The influence of the application of paperrater on the second year students use of conhesive devices in an english class at university of science, vietnma national university ho chi m

Recently, it is said that the introduction of Automated Writing Evaluation AWE, along with its application, has yielded significant results on learners’ language acquisition, and its app

VIETNAM NATIONAL UNIVERSITY- HOCHIMINH CITY UNIVERSITY OF SOCIAL SCIENCES & HUMANITIES

FACULTY OF ENGLISH LINGUISTICS & LITERATURE

THE INFLUENCE OF THE APPLICATION OF PAPERRATER

ON THE SECOND-YEAR STUDENTS' USE OF

COHESIVE DEVICES IN AN ENGLISH CLASS AT

UNIVERSITY OF SCIENCE, VIETNAM NATIONAL UNIVERSITY, HO CHI MINH CITY

Submitted to the Faculty of English Linguistics & Literature

in partial fulfillment of the Master’s degree in Teaching English to Speakers of Other Languages

By HOÀNG KIM MAI KHÔI Supervised by NGUYỄN QUANG TIẾN, Ph.D

Ho Chi Minh City, August, 2017

APPROVAL SHEET

This is to certify that this thesis entitled THE INFLUENCE OF THE APPLICATION

OF PAPERRATER ON THE SECOND-YEAR STUDENTS' USE OF COHESIVE DEVICES IN AN ENGLISH CLASS AT UNIVERSITY OF SCIENCE, VIETNAM NATIONAL UNIVERSITY – HO CHI MINH CITY has been approved by the

supervisor or further approval by the Board of Examiners

My sincere thanks go to the teachers and staff at the English Centre at University of Science, Vietnam National University, Ho Chi Minh city, for providing me with favourite conditions to conduct my thesis at this site I also would like to show my gratitude to my colleagues at the centre for their continuous encouragement and assistance during the process of conducting this thesis

I also would like to express my gratitude to the teachers and staff at Faculty of International Languages and Cultures, University of Economics and Finance, and Faculty of English Linguistics and Literature of University of Social Sciences and Humanities, Vietnam National University, Ho Chi Minh city, who gave opinions in informal discussions and provided me with useful guidelines and information

I am indebted to my parents, who always encourage and believe in me My special thanks to my brother, who helped me to assemble work of references and supported me

at every stage of this thesis

I also want to say thanks to my best friends for their valuable advice and continuous support

STATEMENT OF ORIGINALITY

This is to certify that the thesis entitled THE INFLUENCE OF THE APPLICATION

OF PAPERRATER ON THE SECOND-YEAR STUDENTS' USE OF COHESIVE DEVICES IN AN ENGLISH CLASS AT THE UNIVERSITY OF SCIENCE, VIETNAM NATIONAL UNIVERSITY – HO CHI MINH CITY is the product of my own work

The thesis has not been submitted for the award of any degrees or diplomas in any other institutions

The thesis’s content, except where clearly acknowledged within the text, has not been published by other authors

I hereby state that I approve the requirements of the University for the retention and use

of my thesis and that it is accessible for the purposes of study and research

Ho Chi Minh City, August, 2017

Hoàng Kim Mai Khôi

ABSTRACT

Corrective feedback is one of the hallmarks that can help boost students’ performance

in writing Yet, it is time-consuming and sometimes imposes negative effects on both teachers and students Recently, it is said that the introduction of Automated Writing Evaluation (AWE), along with its application, has yielded significant results on learners’ language acquisition, and its application helps teachers reduce the workload and assists students to improve their performance in writing However, it is still unclear

to what extent the application of such programmes could facilitate learners in a specific aspect of writing in Vietnamese context This study is aimed at investigating the impacts of PaperRater, an AWE programme, on Vietnamese students’ learning of writing skills In this study, the subject were 58 second-year students from University

of Science, Ho Chi Minh city, Vietnam National University To discover the overall effects of using AWE on the improvement of students’ writing in terms of the proper use of cohesive devices and their attitudes, both quantitative and qualitative methods were applied The data used for this study were collected through sets of a pre-test and

a post-test together with a student questionnaire and interviews with six participants After the treatment, the findings of the study revealed a significant difference between the experimental and control groups in terms of their writing improvement Furthermore, the study also received some positive feedback from the participants Although the students pointed out some unsatisfactory features, they still agreed that this teaching method could be used effectively to direct feedback instruction Based on this, the pedagogical implications for applying this method in writing process to enhance learners’ performance arise In short, the study concludes that the application

of PaperRater imposes positive effects on students’ use of cohesive devices, students’ attitudes toward this corrective feedback techniques and learner autonomy and that its limitation could be overcome by the use of an eclectic approach to teaching and learning English academic writing

Key words: AWE, cohesive devices, corrective feedback, teaching writing

city

TABLE OF CONTENT

APPROVAL SHEET i

ACKNOWLEDGEMENT i

STATEMENT OF ORIGINALITY ii

ABSTRACT iii

LIST OF ABBREVIATIONS iv

TABLE OF CONTENT v

LIST OF TABLES x

LIST OF FIGURES xi

CHAPTER 1: INTRODUCTION 1

1.1 The Educational Issue 1

1.2 Rationale for the Study 4

1.3 Statement of Purpose 6

1.4 Research Questions 6

1.5 Research Hypotheses 6

1.6 Significance of the Study 7

1.6.1 Writing Teachers 7

1.6.2 Students 7

1.7 Scope of the Study 8

1.8 Organisation of the Study 8

Summary of Chapter 1 9

CHAPTER 2: LITERATURE REVIEW 10

2.1 Approaches to teaching writing 10

2.1.1 Product-oriented approach 10

2.1.2 Process-oriented approach 11

2.2 Cohesive devices 13

2.2.1 Cohesion in text 13

2.2.2 Categories of cohesive devices 14

2.2.2.1 Reference 15

2.2.2.2 Conjunctions 17

2.3 Errors Theory 18

2.3.1 Definition of Errors 18

2.3.2 Classification of Errors 19

2.3.3 Cohesive Errors 20

2.4 Corrective Feedback Theory 21

2.4.1 Definition of Corrective Feedback 21

2.4.2 Classification of Written Corrective Feedback 22

2.4.2.1 Focused Direct Written Corrective Feedback 23

2.4.2.1.1 Focused Written Corrective Feedback 23

2.4.2.1.2 Direct Written Corrective Feedback 24

2.4.2.2 Electronic Corrective Feedback and Automated Writing Evaluation 25

2.4.2.3 Writing Conferences 26

2.5 Theoretical Framework 27

2.5.1 Noticing Hypothesis and Corrective Feedback 27

2.5.2 Vygotsky’s Zone of Proximal Development, Scaffolding and Corrective Feedback 29

2.5.3 Objection Against the Application of CF 30

2.6 Synthesis of Empirical Work on AWE 31

2.6.1 Previous Studies on the Effects of Automated Writing Evaluation on Students’ Writing Performance 31

2.6.2 Studies on Student Attitudes towards AWE in Writing 33

2.6.3 Discussion of Gaps on Previous Studies 34

2.7 Conceptual Framework 35

Summary of Chapter 2 37

CHAPTER 3: RESEARCH METHODS 38

3.1 Research methodology 38

3.1.1 Research Design 38

3.1.2 Data Collection and Analysis 39

3.2 Research site 40

3.3 Participants 41

3.4 Research instruments 43

3.4.1 Writing tests 43

3.4.1.1 Test design 44

3.4.1.2 Test administration and scoring 44

3.4.2 The student questionnaires 45

3.4.2.1 Design and construct 45

3.4.2.2 Piloting and reliability 47

3.4.3 The student interviews 48

3.4.4 The Automated Writing Evaluation Website PaperRater 51

3.4.4.1 Introduction to the Automated Writing Evaluation Website PaperRater 51

3.4.4.2 The Instructions on How to Use Automated Writing Evaluation PaperRater Website 52

3.5 Data Collection Procedure 53

3.5.1 Pre-treatment Phase 53

3.5.2 During-treatment Phase 54

3.5.2.1 The Pilot and Experimental Groups 55

3.5.2.2 The Control Group 55

3.5.3 Post-experiment Phase 56

3.7 Data Analysis Procedure 57

Summary of Chapter 3 59

CHAPTER 4 FINDINGS AND DISCUSSION 60

4.1 Findings 60

4.1.1 Preconditions for the Independent Samples and Paired Sample T-tests 60

4.1.2 Results of the Writing Tests 61

4.1.2.1 Statistical Analysis of the Pre-test Results 61

4.1.2.2 Statistical Analysis of the Post-test Results 62

4.1.2.3 Within-group Comparison 62

4.1.2.3.1 The Control Group 64

4.1.2.3.1 The Experimental Group 65

4.1.2.4 Between-group Comparison 66

4.1.3 Findings of the Questionnaire and Student Interviews 68

4.1.3.1 Questionnaire Results 68

4.1.3.1.1 General Opinions about the Application of PaperRater 68

4.1.3.1.2 Students’ Attitude towards CF on Cohesive Devices from PaperRater 71

4.1.3.1.3 Suggestions for Further Improvement when Applying PaperRater74 4.1.4 Interview Findings 74

4.1.4.1 Students’ Attitude towards PaperRater 74

4.1.4.1.1 Students’ Attitude on the Strong Points of PaperRater 74

4.1.4.1.2 Students’ Attitude on the Weak Points of PaperRater 75

4.1.4.2 Students’ Attitude towards CF on Cohesive Devices 77

Summary of main findings 78

4.2 Discussion 78

4.2.1 The Proper use of Cohesive Devices 79

4.2.1.1 Within-group Comparison 79

4.2.1.2 Between-group Comparison 81

4.2.2 Attitudes towards PaperRater 83

4.2.2.1 Attitudes towards PaperRater in general 83

4.2.2.1.1 In Comparison with Previous Studies 83

4.2.2.1.1 In Relation to SLA Theories 85

4.2.2.2 Attitudes towards CF on Cohesive Devices from PaperRater 87

Summary of the Discussion 88

Summary of Chapter 4 89

CHAPTER 5: CONCLUSION, IMPLICATATION, LIMITATIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH 90

5.1 Conclusion 90

5.2 Implications 92

5.2.1 For Curriculum Designers 92

5.2.2 For Language Teachers 92

5.2.2 For Language Learners 93

5.3 Limitations of the Study 94

5.4 Recommendations for Future Research 96

BIBLIOGRAPHY 98

APPENDICES 109

APPENDIX 1 109

APPENDIX 2 110

APPENDIX 3 111

APPENDIX 4 112

APPENDIX 5 115

APPENDIX 6 116

APPENDIX 7 117

APPENDIX 8 118

APPENDIX 9 122

APPENDIX 10 130

APPENDIX 11 132

APPENDIX 12 133

LIST OF TABLES

Table 3 - 1 Demographic characteristics of the participants 42

Table 3 - 2 The content of the questions in the questionnaire 47

Table 3 - 3 Reliability of the statements of the questionnaire 48

Table 3 - 4 The content of questions in the student interview 50

Table 3 - 5 Data collection process 57

Table 4 - 1 Shapiro-Wilk of normality for the scores of the two groups 60

Table 4 - 2 Results of the Independent sample t-test for the pre-test 61

Table 4 - 3 Summary of paired-sample t-test for the Control Group post-test 64

Table 4 - 4 Within-group comparison results for the control group 64

Table 4 - 5 Summary of paired-sample t-test for the Experimental Group post-test 65

Table 4 - 6 Within-group comparison results for the control group 65

Table 4 - 7 Summary of Independent Sample t-test for the post-test 66

Table 4 - 8 Between-group comparison for the post-test 67

Table 4 - 9 The students’ general opinion about the application of AWE 70

Table 4 - 10 The students’ attitudes towards CF on cohesive devices from PaperRater 72

LIST OF FIGURES

Figure 2 - 1 The stages of writing process from product-oriented approach (Badger &

White, 2000) 11

Figure 2 - 2 The recursiveness of writing process (Coffin et al 2003, pg 34) 12

Figure 2 - 3 Example of Focused WCF (Ellis, 2008, p 101) 23

Figure 2 - 4 Example of Direct WCF (Ellis, 2008, p 99) 24

Figure 2 - 5 Example of Indirect WCF (Ellis, 2008, p 100) 24

Figure 2 - 6 Conceptual framework for the study 35

Figure 3 - 1 A pre-test – post-test non-randomized control-group design 38

Figure 4 - 1 Students' general attitude towards PaperRater 69

CHAPTER 1: INTRODUCTION

1.1 The Educational Issue

In language teaching and learning, writing is usually thought to be one of the most difficult and challenging skills This is understandable since it is cognitively demanding, even for native speakers (Flower & Hayes, 1980; 1981; Van den Berg & Rijlaarsdam, 1996; cited in Van Beuningen et al., 2012) In writing, it is required of writers not only to use the language appropriately but also to express themselves clearly with adequacy of ideas and content formulation These requirements cause many difficulties for learners, as some complain that they usually struggle to find appropriate words or sentence structures to express their ideas, whereas others are afraid of making errors in grammar, word-choice, etc in their writing These cases emphasise the role of writing teachers in not only helping learners develop their ideas but also raising their awareness of the appropriate language structures and usage The role of writing teachers, hence, is demonstrated through the practice of error correction and feedback provision and it is considered to be one of the hallmarks in teacher’s instruction in writing classes Nevertheless, in reality, the provision of effective correction or corrective feedback (henceforth referred to as CF) in the process of writing does not receive adequate consideration Studies have shown that although teachers acknowledge its benefits, they have little knowledge about CF as well as CF forms to deliver to students (Vasquez & Harvey, 2010) Similarly, in a small survey, Lee (2004) notices that teachers apply a limited range of CF strategies, and their correction is not always accurate

In research on writing error correction, two questions are usually addressed: what to correct and how to correct The first question is to investigate the types of errors that should receive teachers’ commentary and attention A wide range of issues on what to correct have been raised in the literature, such as content, organisation, language,

mechanics, and styles Specifically, recent researches evolve around two types of errors, i.e., content and form While the former refers to the surface errors, such as ideas and language, the latter concerns the organisation and the mechanics of writing Regardless

of the forms of errors, Ferris indicates that most researchers have agreed that correcting errors should be embarked upon learner needs (Ferris, 2004) They believe that individual students might encounter different difficulties, and make some specific types

of errors in their writing; therefore, it requires distinguished focus of error correction

If teachers opt to correct every error in a composition, the number of errors might frustrate, even discourage students and this results in less improvement

It is also important to notice the mismatch between teachers’ and students’ belief about

CF It is found that besides the common practice of giving comments on grammar and mechanics, students generally would like to receive CF on all areas of writing, including content and organisation (Hedgcock & Lefkowitz, 1994; Leki, 1991; Oladejo, 1993; Saito, 1994; cited in Lee, 2008) Moreover, the degree of error correction given in each writing depends on learner’s language competence From the researcher’s perspective, beginner students frequently have problems with grammatical accuracy in their writing, whereas intermediate and advanced students might encounter problems with the mechanics and styles This idea contrasts with students’ belief and preference towards

CF Studies have shown that students in general prefer correction on linguistic features such as grammar, word-choice, etc to feedback relating to meaning and ideas issues (Cohen, 1987; Hedgcock & Lefkowitz, 1994; cited in Lee, 2008) Interestingly, in a study about student preference towards errors, even upper-intermediate students want

to receive feedback on grammar errors (Orts Soler, 2015)

Besides the concern for the types of errors to correct, the modes of delivering CF for the errors, is another concern and is the central issue in research about writing (Hyland, 2010) Regardless of its various forms, previous studies have agreed upon the outperformed effects of explicit CF over some other kinds of CF, e.g., implicit CF (Ellis

et al., 2006; Bitchener, 2008; Al-Mansour & Al-Shorman, 2012) Yet, amongst different types of correction, there exists a lack of a consensus on the most effective type(s) of CF Particularly, recent studies have proven that though CF can be effective

to some extent, it has some drawbacks First, it is time-consuming, as teachers have to spend many hours correcting and giving comments on each student’s writing (Truscott, 2007; Wang et al., 2013) For many teachers, they have claimed to suffer from the monotony and boredom of correcting essays (Hyland, 1990; Hyland, 2006; Yang et al., 2006) The situation becomes worse when class sizes continue to grow and the writing tasks are more complicated Furthermore, it yields a mixed rate of success on learners’ development (Hyland, 2006) For instance, Ferris (1997, cited in Hyland, 2006) finds that although students take teachers’ comments into consideration in their drafts, only half of the participants in the study show their improvement in their subsequent writing Truscott (2007) also claims that students might avoid making the corrected errors; thus, the practice of CF is harmful to the development of students’ writing skills to some extent

Therefore, recently, a new CF method has been introduced, aiming to overcome the limitations of those traditional CF methods This new type of CF is called computer-generated feedback (henceforth CGF), and one popular device to provide this new CF form is automated writing evaluation (henceforth AWE) To be more specific, AWE is

a computer programme designed to assign grades and give feedback on essays It was first introduced by Ellis Batten Page in 1966 Since then, it has been developed into different models and is able to measure different features of an essay, making it a great tool for giving instructions in writing classes

Recent studies (Ware, 2005; Warschauer & Ware, 2006) in this field have claimed that AWE can help reduce the burden for teachers and give students more opportunities to receive CF immediately; thus, it suggests a novel manner of learning and teaching (Wang et al., 2013) However, regardless of its crucial role, the potential of employing

AWE in writing is still being observed On the one hand, immediate feedback can promote the quality of learners’ writing skills Due to the attention to the errors and their explanations granted by these systems, the accuracy in writing can be enhanced (Kern & Warschauer, 2000; Milton, 2006; Ware & Warschauer, 2006; cited in Wang

et al., 2013) On the other hand, recent studies (Chen & Cheng, 2008; Lai, 2010) have casted doubt on the influence of this kind of feedback; Chen and Cheng (2008) even state that the reliance on such software can negatively affect the process of language acquisition, particularly in learning writing In other words, the advantages of AWE over the correct use of language in writing remain unclear, with inconclusive evidence

in the literature Thus, the application of this software is worth researching

Taking those research gaps into account, this study is an attempt to examine one type

of CF which can be used in teaching and learning English writing skills The present study was conducted to investigate the relative influence of CF from AWE on the improvement of students’ writing in English

1.2 Rationale for the Study

Given all the above-mentioned points, it can be concluded that the types of errors and modes of delivering error correction are evolving issues that need further consideration

in giving and directing CF in the researcher’s context, especially in the university where this study was conducted

First, from the researcher’s perspective, university students frequently make errors in the use of cohesive devices – an essential component in creating a meaningful text Specifically, it seems that these students can formulate a complete sentence, with correct use of grammar and word choice However, their pieces of writing are merely a combination of sentences of a given topic, without any connections at both sentence and paragraph levels This issue is also highlighted in Jogthong’s (2001) study, which shows that learners with high language proficiency frequently encounter problems with

cohesion Similarly, in one of his studies, Pilus (1996) highlights the lack of clear understanding of cohesion, which results in incorrect composition at the sentence level

in the revision of students’ writing Specifically, studies in the literature report that of the five ties of cohesion, i.e., reference, conjunctions, substitution, omission and lexical cohesion, students frequently make errors in reference and conjunctions (Ong, 2011; Hidayati, 2014; Sesriyani, 2015) Although many researchers have attempted to investigate the use and method to promote the use of proper reference and conjunctions

in students’ compositions (Taboada, 2004; Tangkiengsirisin, 2010), as observed, studies in the literature on Vietnamese students’ use of reference and conjunctions and

on CF on these ties have not been found adequately (Hung & Thu, 2014)

Second, AWE has emerged as a form of CF and is assumed to benefit teachers and students (Hirvela, 2005; Huot, 1996; Tuzi, 2004; Warschauer & Ware, 2006; cited in Wang et al., 2013) Despite the popular use of such innovative application in ELT around the world, many recent studies have been conducted with software copyright, such as Criterion e-rater, WritetoLearn and Writing Roadmap2.0 Obviously, feedback provided by these applications varies, ranging from correction of word-choice to that

of stylistic features Plus, they can be employed at the administrative level, which allows the institution or university to keep track of the development process of learners Notwithstanding their effectiveness on the development of learners, there exist some drawbacks: they are costly as the use of such software requires specific training as well

as suitable equipment These two requirements seem impossible in the context of English learning in Vietnam, where students have little opportunity to access high-tech due to the limitation of facilities and finance Therefore, suggesting a free-but-effective tool like PaperRater is a choice to bring benefits to both teachers and learners of English

in Vietnam

1.3 Statement of Purpose

The main objectives carried out in this study are (1) to investigate the influential aspects

of PaperRater on learners’ use of cohesive devices, particularly on reference and conjunctions and (2) to find out student attitudes toward the feedback delivered by PaperRater

1.4 Research Questions

To achieve the purpose of this research, the central research question is “What is the influence of PaperRater on learners’ use of cohesive devices, particularly reference and causal conjunctions, at the University of Science, Ho Chi Minh City, Vietnam National University (henceforth HCMUS, VNU)?”

The central research question is addressed by the following sub-questions:

1 To what extent does the students’ use of cohesive devices, particularly reference and causal conjunctions, improve after they have applied the feedback from PaperRater?

2 What are the students’ views on PaperRater as a mode of delivering corrective feedback?

1.5 Research Hypotheses

The present study is guided by the two hypotheses below:

1 The application of corrective feedback from PaperRater does not help students improve the use of cohesive devices, particularly reference and causal conjunctions, in their writing

2 Students do not have positive views towards the use of PaperRater

1.6 Significance of the Study

AWE is amongst the feedback strategies which have been employed in a number of countries around the world, such as in the USA and China, over the past 20 years; however, as observed, Vietnamese teachers seem to be unaware of its application Therefore, they have not applied it widely in teaching English writing In the light of this gap, the researcher found a need to conduct a study which is aimed at investigating its possible influence on students’ writing performance in terms of the use of cohesive devices as well as their attitudes for this feedback strategy Given the objectives just mentioned, the present study is of much significance to the following stakeholders:

1.6.1 Writing Teachers

First, this study provides writing teachers with opportunities to understand how AWE contributes to their instructions in writing class They may consult the findings of this research to devise tasks to teach writing, especially in the editing and revising phases Second, the use of AWE can help reduce the daunting amount of work of error correction, so it gives way for correcting other aspects of writing, such as idea development Also, the results could encourage them to apply more Information and Communication Technology (henceforth ICT) and Computer-assisted Language Learning method (henceforth CALL) in their teaching context Third, as the study focuses on the use of cohesive devices, it shows teachers how to apply AWE flexibly

in writing courses to improve students’ cohesion in English writing

1.6.2 Students

This study aims to provide students with an effective tool to gain a better understanding

of the use of cohesive devices and apply them properly in the editing phase in writing compositions If it is proved to be practical, AWE can grant students a better approach

to develop their understanding of cohesive devices and to enhance their writing skills

1.7 Scope of the Study

This study was conducted to investigate (1) the correct use of reference and causal conjunctions and (2) students’ attitudes towards one strategy of CF in two “English 3” classes at USHCM, VN Hence, the ties of cohesive devices investigated in this study are reference and causal conjunctions, which are ties students frequently make errors Alternative ties of cohesion, such as substitution, omission, and lexical cohesion are beyond the scope of this study

Furthermore, this present empirical work was an attempt to justify the influence of PaperRater on learners’ use of language In other words, this study is an experimental research Because of the potential intrusion of this study on the participants, the sample could not be chosen randomly This study, then, utilised the quasi-experimental method with convenient sample to answer the main and sub research questions

Finally, as this is a small-scale study, it is impossible to ask all the students and teachers

to participate in the study Thus, the researcher could merely invite a small group of 58 non-English major sophomores at HCMUS, VNU to be the participants of this study

In addition, due to time constraint, the application of the AWE only lasted one semester

1.8 Organisation of the Study

The study consists of five chapters Chapter 1 provides an overview of the study

including the background, the rationale, the statement of purpose, the research question, the research hypotheses, the significance of the study, the study’s scope and the

organisation of the study Chapter 2 is concerned with the literature review related to

the study It gives an overview of writing process, cohesion, cohesive devices and cohesive errors in writing It also includes the definitions and models of feedback from various sources, especially from AWE and writing conferences Moreover, it will discuss the theoretical framework for applying CF in teaching writing The last part of this chapter presents the conceptual frameworks used to guide the present study

Chapter 3 presents the research question and hypotheses, research methodology, the

setting, the participants, the research instrument and the procedure of data collection

and data analysis Chapter 4 deals with data analysis and findings In this section, the

analysis of the data collected from the sets of tests, questionnaires and the interviews will be discussed The major findings of this research will be presented in the third part

of this chapter Chapter 5 presents the conclusions of the study, some implications for

curriculum designers, teachers and learners concerning the actual use of the CF from AWE in the future Limitations and some recommendations for further research are also presented in this part

Summary of Chapter 1

This chapter includes the educational issue, the rationale, the statement of purposes, the research question, the research hypothesis, the significance, the scope of the study, and finally the organisation of this study

CHAPTER 2: LITERATURE REVIEW

This section is aimed at presenting the operational discussion and theories relating to this current study To construct the theoretical background for the study, this chapter first provides the overview of the approaches to teaching writing, the theory of cohesive devices, the theory of errors, and the theory of CF This chapter also consists of the theoretical framework, which explains the necessity of CF in teaching writing This framework includes Schmidt’s Noticing Hypothesis (1990) and Vygotsky’s theory of Zone of Proximal Development and scaffolding (1978) Then related studies are reviewed in order to find out gaps in previous studies Finally, the conceptual framework which guides the whole studies is presented

2.1 Approaches to teaching writing

In the literature, there is no consensus on how to teach writing, as the answers largely depend on many factors such as individual teaching and learning styles (Raimes, 1983; Hyland, 2002) Thus, prescribing an optimal approach to teaching writing seems impossible Within the scope of this study, the process-oriented and product-oriented approaches are examined To put it simply, the former focuses on the finished text, and the latter pivots around the act of writing itself (Thompson & Adams, 2010) Below is the discussion of these two approaches in teaching writing

2.1.1 Product-oriented approach

As suggested by its name, the product oriented-approach indicates that the outcome of the writing process, i.e., the composition, is the most important component in teaching writing (Sommer, 1982) Also, in language classroom, the lessons usually revolve around the writing tasks in which learners are asked to imitate, analyse and transform a model composition to create a similar text (Nunan, 2001; Klimova, 2014) In other words, students are given a framework following a pattern of organisation, and are asked to put their ideas into this framework, using the language structures and

vocabulary provided by teachers or books The underlying belief of this approach is when both the content and linguistic forms are strictly constrained, students are prevented from making errors As a result, teachers tend to instruct students to create a text with accurate grammar and written forms rather than to see how the ideas are conveyed and transferred through the text (Hyland, 2015)

A product-oriented writing class typically undergoes four following stages: familiarization, controlled writing, guided writing and free writing These stages are described in Figure 2 - 1 below

Figure 2 - 1 The stages of writing process from product-oriented approach (Badger &

White, 2000)

2.1.2 Process-oriented approach

While language accuracy is the ultimate goal of writing in product-oriented approach, learner’s creativity and their writing development through time are the focus in the process-oriented approach (Tribble, 1996) Nunan (2001) also further claims that the process-oriented approach does not require learners to create an error-free text, and an imperfect composition is allowed In fact, students are encouraged to produce a meaningful text instead of drawing much attention to formal or linguistic features of a text (Sum & Yang, 2005)

Learners compose a text with the grammar and vocabulary points presented in

Learners write on their own

Therefore, a written composition is evaluated based on not only itself, but also the process of writing, such as brainstorming, editing and revision Plus, it is learners who initiate the ideas, compose the writing and revise their drafts several times before handing the final work to teachers To help learners achieve perfection in their writing, teachers can apply a wide range of classroom activities and tasks during the writing time, such as peer interaction and group discussion

It is also noticed that process-oriented approach applies a different model contrasting with product-oriented approach: it does not follow a linear model, but it happens through different patterns, and is independent of learners’ writing skills and language proficiency (Stein, 1986) In fact, writing is characterised as a number of recursive processes, which are shown in Figure 2 - 2 below

Figure 2 - 2 The recursiveness of writing process (Coffin et al., 2003, pg 34)

From the above discussion, it can be noticed that these two approaches distinguish from

each other; their comparison is presented in Appendix 1

Although they are different, and teachers can opt for the type of approach suitable for their students, it is important to note that these two approaches are not incompatible Linguists and researchers also suggest that teachers can combine and integrate these two approaches into their teaching instruction to help develop students’ writing ability (White & Arndt, 1991; Curry & Henning, 2003)

2.2 Cohesive devices

One of the main aims of this study is to investigate students’ use of cohesive devices in their writing Therefore, it is worth mentioning its definition in this section However, the definition of cohesion in text should be reviewed, as it is the overarching term describing the functions of cohesive devices

2.2.1 Cohesion in text

Regardless of the approaches teachers might apply, cohesion is one of the essence features that should receive attention in teaching writing (Witte & Faigley, 1981; Mather et al., 2009) Interestingly, although cohesion has been long brought to light by Halliday and Hasan (1976), it is not well defined in the literature (Van Dijk, 1977; cited

in Pilus, 1996)

Preliminary, Halliday and Hasan coined the first description of cohesion, suggesting that it “refers to relations of meaning that exist within the text, that define it as a text” and “occurs when the interpretation of some elements in the discourse is dependent on that of another” (Halliday & Hasan, 1976, p 4 - 5) For instance:

(1) The schoolmaster was leaving the village, and everybody seemed sorry The miller at Cresscombe lent him the small white cart

In this example, “him” in the second sentence refers to the “schoolmaster” mentioned

in the previous sentence In other words, there exists a tie which connects the two

sentences as one meaningful unit in this text Such ties are called cohesion or cohesive relation in a text

This viewpoint has been analysed and investigated over the past 40 years, building up the foundation for further elaboration in the field Hinkel, for instance, conceptualises

it as “the connectivity of ideas in discourse and sentences to one another in text, thus creating the flow of information in a unified way” (2003, p 279) Similarly, Wodak and Krzyzanowski propose that it “concerns the components of the textual surface that signal the ‘text-syntactic’ connectedness” (2008, p 204) These above definitions seem

to emphasise the complex sides of cohesion, yet they all agree on some common aspects, i.e., it is the close relationship amongst elements in a text, which helps create

a unified text

It is also noticed that cohesion demonstrates a strong connection to coherence, which is also an important part in discourse Although they can be used interchangeably on some occasions, each has its distinctive features and internally links to each other To be more specific, linguists and scholars consent that coherence exists when a text is meaningful and unified, and can be explained based on the conceptual or meaning relations between ideas In contrast, cohesion relates to the meaning of the text and can be realised through grammatical and lexical ties (Halliday & Hasan, 1976; Tanskanen, 2006; Baker, 2011) These ties can be achieved by employing different types of devices which are called cohesive devices (Halliday & Hasan, 1976; Fasold & Connor-Linton, 2014) The

following section elaborates these items in a more careful manner

2.2.2 Categories of cohesive devices

According to Halliday and Hasan (1976), the dependence of one element on the others within a text can be achieved through five major categories of devices: reference, substitution, ellipsis, conjunction and lexical cohesion These devices have been further modified, and some linguists have argued that ellipsis and substitution are only one kind

of devices due to their close relationship and interchangeable roles on many occasions (Halliday, 1994; Biesenbach-Lucas & Weasenforth, 2001; cited in Yang & Sun, 2012)

In line with this, McCarthy (1991) proposes four main categories of cohesive devices,

in which ellipsis and substitution are classified under one category However, he does not provide a comprehensible and convincing evidence for his classification Plus, from the researcher’s viewpoints, it seems these studies merely consider substitution as one subclass of omission, whereas the terminology used for substitution remains intact Therefore, Halliday and Hasan’s taxonomy (1976) is adopted in this study, as it has been applied in many studies in the literature due to its well-developed nature The classification of all cohesive devices suggested by these two linguists is fully presented

in Appendix 2

As discussed in Chapter 1, only reference and conjunctions in the students’ writing were examined, as they are ties of cohesive devices with which students frequently encounter problems in their writing (Ong, 2011; Sesriyani, 2015; Hidayati, 2014) These ties are

presented in the following parts

2.2.2.1 Reference

According to Halliday and Hasan, a reference is “the relation between an element of the text and something else by reference to which it is interpreted in the given instance” (1976, p 308) For further elaboration, recent researchers have specified that reference includes the set of grammatical phrases and words which allow speakers/writers to state

if something is already or not mentioned in the text yet (Yang & Sun, 2012) This view

is in line with Eggins’s (1994) theory, stating that the reference can be either presenting (i.e., has been introduced in the text) or presuming (i.e., can be tracked throughout the text) It, hence, helps create cohesion by connecting the items mentioned in the text (Eggins, 1994)

As explained by Halliday and Hasan (1976) and McCarthy (1991), reference is further classified as anaphora – looking backward where the referent has already mentioned in the text, cataphora – looking forwards where the referent will appear in the upcoming text, and exophora – looking for assumed referent outside of the text The following examples can illustrate the differences amongst these types:

(2) Anaphora: Peter was generally not a hater He spoke tolerantly of his foes

(The personal pronoun ‘he’ refers to Peter.)

(3) Cataphora: When he arrived, John noticed that the door was open (The

personal pronoun ‘he’ refers to John.)

(4) Exophora: Did the gardener water those plants? (Both the listener and speaker mutually understand what ‘those plants’ refer to.)

Besides this classification, Halliday and Hasan (1976) also propose three types of referential cohesion, i.e., personal, demonstrative, and comparative reference Regarding the personal reference, they state it can be expressed through personal pronouns (I, he, she, etc.), possessive determiners (my, your, his, etc.) and possessive pronouns (mine, yours, his, etc.) (Halliday & Hasan, 1976) For instance, in example (2) above, the pronoun ‘he’ refers back to Peter, the man mentioned in the preceding sentence

Similarly, demonstrative reference also points to other elements, including words, phrases and even chunks in the text (Halliday & Hasan, 1976; Sanczyk, 2010) Such

reference consists of determiners and adverbs, such as this, these, that, here, and then

Example (5) below shows a determiner with demonstrative reference, which refers to Hungary – the country mentioned in the preceding sentence:

(5) I visited Hungary last year That was my first visit

Finally, as for the comparative reference, it is employed when there exists a comparison between items (Sanczyk, 2010) It is important to notice that the comparison can be made “either to general features of identity, similarity and difference or to particular features of quality and quantity” (Halliday, 2014, p 560) For instance, in (6) below, the comparative tie refers to the similarity between the items and is created by linking

to Katherine’s story:

(6) Otherwise his story is the same as Katherine’s

A complete list of pronominal, demonstrative, and comparative reference is presented

in Appendix 3

2.2.2.2 Conjunctions

Being another form of cohesive ties, conjunctions are frequently used in writing to connect ideas Unlike reference, conjunctive relation refers to the relationship between one sentence/clause and another (Cook, 1989) In the same vein, Thompson notes that conjunctions are “the semantic links that combine any two textual elements into a potentially coherent complex semantic unit” (Thompson, 2013, p 225) To put it in a simpler manner, conjunctive relation uses appropriate words to connect two components in a text

Halliday and Hasan (1976) propose a scheme for the classification of conjunctive ties, which capture different relations: addition, adversative, causal, and temporal conjunctions The first type – additive conjunctives – adds, elaborates or exemplifies more information to the elements that have already mentioned, as demonstrated in example (7) In contrast, adversative conjunctives contrast or create dissimilarity between the old and new information, as in example (8) Causal conjunctives, as in example (9), relate the new information to the successive elements in terms of causes

or effects In other words, it helps to build up a cause-effect relationship between

elements Finally, the temporal conjunctives put the elements in time order, as shown

in example (10)

For the whole day, he climbed up the steep mountainside, almost without stopping

(7) And in all this time he met no one

(8) Yet he was hardly aware of being tired

(9) So by night time the valley was far below him

(10) Then, as dusk fell, he sat down to rest

Besides these types, Halliday and Hasan (1976) also propose other two types of conjunctive relation, i.e continuatives and intonation, and they are frequently used in spoken form of communication However, as this study merely focuses on written discourse, they are not introduced in this section

The summary of conjunctions offered by Halliday and Hasan (1976) is presented in Appendix 5

2.3 Errors Theory

2.3.1 Definition of Errors

The word “error” has various meanings and usages, which is relative to how it is perceived and in what field it is considered In linguistics, the meaning of errors has been widely discussed, which entails a wide range of definitions and perception

In Oxford Online dictionary, an error is defined as a mistake However, in one of his studies, Ellis (1997) thinks this definition fails to distinguish between learners’ competence and their performance Specifically, he states that errors “reflect gaps in a

learner’s knowledge; they occur because the learner does not know what is correct” (Ellis, 1997, p 15) In contrast, mistakes are considered deviations in usage that reflect the learners’ ability to use what they actually know of the target language In other words, the failure of performance happens when there are slips of tongue, changes of mind, confusion, etc., which then lead to mistakes (Ellis, 1997)

Sharing the same viewpoint, Catalán (1997) further elaborates that errors are changes

of the rules which conform to the norms in the target language This perspective is in line with Thornbury, who identifies errors as:

An instance of the learner’s language that does not conform to accepted norm of usage, and which is attributed to incomplete or faulty learning These norms by which errors are judged are usually defined in terms of adult native speakers of Standard English (of a certain dialect) (Thornbury, 2006, p 163)

By drawing on the concept of errors, Thornbury provides three important aspects of errors that should be examined carefully in a study First, it highlights the incorrect features of an utterance in language usage Second, it happens due to incomplete or faulty acquisition, i.e., it concerns learners’ incapability to know what the norms of English are Third, it defines language norms discreetly, which confines to the rules of Standard English, thus eliminating the incidence of any un-standardised English utterances Hence, this definition is employed in this study, since it encompasses the necessary aspects of errors, and clearly distinguishes errors from mistakes

2.3.2 Classification of Errors

Depending on the criteria on which the classification of errors is based, different methods have been proposed to classify errors An error can be classified in terms of linguistic categories, as was done by Corder (1967), or can be categorized based on surface strategy taxonomy by Dulay et al (1982) The former refers to errors relating

to linguistic features, such as incorrect use of articles, tenses, etc in a written text, and

the latter “highlights the ways surface structures are altered” (Dulay et al., 1982, p 150) As this study merely pivots one linguistic category, i.e., cohesive devices, Corder’s classification is not applicable in this case In contrast, Dulay et al.’s taxonomy (1982) provides a broader view of learners’ errors, which is independent of the linguistics features Hence, within the scope of this study, this taxonomy is employed

In this taxonomy, Dulay et al (1982) suggest assigning errors into four categories: omission of some required elements, addition of some redundant elements, selection of

an incorrect element and misordering of the elements The table below briefly exemplifies this:

Table 2 - 1 The surface strategy taxonomy of errors (Dulay et al., 1982)

Categories Description Examples

Omissions The absence of an item that must

appear in a well-formed utterance

* She sleeping

Additions The presence of an item that must not

appear in well-formed utterances

* We didn’t went there

Misinformations The use of the wrong form of the

morpheme or structure

* The dog ated the chicken

Misorderings The incorrect placement of a

morpheme or group of morphemes in

study They state that cohesive errors happen when the writer fails to provide “a complete unambiguous tie/referent pair” (Spiegel & Fitzgerald, 1990, p 52) In other words, a case is considered a cohesive error if it reflects the misuse, incorrect ordering, wrong omission or wrong addition of one of the five cohesive ties; thus, it makes the text unclear or hard to understand For instance, in (11), the writer might use an improper pronoun in a text, making the whole structure ambiguous:

(11) Fred and Jack went to the beach * He went swimming

As shown in this example, it is unclear whether the reference “he” refers to Fred or Jack One might even think that the writer attempts to point to both of them, but using

a wrong personal pronoun It, therefore, is an erroneous instance of cohesive devices

In this study, Spiegel and Fitzgerald’s definition (1990) was employed as it reflected the nature of cohesive errors and it was widely used in studies relating to the same issue (e.g., Huang, 2005; Yang & Sun, 2012) Types of cohesive errors that confine to the cohesive ties of reference and conjunctions are presented in Appendix 5

2.4 Corrective Feedback Theory

2.4.1 Definition of Corrective Feedback

In teaching writing, the question of what errors should be corrected is not as important

as how to correct and prevent such erroneous instances in students’ compositions At this point, the notion of error correction or corrective feedback emerges and plays an essential role in helping to develop learners’ performance and competence in writing

In the literature, the issues of CF have been mentioned, adopted and adapted by many linguists and scholars (e.g., Kroll, 1990; and Bitchener & Ferris, 2012) Each proposes different angles and aspects, leading to a diverse range of definitions in the field In this current study, this concept is defined as “an indication to a learner that his/her use of

the target language is incorrect, and includes a variety of responses that a language learner receives” (Gutierrez & Atkinson, 2011, p 6147) From this definition, feedback implies not only the identification of the incorrectness of the students’ utterances but also the reader’s reaction for such errors Further elaborating this idea, Bitchener and Storch (2016) add that CF seeks to correct the inaccurate usage or provide information about where and why the errors have occurred, and how they may be corrected

To extend this definition to writing class, which is the heart of this study, written corrective feedback (henceforth WCF), is defined as written methods with which a reader can respond to a text by showing that some usages in the writing do not conform

to the norms of the target language (Van Beuningen, 2011)

2.4.2 Classification of Written Corrective Feedback

A number of researchers (e.g., Bartram & Walton, 1991; Lyster & Ranta, 1997;) have categorised CF into different types Of all classifications, Ellis’s typology (2008) of 5 options for correcting errors in students’ written work is presented in this section The brief elaboration of the popular strategies of CF that are frequently employed in teaching writing is presented in Appendix 6

From this classification, Ellis highlights the differences between (1) focused and unfocused CF, (2) direct and indirect CF, (3) metalinguistic CF, (4) electronic CF, and (5) reformulation Although Ellis presented different strategies for WCF, recent researches have revolved around the two dichotomies: between focused and unfocused WCF; and between direct and indirect WCF From these bases, the researcher selected and employed two strategies in this study: focused direct WCF and electronic CF The definitions and reasons for selecting these two types of feedback are presented in the following parts of this paper

2.4.2.1 Focused Direct Written Corrective Feedback

2.4.2.1.1 Focused Written Corrective Feedback

As presented in Appendix 6, the focused WCF refers to the comprehensiveness of the correction methodologies and is viewed in relation with unfocused WCF (Ellis, 2009) Specifically, from the focused approach, the feedback targets around the linguistic features that should receive attention, and other types of errors outside the focus remain intact and are not corrected This approach contrasts with the unfocused CF, in which all errors in a written text are corrected, irrespective of the types of errors and the levels

of learners (Ellis, 2009) For instance, in the following example, the teacher applies the focused CF and choose to correct just article errors

Figure 2 - 3 Example of Focused WCF (Ellis, 2008, p 101)

With respect to the relative effectiveness of focused WCF, studies have shown 2 reasons for its benefits over unfocused WCF First, Bitchener (2008) argues that unfocused WCF asks students to deal with errors from a broad range of features and might make them overload Therefore, this method inhibits the process of correction Moreover, Ellis (2008, 2009) highlights that learners can notice the corrections through focused WCF Once they pay more attention to the erroneous instances, they can understand and acquire the target structures better

2.4.2.1.2 Direct Written Corrective Feedback

The dichotomy between direct and indirect CF is also another method used in writing instruction In the literature, they are briefly defined as:

[d]irect correction involves supplying learners with the correct form or reformulating the entire text; indirect correction involves indicating that an error has been committed either in the margin of the text or within the text where the error occurs (Ellis & Sheen, 2011; cited in Pawlak, 2013, p 145)

Below are examples of direct and indirect WCF, respectively:

Figure 2 - 4 Example of Direct WCF (Ellis, 2008, p 99)

Figure 2 - 5 Example of Indirect WCF (Ellis, 2008, p 100)

As noted in some experimental studies about language acquisition, direct WCF is more preferable, as it yields a higher success rate in students’ use of some specific language structures (Ferris & Roberts, 2001; Van Beuningen, 2011) Linguists also believe that when the location of the errors is shown, it provides clearer information about the

structure(s), which helps learners acquire the language easier (e.g., Bitchener and Knoch, 2010b; Manchón, 2011; Van Beuningen et al., 2008; Van Beuningen et al., 2012; cited in Ferris et al., 2013)

Guided by these findings, which highlight the effectiveness of some types of error correction on language acquisition, the WCF strategies provided to the learners in this current study were focused and direct

2.4.2.2 Electronic Corrective Feedback and Automated Writing Evaluation

As Ellis (2009) mentions in his study, electronic CF is a new approach to teaching writing and has emerged recently as one of the most effective strategies to provide instant CF to students It has many forms, depending on the types and modes of delivering CF One type of electronic CF that has been used widely and is also one of the instruments of this study is AWE In the literature, AWE software is defined as an electronic programme which generated feedback on writing (e.g., Dikli, 2006; Philips, 2007; Shermis & Burstein, 2003; Yang, Buckendahl, Juszkiewicz & Bhola, 2002; cited

in Steveson & Phakiti, 2013)

Since its first appearance in 1960s, many forms and functions of AWE have been proposed and developed Initially, it was designed to provide feedback in grammar, usage and mechanism (Burstein et al., 2004; Blake, 2012) To be more specific, it focused on linguistic errors in students’ written compositions, such as tenses, verb conjugation, etc Thanks to the development of technology, the focus has been expanded, and students’ essay can be evaluated based on pre-determined criteria, including organization, styles, development, etc (Burstein et al., 2004; Wilken, 2013) Besides being used for assessment, it can also be employed as a writing assistance, which can be similar to one-to-one tutor for students (Gutierrez & Atkinson, 2011) Specifically, when students receive the feedback, explanations and extra tasks are also given so that they can practice on aspects which they have not successfully acquired

In other words, it can provide a wide range of feedback, granting students more opportunities to practice and improve their writing ability

Although the application and the assistance of AWE software in teaching writing is undeniable, it is not “recommended as a substitute for other forms of classroom instruction” (Ware & Warschauer, 2006, p 116) Also, many researchers have suggested that AWE should not be used alone, as computers and machines cannot replace the role of teachers in giving CF (Shermis & Burstein, 2003; Ware, 2005; Warschauer & Ware, 2006; Al-Mansour & Al-Shorman, 2012) Therefore, it is necessary to apply AWE with other forms of CF generated by human in order to maximise its potential and to provide social interaction for communicative purposes (Warschauer & Ware, 2006) It is the necessity of interaction that highlights the role of conferencing in teaching writing, and this method is presented in the following part

2.4.2.3 Writing Conferences

In addition to WCF, another form of CF is also used in teaching writing: writing conferences Preliminarily, writing conferences were defined as a conversation involving two parties, a teacher and a student, about the student’s composition at the stage of revision in the writing process (Carnicelli, 1985; Feehan, 1989) Further elaborating this idea, Freedman and Sperling (1985) and Williams (2002) claim that this detailed discussion about a graded or revised composition can be generated between a teacher and a group of students In other words, this definition highlights two main traits of a writing conference: the participants and its position in a writing process Therefore, it has provided useful information for researchers and educators to decide to apply this in teaching

Yet, the application of AWE also yields the questions of the actual practices evolving during the conversation, from which two main perceptions have emerged Particularly, Clark (1985) emphasizes that the dialogues are merely another form of teacher

instruction whereas Murray argues “[Conferences] are not mini-lectures but the working talk of fellow writers sharing their experience with the writing process” (1985,

p 148) From the traditional approach of Clark, the primary roles in the revision phase lie in the hands of the teachers, and students have little room to express their ideas or raise questions Therefore, most of the literature from this approach is directed to teacher-centred teaching styles Though it highlights the roles of the participants in a conferencing session, Belk (2012) criticises that this approach makes students bored and stressed Plus, to some extent, it is similar to meta-linguistic CF, i.e., feedback provides to learners in the form of linguistic explanation (e.g., Nassaji & Fotos, 2011) Only the mode of delivery is different

Distinct from this approach, Murray’s view focuses on two participants in the dialogue,

in which both students and teachers are given opportunities to express their ideas, concerns and questions This view is similar to that of Park (2012), stating that the students’ opinions and participation are more important than the feedback initiated by the teachers He also emphasizes that during the conference, teachers should give students suggestions which focus more on ideas and meaning rather than grammar and punctuation (Park, 2012)

2.5 Theoretical Framework

Generally, CF can help develop accuracy in writing and foster the process of language acquisition (Ellis, 1997; Ferris, 2003; Hyland & Hyland, 2006) Many theories have presented different rationale for the application of error correction in teaching Some of the grounding theories are rooted in Schmidt’s Noticing hypothesis (1990) and Vygotsky’s key concept of Zone of Proximal Development and scaffolding (1978)

2.5.1 Noticing Hypothesis and Corrective Feedback

In the field of second language acquisition (henceforth SLA), it is believed that there is

a close connection between error correction and Schmidt’s noticing hypothesis