclusions may be reached after analyzing all transcripts, rather than only selected case-study transcripts.
From the perspective of research on corrective feedback, input is divided into positive and negative evidence. In this study, students received positive and negative evidence, but the frequency of negative evidence was low. In addition not all of the positive evidence was target-like, due to the fact that the pairs and groups consisted of non-native speakers.
As had already been mentioned in chapter two, the differentiation between explicit and implicit feedback posed some challenges in regards to the feedback moves during the chat activities. According to Panova and Lyster (2002), explicit feedback is feedback that contains an explicit signal. In the chat environment, repetitions were observed both with an overt signal and without. While repetitions are traditionally considered as implicit feedback, in this study, due to the overt signal, they may be considered explicit feedback. It was decided that the differentiation between explicit and implicit feedback is not as easily applicable to a written medium such as chatting;
furthermore it did not appear to play a role in differentiating the feedback styles of the two teachers, because EveningTeacher seemed to have no clear preferred feedback style based on the case study transcripts. This issue of traditionally implicit feedback forms used with overt signals was also addressed in regards to recasts by Nicholas, Lightbown, and Spada (2001). Similarly to Ene, Goertler, and McBride (2005), this study also found that the use of “*” (asterisk) to mark corrections was common, at least among the students.
Previous research as discussed in chapter two, found that explicit feedback was infrequent (see for example Chun, Day, Chenoweth, & Luppescu, 1982; Seedhouse, 1997;) and implicit feedback was frequent (see for example Morris, 2002;). Since the definitions of implicit versus explicit do not seem applicable for the CMC context, this research can neither be confirmed nor questioned.
Looking further at the definitions of corrective feedback, repetitions and recasts have been differentiated. Repetitions are then repetitions without change and recasts are repetitions that correct the ill-formed part of the phrase (Panova & Lyster, 2002). Yet, Morris (2002) added more features to recasts, as discussed in chapter two. During the review of previous research, the definitions of recasts seemed too varied, and I, hence, adopted the terminology of repetitions with and repetitions without correction. What Nicholas, Lightbown, and Spada (2001) referred to as recasts are referred to in this study as the repetitions with correction introduced with praise used by the MorningTeacher.
Lyster (1998b) reported that recasts (called ‘repetitions with correction’ in this study) were often accompanied with approval of meaning (called ‘praise’ in this study).
Doughty and Varela (1998) described repetitions with corrections as an unintrusive, low affective filter form of feedback. MorningTeacher mentioned the same reasons for preferring such a feedback style.