«Item type text; Electronic Dissertation Authors Goertler, Senta Publisher The University of Arizona. Rights Copyright © is held by the author. ...»
EveningTeacher, on the other hand, used teacher turns to expand the tasks. Based on my qualitative analysis of the teacher moves in the case study subjects’ transcripts, I argue that the MorningTeacher kept the tasks more narrowly focused, while EveningTeacher made them broader.
Table 4.14 Teacher Moves
In summary, there were differences between the two teachers in how they participated in the chatting. MorningTeacher experienced more problems than the EveningTeacher. She walked around more in the physical space than EveningTeacher and made less commentary in the physical space. A group difference was found in regards to teacher output, with ESC students being exposed to the most and NSC students exposed to the fewest teacher words per minute. Since there was a difference between groups, it cannot be said whether the significant difference between the two teachers is due to a teacher difference, a group difference, or a combination of the two, MorningTeacher used significantly more target language than EveningTeacher, and had significantly fewer errors in the transcripts analyzed. Furthermore, EveningTeacher’s errors were mostly systematic, whereas MorningTeacher’s usually were not. In terms of the language addressed to the students, MorningTeacher used significantly more feedback in response to errors in the transcripts analyzed and had a clear highest frequency feedback style. EveningTeacher used less feedback and had no clear feedback style. In addition, MorningTeacher’s teacher moves focused on keeping the conversation narrowly defined within the task, whereas EveningTeacher’s teacher moves modeled a broad task definition.
While some results were discovered in response to research question 1 a, the data analysis faced limitations. First of all, this study is a case study, and was only investigating two teachers; no generalizations can be drawn from the results.
Furthermore, some of the data was only taken from the six case study subjects and not from the other students in the class. While in reviewing the other transcripts, the same patterns appear to be present, this can only be determined by analyzing all transcripts from all students. Another challenge for this case study was the loss of transcripts in NSC and SSC, which resulted in an incomplete picture. Additionally, as is the nature with observations, the teachers and students may have acted differently in my presence than during the chat sessions, when I was not present. Despite these limitations, the data provided an in-depth impression of how the two teachers interacted in the transcripts of the six case study subjects.
As has been discussed, differences have been discovered between the teachers and also between the different levels of support such as interaction style and student output.
However, due to the discussed limitations, it is difficult to identify a definite source for these differences.
4.2.2 Research Question 1b The research question is: What appears to be the teacher’s definition of her role, as evidenced by participation style? As mentioned in response to the previous question, the two teachers employed different feedback and participation styles. Based on these findings, I argue that MorningTeacher saw her role primarily as a teacher, i.e., assisting where necessary, allowing students room to explore, yet holding them accountable to her implied guidelines, and facilitating learning and conversation where applicable.
EveningTeacher, on the other hand, appeared to see herself more in the role of conversation participant, i.e., sharing her own experiences, expanding the topic, and focusing less on limiting the conversation of the students.
This implied definition of the teacher roles poses some problems in consideration of prior research on CMC. One factor discussed in chapter two is the democratization or equalization effect of CMC. One could argue that since MorningTeacher limited the content and language of the students, and served as a source of feedback, she was taking on a teacher role that made use of power relationship usually found in traditional classroom settings. Furthermore, since she used the “invisible” function, essentially “spying” on her students’ chat sessions, she could also be considered to be exerting a type of power granted only to her as a teacher in this context. However, due to her more limited output, she allowed the students to explore their own language, which could be an argument for her working to equalize participation. EveningTeacher on the other hand, could be argued to be equalizing participation, for she primarily was a conversation partner to the students. However, since she produced more output, she may have limited the opportunity for more equalized participation by the students. Furthermore, the use of
the “to all” function could also be considered a sign of power for the following reasons:
(1) only the teacher can use this function, and (2) messages posted using the “to all” function most likely interrupt the flow of conversation in the chat rooms. Hence both teachers, though in two different ways, had a participation style that on the one hand attempted to equalize participation, on the other hand, still established their authority role in the classroom. The complexity and pedagogical ramifications of these uses of authority and attempts to equalize participation will be discussed in more depth in Chapter 5.
4.2.3 Research Question 1c The research question is: What form does corrective feedback take during chatting in this study? To address this research question, I will refer back to the findings of teacher feedback discussed in research question 1 a, and also discuss the findings of peer and self-correction in the transcript sets of the six case study subjects. First, the error sources and the feedback sources are explained. Then the feedback rate by feedback and error source will be presented and discussed. Thirdly, the forms of feedback will be presented with examples, and finally the forms of feedback used by different sources (teacher, self, or peer) will be discussed. All feedback rates are only taken from the transcript sets of the case study subjects.
In a transcript, feedback can is evidence in response to an error made by the case study subject, one of the case study subject’s peers, or the teacher. Furthermore, feedback can be given by oneself, another student, or the teacher. Feedback given by other students in response to a peer mistake were divided into feedback given by other peers, and feedback given by the case study subject; the same was true for feedback given to the teacher by a student. Errors made by the teacher or a peer are labeled as observed errors from the perspective of the case study subject. By the same token, feedback given to other students or to the teacher by a peer or the teacher, are labeled as observed corrections from the perspective of the case study subject.
Feedback rates were calculated by counting the number of mistakes, and the number of instances of feedback. Percentages of errors receiving feedback were established. The feedback rates are listed separately according to source of error and source of feedback to provide a more detailed picture (see table 4.15.) and are discussed in the following section.
Table 4.15 Feedback Rates by Transcript Set
Overall corrective feedback moves were infrequent in comparison to the number of errors. 5.71% of NSCDanielle’s errors received corrective feedback (0.95% by the teacher, 3.81% by a student, and 0.95% as self-correction) and 11.32% of her observed student errors received corrective feedback (2.83% by the teacher, 0% by a student, 7.55% by NSCDanielle, and 0.94% as self-correction). 8.77% of NSCJennifer’s errors received corrective feedback (2.63% by the teacher, 3.51 by a student, and 2.63% as selfcorrection) and 10.77% of her observed student errors received feedback (2.31% by the teacher, 0% by a student, 7.70% by NSCJennifer, and 0.77% as self-correction). Neither of the two teacher errors observed by NSCJennifer received feedback.
29.17% of SSCEmily’s errors received corrective feedback (10.42% by the teacher, 2.08 by a student, and 16.67% as self-correction) and 12.61% of her observed student errors received feedback (5.04% by the teacher, 0.84 by a student, 4.20% by SSCEmily, and 2.52% as self-correction). 100% of all teacher errors received corrective feedback by SSCEmily. Below is an example how SSCEmily uses modeling to correct her teacher.
MorningTeacher: Wenn ich keine Zeit habe, *haben ich leider auch kein Frühstück.
SSCVeronica: ich auch.
SSCEmily: Ich auch. WEnn ich keine Zeit zum Frühstück habe, esse ich ein frühes Mittagessen.
3.51% of SSCGina’s errors received corrective feedback (1.75% by the teacher, 0% by a student, and 1.75% as self-correction) and 8.70% of her observed student errors received feedback (2.90% by the teacher, 0.73% by a student, 2.90% by SSCGina, and 2.17% as self-correction) of the time.
2.60% of ESCAmanda’s errors received corrective feedback (0.65% by the teacher, 0.65 by a student, and 1.30% as self-correction) and 2.37% of her observed student errors received feedback (0.40% by the teacher, 0.79% by a student, 0.79% by ESCAmanda, and 0.40% as self-correction). None of the teacher errors received corrective feedback. 6.67% of ESCVictoria’s errors received corrective feedback (2.96% by the teacher, 2.96 by a student, and 0.74% as self-correction) and 3.88% of her observed student errors received feedback (0% by the teacher, 2.7% by a student, 0% by ESCVictoria, and 1.16%% as self-correction). 8% of the teacher errors received corrective feedback from one of the other students.
Overall corrective feedback was infrequent. Teacher feedback ranged from 0% to 10.42%, combined peer feedback (combining peer feedback by other students and the case study subject) ranged from 1.58% to 7.78%, and student self-correction from 0% to 16.67% in the case study subjects’ transcripts. Other-initiated feedback (combined peer feedback and teacher feedback) ranged from 1.30% to 12.50%. This means that even in the highest frequency, a subject’s errors only received 25 corrective feedback moves for every 200 errors as exhibited in the case study subjects’ transcripts.
After discussing the rate to which feedback was given, I now discuss the feedback forms that were given. Table 4.16 illustrates the different feedback forms with examples.
Corrective Feedback Moves
As can be seen in table 4.16. a variety of feedback forms were used. However, different feedback types were used by different feedback sources (see table 4.17. for an overview). As mentioned in response to question 1a, MorningTeacher used more feedback than EveningTeacher, though still a low amount of feedback considering the number of student errors. Furthermore, MorningTeacher used one feedback style most frequently, which was assumed to be her preferred feedback style (repetitions with correction). EveningTeacher, on the other hand, displayed a larger variety of feedback types and thus did not appear to have a preferred feedback style.
Feedback Form by Source
When correcting themselves, students most frequently used repetitions (25), especially attempted (7) and successful (7) marked partial repetitions. However, when correcting peers students most frequently used attempted (4) and successful (28) models.
One way of interpreting these raw numbers from the case study subjects’ transcripts is to argue that students want to make sure that others recognize when they have discovered their own mistakes, by marking the correction, usually with the messaging convention of “*,”—asterisk. However, when correcting others they may want to use an unintrusive form of feedback such as models. In reviewing the data, one should point out that not all models may have been intended as corrections by the writer, because the repetitive nature of many of the tasks simply made models likely. For example, the previously mentioned activity of answering and asking questions about historic events in the past tense required the same structures in each question and answer. Hence, a writer may model the correct form following an error, but it may just be a sign of answering or asking the next or even the same question, rather than providing corrective feedback, as illustrated in Table 4.16 above. Despite this caution, it can be argued that the most frequent form of self-correction is all forms of repetition, but especially marked partial repetition with correction. The most frequent peer feedback form was modeling. These patterns of feedback were the same across all three classes.
In conclusion, rate of feedback from all sides was low considering the number of errors made by the students. However, MorningTeacher provided more feedback than EveningTeacher. In addition, MorningTeacher used repetitions and models more frequently than any other feedback form and is therefore referred to as having a systematic feedback style. EveningTeacher’s feedback style, on the other hand, is considered unsystematic, as she used a variety of feedback forms with no one style predominating. Self-corrections were mainly in the form of marked repetitions, while peer feedback was mostly in the form of models. These results have to be viewed with some caution, because only the case study subjects’ transcripts were consulted. While the transcripts reflect both feedback given to peers and to the case study subjects, the feedback given to the peers is only a sample and not representative of everyone of their chat transcripts. However, while no quantitative data was taken from the other transcripts, detailed readings from the other transcripts provided the same impression, though it cannot be confirmed with statistical analysis.
4.3.1 Research Question 2a The research question is: What influence does corrective feedback style have on students’ learning, as perceived through language production during chat as measured through word count? To address this research question, the word count from all subjects in all classes is considered in relation to the dominant feedback style used by the teacher.