Corrective feedback and student reaction. An observational study on communication classrooms

Table of Contents

I. Introduction

II. Database
II.1 Classes
II.2 Recording Procedure
II.3 Teachers
II.4 Students

III. Data Analysis
III.1 Student Errors
III.2 Corrective Feedback
III.2.1 Recast
III.2.2 Elicitation
III.2.3 Clarification Request
III.2.4 Explicit Correction
III.2.5 Repetition
III.3 Student Reaction

IV. Results
IV.1 Distribution of types of corrective feedback
IV.1.1 Comprehensive School Results
IV.1.2 Gymnasium Results
IV.2 Student Reaction

V. Discussion
V.1 Distribution of Different Types of Corrective Feedback
V.2 Student Reaction and Negotiation of Form

VI. Conclusion

VII. Bibliography

VIII. Appendix

I. Introduction

Cuius vis hominis est errare, nullius nisi insipientis in error perseverare.

(M.T. Cicero, 44 BC)

Every man can err but only the fool will insist upon it. When Marcus Tullius Cicero made this statement in his first Phillipica in 44 BC he did not know that there would be a time in which errors are no longer perceived as an unwanted, although widely spread, part of human existence.

On the contrary, in 1978, James Hendrickson stated that errors should be accepted as a “natural phenomenon integral to the process of learning a second language.”^[1] By this responsibility is passed to the teacher in charge of his students’ progress and who has to ensure that the errors which occur are used properly to enhance their individual learning process. In order to support teachers in this respect considerable research has been conducted particularly on the questions: ‘Should learner errors be corrected? If so, when should learner errors be corrected? Which learner errors should be corrected? Who should correct learner errors? And how should learner errors be corrected?’^[2]

The main focus of this study is how learner errors should be corrected. To contribute to answering this overall question we developed two research questions. Firstly, what is the distribution of different types of corrective feedback in L2 classrooms (communicatively oriented lessons) in England and Germany?; secondly, what is the distribution of student reactions to different types of corrective feedback?

The presentation of our study begins with the description of the database, followed by the analysis of the data according to errors, feedback and student reaction. In the subsequent chapter the results will be presented. In a fourth step the results will be discussed and linked to other studies. Taking into account our findings in addition to the findings of further studies we will come to a conclusion in which possible implications for teachers’ feedback behaviour are summarized.

II. Database

The Data presented here have been raised in a quantitative classroom observation study at a Christian comprehensive school in England, LEA Bolton, and at a German Gymnasium in Nordrhein-Westfalen.

According to Kasper, ‘teachers’ corrective feedback behaviour differs significantly depending on the communicative focus of the lesson.’^[3] For this reason we observed 18 particularly communicatively oriented lessons, 10 (60 min each) at the comprehensive school and 8 (45 min each) at the Gymnasium.

II.1 Classes

At the comprehensive school in England we selected two Year 11 classes, whose students had been learning German as their second language for 4.5 years at secondary school level. The first class (Y11.I) comprised 25 students of set 1 level and the second one (Y11.II) comprised 22 students of set 3 level. That means that Y11.I performed better in previous centrally administered examinations than Y11.II.

At the Gymnasium in Germany two classes of Stufe 10 had been selected whose students were in their 6th year of learning English as their second language. The performance level of these two classes could not be evaluated.

II.2 Recording Procedure

In each study one observer coded the teachers’ corrective feedback and the resulting student reaction in a live classroom observation on the Feedback-Reaction-Sheet^[4].

Both corrective feedback and student reaction were coded according to a “category system”^[5] of different feedback types and different types of student reaction.

II.3 Teachers

At the English comprehensive school as well as at the German Gymnasium the teachers were selected on grounds of their willingness to have their lessons observed. They were informed about the general purpose of our observation, namely to study the teachers’ feedback behaviour. They neither knew our Feedback-Reaction-Sheet nor the categories used to code their corrective feedback.

The first British teacher (BT1) was male and 35 years old. At the time of the survey he had been teaching German for 9 years. In our judgement as native speakers of German he spoke the target language at native standard.

The second British teacher (BT2) was female and 37 years old. She had been teaching German for 12 years and was Head of German at the MFL-Department of that school. In our view her target language proficiency level was slightly lower than that of BT1.

The first German teacher (GT1) was male and 52 years old. He had been teaching English for 25 years. The fact that his second subject, sport, had always been in demand more often than English, revealed itself in that his knowledge of English seemed less developed than that of GT2.

The second German teacher (GT2) was male and 49 years old. As he was responsible for further teachers’ training in English, his methods in our opinion corresponded with the current teaching practice, despite his 21 years of teaching. Several stays in Great Britain had made his English use and pronunciation seem of native standard.

II.4 Students

Most students of both Year 11 classes came from rather conservative and Christian family backgrounds. There was no student with German as his or her mother tongue. Most of the students were learning French or Spanish as a second foreign language.

Most students of the two German classes came from the particularly rural Niederrhein area. None of them was a native speaker of English. About 60% of the students were learning French as their second foreign language and about 40% did Latin.

The database, presented above, requires further analysis of the error treatment sequence, which was at the centre of our observation. It contains student errors, corrective feedback by the teachers and the student reaction provoked by the teachers’ feedback. This sequence will be explained in more detail in the subsequent chapter.

III. Data Analysis

Adopting parts of Lyster and Ranta’s E rror Treatment Sequence^[6] presented in 1997, we developed a Coding-Sequence^[7] containing learner error, teacher feedback and student reaction.

The sequence starts with an erroneous utterance of a student followed by the teacher providing corrective feedback, to which the student reacts. The observer codes only the teacher’s corrective feedback and the student’s reaction. Although Chaudron points out that “learners derive information about their behaviour [both] from the teacher’s reaction [and the] …lack of one,”^[8] student utterances containing an error which does not trigger any kind of corrective feedback is not coded on the Feedback-Reaction-Sheet.

In borderline cases, where the real-time classroom observer could not fit either the corrective feedback or the student reaction into one category only, more than one code was assigned to the behaviour (multiple coding).

illustration not visible in this excerpt

Figure 1 Coding-Sequence used for live classroom observation

III.1 Student Errors

As can be seen in Figure 1, errors and in this case particularly student errors are the starting point for our Coding-Sequence. This raises the questions how errors can be defined in context with this study and how they can be coded.

When Carl James rhetorically asks, “are not these learners unstinting in their invention of new ways of getting it wrong?”^[9], he lays his finger on the crucial problem of defining the meaning behind the term ‘error’. Nevertheless some distinguishing features embracing certain groups of deviances, mistakes, slips etc. can be identified.

Corder points out that it might be useful “to refer to errors of performance as mistakes, reserving the term error to refer to the systematic errors of the learner…i.e. his transitional competence.”^[10] We followed Corder’s differentiation between failures in performance and failures in competence although we left it to the teacher to decide whether the occurring deviance from native-like use is worth being treated.

Therefore, there has been no direct error coding. All errors including both failures in competence and in performance have only been coded indirectly via the teachers’ corrective feedback. Consequently, the teachers’ feedback could have been triggered by lexical, phonological and grammatical errors, mistakes or slips. Errors of meaning, however, have not been coded, although they often lead to corrective feedback, which will be dealt with in the subsequent chapter.

[...]

^[1] J.M. Hendrickson, “Error Correction in Foreign Language Teaching: Recent Theory, Research, and Practice,” The Modern Language Journal 62 (1978): 388.

^[2] Compare: Hendrickson 389.

^[3] Compare: G. Kasper, “Repair in Foreign Language Learning,” Studies in Second Language Acquisition 7 (1985): 215.

^[4] See: Figure 2.

^[5] C. Chaudron, Second Language Classrooms. Research on Teaching and Learning (New York: Cambridge University Press, 1988): 18.

^[6] Compare: R. Lyster and L. Ranta, “Corrective Feedback and Learner Uptake. Negotiation of Form in Communicative Classrooms,” Studies in Second Language Acquisition 20 (1997): 44.

^[7] See: Figure 1.

^[8] C. Chaudron, Second Language Classrooms. Research on Teaching and Learning (New York: Cambridge University Press, 1988) 133.

^[9] C. James, Errors in Language Learning and Use. Exploring Error Analysis (London: Longman, 1998) 19.

^[10] S.P. Corder, “The significance of learners’ errors,” Error Analysis. Perspectives on Second Language Acquisition, ed. J.C. Richards (London: Longman, 1974) 25.