Peter Beech

Greece 1995-96

 Teacher Development Assignment  

 Learners Evaluate Teachers Repair  

MSc in Teaching English
Language Studies Unit
Aston University

 January 1997  

 Contents

1. Introduction                                                  

2. Description of the survey                             

3. Discussion of the results                               

4. Conclusion                                                  

5. References                                                  

6. Appendix  - Specimens of the survey forms and results

1. Introduction

My experience of the Aston MSc, combining reading, teaching practice and research, combines what Edge (1992: 6) refers to as intellectual comprehension and experiential understanding. The theory that we adopt  is  integrated  into our practice and classroom research into the results of  its implementation provides feedback to refine the theory. If the course is conceived as comprising of linguistic and pedagogic strands, these are most closely intertwined in the Linguistic Varieties component; by applying sociolinguistic studies to our own classroom situation we can examine the type and quality of interactions that take place. 

Underhill (1991: 2) has said that “the psychological learning atmosphere created by a teacher is a deciding factor in the effectiveness and personal significance of that teacher’s lessons”. Coming to the Teacher Development option at the end of the course has given me the opportunity to reflect on my development throughout the course and in particular on how far I have been able to develop my “attitudes, awareness and behaviours” to promote a positive atmosphere. I regard this as an essential condition for effective learning as “The whole person, both in feeling and in cognitive aspects, is part of the learning event” ( Rogers & Freiberg 1969, 1994: 36). For this reason I decided to base this account of my development on a piece of research into my students’ attitudes to patterns of interaction in the classroom. As Wright (1990: 84) states, “An understanding and awareness of the intricacies of the social and psychological processes of the classroom is central to effective teacher development”.

 In many aspects of our work we are obliged to evaluate learners’ performance, and it is essential to do this in ways which will not compromise our “unconditional acceptance of students’ worth” . While we would not want to abandon criteria of achievement and success, we should recognise that the values we bring to evaluation are limited to our own perspectives. To overcome this limitation, some forms of evaluation can be negotiated, thus enhancing the extent to which learning can be learner- or person-centred. But while a  humanistic approach seems most likely  to create a positive atmosphere, Bailey et al. (1996: 16) note “Some of the teachers that  we considered the best were perceived as successful because they were demanding and strict, rather than displaying supportive behaviour.” In any case, evaluation is at the heart of classroom interaction. Of the many possible ways of giving feedback (Richards & Nunan 1990: 7) the most common is evaluation of learners’ contributions by the teacher (Van Lier 1988: 33). Even when we try to minimise the asymmetry of power in the classroom, conversational patterns are not normal, and the three-turn sequence initiation - response - evaluation is prevalent, with evaluation almost always performed by the teacher. Given the inevitability of evaluation it seemed important to be aware of the most effective way of conducting it. In order to explore whether my students concur with those in Freeman & Richards’ data, or if  they respond more favourably to a  person-centred classroom, I returned to a piece of classroom research that I had originally undertaken in the context of the Linguistic Varieties component and asked  students to provide feedback on the various kinds of repair that occurred in my data. 

As Tsui (1996: 156) points out, [ language] “ learners are much more vulnerable to criticism and negative evaluation than in other subjects because the chances of making mistakes in the language class are much greater. A learner may get the answer right in terms of content, but wrong in terms of form or pronunciation.” In order to minimise the threat to the learners’ self-confidence, we should therefore modulate our treatment of the errors, combining it with positive feedback for those aspects that are correct. That is not to say that errors are undesirable; as Widdowson   (1990: 121-122) puts it, “ Learning is not now seen as conformity to the conditions controlled by the teacher but as a self-generating process by the learners themselves. ... Non conformity is positively evaluated as the achievement of an interim interlanguage.” But if our evaluation is intended to be positive, it is important that it should be perceived as such by the learners being evaluated, and in order to be aware of students’ perceptions, it is necessary to conduct research. As Peck ( 1993: 84) points out, “Teachers who become aware of their own techniques of teaching, as a result of research, may become sensitised to the relative effectiveness of different aspects of their own methodology.”

 In the original project I had transcribed repair trajectories from recordings of my lessons and had quantified the relative frequency of various kinds. I had been particularly struck by the unexpected prevalence of other-initiated other-repair, which essentially took the form of the teacher correcting an error rather than prompting the student to correct it herself. Although this kind of repair is highly constrained in social situations, I had concluded that, in Van Lier’s(1988: 189-190) terms, this evaluative didactic repair need not be perceived as disjunctive, with a high likelihood of threat to face. Rather it could be regarded as affirming and enabling; in the interactions I had recorded, repair often took the form of a cooperative effort to reach the correct utterance, analogous to back channels which support the current speaker without taking over the turn.

 However, as Underhill ( 1993: 183) notes, “ One thing that disturbs me is the discrepancy between my experience of myself and my learners’ experience of me. There is constant evidence that the effect I think I am having on my learners ... is different from the effect I am actually having on them.” I thus realised the necessity of testing how far my students’ perceptions of the sensitivity and helpfulness of the various kinds of repair concurred with my own.Allwright & Bailey (1991: 103) quote research which found that learners actually want more correction than they get, and I have often been told by adult learners that they want all their mistakes to be corrected. I thus decided to design a new piece of research to survey the learners own attitudes to errors and correction. As Clark (1992: 78) points out, we need to make our beliefs and theories explicit in order to perceive the frames through which we interpret experience. At the time of the original research, the focus was on Conversation Analysis rather than the affective needs of learners, but after my experience of the Teacher Development component I realised that in order to understand whether the repair strategies I use are regarded as threatening or enabling, I should ask the learners themselves. In that way I understood what Edge (1992: 81) means when he says “ Classroom research is teacher development made explicit”.

 2. Description of the Survey

The first phase of the survey ( see appendix) explores learners’ attitudes to errors and corrections. It aims to establish attitudes to making errors through items A-C, and then

if students want to be corrected (items D-H)

how students want to be corrected (items I-L)

when students want to be corrected (items M-O)

by whom students want to be corrected (items P-Q). 

The second phase takes up the question of how students want to be corrected through examples of various treatments. The second phase was administered immediately after the first, and each student’s completed forms were stapled together so that the responses could be correlated. In particular, in order to assess the reliability of the data, I wanted  to see how far each student’s responses in phase two exemplified the preferences expressed in phase one I-L. The survey was completed by students from four classes:

9 elementary, represented as B

11 pre-intermediate, represented as C

11 intermediate, represented as D

8 upper intermediate, represented as E.

In the tables of results, as in the discussion, capital letters refer to classes and small letters denote individual students; Bf is student f in class B.   

The authentic transcript from my original data is D, and the other versions are designed to illustrate various kinds of repair trajectories as follows:

In A, B, C and D the teacher initiates repair which is done by the original student.

A simply indicates that the error exists, after the Transition  Relevant Point

Corresponds to phase one I + N

B indicates the location of the error, after the Transition  Relevant Point

Corresponds to phase one J + N

C indicates the location of the error, interrupting the turn

Corresponds to phase one J + M

D explains the error

Corresponds to phase one K + N

In E, F and G, the teacher performs the repair.

E after the TRP

Corresponds to phase one L + N

F before the TRP

Corresponds to phase one L + M

G in response to the self-initiation by the student

Corresponds to phase one L + N

H is self-initiated self-repair

 No equivalent in phase one

I is repaired by a fellow student

Corresponds to phase one Q + M

It was clearly impractical to try to illustrate all the possible repair trajectories, so I decided to restrict myself to the number that would fit on one page. This meant that there is only one example of peer correction, and no instance of the error being left untreated. It is also likely that students’ evaluation of appropriate repair might depend on the kind of error being treated, but it was not possible handle the proliferation that a variety of errors would have entailed. It is clear from the transcript that the error is grammatical, but there is no wider context to show if it arose during a high structure activity such as a grammar exercise or a low structure one such as a discussion. In my original data, initiation was often performed by drawing attention to the trouble source by means of a question (“What’s the correct preposition?”) or explanation (“What you need is a singular noun”), a technique which obviously lends itself to grammar exercises. While the type of error is probably the most important criterion for the decision to repair ( Allwright 1988: 205), that might also depend on its relevance to the pedagogic focus. Time constraints might also influence the treatment of errors, as would the phase of the lesson ( Allwright & Bailey 1991: 90). Additionally, the teacher might be more likely to overlook errors by certain individuals in order to spare them the embarrassment of constant correction.

 3. Discussion of the Results

The order of most positively evaluated statements in phase one was as follows: 

F. I want to be corrected if I make mistakes in an exercise.

C. I think it’s good to make mistakes because you learn from them.

N. I prefer to be corrected after I have finished.

H. I want all my mistakes to be corrected.

K. I like the teacher to indicate where the mistake is and explain it so that I can correct it.

I. I like the teacher to indicate that there’s a mistake and let me find and correct it myself.

G. I want to be corrected if I make mistakes in conversation.

P. I think corrections should be done by the teacher.

J. I like the teacher to indicate where the mistake is and let me correct it myself.

E. I don’t mind being corrected.

A. I hate making mistakes.

Q. I think corrections can be done by other students as well.

M. I prefer to be corrected immediately.

O. I prefer to be corrected later.

B. I don’t mind making mistakes.

D. I hate being corrected.

L. I like the teacher to give me the correction.

 

The order of most positively evaluated repair trajectories in phase two was as follows:

D

Student: I mustn’t crossing the busy road.

Teacher: Remember that modal verbs are followed by the bare infinitive.

Student: I mustn’t cross the busy road

 

A

Student: I mustn’t crossing the busy road.

Teacher: Try again.

Student: I mustn’t cross the busy road.

 

H  

Student: I mustn’t crossing, no, cross, I mustn’t cross the busy road.

 

B

Student: I mustn’t crossing the busy road.

Teacher: I mustn’t ...?

Student: I mustn’t cross the busy road.

 

C

Student : I mustn’t crossing...

Teacher : Not crossing.

Student : I mustn’t cross the busy road

 

I

Student 1 : I mustn’t crossing...                                  

Student 2 : cross

Student 1 : I mustn’t cross the busy road.

 

E

Student: I mustn’t crossing the busy road.

Teacher: I mustn’t cross the busy road.

Student: I mustn’t cross the busy road

 

G

Student: I mustn’t crossing ... crossing or cross?

Teacher: Cross

Student: I mustn’t cross the busy road.

 

F

Student: I mustn’t crossing...

Teacher: I mustn’t cross

Student: I mustn’t cross the busy road.

 

By comparing each individual’s responses across the two phases, we can evaluate the reliability of the data in expressing the students’ real preferences. Thus, where I, J and K are preferred (rated 1 or 2) in phase one, A, B, C and D should be preferred (given a lower average number than E, F and G) in phase two. Where L is preferred in phase one, E, F and G should be preferred in phase two. On this measure, there is a satisfactory degree of reliability. The average of I, J and K is lower than L for 30 of the 39 students. Of these 30, all but 3 gave A, B, C and D a lower average than E, F and G. On the other hand, for the students who expressed a preference for L over I, J and K, 4 out of 6 failed to corroborate that in phase two, indicating that the discrepancy is due to carelessness or misunderstanding rather than a minority preference. This can be borne out by examination of the responses of individuals Bd, Cb, Cc, Cd, Cf, Cg, Eb, Ed and Eh. 

There is a high level of agreement that “ it’s good to make mistakes because you learn from them”, even though nine individuals strongly agreed that they hate making mistakes, and a very low rating was given to “ I don’t mind making mistakes”. It seems that while, in theory at least, students recognise that errors are an essential part of the learning process, they have very negative feelings about actually making them. Nevertheless, they do want their errors to be corrected, and while correction of mistakes that occur in conversation was given slightly less approval than correction of mistakes made in exercises, the difference was not great enough to be considered significant. The high placings of K, I and J and the last position of L in phase one bear out the dispreference for the teacher providing the correction which is clearly indicated in phase two, while the high placing of “I prefer to be corrected after I’ve finished” is also clearly supported by the ordering of the repair trajectories. On those two points there is clear evidence of strong student preferences. 

It can be seen from the graph which displays the average rating given by all the students that treatments D and A are much more strongly preferred than the others, followed by H, B and C. This means that self-repair is clearly perceived as preferable. The ratings for the other treatments, I, E, G, F are significantly lower, with F, where the teacher interrupts and gives the correct answer, being perceived as the least desirable. The most highly rated treatment of the error is for the teacher to explain  it so that the student can correct it (D), followed by encouragement to try again, prompting and pointing to the trouble source. The rating of H, where the student corrects her own error, varies enormously amongst individuals, but for the most preferred treatments, A and D, there was a very high degree of agreement. With the exception of the cases where they were pushed down one rank by H, D and A occupied first and second placings in all classes. 

In phase one, the question of who should perform the repair is a divided one. There was a very slight tendency to agree (2.7) that corrections can be done by other students, and a rather stronger tendency to agree (2.2) that corrections should be done by the teacher. For any average figure, the question arises if a lot of students are neutral or if the group contains individuals with various strong preferences that cancel each other out in the averaging. As we are concerned here with the attitudes of individuals, I felt it was important to resolve that point. If we look at the individual ratings for statement Q, we find the following number of occurrences of each rating:

 

CLASS         B     C     D     E     TOTAL
RATING

1                   2      1     1      0           4

2                   2      5     2      5          14

3                   4      3     5      1          13

4                   0      2     2      1            5

5                   1      0     1      1            3

 

This shows that the great majority of individuals either agree weakly or are neutral on the question of whether corrections can be done by other students, and the only example of such a repair trajectory in phase two was given  a low ranking, presumably because it constitutes an interruption. As Allwright & Bailey (1991: 108) note, “Porter found that second language learners interacting either with native speakers or with other second language learners were able to accurately correct other learners’ errors, though they rarely did so. A possible disadvantage of peer correction is that it deprives the speaker of the opportunity to correct the error; in my data, self-repair initiated by other students did not occur, which is natural as students cannot be expected to employ didactic techniques. On the other hand, there is evidence that error correction by peers may be more likely to lead to learning. Block (1996: 170) suggests that “...it would appear that teacher-generated discourse is less memorable than learner-generated discourse”. In addition to possible cognitive advantages, greater student participation in correction might also bring the affective benefits associated with peer teaching. In a situation described in Rogers & Freiberg (1969, 1994: 201), the tutees “showed greater confidence, more motivation to work, and an improved attitude towards mathematics....Tutors gained in their own self-assurance and their willingness to assume responsibility.”  

On the question of when repair should be carried out, my idea that repair before the TRP might be regarded as enabling rather than threatening was not borne out by the data, although in order to test that hypothesis I would need to modify the design of the research to include more extensive transcripts exemplifying that kind of support. As Van Lier (1988: 235) notes, “ When control is shared, a type of repair is favoured which is proactive, i.e. which aims to further the interaction as it is developing.”  The corollary of this is that there will be a greater proportion of repair or initiation before the turn is completed, as in my original data. On the present evidence, however, it seems more important that, as Nunan & Lamb (1996: 76) point out, “Correction methods which encourage purposeful learner involvement by allowing opportunities to self-correct or analyse the errors facilitate learning.” So repair trajectory C, although it had a rather low rating because of the interruption, still rates significantly higher than those where the correction is provided. 

By using concrete examples in phase two of the survey, I hoped to increase the reliability of the responses, but the most reliable way to gauge attitudes to repair is through observing its actual effects. The uptake of repair varied greatly between individual students in my original data. Having modified my practice in line with the conclusions of the present research, it will be interesting to see if the incidence of uptake is improved, on the occasions where repair is provided, and if the incidence of repair is improved on the occasions where initiation is provided; initiation which is more informative, in explaining the nature of the error should clearly be more likely to lead to successful repair by the learner.        

4. Conclusion

The results of the survey, in conjunction with the transcripts from my original research, constitute an evaluation of myself as a teacher ( Nunan 1988: 147-148, Allwright 1988: 200). I can now develop a new cycle of action research by observing the effects of modifying my practice in line with my students’ expressed preferences. The particular focus of this research is clearly only one of many aspects of classroom practice which should be made explicit, but it highlights the importance of combining essential evaluation with sensitivity to the learners’ affective domain. As Leithwood (1992: 91) notes, “At the highest stages of psychological development, teachers are inner-directed but appreciate the interdependent nature of relationships in a social setting such as a classroom....these teachers have achieved more of a synthesis in their classrooms between an emphasis on achievement and an interpersonal orientation.” 

My own development throughout the past two years has been characterised by the growing understanding that the learner is at the centre, and that the focus for my development must be a focus on the learner. The emphasis placed on experiential understanding by this course means that our readings are mediated through our own research in our own classrooms. So, as Oberg & Underwood (1992: 163) express it, “The theory that we thus create is not separate from our practice, either as source or product. It is simply how we speak of what we do when our actions are in tune with our intentions.” 

5. References

Allwright, D.  1988. Observation in the Language Classroom  London: Longman.

Allwright, D. & Bailey, K. 1991. Focus on the Language Classroom  Cambridge: Cambridge University Press. 

Bailey, K. et al. “The language learner’s autobiography: Examining the                    ‘apprenticeship of  observation’.” In Freeman & Richards (1996) : 11-29. 

Bailey, K. & Nunan, D. (eds.) 1996. Voices From the Language Classroom  Cambridge: Cambridge University Press. 

Block, D. “A window on the classroom: classroom events viewed from different angles.” In Bailey & Nunan (1996) : 168-194. 

Clark, C. “Teachers as Designers in Self-directed Professional Development.” In Hargreaves & Fullan (1992) : 75-84. 

Edge, J. (1992) Cooperative Development  Harlow: Longman 

Edge, J & Richards, K. (eds.) (1993) Teachers Develop Teachers Research  Oxford: Heinemann. 

Freeman, D. & Richards, J. (eds.) (1996) Teacher Learning in Language Teaching   Cambridge: Cambridge University Press. 

Fullan, M. & Hargreaves, A. (eds.) (1992)  Teacher Development and Educational Change London: The Falmer Press. 

Hargreaves, A. & Fullan, M. (eds.) (1992) Understanding Teacher Development London: Cassell. 

Leithwood, K. “The Principal’s Role in Teacher Development.” In Fullan & Hargreaves (1992) : 86-103. 

Nunan, D. (1988) The Learner-Centred Curriculum  Cambridge: Cambridge University Press. 

Nunan, D. & Lamb, C. (1996) The Self-Directed Teacher  Cambridge: Cambridge University Press 

Oberg, A. & Underwood, S. “Facilitating Teacher Self-development: Reflections on Experience.” In Hargreaves & Fullan (1992) : 162-177. 

Peck, A. “The relationship between classroom process analysis and interventionist strategies in foreign language teaching.” In Edge & Richards (1993) : 83-93. 

Richards, J. & Nunan, D. (eds.) (1990) Second Language Teacher Education  Cambridge: Cambridge University Press. 

Rogers, C. & Freiberg, H. (1969, 1994) Freedom to Learn  New York: Macmillan 

Tsui, A. “Reticence and anxiety in second language learning.” In Bailey & Nunan (1996) : 145-167. 

Underhill, A. Plenary talk in IATEFL Teacher Development Group newsletter 17/1991. 

Underhill, A. “Awareness: the instrument and the aim of experiential research.” In Edge & Richards (1993) : 183-187. 

Van Lier, L. (1988) The Classroom and the Language Learner  London: Longman 

Widdowson, H. (1990) Aspects of Language Teaching  Oxford: Oxford University Press. 

Wright, T. “Understanding classroom role relationships.” In Richards & Nunan (1990) : 82-97.

 

6. Appendix: specimens of the survey forms and results.

Learners’ Attitudes to Errors and Correction

Please indicate how strongly you agree or disagree with each of the following statements by giving it a number 1-5:

1 = I strongly agree

2 = I agree

3 = Neutral

4 = I disagree

5 = I strongly disagree

 

A. ___ I hate making mistakes.

B. ___ I don’t mind making mistakes.

C. ___ I think it’s good to make mistakes because you learn from them.  

D. ___ I hate being corrected.

E. ___ I don’t mind being corrected.

F. ___ I want to be corrected if I make mistakes in an exercise.

G. ___ I want to be corrected if I make mistakes in conversation.

H. ___ I want all my mistakes to be corrected. 

I. ___ I like the teacher to indicate  that there’s a mistake and let me find and correct

           it myself.

J. ___ I like the teacher to indicate where the mistake is and let me correct it myself.

K.___ I like the teacher to indicate where the mistake is and explain it so that I can

            correct it.

L. ___ I like the teacher to give me the correction.                            

M. ___ I prefer to be corrected immediately.

N. ___ I prefer to be corrected after I have finished.

O. ___ I prefer to be corrected later. 

P. ___ I think corrections should be done by the teacher.

Q.___ I think corrections can be done by other students as well.

 

Please add any other comments you would like to make.

 

_____________________________________________________________________

 

_____________________________________________________________________

 

_____________________________________________________________________

 

Peter Beech       Language Studies Unit, Aston University       1996 - 97


Kinds of Correction

Please number these dialogues in order from the best (1) to the worst (9) way of correcting mistakes.

A. ___

Student: I mustn’t crossing the busy road.

Teacher: Try again.

Student: I mustn’t cross the busy road.

 

B. ___

Student: I mustn’t crossing the busy road.

Teacher: I mustn’t ...?

Student: I mustn’t cross the busy road.

 

C. ___

Student : I mustn’t crossing...

Teacher : Not crossing.

Student : I mustn’t cross the busy road

 

D. ___

Student: I mustn’t crossing the busy road.

Teacher: Remember that modal verbs are followed by the bare infinitive.

Student: I mustn’t cross the busy road.

 

E. ____

Student: I mustn’t crossing the busy road.

Teacher: I mustn’t cross the busy road.

Student: I mustn’t cross the busy road

 

F. ___

Student: I mustn’t crossing...

Teacher: I mustn’t cross

Student: I mustn’t cross the busy road.

 

G. ___

Student: I mustn’t crossing ... crossing or cross?

Teacher: Cross

Student: I mustn’t cross the busy road.

 

H. ___

Student: I mustn’t crossing, no, cross, I mustn’t cross the busy road.

 

I. ___

Student 1 : I mustn’t crossing...                                  

Student 2 : cross

Student 1 : I mustn’t cross the busy road.

 

Peter Beech       Language Studies Unit, Aston University       1996 - 97