«Item type text; Electronic Dissertation Authors Goertler, Senta Publisher The University of Arizona. Rights Copyright © is held by the author. ...»
4.4.2 Research Question 3b The question is: What patterns occur in the data between error type and error uptake? One of the fears of CMC is that due to the written format, it may lead to more error uptake. In response to research question 3b, I will first discuss the amount of error uptake in the case study subjects’ transcripts and then discuss any potential patterns between error uptake and the error type. Error uptake is defined in this study as instances in which one student uses the same error(s) with the same word within the same transcript as another student or the teacher. For example, Jennifer: Meine Mutter heist A. und sie hat zwei Kinder. Die Kinder der Mutter sind mich und meinem Bruder N.
Cowboy: Die kinder der Mutter sind mich und meinem brooders J., C. und meiner schwester M.
Error uptake in this study, while still low, was more than initially expected. In NSCDanielle’s transcripts one of her own errors (0.95% of her errors) resulted in error uptake, while 12 observed errors (11.32% of the other’s errors) resulted in uptake (10.38% by NSCDanielle, and 0.94% by other students). In NSCJennifer’s transcripts 7 of her own errors (6.14% of her errors) resulted in error uptake, while 4 observed errors (3.08% of the other’s errors) resulted in error uptake (all by NSCJennifer). In SSCEmily’s transcripts one of her own errors (2.08% of her errors) resulted in error uptake, and 1 error by others (0.84% of the other’s errors) resulted in error uptake by another student. In ESCAmanda’s transcripts 2 of her own errors (1.30% of her errors) resulted in error uptake, while 13 observed errors (5.14% of the other’s errors) resulted in uptake (2.77% by ESCAmanda, and 2.37% by other students). In ESCVictoria’s transcripts none of her own errors resulted in error uptake, while 20 errors (7.75% of the other’s errors) resulted in uptake (5.43% by ESCVictoria, and 2.33% by other students).
In reading through the transcripts, it appeared that weaker students are more susceptible to error uptake for they use other people’s phrases to provide answers to questions in the conversation and are not able to correct the errors. This is a qualitative observation only and at this point is not confirmed by quantitative findings; however, it should be explored in the future.
None of MorningTeacher’s errors resulted in uptake, which can be attributed to the fact that they were not crucial words for the activities. However, 1 of EveningTeacher’s errors in ESCAmanda’s transcripts (4.35% of her errors) resulted in uptake by a student and 7 of EveningTeacher’s errors (28% of her errors) in ESCVictoria’s transcripts resulted in uptake by students (8% by ESCVictoria and 20% by other students). This can be attributed to the fact that EveningTeacher was encouraging the students to use the incorrect words or plural forms that she provided (see example below).
EveningTeacher: Was musstet *(PN) euch für Hausaufgaben machen? Hattet *(PN) euch viele Hausaudgaben? Waren die *(PL) Lehreren böse oder nett?, interessant oder langweilig. … ESCVictoria: Meine *(PL)Lehreren *(SVA) war nett.
ESCVictoria: Meine hehreren *(SVA) war interessant auch ESCVictoria: *(PL) lehreren… EveningTeacher: *(PL)Lehreren = *(ENG) teachers (plural) … ESCVirginia: Meine *(PL) Lehreren *(SVA) war interessant.
Table 4.27: Error Type and Error Uptake
As table 4.27 illustrates, a total of 82 of the 1670 student and teacher errors resulted in error uptake (4.
91%). Since this is a low amount of error uptake, patterns between error type and error uptake are hard to establish. However, based on the case study subjects’ transcripts it could be suggested that lexical and morphological errors that are part of German word learning are more likely to lead to uptake, such as plural forms, word choice, noun gender, and verb forms. However, these findings have to be viewed with caution since morphological errors such as the ones discussed above are also determined syntactically (e.g., case), or they are unclassifiable, apparently random, ending errors. Furthermore, no statistical analysis was performed between error type and corrective feedback, and in the above-mentioned chart, only raw numbers of errors are represented, without consideration of the number of instances of such error type. While no quantitative analysis was performed on the remaining transcripts sets, during the reading the error types discussed above appeared to lead more frequently to uptake than other errors such as word order or subject verb agreement errors. Yet another word of caution is the fact that what appears to be error uptake may simply be signs of learners being at a similar interlanguage developmental stage.
4.4.3 Research Question 3c The question is: What patterns occur in the data between source of error and error uptake? Since both teachers are non-native speakers of German, it was assumed that both would make errors, which was confirmed in the analysis of the case study subjects’ transcripts. One of the hypotheses formed was that teacher errors would be more likely to be uptaken than student errors.
To review, the students made 1617 errors, MorningTeacher made 5 errors and EveningTeacher 48 errors in the transcripts analyzed. Of the student errors, 74 (0.31%) resulted in error uptake, none (0%) of MorningTeacher’s mistakes resulted in error uptake, and 8 (16.67%) of EveningTeacher’s errors resulted in error uptake. Below in table 4.28 the numbers are reflected again and examples are given from each error source.
Table 4.28 Source of Error and Uptake
Since the exact number of student subjects cannot be established in the report of this data (as the errors were made by both the case study subjects and their partners), a non-parametric measure had to be used to understand whether there was a significant difference between student and teacher feedback. In this case a chi squared analysis was performed, and it was found that there was a significant difference in the amount of error uptake for student errors, MorningTeacher’s errors and EveningTeacher’s errors. In the transcripts analyzed, the mistakes most likely to be uptaken were the ones made by EveningTeacher and the least likely the ones by MorningTeacher.
While it was initially thought that any teacher error would be more likely to be uptaken, this was not confirmed by the chi square analysis of the instances of error uptake
MorningTeacher’s errors did not lead to uptake may be that there was no need to use the structure or lexical item again. For example, by not using the form in a question or highlighting it through a translation, she may not have indicated the importance of a word. One may compare the examples from EveningTeacher’s (see those quotes above table 4.27) errors with the following MorningTeacher’s error below to indicate that error uptake may not have been encouraged.
MorningTeacher: Was ist *(WW) mit dem Bruder los? Er ist krank, nicht wahr?
4.5.1 Research Question 4a The question is: How do students perceive the teacher’s role in the chat room and in the physical space? To investigate the question, the findings from the survey and the self-report forms are discussed. To provide a quick overview, all items of the survey are presented in see Appendix 14, even those that do not apply to this question. In the chart, the original statement is reiterated, and the percentage of students who either answered “agree” or “strongly agree” for opinion statements and “sometimes” or “always” on frequency items is listed. Then the results for both class and teacher differences are listed (as calculated with ANOVAs). For the statistical analysis the actual values from the Likert scale were used and not a combined value. The repeated measures ANOVAs were used to establish a potential significant change in attitude, a difference between classes or teachers, or an interaction between the pre- and post-survey with the factor teacher or class. In cases where the item only appeared on either the pre or the post-survey a between subjects ANOVA was used to establish differences between classes or teachers.
Some items were not answered by all students and were therefore disregarded for statistical analysis. To answer research question 4a, both the quantitative and the qualitative results from the survey will be reviewed. Furthermore, in the results both the preferred role of the teacher (i.e., how the teacher chose to interact), and the experienced role of the teacher (i.e., how the students experienced the teacher interacting) will be discussed based on the students’ responses on the survey.
Despite the observed different teacher roles in the transcripts, there were only three items with significant differences in the quantitative analysis of the survey. This suggests that, for the most part, students share similar opinions and do not change their opinion over the course of a semester significantly, regardless of prior experience, the teacher or the level of technolological support.
The items with a significant difference were items 9, 10, and 17. Item 9 states: “I don’t think my classmates should correct me.” Item 9 exhibited a statistically significant interaction between pre and post-survey and the class as the factor. Looking at the percentages you can see that the approval rating of peer feedback went down in the SSC and up in the ESC and NSC. However, it has to be admitted that students were challenged by the negative formulation of the item, often contradicting themselves between the circled answer and their comments.
The second significant difference was the interaction between pre and post survey and the teacher as a factor on item 10, which states “When I say something wrong, I like it when the teacher writes the correction on the board.” While MorningTeacher’s classes either increased the approval rating or stayed the same, EveningTeacher’s class went down in approval rating. This could suggest a difference in frequency of this form of feedback or a difference in implementation with either of these differences resulting in a more negative experience for the ESC students.
The third item with significant differences was item 17 which states: “In a foreign language class using computers gets in the way of really learning the language”. A statistically significant difference was observed between-subjects for both the class effect and the teacher effect. Given the different levels of support and the difference in teacher comfort with technology, it was to be expected that the ESC rated technology the highest on this item, and the SSC the lowest on this item. However, this same difference was not observed in the positively formulated item 13 (“I believe using technology in the language classroom is beneficial for language learning”), or any of the items that asked about chatting specifically.
Even though there were minimal quantitatively significant differences between pre- and post-survey, between teachers, and between groups, the analyses still provide a general picture of the preferences and experiences of the students, which will be summarized in the following.
To summarize the various items on the survey, students generally like corrective feedback both from peers and from teachers. Furthermore, students reported liking technology in the classroom, and considered chatting to be both fun and beneficial for language learning. Students also reported liking active participation in chat by the teacher and error correction from the teacher during chat. Interestingly, while most students reported seeing their own and others’ mistakes, only about half of them reported providing feedback to their peers.
The majority of students reported having received feedback from their teachers in the past; however, their reports on the frequency of feedback from their current teacher were more mixed. While this could mean that they experienced less feedback from the two teachers in this study than their prior teachers, this cannot be argued comfortably due to problems with the writing of the items. The addition of “only” in these items, an unfortunate wording, made the items unclear (e.g., “My teacher mainly corrects only certain kinds of mistakes I make during class”). It appears that some students interpreted the item to mean exactly the opposite of how it was intended. This was apparent, since several students answered “never” in response to the statement, but then wrote in the comment section “she corrects ALL mistakes.” Therefore, it is important to analyze the qualitative findings on these items, rather than only the quantitative results.
In the table (in Appendix 15), the comments made by the students are listed to provide a detailed picture of their responses. Again, first students’ preferred participation style will be discussed and then the students’ comments on their experience will be discussed.
From the qualitative analysis of the survey, it appears that students prefer a teacher who provides corrective feedback in a discrete yet noticeable form. Some of the reasons students provided for the desire for feedback were that: (a) feedback is necessary for learning; (b) feedback provides an opportunity for all students to learn; and (c) it shows that the teacher is listening to them. On the other hand, several students cautioned that corrective feedback can also be embarrassing. Interestingly SSCOphelia pointed out that she thought that error correction during chat was not necessary.
In terms of the students’ preference for teacher participation during chat, only a few students made comments (4 in NSC, 1 in SSC, and 6 in ESC). NSCDanielle, NSCLaura, NSCMarkus, and ESCBarbara pointed out that an active teacher can help move the conversation in the chat rooms along. NSCGeorge and ESCTiffany mentioned that an active teacher will help them focus. SSCAdrienne, on the other hand, stated that she likes student-student interactions. ESCDominique, though she preferred an active teacher during chat, added that she only appreciates such interaction if it is related to the topic.
As mentioned earlier, the items about the teacher’s use of corrective feedback were stated confusingly, and it is necessary to turn to the comments to understand the
students’ true opinion. In regards to error correction by the MorningTeacher during class: