UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effects of tiered corrective feedback on second language academic writing Anderson, Timothy 2010

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2010_fall_anderson_timothy.pdf [ 620.31kB ]
JSON: 24-1.0071093.json
JSON-LD: 24-1.0071093-ld.json
RDF/XML (Pretty): 24-1.0071093-rdf.xml
RDF/JSON: 24-1.0071093-rdf.json
Turtle: 24-1.0071093-turtle.txt
N-Triples: 24-1.0071093-rdf-ntriples.txt
Original Record: 24-1.0071093-source.json
Full Text

Full Text

          THE EFFECTS OF TIERED CORRECTIVE FEEDBACK ON SECOND LANGUAGE ACADEMIC WRITING by TIMOTHY ANDERSON B.A., The University of Saskatchewan, 1999 B.Ed., The University of Saskatchewan, 2005    A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF ARTS   in   THE FACULTY OF GRADUATE STUDIES (Modern Languages Education)    THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)    July 2010  © Timothy Anderson, 2010      ii    Abstract The impact that corrective feedback (CF) has on second language (L2) writing has garnered significant attention in the field of second language acquisition (SLA). The many varieties of written CF available mean that while there is comfort in the potential options, there is often confusion in the choice; teachers still struggle to implement the most effective method of responding to students’ writing and students still struggle to effectively implement the feedback. This study reports on the effects of tiered focused metalinguistic CF on the reduction of grammatical errors in 39 intermediate adult ESL students’ academic writing at a major Canadian university. Student’s t-tests for paired samples were used to measure the influence of CF on the reduction of targeted error, and a repeated measures design was used to calculate the between- subject effects of the CF treatment to determine the result of feedback versus no feedback on the different dependent variables. Using a theoretical framework based on Vygotsky and Piaget’s social and cognitive constructivism, this thesis valued the role of self-discovery in the implementation of corrective feedback. This study also used questionnaires and semi-structured interviews to inquire into students’ feelings and perceptions regarding the feedback process. Data from the quantitative statistics suggest that the corrective feedback used in this study resulted in a reduction of targeted error and that this reduction was statistically significant. Results also indicate that as the corrective feedback lessened in focus it appears to have lessened in effectiveness. Finally, this thesis concludes with suggestions for further research and a discussion of the quantitative and qualitative findings. The results demonstrate that the feedback treatment outperformed no feedback in statistically significant ways, although the students reported with high frequency their distaste regarding the type of CF used in this study.     iii    Preface The following thesis is based on a manuscript model and contains four chapters, one which was structured for future publication in an academic journal. A manuscript-based thesis is organized based on one or more “manuscripts” that can potentially be publishable in academic journals with little revision (in terms of format). Readers will therefore notice some repetition within this thesis, most notably with respect to the literature review.  The third chapter of this thesis is written in the form of a publishable article, although it has not yet been published. This thesis was written entirely by Timothy Anderson and was not co-authored. The approval to conduct the research contained within was issued by the University of British Columbia’s Behavioural Research Ethics Board, number H09-02220.                  iv    Table of Contents Abstract..........................................................................................................................................  ii Preface..........................................................................................................................................  iii Table of contents..........................................................................................................................  iv List of tables................................................................................................................................  vii List of figures..............................................................................................................................  viii Acknowledgements......................................................................................................................  ix  Chapter 1: INTRODUCTION........................................................................................................  1  1.1 1.2 1.3 1.4 1.5 1.6 Introduction....................................................................................................................  1 Purpose of the study.......................................................................................................  2 Research problem...........................................................................................................  3 Significance of the study................................................................................................  4 Research questions.........................................................................................................  5 Thesis organization........................................................................................................  6  Chapter 2: LITERATURE REVIEW.............................................................................................  8  2.1 2.2 Introduction....................................................................................................................  8 Second language acquisition..........................................................................................  8 2.2.1 2.2.2 2.2.3 Behaviourism and structural linguistics...............................................................  8 Cognitive approaches (innatism/nativist).............................................................  9 Constructivism and sociocultural theory............................................................ 11 Constructivism..........................................................................................  12 Sociocultural theory.................................................................................. 13 2.3 Theoretical framework.................................................................................................  15 2.3.1 Constructivism and corrective feedback............................................................  15 2.4 2.5 Corrective feedback and second language acquisition................................................  17 Second language writing.............................................................................................. 18 2.5.1 2.5.2 2.5.3 Current-traditional rhetoric................................................................................  19 Process...............................................................................................................  19 Post-process.......................................................................................................  20 2.6 2.7 A brief modern history of corrective feedback...........................................................  21 Strategies of written corrective feedback....................................................................  24 2.7.1 2.7.2 2.7.3 2.7.4 2.7.5 2.7.6 Direct corrective feedback.................................................................................. 25 Indirect corrective feedback...............................................................................  30 Metalinguistic corrective feedback.....................................................................  31 Focused and unfocused corrective feedback......................................................  33 Electronic corrective feedback...........................................................................  34 Reformulation.....................................................................................................  35 2.8 2.9   2.10 The importance of grammar........................................................................................ 36 The importance of focused, explicit and consistent grammar instruction...................  39 The relevant grammatical features...............................................................................  43    2.10.1 The most common errors....................................................................................  43     v       2.10.2    2.10.3 The most problematic errors...............................................................................  44 The most treatable errors....................................................................................  46   2.11   2.12   2.13     2.14 The concentration on limited features..........................................................................  47 Focus on form..............................................................................................................  49 The learner...................................................................................................................  51 Summary.....................................................................................................................   52  Chapter 3: THE EFFECTS OF TIERED CORRECTIVE FEEDBACK ON SECOND LANGUAGE WRITING.............................................................................................................  54  3.1 3.2 Introduction..................................................................................................................  54 Literature review..........................................................................................................  54 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.2.6 Corrective feedback and SLA............................................................................  54 A brief modern history.......................................................................................  56 Strategies of written corrective feedback...........................................................  58 The role of grammar...........................................................................................  61 Targeted features................................................................................................  62 Students’ views on corrective feedback.............................................................  63 3.3 Method.........................................................................................................................  64 3.3.1 3.3.2 3.3.3 3.3.4 Design................................................................................................................  64 Context...............................................................................................................  65 Participants.........................................................................................................  66 Operationalizations.............................................................................................  68 3.4 Intervention……………………………………………………………………….…  69 3.4.1 3.4.2 Metalinguistic feedback…………………………………………………….…  69 Writing tasks…………………………………………………………………..  70 3.5 Data..............................................................................................................................  71 3.5.1 3.5.2 Questionnaires....................................................................................................  74 Interviews...........................................................................................................  74 3.6 Analysis.......................................................................................................................  76 3.6.1 3.6.2 The effects of tiered CF on grammatical accuracy…….....................................  76 Corrective feedback and the learner...................................................................  77 Questionnaires..........................................................................................  77 Interviews.................................................................................................  78 3.7 Results..........................................................................................................................  79 3.7.1 Findings based on Student’s t-tests....................................................................  79 3.7.2 Repeated measures design.................................................................................   82 3.7.3 Findings based on questionnaires and interviews..............................................  82 Questionnaires..........................................................................................  82 Interviews.................................................................................................  84 3.8 3.9 Discussion....................................................................................................................  85 Conclusion...................................................................................................................  87  Chapter 4: DISCUSSION............................................................................................................  89  4.1 Discussion....................................................................................................................  89     vi    4.2 4.3 4.4 4.5 Limitations....................................................................................................................  92 Pedagogical implications..............................................................................................  95 Suggestions for further research...................................................................................  96 Summary......................................................................................................................  98  References..................................................................................................................................  101  Appendix A: Writing prompts for three in-class compositions................................................. 113 Appendix B: Form of grammatical codes..................................................................................  116 Appendix C: Table of questionnaire themes..............................................................................  117 Appendix D: Student questionnaire of written corrective feedback..........................................  118 Appendix E: Student interview protocols..................................................................................  124 Appendix F: Informed consent form..........................................................................................  126 Appendix G: Feedback sample..................................................................................................  130 Appendix H: UBC research ethics board certificate of approval...............................................  131                      vii    List of Tables  Table 2.1 Categories of written corrective feedback...................................................................  25 Table 3.5 Description of error types............................................................................................  72                         viii    List of Figures  Figure 3.1 Sample grammatical codes handout...........................................................................  73  Figure 3.2 The mean error rate of articles for group one.............................................................  80 Figure 3.3 The mean error rate of articles and lexical categories for group two.........................  80 Figure 3.4 The mean error rate of articles, lexical categories and subject-verb agreement for group three..............................................................................................................................  81 Figure 3.5 The mean error rate of articles, lexical categories and subject-verb agreement for group four....................................................................................................................................  81                              ix    Acknowledgements  I owe my deepest appreciation to the many people who have supported me both personally and academically during the completion of this thesis. Without their help and kindness none of this would have been possible.  To my patient and caring advisor, Dr. Monique Bournot-Trites, I want to express my deepest gratitude. Your help and guidance over the last few years have been invaluable. I feel very fortunate to have had the opportunity to work with you. As members of my committee, I would also like to express my sincerest thanks to Dr. Ling Shi and Dr. Steven Talmy who provided me with insight and direction during the research and writing of this thesis.  I would also like to acknowledge Dr. Patricia Duff for her assistance and support over the years and for inclusion into the exciting research we are doing with the Centre for Research in Chinese Language and Literacy Education. I look forward to our future work together. Dr. Geoff Williams, both as the head of our department, LLED, and also an inspirational teacher and scholar, has helped me in many ways since first coming to UBC. My deepest thanks also go out to Dr. Bernie Mohan. I will always be grateful for his kindness and generosity.  There are many others at UBC who I’d like to recognize and who have each helped me in various ways throughout the course of this degree. To Dr. Jan Hare I extend my gratitude for having confidence in me and for giving me the opportunity to conduct research dear to my heart. Dr. Gaalen Erickson and my wonderful new colleagues from Thailand gave me new insight into L2 academic writing as well as new friendships. I’d also like to thank Anne Eastham, Theresa O’Shea, and Anne White for their patience and dedication to our department and for their invaluable help over the years. My appreciation also extends to Andrew Scales, Sarah ter Keurs, and all of the students and teachers who were involved with this research.  I would also like to acknowledge the members of CRCLLE, based initially out of UBC but quickly spread across cities, countries, and continents: Roma, Elliot, Ella, Rachel. My inclusion in the group has been both intellectually exciting and also a source of great inspiration in my own long (and seemingly never-ending) language study.  My good friends, spread around the world, helped me recharge with basketball and music, good conversation and a kind of permanence and safety amongst turbulence and disruption that is deeply appreciated and needed. Burt, Garry, Chad, and Cam, thank you.  My wonderful family has always supported and helped me, and it’s difficult to imagine how this would have been possible without their unconditional love and support. My parents, Rick and Brenda, and my sister, Corine, have always been there for me, and have cheered me up in tough times, and have made me laugh when I needed it the most.  Finally, I want to thank my wonderful partner, Liz, for her love, support, kindness, and (above all) patience. I feel blessed to have you in my life.       1    Chapter 1: Introduction  1.1     Introduction As a teacher and a graduate student, I am fortunate to live in two worlds simultaneously. One is the “novice” as I continue to grow and learn about language, and one is the “expert” as I work to help my students progress. Through these interconnected roles, I have learned that academic writing is a process of struggle and catharsis. Many factors contribute to the success or failure of academic texts. The obstacles that all writers face in producing these texts can be challenging as they navigate their way through a multitude of complex factors in the quest to produce scholarly, grammatical, and interesting writing. For many English L2 speakers, this process is especially tricky. As they engage in producing academic discourse, their writing is often the face of their personality: the ultimate predictor of success or failure. For myself, both as a student writer and as a writing teacher, this is no less true. The possibilities that are available for writers in terms of rhetorical, grammatical, and lexical choices are equally limitless and daunting, but also exciting and inspiring. And while this means that in many ways writing is truly without borders, it also means that a solid grammatical foundation is crucial in the longer journey. For English as a second language (ESL) students, attaining grammatical accuracy in their academic writing is an important part of greater academic success. While the goal for students is to achieve this success, the goal for teachers is how best to help; how best to facilitate student improvement and how to involve and implement relevant pedagogy and methodologies so that students can grow and succeed. Although this process is a long and complicated one, the feedback that teachers provide on students’ writing plays an important role in this progression. The precise role of feedback on writing is still, however, subjective and inconclusive. Teachers     2    are burdened by time constraints, and students often fail to incorporate corrections. Despite the difficulties and ambiguities, corrective feedback maintains a significant role in the development of students’ academic writing, and, as such, it is important to continue the quest for best practices to achieve best results. This study will therefore investigate a way to be effective in providing corrective feedback while paying proper attention to the needs and expectations of the students and the limitations placed on teachers.  1.2     Purpose of the study The purpose of this study is to investigate the role of written corrective feedback on the grammatical accuracy of English as a second language learners’ academic writing. The area of corrective feedback has indeed come a long way in the last decade, and recent work by some of the field’s noted scholars have called for further study in CF with a careful approach paid to appropriate methodology (Guenette, 2007; Sheen, 2007; Ferris, 2004). This study, drawing on those sensibilities, will seek to add to current scholarship regarding written corrective feedback. This study will also look at the writers’ themselves and their relationship to corrective feedback; or rather, how they understand the function of CF and what their feelings are towards it. It has been clearly established that successfully learning an additional language requires many moons to align at the same time, and a learner’s affect and motivation play crucial parts in this process. Without students understanding and accepting how their learning takes shape, their successful acquisition of the language may suffer. Although peers play a vital role in the language classroom as companions, study partners, socializers and scaffolders, this study has not included peer feedback in its design, and instead has focused solely on the impact of teacher-directed corrective feedback on students’ writing.      3    1.3     Research problem Despite years of research, there is still considerable debate about the precise role that corrective feedback has on the development of writing for second language learners (Ferris, 1999; Ferris, 2004; Truscott, 1996; Truscott, 1999). While much study has been done on CF, recent work has shown that a concentration on one linguistic feature, focused CF, may provide the best means for real improvement (Han, 2002; Nicholas, Lightbown, & Spada, 2001; Sheen, 2007). Research has also demonstrated that grammatical accuracy remains a major concern for both teachers and students as it pertains to the academic writing of English L2 speakers (Rosenfeld, Leung, & Oltman, 2001). The debate regarding the worth of written corrective feedback has presented interesting and compelling arguments on both sides. As will be outlined in greater detail in the forthcoming chapters, the Truscott-Ferris debates beginning in the 1990’s ignited a significant increase in interest and attention paid towards investigating the role of feedback on students’ writing. Aside from questions regarding the basic worth of CF were also critiques surrounding the methodologies used in many of the studies investigating feedback. Truscott (1996) was vocal in questioning the design of various studies and the problem with comparing empirical studies which were dissimilar. Indeed, much of the fervour that was sparked by the Truscott/Ferris debates of the 1990’s has subsided, but this in turn has served a valuable purpose: to expand and extend the scope of research into corrective feedback. Today, the question is less if CF works, but how it works best. The aims of this study are to investigate any potential differences that concentrating on one, two, and three distinct linguistic features may have on the improvement of that feature or features in the participating students’ academic writing. I call these different ways of giving     4    feedback, tiered feedback. More precisely, I define tiered metalinguistic written corrective feedback as the various stages of focused corrective feedback beginning with the concentration on one grammatical feature and proceeding upwards in increments of one with the use of abbreviated errors codes. The notion of tiered corrective feedback is based on recent work that finds focused CF on specific linguistic features to be more effective in promoting grammatical accuracy than unfocused CF on all linguistic features (Han, 2002; Nicholas, Lightbown, & Spada, 2001; Sheen, 2007). Further, I hope to find the balance, if one exists, between focused and unfocused metalinguistic corrective feedback through this tiered approach. The place that written corrective feedback has in students’ lives, including their perceptions, feelings, and attitudes, will also be investigated.  1.4     Significance of the study As L2 writers have and will continue to have a major presence in the English academic community, it is paramount that feedback on academic writing serve a valuable purpose for promoting grammatical accuracy and that learners’ opinions are factored into the overall process. The results of this research will have a direct impact on teachers’ and students’ lives as both can benefit from increased knowledge of how best to give and receive corrective feedback and the role CF has on achieving grammatical improvements in academic writing. This study seeks to build on past research and contribute to a greater understanding of written corrective feedback, particularly as it relates to the degrees of feedback focus and the role of metalinguistic feedback on students’ academic writing. There have been various studies that have looked at both focused (Sheen, 2007) and unfocused feedback (Chandler 2003; Ferris 2006), but, as Ellis notes, “…to date, there have been no studies comparing the relative effects of focused and unfocused CF. This is clearly a distinction in need of further study” (2009, p. 102).     5    In general terms, this focused/unfocused dichotomy can be spoken of in exaggerated ways to create a clearer distinction: addressing one error is focused feedback, and addressing everything is unfocused feedback. As a teacher, this distinction is useful, but not very helpful. While focused feedback appears to be quite effective in the improvement of the focused features (Sheen, 2007), it is not practical to ask teachers to ignore all the other problems (be they grammatical, rhetorical, or otherwise) and focus solely on one targeted feature in the confines of many other priorities. Especially for content teachers who still wish to address linguistic concerns alongside regular class content, the use of very focused feedback on the two or three essays per term is impractical at best and irresponsible at worst. Finding a line where students still have sufficient focus to be able to more successfully process errors and make more substantive and longer lasting improvements could better meet the needs of teachers and students in more productive ways. This line between more and less effective tiers of feedback could help maximize time and effort for teachers and value for students. Undoubtedly, the cognitive loads that students can bare will vary across individuals and this line will likely never be completely uniform, but findings of this kind can at the least provide a basis from which to proceed. If addressing two features shows improvements in writing, but addressing three shows a decrease in efficacy, then teachers and students can benefit from knowing which tier of corrective feedback is most useful to them.  1.5     Research questions The many and somewhat contested methods of giving feedback on students’ writing show the importance of the process: teachers and students alike want the best possible ways to help improve writing quality. It is clear that, especially as language students enter into mainstream content classes, surface level errors must be minimized to achieve greater academic success.     6    There has also been significant research into the frequency and severity of these surface level errors and the potential for CF to have a greater impact on error reduction. It is based on these assumptions that the following research questions have been developed to investigate the effect of CF on reducing error. The research questions which guided this study are: Q1: What is the effect of focused metalinguistic corrective feedback on the accurate use of the targeted linguistic features? Q2: At what point does focused metalinguistic corrective feedback become unfocused? Q3: Is there a difference in the effect of varying intensities of written corrective feedback (on one, two, or three specified linguistic features) on students’ accuracy in the production of those features? Q4: What do NNS expect from their instructors as far as feedback on their academic writing is concerned? Q5: How do students perceive different types of corrective feedback?  1.6     Thesis organization Chapter One introduces the topic of the thesis as well as presenting the purpose, research problem, research questions, and significance of the study. Chapter Two outlines the theoretical framework and reviews the literature relevant to this study. Chapter Three outlines the specific research methodologies used in this study and presents a discussion of the data and the results. Chapter Four outlines the limitations of the study as well as the pedagogical implications that this research will have on the application of written corrective feedback on English L2     7    learners’ academic writing. It will also propose ideas for further research in the domain of corrective feedback.                                     8    Chapter 2: Literature Review  2.1     Introduction This chapter provides a review of the literature related to written corrective feedback and second language acquisition, as well as the theoretical framework used in this study.  I use the term corrective feedback in the same vein as Lightbown and Spada (2006) who define it as: Any indication to the learner that his or her use of the target language is incorrect. Corrective feedback can be explicit (for example, in response to the learner error ‘He go’ – No, you should say “goes”, not “go”) or implicit (for examples, ‘Yes, he goes to school every day’), and may or may not include metalinguistic information (for example, ‘Don’t forget to make the verb agree with the subject’).   (p. 197)  2.2     Second language acquisition This section will present a brief outline of the major SLA theories and their origins and will then transition to the theoretical framework which guided this research.  2.2.1    Behaviourism and structural linguistics From psychology, the theory of behaviourism was applied to both first and second language learning to describe the ways people learn a language as a process of conditioning and response to stimuli, and a language was learned wholly dependent upon the type of stimuli available to be inputted. Language acquisition was therefore not rooted in cognitive processes but rather as a response to these available stimuli and the subsequent punishment or reinforcement given to the learner. This approach to learning had a strong impact on the field of SLA, both theoretically and pedagogically. In order to learn an L2 properly, repetition of correct forms was encouraged and feedback was very important to help condition the learner (through     9    the use of punishment and/or reinforcement). The audio-lingual methods of the 1950s and 1960s were based on the theory of behaviourism and the process of memorization and repeated practice.  SLA was also heavily influenced by structural linguistics (structuralism) at this early stage. Structural linguistics viewed language as a discrete set of language forms and patterns which could be analyzed in each constituent part. This focus on analyzing these discrete units of language became influential in many L2 classrooms since language (it was thought) could be broken down, learned at its most basic level, and then built back up. Behaviourism and structural linguistics were highly influential during this period and were also viewed as complementary theories to achieve the same goal. VanPatten and Williams (2007) describe this relationship well. “Because structural linguistics portrayed language as based on a discrete and finite set of patterns, it blended easily with behaviourism, which viewed learning as the acquisition of a discrete set of behaviours” (p. 20). Although both of these theories continue to be somewhat influential in the SLA world today, they were critiqued for not recognizing the dynamic reality of language and the recognition that: (1) language is not simply a sum of its parts, and (2) a person cannot learn a language only by observing, repeating, and reacting to stimuli. This reaction to the constraints of behaviourism led to other advances in SLA theory, most notably the cognitive approaches largely inspired by Chomsky (1975) and his research into first language acquisition.  2.2.2     Cognitive approaches (innatism/nativist)  There have been a variety of second language acquisition theories that have emerged since the 1970’s, most, as Zuengler and Miller (2006) note, cognitive in their approach to describe and explain second language learning. These cognitive based theories of SLA have had a dominating     10    presence in the field (Firth & Wagner, 1997; Zuengler & Miller, 2006). Noam Chomsky’s (1975) work into L1 acquisition has had a significant influence on SLA following the recognition of the limitations of behaviourism and structuralism to account for language acquisition. Chomsky argued that humans have an innate ability to learn a language, facilitated by a metaphoric “black box”, or language acquisition device (LAD). This LAD is specific to humans and specific for language. Chomsky’s subsequent transformational and generative grammar asserted that the infinite number of language combinations possible made the basic tenets of behaviourism (in terms of language acquisition) highly questionable. Also important to Chomsky’s theory was the presence of the grammatical foundation, universal grammar, that is said to exist in all languages and that all humans posses. This concept of universal grammar is key to Chomsky’s theory as well as the cognitive SLA theories it spawned. According to Chomsky (and referring to an L1 context) a regular child needs only to be placed in the appropriate context in order to facilitate successful language acquisition. This proved to be very influential for early cognitive based L2 acquisition theories, such as Krashen’s monitor model (1981, 1982, 1985) which was heavily influenced by Chomsky and his theories of first language acquisition. Krashen’s five hypotheses to L2 learning are as follows.             1.   Acquisition-learning hypothesis: According to Krashen, “learning” and “acquiring” were two different concepts. Learning as a child does in a natural way was called “acquiring” whereas “learning” requires a conscious effort on the part of the learner. 2.   The monitor hypothesis was Krashen’s way to reconcile the distinction between “learning” and “acquisition”, particularly how learning influences or “monitors” acquisition. Providing the learner has sufficient time and understands the particular rule, this monitoring will allow the correct language form to be produced.     11    3.   The natural order hypothesis was based on the notion that L2 acquisition, similar to L1 acquisition, unfolds in an ordered sequence and that teachers need to be cognizant of this when they are teaching. Although teaching out of sequence may result in partial or short term use of the particular feature, if it is far removed from its place in the natural order, students cannot successfully acquire it. 4.   The input hypothesis, or  “i +1”, refers to the comprehensible input that is required for successful acquisition to occur. If the language being inputted is too complex, it is outside of the learners’ abilities (such as i + 2 or i + 3) and cannot be successfully acquired. 5.   The affective filter hypothesis: the affective filter is a metaphorical wall that prevents learners from learning due to a variety of emotional states.  Krashen’s theories, while much lauded for moving the field of SLA forward, have been widely criticized for assuming L2 acquisition was the same as L1, particularly his focus on the natural approach to language learning at the expense of explicit grammar instruction.  2.2.3     Constructivism and sociocultural theory  DeKeyser and Juffs (2005) note that: “Nobody would doubt that language, whether first or second, is an aspect of human cognition” (p. 437). The assertion that cognition plays a role in language acquisition is difficult to argue with. However, to claim that language learning is solely (or even primarily) a cognitive process has come under increasing critique in the last decade. Particularly since Lantolf (1994), we have seen a drastic increase in the focus paid to more social approaches to L2 acquisition. Although sociocultural theory has been developed largely from the theories of Lev Vygotsky, the following will outline constructivism in a more general sense,     12    including both social and cognitive models of constructivism, and will end with a discussion applying sociocultural theory to an L2 context.     Constructivism Constructivism, referred to as a philosophy, epistemology, and a theory of communication (Kaufman, 2004), currently plays an important role in how teachers are trained and how students are taught. As the central paradigm in western-based models of education, constructivism has its roots in two distinct but deeply interconnected sources: Piaget’s cognitive constructivism and Vygotsky’s social constructivism. Piaget’s (1950, 1953) cognitive constructivism affirms that a learner cannot actively use new information unless it has been first incorporated into previous schemata; a learner builds on previous experiences in the accumulation of new knowledge. This construction of knowledge takes place as a person interacts with and explores the world, but ultimately “cognition occurs in the head of the individual and … learners make intellectual sense of the materials on their own” (Felix, 2005, p. 86). This process has three interrelated states: assimilation, accommodation, and equilibrium. According to Piaget, humans have two ways of adapting to the external world and the acquisition of knowledge: first, new information and experiences are incorporated into pre- existing schemata (assimilation) or; second, if the pre-existing schemata do not exist or do not correspond to the experience, new ones must be created or old ones must be changed to accommodate the new information (accommodation). Through assimilation and accommodation, the internal mind and the external world find a balance (equilibrium). Vygotsky (1978) extended Piaget’s theories by recognizing that the social and cultural worlds we inhabit play a significant role in how we view the world and acquire new knowledge; essentially, meaning and understanding of the world are products of social interaction and     13    discovery. Instead of being passive vessels waiting to be filled with knowledge, people are social creatures who gain knowledge by their interactions with their environments. Unlike Piaget, who stressed the role of cognition as being central, Vygotsky felt that “cognitive operations originate in social interactions and [he] emphasized the role of language and culture in cognitive development as frameworks through which humans experience, communicate and understand reality” (Simina & Hamel, 2005,  p. 219). Although it was Wood, Bruner, and Ross (1976) who originally coined the term scaffolding, it is a concept most frequently associated with Vygotsky’s social constructivism and his theory of the zone of proximal development (ZPD). According to Wood, Bruner, and Ross, scaffolding is a “process that enables a child or novice to solve a problem, carry out a task or achieve a goal which would be beyond his unassisted efforts” (p. 90). The ultimate goal of scaffolding is to provide this initial guided help which then leads to greater self-sufficiency in future tasks. Scaffolding, and constructivism as a whole, is both internally and externally driven. The learner receives new knowledge from external sources and incorporates it into his or her ways of knowing. As a learner is scaffolded by a more able source, the process of guided learning to acquisition of new knowledge begins. Learning, according to Vygotsky (1978), takes place within the ZPD and the space between what is known and what is not yet known. To bridge this gap, a more knowledgeable teacher, parent, or peer helps facilitate learning as students build up new forms of knowledge. It is within this gap, between the actual and potential, that learning occurs.     Sociocultural theory In the L2 context, Lantolf (1994, 2000, 2006; Lantolf & Thorne, 2006b) has written extensively on sociocultural theory (SCT) and its application to second language learning.     14    Informed by Vygotsky (1986) and associates, Lantolf argues that the tenets of SCT, namely the use of an L2 to mediate mental and social activity, occupy important places in the discussion of SLA. In particular, Lantolf (2006), based on Vygotsky’s sociocultural theory of mind, notes two primary constructs of SCT in L2 development: mediation and internalization. Vygotsky argues that humans rely on both physical and symbolic (or psychological) tools to interact with the world, both of which are created by different cultures over the course of human history. Just as humans use physical tools to assist in the completion of tasks, we also use symbolic tools (or artifacts) to interact in the world and mediate our place in society, such as music, art, and most notably, language. These symbolic artifacts (like language) are thus appropriated by individuals to regulate their mental activity, also called internalization. Of particular interest to Lantolf (1994, 2000, 2006) is the place that a second language has in this process of mediation. Just as a first language is an important symbolic artifact (in fact, the most important artifact) so too is a second language. A learner’s L2 is an important tool used to mediate the relationships between the self and his or her world. This sociocultural approach as a theory of SLA best recognizes the role of this outside world as essential to learning a second language. The world is not something which exists separate from the learner, but this world is formative in knowledge construction (or language acquisition) and is the ultimate source of development (Lantolf and Thorne, 2006a). In the practical sense, sociocultural theory affirms the active role of language learners in construction of their own knowledge. It recognizes that although cognition has a role in language learning, this cognitive element is insufficient to completely explain or address the complex nature of learning a language.         15    2.3     Theoretical framework The framework for this study is based primarily on the work of Lev Vygotsky and Jean Piaget and their respective forms of social and cognitive constructivism. A discussion of constructivism and how it applies to corrective feedback and this research follows.  2.3.1     Constructivism and corrective feedback The basic tenets of constructivism encourage the active role of the learner in the (co)construction of knowledge. Metalinguistic corrective feedback provides this opportunity and encapsulates the tenets of constructivism by encouraging learners to build their own knowledge in a scaffolded manner. By providing clues, and not direct answers, metalinguistic CF allows for self-discovery while still providing guidance within a student’s ZPD. Metalinguistic feedback, particularly focused metalinguistic feedback, allows teachers to better guide their students’ learning by being active participants in the decision to concentrate on specific and problematic language features. Since the development of an L2 learner’s interlanguage is not always linear, concentrating on various features at different points in students’ learning should provide teachers with an appropriate balance of being able to focus on a variety of problematic language features at various times depending on specific learning needs. Although the teacher provides the external means for the acquisition of knowledge by giving feedback on targeted language features, the students must construct their own knowledge by internalizing the linguistic clues and adapting this new knowledge into old schemata. Specifically related to constructivism and SLA, Cobb (2005) notes that, while constructivism as an explicit concept may not be a dominant paradigm of second language pedagogy, many of the common-sense approaches to learning an additional language that teachers frequently employ are precisely constructivist in their approach.     16    The Ministry of Education in Quebec, Canada, has recently launched a major reform in the public school system. Everything from textbooks and teaching methods to classroom organization has been affected. The word constructivism is writ large across the many documents accompanying this reform; we are effectively living through a constructivist revolution. And yet, as many of the province’s language educators point out, the “new” recommendations for science and mathematics classrooms seem to be about doing precisely what ESL (English as a Second Language) and FSL (French) language educators have been doing for a long time: group work, project work, emphasis on active use rather than passive understanding, emphasis on what the learner does rather than on what the teacher does, concern for the key roles of motivation and prior knowledge— and  many others in the constructivist line-up. (Cobb, 2005, n.p.) It is well recognized that learning a second language is not a straight and error-free path from lesser to greater fluency, and that making mistakes, especially with previously learned concepts or features, is a natural part of language acquisition as learners continue to restructure their interlanguage. However, corrective feedback can help to provide the continued focused assistance that learners require to overcome language errors and move closer to greater proficiency. Specifically, the primary goal of providing metalinguistic clues to language learners is the hope that they will utilize this knowledge in a form of self-discovery and self-correction. Metalinguistic corrective feedback, perhaps more so than other more direct or obscure CF types, fits well into a constructivist learning theory where learners are guided to bridge the gap between the unknown (or forgotten) to the known (or remembered). With the use of metalinguistic feedback, students are active in their reconstruction of knowledge; they rely on the teacher as a guide but not as the ultimate source of information.     17    This study uses the term constructivism in the spirit of both Vygotsky and Piaget’s social and cognitive constructivism. Similar to Simina and Hamel (2005), this thesis takes “a complementarist position” (p. 220) in recognizing the importance of both individual cognition and the sociocultural world in the acquisition of an additional language. A synthesized view of cognitive and social constructivism views these two paradigms as interacting and thus complementary. Both Vygotsky and Piaget outline their theories of learning as the learner being central to the creation of knowledge. The differences in how this occurs, however, are substantial. For Piaget, the individual is primary in this construction of knowledge, and for Vygotsky, culture and social interactions are where knowledge develops. A synthesis of these two paradigms accounts for and values both cognition and the important role of individual learning in L2 acquisition, as well as the convincing argument that knowledge is also a shared experience which develops from social activity, or as Felix (2005) notes, “knowledge is constructed individually but mediated socially” (p. 86). The position taken in this thesis is thus one of symmetry; a recognition that both the individual and the social play, if not equal, then at least significant roles in the acquisition of a second language.  2.4     Corrective feedback and second language acquisition Teachers, students, or researchers who have criticized corrective feedback as being ineffective, or even harmful (for example: Truscott, 1996; 1999), may be missing the point. CF is not an island where all errors can be fixed in isolation. It is a part of the much larger SLA process, and like anything else in language acquisition, CF takes time to be effective. Truscott’s assertion that CF was at best a waste of time and at worst a detriment is intriguing, but it also might be at odds with what is known in second language acquisition theory. Hyland and Hyland (2006) outline this discrepancy.     18    SLA studies indicate that second language acquisition takes place gradually over time and that mistakes are an important part of the highly complex developmental process of acquiring the target language…. We cannot, in other words, expect that a target form will be acquired either immediately or permanently after it has been highlighted through feedback. Even though explicit feedback can play an important role in second language acquisition, it needs time and repetition before it can help learners to notice correct forms, compare these with their own interlanguage and test their hypotheses about the target language.… [M]any studies of feedback on error have ignored how language acquisition occurs, although the influence of feedback on the learner’s long term writing development fits closely with the SLA research.... (Hyland & Hyland, 2006, p. 85) Corrective feedback is not intended to be used as a replacement for regular classroom instruction, but rather as an addition; WCF is a supplement to already established classroom activities. It deals with error in real situations and embraces the concept that acquisition takes time and mistakes are a natural part of learning. However, the only way adult learners can use the language more accurately is by recognizing that they are making mistakes. Written corrective feedback can help provide this recognition.  2.5     Second language writing Much like SLA theory, approaches to writing instruction in both first and second language contexts have undergone some major changes. And like SLA theory, the early approaches to L2 writing were heavily influenced by research in first language acquisition. For the purposes of this thesis, I will focus on three major approaches to writing: current-traditional rhetoric, process, and post-process.      19    2.5.1     Current-traditional rhetoric Current-traditional rhetoric (CTR), or product-centered pedagogy (Matsuda, 2003), is an approach to writing instruction born in the 1960s as a response to early and mid-twentieth century writing theory and pedagogy which were rooted heavily in behaviourism and structural linguistics. CTR concerned itself with giving students the ability to write extended discourse, in particular academic discourse, and argues convincingly that writing is much more than a combination of discrete language features. According to Silva (1990), a current-traditional approach to composition studies assumes that “writing is basically a matter of arrangement, of fitting sentences and paragraphs into prescribed patterns. Learning to write, then, involves becoming skilled in identifying, internalizing, and executing these patterns” (p. 14). Current-traditional rhetoric came under scrutiny for its focus on the final product over the broader composing process. Matsuda (2003) sums up the discord felt by many critics of this writing approach. In the bad old days of current-traditional rhetoric, the story goes, students learned modes of discourse and applied them to write their five-paragraph themes on topics assigned by the teacher, which were then graded without the opportunity to receive feedback or to revise. (p. 67)  In reaction to the limitations of current-traditional rhetoric, other approaches sought to better account for the unique challenges of second language writing.  2.5.2     Process Out of the critiques of previous rhetorical traditions came the process approach, which has been highly influential in both first and second language composition studies. While current- traditional rhetoric was criticized for not encouraging creative thinking or creative writing (Silva,     20    1990), a process approach was claimed to better address these problems and encourage a more exploratory and inventive approach to composing. Process approaches to writing focus less on the final product and more on the various, and valued, stages of getting there. As Matsuda (2003) notes, the assumptions behind process pedagogy were: …of helping students discover their own voice; of recognizing that students have something important to say; of allowing students to choose their own topic; of providing teacher and peer feedback; of encouraging revision; and of using student writing as the primary text of the course. (p. 67) In the pedagogical sense, a process approach differs greatly from previous writing theory and instruction in that it allows students the ability to chart their progress, usually with the use of multiple drafts of the same writing task, and the scaffolded use of teacher feedback across these drafts. Arguments for this approach claim that considering writing as a process is more contextualized and better reflects the needs of L2 students. Critics argue that process pedagogy “oversimplifies the multiplicity of perspectives that have always been present throughout the 20th century” (Matsuda, 2003, p. 67), as well as possibly leading students to believe that writing is only about the process at the expense of a finished product.  2.5.3     Post-process Post-process was a term first used by Trimbur (1994) to describe the “social turn” which was currently in progress in L1 composition studies. This social turn, according to Trimbur, was “a post-process, post-cognitivist theory and pedagogy that represent literacy as an ideological arena and composing as a cultural activity by which writers position and reposition themselves in relation to their own and others' subjectivities, discourses, practices, and institutions” (p. 109). Process approaches to writing (and certainly current-traditional rhetoric) were being critiqued as     21    not adequately representing the “socially situated reality of writing itself” (Sinor and Huston, 2004, p. 371). Post-process theory, in recognizing the socially and culturally complex reality of writing, does not reject process approaches, but instead adds to them by recognizing the “multiplicity of L2 writing theories and pedagogies” (Matsuda, 2003, p. 65). In terms of actual classroom application, this theory (as a culmination and not a cull) recognizes the social, historical, and cultural lives of the writers, as well as the discursive nature of the writing process itself and the various stages involved in composing. Atkinson (2003) notes that the benefits of approaching L2 writing through a post-process lens is that it expands the conceptualization of writing past that of a “highly cognitive, individualist, largely asocial process” (p. 10).  2.6     A brief modern history of corrective feedback There is a rich history of written corrective feedback, particularly as it relates to the academic writing of English L2 students. The following well-known debate between Dana Ferris and John Truscott outlines some of the central issues currently surrounding written corrective feedback. John Truscott (1996, 1999, 2007, 2008, 2009) has been the most vocal opponent of error correction, and was responsible for igniting the debate and the subsequent call for further, and more methodologically appropriate, research. Truscott (particularly in his 1996 article) made some valuable critiques about corrective feedback and the potential harm that it does to writers. According to Truscott, error correction brings an air of negativity to the writing process, thus discouraging L2 writers to engage in long texts, or enjoy writing them. He also noted that, regardless of learners’ perceptions that CF is useful, it should not be given; that even though students may want or expect CF, “this does not mean that teachers should give it to them” (1996, p. 359). In Truscott’s (1999) later response to Ferris (1999) he elaborates this same sentiment:     22    How much of students’ false faith in correction is due to the reinforcement it receives from their teachers? To some extent, the argument from students’ beliefs is circular: By using correction, teachers encourage students to believe in it; because students believe in it, teachers must continue using it. (1999, p.116) Truscott’s (1996) claims that error correction did not work were based on some questionable positions. In support of his argument, Truscott used research from first language writing (Hillocks, 1986; Knoblauch & Brannon,1981) that showed error correction had little impact on improving writing. Although Truscott admitted that L1 and L2 research are not always synonymous, he stated that the findings in L1 studies “…certainly provide strong grounds for doubt” regarding parallels with L2 corrective feedback (1996, p. 330). He further used two often referred to empirical studies by Robb et al. (1986) and Semke (1984) to support his assertions that CF was ineffective. Both Robb et al. and Semke failed to find any significant difference between the different error correction groups that they were investigating. Truscott took these data and used them to assert that error correction did not work, instead of noting (as others have since) that all groups in these two studies improved, although the impact that CF had on this improvement was impossible to calculate because of the lack of control groups in both of these studies. To his credit, Truscott did wonder why a true control group in Robb’s study was not added. However, instead of saying this lack of a control would have helped the study’s credibility, he used this to further support his stance that error correction was not useful in the improvement of grammatical accuracy, and hence the improvement of writing. Regardless of the belief in whether Truscott’s claims were completely valid or not, he did raise some crucial questions about the lack of convincing research and literature in the area of CF. In his words,     23    The literature contains few serious attempts to justify the practice on empirical grounds; those that exist pay scant attention to the substantial research that has found correction ineffective or harmful. Most writing on the subject simply takes the value of grammar correction for granted. Thus, authors often assume the practice is effective, without offering any argument or citing any evidence. When someone cites evidence, it generally consists of only one or two token sources, with no critical assessment of them. (1996, p. 328) Of course, this lack of research has changed significantly in the last 13 years, perhaps largely because Truscott so blatantly critiqued an entrenched institution of the L2 writing world. This call for further research, whether completely Truscott’s doing or not, resulted in a refinement of research methodologies and a rise in interest as researchers and teachers alike attempted to find relevant, L2 specific proof for their assumptions and practices. Since Truscott’s 1996 critique, there have been a number of studies that have attempted to refute his position and argue for the long-believed worth of error correction. Dana Ferris (1999, 2004, 2010) took up Truscott’s challenge and responded with counter- arguments based on her own interpretation of studies, as well as her own research. In particular, her 1999 response to Truscott’s 1996 article questioned his cynicism towards CF by referring to problems with Truscott’s positions. “(a) The subjects in the various studies are not comparable; (b) The research paradigms and teaching strategies vary widely across the studies; and (c) Truscott overstates negative evidence while disregarding research results that contradict his thesis. (p. 4)”. In spite of their polar viewpoints, both parties, as the de facto representatives of their respective movements, agreed that more research was needed before any generalizing statements could be said with certainty.     24    2.7     Strategies of written corrective feedback There is a wide variety of potential corrective feedback choices when dealing with written error. And while the choice may be somewhat varied, it demonstrates the creativity and inquisitiveness of teachers and researchers who strive to find the most effective means to give feedback in an attempt to enact the greatest change. What follows is based on Rod Ellis’ (2009) summary of the various types of strategies used in the written corrective feedback that teachers give students. As the following research will outline, WCF can take many forms with varying degrees of success. But the most consistent finding (albeit not without debate) has been that the use of corrective feedback on students’ writing outperforms no feedback at reducing error and improving the overall quality of the texts. The use of terminology and classification with respect to the feedback types discussed in this thesis is based on the following typology of written corrective feedback, as outlined by Ellis (2009).  Ellis has categorized the various types of written corrective feedback into six major categories (see Table 2.1).              25    Table 2.1 Categories of Written Corrective Feedback (adapted from Ellis, 2009) WCF Types Description Direct CF This occurs when the correct form is given in place of an incorrect form. It is the direct correction of error. Indirect CF   1) Indicating only   2) Indicating the specific location Indirect CF occurs when an error is indicated but the correct form is not given. Ellis identifies two types of indirect CF: 1) Indicating only is when an error is noted, such as in the margin, but the exact location is not provided. 2) Indicating the specific location is when the error is underlined or given specific reference. Metalinguistic   1) Error codes 2) Brief grammatical description Metalinguistic feedback occurs when the writer is given a linguistic clue of the error. This can take two forms: 1) The use of abbreviations or error codes 2) A brief grammatical explanation usually given at the bottom of the text or on an attached form. The focus of the feedback  1) Focused   2) Unfocused Feedback can take a variety of forms in the way it is given, such as the level of focus. 1) Focused feedback occurs when a limited number of language features are concentrated on. 2) Unfocused feedback occurs when many or all language features are addressed in the feedback. Electronic Electronic feedback occurs via computer mediated methods when a hyperlink is used to indicate an error has occurred. Reformulation Reformulation occurs when a first language user rewrites or reformulates the targeted second language student’s text.   2.7.1     Direct corrective feedback Direct corrective feedback is the process of providing the correct lexical items, syntax, word choice, or style. Mistakes are crossed out and corrections are made. This can take various approaches depending on the marker, but little variation typically occurs: if the correct form is     26    provided in place of the incorrect form, this is known as direct feedback. The following outlines some important studies which have used direct CF in the investigation of student written error. Chandler’s (2003) study into the effects of four corrective feedback types resulted in mixed findings. The participants were first or second-year music students at an American conservatory and were L2 English speakers. Over the course of one semester, Chandler’s participants were required to write 5 five page autobiographical essays on which the feedback treatments were performed. Chandler used four treatment groups in her investigation into the effects of WCF on student writing, which she labeled: (a) Correction (also called direct correction (Ellis, 2009) where the correct response was either inserted or replaced the incorrect response); (b) Underlining with Description (also referred to as metalinguistic feedback where the incorrect item was underlined and the relevant error code was written in the margin); (c) Description of type only (metalinguistic feedback where an error code was written in the margin next to the line the error was in); and (d) Underlining (indirect feedback where the error was underlined but no other description was given). Two of these methods were reported by Chandler as being equally effective, Correction and Underlining with Description, with the author reporting less effective results for the other two treatment groups. Chandler’s findings, that direct feedback had equal benefits with indirect metalinguistic feedback, and that the two were superior to the others, may seem contradictory to constructivist or student-centered approaches to teaching and learning. With direct correction, there is inherently less opportunity for self-discovery. And yet, Chandler’s participants showed equal gains with the least student-centered approaches. This may have been in part due to the English proficiency of Chandler’s participants. Ferris and Roberts (2001) note that students often do not possess the linguistic and metalinguistic knowledge needed to process and react correctly to various corrective feedback types. Ellis (2009) further     27    adds: “[d]irect CF has the advantage that it provides learners with explicit guidance about how to correct their errors. This is clearly desirable if learners do not know what the correct form is (i.e. are not capable of self-correcting the error) (p. 99). Chandler sums up the strengths of both the correction (direct) and underlining (indirect) approaches: Direct correction is best for producing accurate revisions, and students prefer it because it is the fastest and easiest way for them as well as the fastest way for teachers over several drafts. However, students feel that they learn more from self correction, and simple underlining of errors takes less teacher time on the first draft. Both are viable methods depending on other goals. (p. 270) Chandler’s findings are at times difficult to understand because of the design of the study. Each group of students received all four types of corrective feedback at different stages of the treatment process. For example, if two students were part of the Correction group one week, then the following week (the following treatment) they would be part of the Underlining with Description group, and the following week they would be part of the next group. After receiving one type of feedback, students were then asked to revise that piece of writing (making the necessary corrections). Following this, on the next writing assignment, is when the impact of the previous week’s corrective feedback would be measured on the new writing sample. This approach becomes problematic for a variety of reasons. First, having received such a variety (and inconsistent) set of CF treatments, it is possible that students were either confused, disoriented, or influenced by one or more of the types of CF. By not measuring one type of CF using one specific (and unchanging) experimental group, it is difficult to predict the impact of that CF type on the participants’ writing. Further, by requiring students to revise corrected drafts, this study was not solely about corrective feedback, but about error correction. Although teachers most     28    likely hope that students always correct the errors which are given corrective feedback, this is likely less than comprehensive in the average L2 classroom. Unless teachers require students to correct the errors targeted by CF, it is likely that many students fail to do this (at least, all the time and with all errors). For Chandler’s study, it is therefore difficult to make any definitive conclusions about the impact of WCF on student writing when this impact could be from either the corrective feedback, the error correction, or both (which is most likely). Finally, Chandler failed to include a control group for this study, so the conclusions she draws have no comparison. Sheen (2007) also approached CF from a direct and metalinguistic stance in her investigation into the effects of focused CF on 91 American-based college ESL students’ writing. Her study investigated the effects of direct corrective feedback on the acquisition of English articles, with the use of three groups: a direct-only CF group, a direct with metalinguistic clues CF group, and a control group. Both treatment groups used a focused approach by only giving feedback on one language feature, articles. Two CF treatments were conducted over the course of the study. The participants were required to read and listen to a short story, and then asked to reproduce the story on their own. Following this the researcher used one (or none for the control group) of the CF treatments, depending on the students’ assigned groups. After receiving the corrective feedback, students were asked to look over their writing samples to see where their errors occurred, but were not required to make revisions, nor were the errors counted on these writing samples. To calculate the impact of the corrective feedback, Sheen used a quasi- experimental research design with pre-tests, post-tests, and delayed post-tests (three to four weeks after the final CF treatment). Her findings showed that both approaches resulted in significant improvement in the linguistic accuracy of students’ writing, but the method using a     29    metalinguistic approach showed greater long term improvements as evidenced by the delayed post-test results. The differences between Sheen (2007) and Chandlers’ (2003) findings may be a little confusing if the only comparison made is between the feedback types used. While both studies did investigate the differences between direct and metalinguistic approaches of WCF, they differed in important ways. Sheen’s study took a focused approach by concentrating on only one linguistic feature, articles, while Chandler’s study was very much unfocused, addressing a multitude of issues in writing instead of just one specific feature. Further, while Chandler did address what exactly she considered metalinguistic feedback to be (a simple notation of the error in a coded form), Sheen’s article did not include a thorough explanation of the metalinguistic descriptions used for her participants. An explanation, particularly regarding the explicitness of her metalinguistic descriptions, would have been particularly useful as this may have played an important role in the reported long-term success of the metalinguistic feedback group in her study. Robb et al (1986), in their much cited study of 134 Japanese college EFL students, investigated the effects of four types of feedback which they labeled as: 1) the correction group (also referred to as direct CF (Ellis, 2009) where the teacher fixed the error or inserted the correct form); 2) the coded feedback group (a metalinguistic approach where error codes were used and an accompanying chart assisted learners to decipher the codes); 3) the uncoded feedback group (an indirect CF approach where errors were highlighted but no clue was given as to why it was incorrect) and; 4) the marginal feedback group (another indirect approach, but more indirect than the previous. Here the total errors per line were calculated and the number written in the margins). These four groups outlined a descent from direct to indirect in the amount of feedback     30    given to students, ranging from the complete correction of every lexical, syntactic, and stylistic error (the correction group) to only making note of the number of errors per line and writing the number in the margin (the marginal feedback group). The authors found that there was not enough difference to be statistically relevant between the groups and thus concluded that no approach was preferable to the other. In fact, their findings led the authors to report that because of the similarities in outcomes for all treatment groups, this was evidence against direct error correction, stating that more indirect (and therefore quicker and easier for the average teacher to use) approaches are sufficient to effect change. Robb et al.’s study, while crucial to the subsequent development of WCF, has been criticized as having a major design problem by not including a control group (Sheen, 2007).  2.7.2     Indirect corrective feedback Indirect corrective feedback (ICF) occurs when an incorrect form is made note of, but no direct correction is made. Often this takes place by underlining or otherwise highlighting the item in question, but can also occur by making a note in the margin without exact indication of where the error occurred. For teachers believing in the importance of student self-discovery, this method perhaps has the most appeal. Since correcting and giving feedback on student writing plays an integral role in the ESL or EFL classroom, indirect methods of giving feedback are often the quickest and easiest to do. At the same time, this approach allows the writer to become cognizant that an error exists without being directly given the answer. One potential problem with indirect CF is if students do not possess the linguistic knowledge necessary to understand why (or where) the error occurred. This can be especially true for the “untreatable” (Ferris, 1999) or the problematic and ambiguous errors that English L2 students often make. If, for example, a student uses an inappropriate or non-academic word that expresses the same meaning as its     31    academic counterpart, an indirect approach to error correction with no explanation will do little good to improve the student’s writing. Ferris and Roberts (2001) and Lalande (1982) provide support for indirect corrective feedback by emphasizing the importance of self-discovery and the impact this has on long-term sustainable improvement in writing. Both of these studies show indirect approaches as being equal to or superior to direct approaches, as well as suggesting that ICF facilitates more substantial long term gains. Ellis (2009) states: In accordance with the general line of argument by Ferris and Roberts, it might be claimed that indirect feedback where the exact location of errors is not shown might be more effective than direct feedback where the location of the errors is shown…as students would have to engage in deeper processing. (p. 100) These findings have direct implications for SLA pedagogy, as indirect approaches are much quicker to perform, and might be more attainable or attractive options for many writing teachers.  2.7.3     Metalinguistic corrective feedback Metalinguistic feedback is the process of providing a linguistic clue for the targeted error(s). This explanation can take the form of error codes, as Ellis (2009) states is most often the case, or can come in the form of a longer and more detailed explanation. Bitchener (2008) published results of a study that discussed the effects of various combinations of direct WCF on the acquisition of English articles. The groups were: (1) direct corrective feedback with written and oral metalinguistic explanation; (2) direct corrective feedback and written metalinguistic explanation; (3) direct corrective feedback only; and (4) no corrective feedback. The research followed a pre-test, post-test, delayed post-test design with each participant, including the control subjects, required to produce three writing samples. The     32    treatment was as follows. Two weeks after the first writing assignment, and one week after the second and third, the researcher returned the students’ texts with the assigned CF treatments. The students were asked to study their papers for approximately five minutes. Depending on the students’ assigned group, they were asked to do the following: (1) The first group was given a 30 minute oral lesson (oral metalinguistic explanation). Immediately following this, students wrote their second piece of writing. (2) After looking over their returned papers, which had written metalinguistic explanations, students in the second group were asked to write a second short essay. 3) The third group was only required to look at the corrections (with no metalinguistic explanations). Following this they immediately wrote their second piece of writing. 4) The fourth group, the control, received no feedback but was still asked to write another sample to account for normal improvements or regression that may have been made by the participants. Like most other studies reporting on WCF, Bitchener (2008) reported significant improvements of all groups compared to the fourth group which received no corrective feedback. Although the improvements with all three groups receiving some sort of CF were generally quite similar, the third group, which included no metalinguistic feedback, showed the least gain and the first group, which had the most metalinguistic feedback (both oral and written) showed the most. Although the differences were too close to be statistically significant, it could be suggested that the greater amounts of linguistic knowledge that are available to students provide greater benefits (and this is indeed what Bitchener et al. (2005) had concluded in a previous study). While Bitchener’s (2008) work is indeed beneficial, it does not address the concerns that many teachers have in simply providing students with the correct answers without the need for self discovery. Regardless of whether a direct approach is more or less effective than indirect approaches (which Bitchener (2008) and Bitchener et al.’s (2005) studies did not consider), the     33    reality for most classroom teachers is that they may feel hesitant to correct written work completely, without students first trying to solve their own problems in a directed, scaffolded manner. As previously mentioned, both Lalande (1982) and Robb et al. (1986) also used metalinguistic approaches in their studies, and both reported improvements in the accuracy of their students’ writing, although the exact level is impossible to say due to methodological inconsistencies. Robb et al. failed to include a control group, and Lalande’s “control” groups still received direct feedback in the form of corrections.  2.7.4     Focused and unfocused corrective feedback Focused and unfocused feedback refers to the degree in which the language features in students’ texts are targeted. Focused strategies, for example, could take the form of concentrating on one specific feature, regardless of the other errors that may occur in the writing. Unfocused CF occurs when many or all of the errors are addressed in the student’s text. Sheen (2007) used a focused strategy with great success (see above) by using WCF in trying to influence the correct use of definite and indefinite articles. The benefits of focused strategies, she asserts, are significant in that they are able to better pinpoint problem areas, and thus reduce the potential confusion and cognitive overload of the writers/students. Written CF is complex. It addresses different aspects of writing— content, organization, rhetoric, and mechanics, as well as linguistic accuracy. The question arises, however, whether written CF should deal with all these aspects at the same time or address different aspects selectively when correcting different pieces of writing. L2 learners have limited processing capacity and asking them to attend to corrections that address a range of issues at the same time may tax their ability to process the feedback. One reason that     34    previous studies of written CF have failed to demonstrate any effect on students’ accuracy in subsequent writing may simply be that the linguistic feedback was not sufficiently focused and intensive. (p. 278) Unfocused corrective feedback, which deals with a variety of features, appears to be the norm for research and practice in WCF, and understandably so. Most teachers (and researchers alike) might have a difficult time ignoring large segments of problematic areas by concentrating on one specific feature, over a significant period of time. Practically, concentrating on limited features in the classroom may be difficult because of the students’ (and all other stakeholders’) expectations. And yet, as Sheen’s study asserts, there may be some important lessons to be learned from the focused-unfocused dichotomy in WCF and from further investigation into the impacts each one has on the improvement of writing.  2.7.5     Electronic corrective feedback Electronic feedback, as its name implies, uses computer-based means to draw attention to written error. “The teacher indicates an error and provides a hyperlink to a concordance file that provides examples of correct usage” (Ellis, 2009, p. 98). As the use of technology in the language class continues to grow, electronic means to provide feedback on writing will no doubt follow suit. There have been a variety of studies that have explored the use of computer mediated means of providing feedback, and have reported positive findings with respect to marked improvement of targeted areas. For example, Yeh and Lo’s (2009) study of 50 Taiwanese college students’ use of electronic feedback found that an online feedback annotation system slightly outperformed traditional paper-based feedback. Over the course of two writing assignments, the students in the two groups (the experimental group receiving online annotations and the control group receiving paper-based feedback) studied the feedback they received on both writing tasks.     35    Following the second writing assignment, all participants were given the same sample piece of writing which contained student errors and were asked to identify where the incorrect items were. The authors reported that the online annotation group had slightly better results and thus concluded that online annotations could be effective in treating student error. Nagata (1993, 1997) also reported positive findings in her study of 14 second year Japanese students’ acquisition of Japanese particles. Students were divided into two groups: one receiving online metalinguistic feedback by means of online particle exercises, and the other receiving translation feedback on the same online particle errors. Nagata reported that the group which received metalinguistic electronic feedback outperformed the translation group. Although both of these studies could have benefited from a true control group, their findings show that the use of online corrective feedback is an area which will continue to grow. Necessary for this growth, however, is the need for students to have consistent access to computers for their writing and subsequent analysis. For most teachers with limited resources or desire, this is likely an issue. In reference to my study, students did not have access to computers within class nor did I want them to use any sort of a word processor that could potentially modify their compositions with various automated features like spell-check or grammar-check.  2.7.6     Reformulation This approach to error correction occurs when an L2 student’s text is rewritten by an L1 speaker who maintains the general tone and content of the original work. The hope here is that, through seeing an L1 speaker’s rendition of the text with proper syntax, lexical choices, and rhetorical structure, the L2 writer can learn from this correct(ed) model and appropriate the forms and approaches into his or her own work. Sachs and Polio (2007) used reformulation in a study and compared it to a more traditional direct error-correction approach. While both     36    reformulation and direct error-correction outperformed the control group, the direct approach was consistently the most effective way to deal with error in this study. Regardless of the findings, it is clear that reformulation as a widespread practice in error correction is impractical, if not impossible. The time demands on teachers needing to rewrite entire compositions would be tremendous, especially given Sachs and Polio’s findings that this approach underperformed direct error correction, which is a much less demanding task for teachers.  2.8     The importance of grammar For older language learners, learning a language might require greater explicit focus of specific features (Long, 1990). Ferris (2002) discusses the importance of grammar instruction in detail, and states that the unique differences between the L1 and L2 make grammar instruction a necessity for properly developing L2 writers. She writes, One of the inescapable differences between L1 and L2 student writers is that the nonnative speakers make errors related both to negative transfer from their L1s and to incomplete acquisition of the L2. Though L1 student writing is obviously not error free, the errors made are different in quantity and nature. Because L2 students, in addition to being developing writers, are still in the process of acquiring the L2 lexicon and morphological and syntactic systems, they need distinct and additional intervention from their writing teachers to make up these deficits and develop strategies for finding, correcting, and avoiding errors. (p. 4) Silva (1993) also notes that the unique needs of L2 writers make L1 pedagogy inappropriate for unquestioned transfer. Despite noting that the “general composing process patterns are similar (p. 669)”, he outlines the major differences between L1 and L2 writing.     37    L2 writing specialists need to look beyond L1 writing theories, to better describe the unique nature of L2 writing, to look into the potential sources (e.g., cognitive, developmental, social, cultural, educational, linguistic) of this uniqueness, to develop theories that adequately explain the phenomenon of L2 writing. (p. 669) According to Silva, L2 writers generally struggle more, and spend more time writing and revising their texts, with less success, than L1 users. Their use of vocabulary is more limited and the clauses in their texts are longer than L1 speakers (a sign of their lack of ability to nominalize in their writing in order to increase density). Perhaps the biggest concern for many students is the ultimate consequence of poor writing: their marks. Silva also identified lower scores due to more errors and less general fluency with L2 writing. Language proficiency plays a major role in the success or failure of international students (Andrade, 2006). To study and, more specific to the context of this research, to write in another language, there needs to be a certain level of proficiency in order to be able to function at the appropriately high level needed to succeed. Johnson (1988) investigated the relationship between the English language proficiency and academic success of undergraduate international students at an English-medium American university (n=196) and found a strong correlation to suggest that English proficiency does have a significant impact on success in school. Her findings further suggest that certain skills may be more crucial for ESL students’ success at the post secondary level. Specifically, the study revealed a notable relationship between the students’ GPA and high scores in the written and reading sections of TOEFL, suggesting the importance of having a strong command of grammar in writing as an important predictor of academic success. Johnson’s study did present some contradictory findings to an earlier 1987 study done by Light, Xu, and Mossop, who found that English proficiency and academic success were not always related.     38    However, as Johnson (1988) notes, the differences in the studies’ subjects regarding their English abilities may have been an important factor in the discrepancies. For higher proficiency students like the ones included in Light, Xu, and Mossop’s study, the subtle differences between the English abilities of the participants did not have an impact on their success at school, whereas in the Johnson (1988) article, which investigated lower level students, this was clearly the case. Celce-Murcia’s 1991 article advocated the importance of integrating grammar instruction into L2 pedagogy. The author stressed the necessity for L2 writers to be grammatically accurate in their academic texts. “The importance of a reasonable degree of grammatical accuracy in academic or professional writing cannot be overstated” (p. 465). Citing an unpublished study from McCirt (1984), Celce-Murcia (1991) claimed that 7.2 surface level errors per 100 words were considered to be a failing piece of work at the university level. Whereas, after the correction of these errors, the same students’ texts were considered to contain strong ideas and content, showing the important role that grammar has in the creation of strong texts, or at least the perception of strong texts. Regardless of the meaning or ideas contained within L2 writing, there appears to be a line that cannot be crossed; in order to succeed, second language students still need to be concerned with surface level grammatical errors. Doughty and Williams (1998), citing lessons learned from immersion and naturalistic learning environments, claim that L2 students, particularly older or adult learners who learn in a naturalistic setting, may never develop certain linguistic features with high levels of proficiency. The following is a comment from the two authors on the differences between adult L2 and L1 learning: Taking the perspective that adult second language learning is, in many respects, fundamentally different from first language learning, we believe that leaving learners to     39    discover form-function relationships and the intricacies of a new linguistic system wholly on their own makes little sense. (1998, p. 11) Without specific and detailed attention to form, these learners may be able to attain communicative levels of language ability, but may not be able to reach the upper echelon needed to perform complex language tasks and write at an advanced level.  2.9     The importance of focused, explicit and consistent grammar instruction If grammar instruction is needed for second language learners to achieve high levels of academic writing proficiency, the question then becomes what kind of grammar instruction, at what times, and in what situations. The concentration on more traditional, teacher-directed, and grammar translation approaches and the subsequent backlash led to a void in effective L2 grammar pedagogy. However, more recent trends have seen the focus on grammar come back into discussion on best practices in L2 pedagogy and back into many classrooms. Since it is common, as Silva (1993) asserts, that many aspects of L1 and L2 writing differ, it is then necessary that these aspects be taught in clear and direct ways for L2 students. This dedication to L2 writing specific research is clearly evidenced by the proliferation of various studies in recent years (see Silva & Brice, 2004 for a comprehensive review). It is important to note that this return to a concentration on grammar is not a return to the past and to decontextualized, segregated instruction. The call to bring grammar instruction back into the second language classroom was prefaced with the need to do it in proper context, instead of as an autonomous and separate endeavor not related to students` actual writing. Long (1991) and Long and Robinson’s (1998) discussion of focus on form (FonF) research values the importance of integrating overt grammar instruction into the communicative language classroom. The authors make a careful distinction between this more contextual approach to     40    grammar (FonF) and the more traditional approach to grammar instruction which they refer to as focus on formS (FonFS). A FonFS approach to grammar instruction often teaches grammar as discrete points separate from the students’ own writing (or speaking) whereas FonF tries to integrate grammar instruction into the students’ individual work or in the context of their own language use. In Long and Robinson (1998), special reference is made to Canadian French immersion students and examples of significant difficulties acquiring L1-like grammatical competency, despite long tenures in immersion classrooms. The authors note that “additional salience for the problematic features seems to be required, achieved either through enhancement of positive evidence or though provision of negative evidence of some kind” (p. 21). This suggests that the needs of L2 learners are unique, requiring special attention in specific ways and that learning a language purely by being immersed in a more natural-like setting is insufficient to attain advanced use. Yip (1994), in her pilot study of the grammatical consciousness raising of “ergative” verbs1 of 10 advanced ESL students at an American university, found that focused and explicit instruction was successful in promoting greater accuracy in the future use of these verbs. Yip investigated the impact that grammatical consciousness-raising (C-R) had on students’ acquisition of ergative verbs. C-R, according to Yip, focuses students` attention of targeted grammatical features, with the ultimate goal being to assist students with self-discovery by directing them to the incorrect forms and eventual acquisition of the correct form. In this empirical study, students were first given a pre-test and were asked to classify a sentence into one of the following categories: clearly grammatical, probably grammatical, probably                                                              1 Ergative verbs are verbs that when used cause the grammatical subject to function as the object. For example: “The window shattered”. Here “window” does not act upon the verb, but rather is being acted upon, as in “The window (is) shattered (by the rock)”. See Yip (1994) for a more thorough explanation of ergative verbs.         41    ungrammatical, and clearly ungrammatical. Following the pretest, two instructional periods (grammatical consciousness-raising sessions) were held to help raise students’ awareness of ergative verbs. Two weeks following the C-R sessions, post-tests were administered which used the same questionnaire as the pre-test. The results of the post-test showed a decrease in errors for the majority of the students. The author argues that these findings are suggestive that this sort of concentrated consciousness raising activity is useful for improving the grammatical accuracy of ergative verbs for advanced ESL students. This pilot study reveals interesting data about the potential effectiveness of focus-on-form instruction, although two major factors must be considered. First, the small sample size of 10 participants means limited conclusions should be drawn from the study. Second, Yip’s use of identical test items in both the pre- and post-tests means the students’ improvement of ergative verb use could be the result of memorization and not true growth. Other authors have provided comprehensive summaries of the values of form-focused instruction (Norris and Ortega, 2000; Hinkel, 2004b). The results of these studies suggest that focused instruction of grammar in L2 contexts is superior to little or no focus. Norris and Ortega’s (2000) analysis of a wide body of research concluded: “In general, focused L2 instruction results in large gains over the course of an intervention. Specifically, L2 instruction of particular language forms induces substantial target-oriented change…” (p. 500), and “focused instructional treatments of whatever sort far surpass non- or minimally focused exposure to the L2” (p. 463). Hinkle (2004b) strongly mirrors Norris and Ortega’s sentiments about the necessity to include grammar, in concentrated ways, into L2 instruction for students to attain high levels of proficiency. She states,     42    Instruction in L2 vocabulary and grammar improves learners’ receptive and productive skills and provides important means of expanding NNS lexical and syntactic repertoires necessary in L2 reading, constructing academic text, listening, and other fundamental facets of functioning in the academy. (p. 13) With the realization that no grammar in the second language classroom was as bad (or worse) than only grammar (or out of context grammar), a re-approach to targeted and explicit, but integrated, grammar instruction became accepted as important for L2 learners to become highly proficient in their target language. The understandable backlash against out of context grammar instruction, tedious grammar worksheets, and parsing sentences is understandable. It is difficult to believe that this approach to grammar can translate successfully into students’ actual language production. And yet the Krashen (1981, 1982, 1985) and L1 pedagogy influenced anti- grammar movement was equally harmful to L2 learners and their quest to achieve advanced English proficiency. Corrective feedback is one part of a much larger system where grammar instruction plays an important role in acquiring a second language. In particular, as has been learned (I am sure somewhat painstakingly) by teachers and students over many years, this instruction is best served in relation to the students’ actual work and not as separate drills with no association to their actual writing. Classroom instruction of grammar and students’ individual practice, whatever form this takes, should combine with CF to form a union of language teaching that is needed to successfully flood the language learner with appropriate input to positively affect their output and help to restructure their interlanguage. Although there is still no consensus about what types of WCF may be most effective, recent research into the efficacy of focused feedback is encouraging, as well as the longer term gains of metalinguistic feedback (Sheen, 2007).     43    2.10     The relevant grammatical features The choice of the targeted linguistic features used in this study has been informed by a large body of research on: (a) the most common errors in English L2 academic writing, (b) the most problematic errors in English L2 academic writing as reported by relevant stakeholders (primarily teachers, professors, and researchers) and, (c) the “treatable” grammatical errors in L2 writing (Ferris, 1999).  2.10.1     The most common errors Depending on the level of the language learner, with the obvious exception of highly proficient L2 users, certain errors are more common than others and appear regularly in students’ work. Santos’1988 study of 178 professors at the University of California, Los Angeles used two compositions written by two L2 university students to measure the professors’ reactions towards various error types. The study identified the following as being most common in ESL writing, informed by previous studies by Tomiyama (1980), Vann and Meyer (1984), and Sheorey and Ward (1984) and further corroborated by his findings: articles, lexical items, relative pronouns, conjunctions, possessives, prepositions, singular/plural, subject-verb agreement, verb tense, and word form. Santos’ study revealed that omission or misuse of articles and verb problems were amongst the leaders in terms of what professors were most critical of. And although the professors generally agreed that the errors did not impede the comprehension of the students’ texts, they viewed the frequency of these error types to be academically inappropriate. These findings regarding the important impact of surface level errors on students’ writing are further corroborated by Hinkel (2004b) in her thorough examination of the literature. Hinkel refers to a vast wealth of previous sources in claiming that grammar (and vocabulary) plays a crucial part in how students perform at university. More particularly, for teachers who have worked with ESL     44    students’ writing at any sort of a level, these findings should not be surprising, both in terms of the predictable error types and the potential impact of these errors on students’ academic success.  2.10.2     The most problematic errors All errors are not equal, and this is especially true regarding teachers’ assessment of these errors. Certain problems in ESL students’ writing tend to be considered more or less egregious depending on a multitude of factors. Vann, Meyer, and Lorenz’s (1984) survey of 440 Iowa State faculty investigated teachers’ opinions regarding ESL students’ errors in written work. The study revealed some interesting findings. Generally, older teachers were more lenient with error compared with their younger colleagues. The authors wondered whether this was due to a change in attitude and an increase in tolerance as teachers age, or because of a product of their respective educations compared with professors of other generations. While there were slight variations across the various disciplines, general trends did emerge. A hierarchy of more and less problematic errors emerged in the questionnaires and rating scales the professors were given. Each professor was given 24 sentences with various errors (both global and local2) occurring in each. They were then asked to rate on a holistic scale from one to five with one being the least tolerable in academic situations and five being most tolerable. The errors which were identified as being most problematic were (ranked from most to least tolerant): spelling, articles, comma splice, prepositions, pronoun agreement, subject-verb agreement, word choice, relative clauses, it-deletion in cleft constructions, and word order. With the exception of pronoun agreement and the addition of verb tense, Hinkel’s (2004b) summary of a variety of studies is identical with Vann, Meyer, and Lorenz’s assessment.                                                              2 Global errors are errors that have an impact on the overall understanding of a text whereas local errors, while still errors, do not impede comprehension (Ferris, 2002).         45    Rifkin and Roberts’ (1995) critical review of the research discussed the notions of “irritation” and “acceptability” as it pertains to teachers’ assessment of students’ L2 errors. They caution that evaluators’ social attitudes play a large role in determining perceptions of error, and that error gravity (EG) studies were problematic due to their lack of consistent designs and variability. Despite the occurrence of variation that takes place amongst teachers, institutions, disciplines, and across time and space, certain trends that classify errors as more or less problematic are important to note because of the real consequences that ESL errors can have on English L2 students’ academic writing and the corresponding grades they are awarded. However, as Rifkin and Roberts state, error gravity studies and the claims made by researchers must be viewed carefully and not used to make grand assumptions. Despite this caution, there is still much to be learned from EG studies, namely that the perception of professors is the reality for students; whatever professors feel will be a detriment to student writing should be noted by students and writing programs alike so teachers can focus on helping students improve. Despite these findings, Ellis (1994) cautions against a definitive statement which claims that all teachers view (and grade) errors the same. As with error production itself, this tends to be somewhat subjective depending on the various academic disciplines, the students, and the teachers. Teaching and learning are ultimately very personal endeavors, and while the aforementioned studies confirm some general trends regarding perception of error in L2 writing, they should be seen as a guide to help gain some understanding into what errors occur most frequently, and what errors are usually penalized most heavily. The attempt to classify errors as being less or more serious is also slightly problematic, as Albrechtsen, Henriksen, and Faerch (1980) note that, “[a]ll errors are equally irritating” (p. 395). Therefore, the previously mentioned research needs to be properly contextualized for the purpose of this present study. L2 writers     46    struggle with a variety of factors, some considered more or less treatable than others, and some considered more or less “annoying” or egregious. Somehow, teachers must find ways to address all of these problems, but ultimately the context is very important in this process. The question about whether errors should be addressed should not be if, but rather how: how many, how often, and how best to approach them.  2.10.3     The most treatable errors While there has been much discussion about what features are appropriate for what students to learn (and at what time), it is clear that an all inclusive definitive answer probably does not exist. As Williams and Evans (1998) note that “each individual student has a point of readiness” (p. 155); all students travel along the path of acquisition in slightly unique ways, and it is largely the responsibility of the teacher to be able to gauge which language feature or features should be targeted and addressed. If the targeted features are outside of the learner’s present ability or are developmentally inappropriate, then feedback will likely do no good, whatever form it takes.  If, however, the targeted features are both “treatable” and within the learner’s abilities, then CF appears to be a positive force to impact second language writing. Ferris (1999, 2002), Ferris and Hedgcock (1998), and Ferris and Roberts (2001) all point to the necessity for feedback on student writing to be developmentally appropriate. “…[I]t makes sense for teachers to focus their feedback and instruction on aspects of grammar that can be explained, understood, and generalized to students’ particular writing needs…” (Ferris & Hedgcock, p. 205). Ferris (1999) dichotomizes possible correction areas in students’ writing into two categories: treatable and untreatable. The treatable errors (such as subject-verb agreement, run-ons and comma splices, missing articles, and verb form errors) are so categorized because of the set of explicit rules governing their use. Ferris explains that because of these governing rules,     47    these sorts of errors are most amendable to change. Untreatable errors are errors without such cut and dry explanations, which include lexical choices, word order problems, and non-standard language. Although Ferris (1999) does not argue that untreatable errors should be neglected because “[t]here is no handbook or set of rules students can consult to avoid or fix those types of errors” (p. 6), she does caution that teachers must approach treatable and untreatable errors in different ways. A one size fits all approach to error correction is inappropriate for both the error types as well as the students. Ferris and Roberts (2001) further this discussion in their study of 72 university ESL students and the impact of CF on self-revision. The authors, noting a gap in research regarding levels of explicitness in CF and the impact this has on self editing, investigated two types of corrective feedback, one explicit (the use of error codes or, as Ellis (2009) refers to it, metalinguistic feedback) and one not explicit (no error codes). The categories of errors, verb errors, noun ending errors, article errors, word choice, and sentence structure were chosen based on data claiming that these were the most frequent errors made by ESL writers in a similar situation to the Ferris and Robert’s study (Chaney, 1999). After receiving their essays back with the corrective feedback, students were asked to spend 20 minutes self-editing with help from the CF. The authors report no significant differences between the two CF treatments, and thus conclude that less explicit (and less time consuming) methods of CF are as effective in helping students self edit their writing.  2.11     The concentration on limited features While there is a rich history of research dealing with corrective feedback, recent work has shown that a concentration on one linguistic feature, focused CF, may provide the best means for real improvement (Han, 2002; Nicholas, Lightbown, & Spada, 2001; Sheen, 2007). In particular,     48    Sheen’s (2007) study investigated the effects of focused direct-only feedback and focused metalinguistic feedback on the improvement of intermediate ESL students’ definite and indefinite article use at an American college. Sheen found that indirect methods of CF may not be more effective than direct approaches when dealing with ESL written work, and that the most effective means was a combination of direct and metalinguistic feedback. The cognitive processing and the delay in finding the correct form that indirect CF demands on students may be less efficient than direct approaches in helping students improve their writing (Chandler, 2003). Of particular interest was Sheen’s use of focused CF in her approach to dealing with error. Instead of addressing most or every error found in the writing, Sheen’s concentration on English articles seemed to focus the students’ attention enough to lead to definitive improvements in the post-tests and delayed post-tests. This focus concentrates a learner’s attention on a specific feature in hopes of providing better opportunities to successfully decrease error. Although Sheen notes that both the direct-only and direct metalinguistic feedback groups far outperformed the control group, the metalinguistic group showed greater improvement, particularly over a greater period of time as evidenced by her delayed post-test results. Sheen’s claim, however, that the study focused solely on articles is not entirely accurate. She writes: “To mask the focus of the study, the researcher corrected a few errors other than those involving articles” (p. 264). It is unclear why the author needed to mask the focus of the study or how many “a few errors” were. Since the CF treatment was direct but not entirely focused (with other errors receiving some sort of unspecified treatment) limited claims can be made regarding the role that the concentrating on one linguistic feature had on the learners’ growth. Regardless of this inconsistency, Sheen’s study has drawn attention to some important aspects of the impact that metalinguistic feedback may provide on the successful reduction of error.     49    Ellis et al.’s (2008) study of 49 Japanese EFL students investigated the role of focused and unfocused direct written CF on the accuracy of definite and indefinite articles. Each of the three groups performed three separate but similar writing tasks, with two of the groups being experimental, and the third a control group. The students read and listened to the same short story and were then asked to reproduce the story in writing as accurately as possible. The writing tasks were subsequently analyzed for the accuracy of article use over the three treatments. The authors concluded that the two experimental groups showed significant improvements in their accuracy of articles over the course of the treatment, the post-tests, and the delayed post-tests relative to the control group. They further concluded that neither treatment, focused or unfocused, had any advantage over the other, and although the focused feedback group did continue to show improvements over a longer period of time (between the post-test and delayed post-test) the results were not significant enough to be statistically relevant. This very interesting study evokes a few questions: How genuine was this writing experiment? How often are students required to reproduce a story verbatim? Although these writing tasks were meant to elicit a particular response which encouraged the use of the targeted linguistic feature(s), it seems to be somewhat contrived. In particular, by not allowing or requiring the participants in this study to produce original written work, there are questions regarding the students’ writing and its authenticity.  2.12     Focus on form With the movement towards communicative teaching practices and integrated form- focused instruction, there was recognition that the backlash against grammar teaching was too harsh, however well intentioned it had meant to be (Doughty and Williams, 1998). While teaching grammar which is out of context through drills, memorization, and worksheets was     50    recognized as being problematic, the complete natural learning advocated by Krashen (1981, 1982) and the trend to focus on comprehensible input largely at the exclusion of grammar was equally problematic. Subsequently, the focus-on-form movement and the recognition that grammar knowledge is crucial to attain high levels of proficiency became more prominent in English language teaching discourse and practice. The tenets behind focus on form and a return to grammar teaching merge well with corrective feedback and error correction. The process of corrective feedback is about classroom teachers and students concentrating on grammar in relevant situations, and not as discrete-point grammar instruction. Focused metalinguistic CF provides this opportunity by raising awareness of grammar by using students’ own writing. Key to the assumption that CF is operating within the parameters of FonF is context; drawing attention to linguistic features in appropriate situations. This is compared with the more traditional FonFS which approaches form focused instruction as an autonomous endeavor separate from the students own work. At every step, for every attention to error or correction made, decisions in FonF are based on the learner and his or her individual needs. Doughty and Williams (1998) note that:  It should be kept in mind that the fundamental of focus on form instruction is that meaning and use must already be evident to the learner at the time that attention is drawn to the linguistic apparatus needed to get the meaning across. (p. 4) Although general instruction is a crucial part of second language growth, written CF allows the learner to be able to pinpoint his or her unique error patterns and be able to see them as they occur in his or her own writing. Focus on form instruction and written CF appear to be alike in this philosophy: language in context will provide the best opportunity for growth.       51    2.13     The learner It has been written about extensively that students appreciate CF and find it useful (Radecki & Swales, 1988; Enginarlar, 1993; Ferris, 1995; Zhang, 1995; Diab, 2005) and that they feel upset when teachers do not comment on their errors (Leki, 1991; Hedgcock & Lefkowitz, 1994; Ferris, 1995; Ferris & Roberts, 2001). Radecki and Swales (1988) reported on the corrective feedback views and preferences of 59 ESL students at an American university. Although, by the authors’ own admission, this study is limited in what it can claim due to the small sample size, their findings are still valuable. Most students (87%) appreciated teacher feedback on their writing, as well as expecting direct correction on their errors. The authors make this observation about the problems of students appreciating CF, but also expecting mass correction of all surface errors. “If they [teachers] do not surface-correct but respond to a writer's meaning, their credibility among their students can be impaired” (p. 364). Enginarlar (1993) surveyed 47 freshmen EFL students regarding their attitudes toward corrective feedback on their written work. While most of the students reported moderate opinions about whether feedback was interesting, the vast majority expressed finding CF useful, necessary, and didactic. Zhang (1995) investigated the role of peer feedback in L2 writing and found that students strongly prefer teacher feedback over that of fellow students. In contrast with L1 research that has shown peer feedback can play a positive role in the affect of student writers, Zhang’s study confirms past studies (Semke, 1984) which found that L2 students generally trust and value teachers’ feedback over other types of feedback.     52    Diab’s (2005) case study of two students and their teachers also reinforces these aforementioned studies, and provides a qualitative voice to the statistics. Diab’s student participants reflected their views of the important role that written feedback plays in the writing process. Both students acknowledged that teachers’ comments are essential. According to Vivian, “I read the comments at the end of the paper first, because they are the most important, but also all the comments are important I think.” Moreover, Zeina said, “The more comments I see the more it’s useful to me…. I want to see them because I want to see what I should do, what I’m doing wrong. (pp. 38-39) These two students strongly mirror the sentiments expressed in the aforementioned studies reporting strong, and generally positive feelings, regarding student perception of the values of written feedback.  2.14     Summary This chapter introduced some of the central issues surrounding corrective feedback and the written work of second language learners. It introduced the theoretical framework that is used in this research and presented the typology of written corrective feedback that will be referred to throughout this thesis. There are many viewpoints regarding CF, some of them very opposing and at times contradictory. If one clear result has come out of the disagreement, it is the need to continue investigation into the roles CF has on improving students’ written accuracy, as well as the agreement that there is no easy solution and no conclusive answer to these issues. Largely, written CF is subjective and contextual. It will vary from teacher to teacher, and the impact it has on students will never be completely predictable or uniform. However, this variability does not mean that best practices cannot be further refined. Fortunately, for teachers, researchers and,     53    more importantly, students, the work of corrective feedback continues to grow and will continue to be a rich source of data in the future. Finally, this chapter closed with a brief discussion of the role of student perception towards the feedback process. The following chapter will present on the method used in this study and will also report on the results.                        54    Chapter 3: The Effects of Tiered Corrective Feedback on Second Language Writing  3.1 Introduction The purpose of chapter three is to present a brief review of the relevant literature and to report on the method used and the quantitative and qualitative findings from this exploratory study.  3.2 Literature review 3.2.1 Corrective feedback and SLA Corrective feedback is an integral part of acquiring a second language. In particular, feedback in the language classroom is both ubiquitous and wholly essential to greater success in acquisition. Learners are consistently receiving some sort of corrective input from their teachers (or peers) in both oral and written forms and both play an important role in second language learning. Although both forms of oral and written corrective feedback are important parts of SLA, written corrective feedback is situated slightly differently than oral feedback in that it requires more processing time and usually deals with greater instances of error but poses less of an immediate cognitive load on short term memory. As Sheen (2007) notes, These differences may explain why, in contrast to the SLA research that in general has shown that oral CF is effective, L2 writing researchers have not been able to convincingly demonstrate that written CF leads to improvement in grammatical accuracy in new pieces of writing. (p. 257) That written CF is a challenge comes as no surprise to language teachers. Because it is so taxing and comes in so many forms, it is also not surprising that results of empirical studies (or at least interpretation of results) have led to disagreements as to the efficacy of WCF.     55    The critiques of WCF have pushed the field to further research in order to justify its position in the larger SLA continuum. It is well documented that language learning is not a rigid trajectory of lesser to greater skills and that often the existence of error in language use is a sign of positive growth (Lightbown & Spada, 2006). If a learner can use language correctly one day, he or she may not the next. This is how acquisition occurs; a back and forth process where new errors may be the result of emerging structures. As such, the place of feedback in the classroom cannot be denied: to help learners become aware of the error as they continue to strive for the correct form and the next rung up the SLA ladder. Not all are in agreement that negative evidence (used predominantly in WCF) is useful for language acquisition. Chomsky (1975) and other universal grammarians (UG) believe that the innate mechanisms that all healthy humans have to learn a language mean that corrective feedback plays a very minimal role in language acquisition. UG theorists hold that positive linguistic evidence is needed to effectively change interlanguage grammar, as opposed to negative evidence which only provides a temporary effect on language production and not on the restructuring of interlanguage grammar. In the SLA context, Krashen (1982, 1985), drawing largely on Chomsky and the nativists, applied a similar position to the role of negative feedback in acquiring an additional language. In Krashen’s hypothesis of the differences between (conscious) learning and (subconscious) acquiring, the place of corrective feedback in SLA is not substantial to overall proficiency. Like the UG nativists, Krashen focused on the impact of positive evidence in language acquisition, and argued that corrective feedback may be harmful to overall acquisition because it takes away valuable time needed for comprehensible input of the language. Although the influence UG and Krashen in SLA cannot be understated, both have been     56    criticized as not fully recognizing the differences between first and second language acquisition and the importance of conscious learning, or “noticing” (Schmidt, 1990), in SLA.  3.2.2 A brief modern history The dialogue in the 1990’s between Dana Ferris and John Truscott has been thoroughly discussed, and yet its importance to the study of written corrective feedback is worth reviewing. These debates arguing the benefits and harms of WCF reinvigorated interest and discussion into feedback practices and the place it has in SLA pedagogy. Perhaps most interesting about this dialectic were the passion of Truscott in his skepticism and the honest reply of Ferris (1999) when she stated: I must admit that I was secretly hoping I could agree with Truscott and accept his argument. Like most people, I find responding to students’ written errors time-consuming and mostly tedious. I also find that the time and energy I spend sometimes does not pay off in long-term student improvement. Wouldn’t it be convenient if we could all just agree that grammar correction does not work and that we should spend our in-class and out-of-class time focusing on more engaging issues of content and rhetoric? (Ferris, 1999, p. 2) Providing feedback on writing is a difficult and arduous process. If, as Truscott asserts, WCF does more harm than good, then many teachers might be the first in line to implement his claims. Although the debate still persists in a much more subdued voice today, it is important to look at the recent history of the Truscott inspired debates in more detail to guide present scholarship. John Truscott (1996, 1999, 2007, 2009) has been clear in his opposition to corrective feedback. In 1996, the article that famously started it all, Truscott questioned the “institution” of giving feedback on students’ grammar, stating there was simply not enough evidence to warrant     57    its continued practice and that the research that existed was proof of the ineffectiveness of grammar correction. Truscott claimed that CF is not only ineffective but also harmful. It produces too much stress in learners, thus lessening their desire and will to write effectively and to improve, and it is a waste of time for both the student – to read and respond – and the teacher – to give. Responding to the glut of empirical data and research espousing the benefits and effectiveness of error correction and corrective feedback since 1996, Truscott’s (2007) meta- analysis further claims that if error correction works at all, then it “most likely has small harmful effects on students’ ability to write accurately and that we can be reasonably confident that if it does have any genuine benefits, they are so small as to be uninteresting” (p. 256). While undoubtedly there are others who share Truscott’s concerns regarding CF’s usefulness (Fazio, 2001, for example, found her grade five students actually increased in error production over a four month period), there are certainly none as vocal. In being so vocal, he has in many ways become the face of the anti-feedback movement. Ferris’ (1999, 2004) response to Truscott, as well as her own research into corrective feedback (1995, 2002; Ferris, & Roberts, 2001), offer both refutations of Truscott’s bold claims as well as support for the effectiveness (and students’ preferences) of CF. Ferris (1999) critiqued Truscott’s position as misguided based on his use of incompatible data (such as many L1 studies that may not be transferable to an L2 context), a failure to include research that contradicted his thesis, and by exaggerating claims and empirical data that supported his arguments. Chandler’s (2004) response to Truscott’s critique of her earlier study (2003) mirrors Ferris in her defence of corrective feedback and the stark disagreement that all grammar correction is bad or harmful. An important conclusion to be drawn from this debate is the need for further investigation into the     58    place CF has in effecting positive linguistic growth, particularly as it relates to the kind of feedback being used.  3.2.3 Strategies of written corrective feedback Direct WCF, the act of fixing an incorrect form, and indirect WCF, the act of indicating where an error exists but not correcting it, are two sides of the same feedback coin and one many teachers struggle with. The direct versus indirect CF dilemma is difficult to resolve. For teachers who firmly believe in the value of constructivist teaching and learning, fixing grammatical errors on student writing may not be entirely satisfying. Student centered learning is difficult because in many ways it requires more work for both the teacher and learner. The teacher needs to find more creative ways to guide students, and students are ultimately responsible for their own knowledge acquisition. With direct correction, there is much less ambiguity, and it is a form of WCF that is both widespread and also somewhat contentious in that students are not required to actively discover new knowledge or fix their errors by themselves. Rather, through direct feedback, the correct response is provided without any additional student effort. Despite the potential misgivings by teachers, direct corrective feedback has been shown to be effective, especially for learners who may not possess the appropriate linguistic knowledge to process the errors and understand how to fix them (Ferris & Roberts, 2001). Chandler’s (2003) investigation into four feedback types demonstrated that direct WCF was as effective as indirect WCF in decreasing error. Chandler’s findings, that students with direct feedback outperformed those with more indirect CF on revisions, is not surprising. More interesting is that direct CF was most effective on reducing error in subsequent new writing. Chandler observes that: Perhaps when ESL students can see their errors corrected soon after writing, they internalize the correct form better. Perhaps the greater cognitive effort expended in     59    making their own corrections is offset by the additional delay in knowing whether their own hypothesized correction was in fact accurate. (p. 291) Ferris and Roberts (2001) and Ellis (2009) note the potential benefits of direct feedback for learners who do not possess the necessary linguistic or metalinguistic knowledge needed to adequately benefit from more indirect methods. It should be noted, however, that learners who do possess the linguistic tools needed to make their own corrections (but perhaps only need a prompt or clue) might benefit greater from more indirect methods of CF and may not appreciate or profit from the teacher providing the correct form every time a mistake is made. Although more research is needed into the benefits and drawbacks of both direct and indirect corrective feedback, the aforementioned studies note the interesting role each has in the L2 writing process. Recent research by Sheen (2007) has shown positive evidence that the use of focused3 strategies to giving feedback might better help decrease error by lessening the cognitive overload that may come from unfocused CF. Sheen’s (2007) study investigating the effects of focused CF on L2 writing notes that direct CF may be preferable for improving certain grammatical features. Sheen explored the effects of ‘focused direct’ and ‘focused direct with metalinguistic explanation’ corrective feedback for adult intermediate ESL learners. Although both groups outperformed the control group, Sheen concluded, based on delayed post-tests, that the metalinguistic group had significant longer term gains in the successful acquisition of the targeted feature. The finding that indirect feedback may not be best for reducing student error is not limited to Chandler’s study. Robb et al’s (1986) investigation into WCF concluded that out of the four types studied, no one group emerged as a dominant force. All four of Robb et al.’s feedback                                                              3 Focused corrective feedback is the process of giving feedback on a limited number of errors, typically one or two, although this line has not been clearly defined in the literature.     60    groups, one direct, one metalinguistic, and two indirect, demonstrated statistically equal results for effecting a decrease of written error. Robb et al.’s study has been much cited in various ways, from the conclusions that this was evidence that indirect CF was no more (or less) effective than direct CF (Chandler, 2003; Ferris & Roberts, 2001)  to Truscott (1996) who used this as evidence to claim “grammar correction’s futility” (p. 331). Other studies have also found that the use of metalinguistic explanations or clues have been effective in decreasing error. Bitchener (2008) provides evidence that greater levels of metalinguistic feedback are most effective in reducing error, and that corrective feedback (in various forms) outperforms no corrective feedback. Other studies (Robb et al., 1986; Lalande, 1982) found that metalinguistic feedback was no less effective than other forms of WCF. Ferris’ (2006) findings, that metalinguistic CF was more effective in reducing certain errors than others, is valuable and could relate to her previous discussion on “treatable” and “untreatable” errors in ESL writing (Ferris, 1999) and the argument that certain error types may be more amendable to certain feedback types. Sheen’s (2007) findings mirror Ferris (2006) in that a combination of metalinguistic feedback with other forms (in Sheen’s case, direct WCF) produced the greatest results in encouraging positive grammatical growth in ESL writing. Sheen’s study also corroborates others by suggesting that feedback versus no feedback is preferable in decreasing error and that all forms of CF are more effective than the use of no feedback. Sheen’s study differs slightly from others by using focused feedback on only one language feature (English articles). She suggests that focused approaches to feedback may allow learners to better concentrate on problematic issues without being weighed down by too much information and too many errors to process. Despite Sheen’s conclusions, there has been limited research into the role that focused CF has on     61    writing specifically with respect to degrees of focus and the point where feedback might become less effective.  3.2.4 The role of grammar In certain ways, the focus on grammar in SLA may be seen as somewhat passé. The “social turn” in second language acquisition has meant language teaching and learning are being investigated and understood in wider and deeper contexts than merely surface level problems. This turn has been understandably embraced by applied linguists with good reason. And yet, it is clear that these surface level errors that L2 students contend with are still major sources of personal stress and embarrassment (Leki, 2006). It is also still a very prevalent reality that these sorts of errors in university level writing can have a negative impact on academic performance (Celce-Murcia, 1991) and that language problems are a major contributing factor to the failure of international students (Andrade, 2006). As L2 students transition from learning a language in academic settings to using the language in mainstream classes, the goals and expectations change drastically. Language is no longer the pure focus in the classroom, but is instead a medium to convey the subject content. And yet, ironically, advanced language proficiency may be most important at this stage. Because it is not a language specific class, the expectations of proficiency are assumed, and there is less support and perhaps less latitude for surface level errors. Grammatical errors in university students’ writing are judged harshly and at great expense to the student and should not be neglected in either classroom instruction or corrective feedback choices.          62    3.2.5 Targeted features Although a variety of linguistic features are important to the success of L2 students’ academic writing, the choice to include some and exclude others when using focused CF is important when targeting problematic areas. A variety of studies have identified the most commonly occurring errors in English L2 writing: articles, lexical items, relative pronouns, conjunctions, possessives, prepositions, singular/plural, subject-verb agreement, verb tense, and word form (Santos, 1988; Sheorey & Ward, 1984; Tomiyama, 1980; Vann and Meyer, 1984). And while content has been reported as being of prime importance in university level writing (Santos, 1988), grammatical accuracy is reported as having significant consequences for students’ writing, particularly as it pertains to assessment and grades. Celce-Murcia (1991) reported that 7.2 grammatical errors per every hundred words constituted a failing mark at the post-secondary level. This may seem contradictory to findings that emphasize the importance of content over language in academic writing, but this appears to fit well with the observation that “all errors are equally irritating” (Albrechtsen, Henriksen, & Faerch, 1980, p. 395) and “ESL errors in students texts are costly in terms of grades and overall evaluations of work quality” (Hinkel, 2004, p. 24). Although it is established that learners are unique in the specific ways they acquire various language features, it appears that certain features are uniformly more difficult to acquire than others, and are therefore less amendable to certain kinds of corrective feedback. Because certain features may be more or less responsive to corrective feedback, it is important for CF to be developmentally appropriate and targeted to certain types of errors (Ferris, 1999, 2002; Ferris & Hedgcock, 1998; Ferris and Roberts, 2001). Ferris (1999) categorizes common ESL errors into two categories: treatable and untreatable. Treatable errors are generally those that are more rule-     63    governed (such as subject-verb agreement, verb tense, and articles) and thus, she asserts, are more susceptible to various forms of CF. These factors should be considered when teachers are deciding not only how to give feedback on writing, but also what to give feedback on. The overlap between Ferris’ treatable errors and those errors found to be most frequent and problematic in L2 writing led to my decision to focus on articles and subject-verb agreement in this present study. The third error, which I have labeled “lexical category”,4 was included based on the high rate of occurrence in the students’ first writing task.  3.2.6 Students’ views on corrective feedback Although research into the impact of errors on L2 students’ overall writing proficiency and grades is crucial, it is of equal importance to consider students’ attitudes towards the feedback process. There is strong evidence to suggest that the majority of second language students appreciate and expect corrective feedback on their writing (Diab, 2005; Enginarlar, 1993; Ferris, 1995; Hedgcock & Lefkowitz, 1994; Radecki & Swales, 1988; Zhang, 1995). Further, the absence of corrective feedback has been described as being a “very bad” method of responding to error according to college level ESL students (Leki, 1991).  Leki (1991) and Radecki and Swales (1988) both reported that ESL students prefer receiving more thorough corrective feedback on grammar over content. Hedgcock & Lefkowitz (1994) found that significant differences may exist between various L2 student populations. In their study of 137 foreign language (FL) students and 110 ESL students, the FL students demonstrated stronger preferences towards surface level corrective feedback whereas the ESL students (enrolled in both language and content courses) preferred feedback on content (although grammar was close behind). Others, however, suggest that the expectations of most second language learners are quite clear:                                                              4 Also called word class or parts of speech. In English, there are eight major lexical categories: nouns, verbs, adjectives, adverbs, pronouns, conjunctions, prepositions and interjections.     64    that receiving CF on their surface level errors is paramount (Leki, 1991; Radecki & Swales, 1988; Schulz, 1996, 2001). Fittingly so, this preference matches well with second language teachers’ feedback practices which tend to concentrate on form over content in student writing (Kassen,1995) . Leki (1991) also notes that feedback on content over grammar may lead students to misunderstand the importance of grammar development in writing. This research into the preferences and perceptions of students tells us that regardless of the type of feedback used it seems clear that: (1) feedback over no feedback is vastly preferred and, (2) feedback on surface level (grammatical) errors is both appreciated and expected.    3.3 Method 3.3.1     Design There were three primary steps to this research. The first was to gauge whether the CF treatment had an impact on the students’ grammatical accuracy in their academic writing. This impact can be further classified into two sections: (a) whether the grammatical accuracy of the students improved over the course of the treatment, and (b) whether the degree of focus in the CF had an effect on the grammatical accuracy of students’ writing. This step was measured using error counts to calculate the percentage of correct to incorrect targeted features over the complete treatment of three essays. The second step was the use of questionnaires to gain some feedback from the students about their feelings and opinions on the use of CF in both general terms (as it related to their lives in the macro sense) and specific terms (as it related to this study). There were two sets of questionnaires given: preceding the three treatments of CF and following the three treatments of CF. The post-study questionnaire was primarily given to gain a sense whether the students’ attitudes about CF had changed due to their involvement in this study. The third and final step was the use of semi-structured interviews. These interviews allowed for further     65    information to be gained in greater detail as well as to add to the information collected during the interviews. The interviews and the questionnaires provided two sources of data to help configure any possible findings. To account for any potential improvements in the students’ grammatical accuracy, particularly as it related to their academic writing, it was necessary to include a control group. By including a control in the study, I was better able to account for any across time and space discrepancies, and was better able to pinpoint any consequences that could be interpreted as a result of the treatment. Members of the control group were spread out evenly across the three participating classes.  3.3.2 Context This research took place in the fall of 2009 in the English language school of a large Canadian university. This on-campus language school serves an adult population with mid- beginner to upper-advanced English proficiency, and is in many cases the destination for students looking to transition to mainstream university undergraduate and graduate programs. The various levels and relatively large student population made this setting appealing for my research. Of the seven major levels at this school, this research concentrated on the fourth (mid- to high-intermediate) as it was the stage in the students’ learning when they were beginning to write in more formal academic registers while still having difficulty with their grammatical accuracy. This combination allowed me to study the effects of CF on the grammatical accuracy of various targeted features in their academic writing. Mandatory for all of the students at this level was a course on academic writing. This course took place four times per week, Monday through Thursday, with a duration of an hour and forty minutes per class.     66    The participating students produced three in-class academic compositions over a three week period from mid October to early November, 2009. For each of the three tasks, the students were asked to write approximately 200 words on topics selected by the three participating teachers (see Appendix A for writing prompts). They were given roughly one hour to plan and compose the writing tasks. The teachers were asked to not give specific grammatical help during the writing nor to give any sort of detailed oral or written feedback, with the exception of regular classroom procedural issues. The students were not allowed to use any external resources (i.e. dictionaries, computers, or textbooks) in the composing process. Each of the writing tasks were then given the appropriate “tier” of feedback (depending on their assigned group) and were returned to the students within three days of composing. Typically, the in-class writing took place on Monday and the compositions with feedback were returned by the following Wednesday. This allowed ample time for the students to observe the feedback before the next in- class task would occur.  3.3.3 Participants The participants were selected based on the following criteria: (1) their willingness to participate in the study, (2) their enrollment in the course, Writing 420 and (3) their level of English (intermediate to high intermediate). Preceding the direct recruitment of participants for this study, I first needed to get the cooperation of the classroom teachers. The teachers were contacted by the school’s Program Manager via email to outline the study and to ask for the teachers’ cooperation. Four teachers replied, and three eventually agreed to participate. The fourth teacher preferred if her entire class could join the study but since these conditions could not be guaranteed she then decided to withdraw unless her participation was urgently needed. For the three teachers who agreed to participate, I made individual contact with them during the     67    first week of October and we coordinated times for me to come and speak with their classes about the project, explain the three stages of the research, and hand out the informed consent forms (included in appendix F). The students were asked to consider participating in the study and were then asked to return the consent forms to their teachers and the teachers then returned the forms to the researcher. By October 9th all of the consent forms had been returned and 41 students out of the 48 agreed to take part in the study, with 39 eventually completing it. Of the 39 total participants included in this study, 20 were female and 19 male. The students attended the same intermediate-level writing class (Writing 420) four days a week from 12:10 pm to 1:50 pm as part of their program requirements. Forty-one students initially consented to participate and were divided into four groups. Group one (N = 11) received focused metalinguistic CF on English articles. Group two (N = 10) received focused metalinguistic CF on articles and lexical categories. Group three (N = 11) received focused metalinguistic CF on articles, lexical categories, and subject-verb agreement. And the fourth group (N = 10) served as a control and received no feedback of any kind on their compositions. Of the 41 students who initially agreed to participate, 39 completed the study. The remaining two were absent either one or two of the days during the in-class writing and were thus excluded from the final results. Although there was a mixture of cultural backgrounds represented in this study, the majority of participants, 74%, were Chinese (n = 29). Japanese learners accounted for 13% (n = 5), Saudi Arabians for 5% (n = 2), Koreans for 5% (n = 2) and a Tunisian for 3% (n = 1). All of the students reported having completed high school in their home countries with the majority (83%) noting additional post-secondary education of some kind. Of this majority, 80% of the participants had received undergraduate or graduate degrees before coming to this school to further their English study.     68    The teachers at this school were experienced English L2 teachers. Although they were responsible for initially cooperating with this study and giving the students class time to write the essays and later to look at the feedback, that is where their involvement ended. I did not ask them to focus on any specific content that might have had an impact on the study, nor did I want them to. With the exception of the feedback, all other effects on their grammatical accuracy were the result of normal progression and normal teacher variation. With the inclusion of a control group, this normal progression was accounted for in this study. All three of the participating teachers had masters’ degrees in TESOL or related disciplines and were instrumental in the successful completion of this research by allowing me access to the classes and: (a) time at the beginning of the study to discuss the research with the students (and request their consent) and; (b) time to administer the questionnaires at the end. Out of the three classes (and three teachers), the 39 total participants were formed into four groups to study the effects of various tiers of corrective feedback, ranging from more to less focused, on the grammatical accuracy of the targeted language features.  3.3.4 Operationalizations Focused metalinguistic feedback is operationalized as the act of indicating a single (or limited) grammatical error and providing a metalinguistic clue in the form of an error code above where the error occurred.                                                      LC5 Example: Every day, she accident kick the ball into neighbors yard.                                                                5 Lexical Category     69    Tiered corrective feedback is operationalized as the various stages of focused corrective feedback beginning with the concentration on one grammatical feature and proceeding upwards in increments of one.  Examples:                                              LC 1. Every day, she accident kick the ball into neighbors yard.                                   LC     SV6 2. Every day, she accident kick the ball into neighbors yard.                                                LC      SV                   Art7 3.  Every day, she accident kick the ball into neighbors yard.   3.4 Intervention 3.4.1 Metalinguistic feedback There is evidence to suggest that the explicitness of feedback may not be a crucial step to the grammatical improvement of student writing. Robb et al. (1986) and Ferris and Roberts (2001) suggest that less explicit corrective feedback is just as useful as more explicit feedback in terms of surface level errors. If the same result can therefore be achieved with greater utility, providing detailed metalinguistic feedback may not be justified when time spent is weighed against the writing improvements. Specifically, Ferris and Roberts demonstrated that the use of error codes in the form of simple abbreviations were successful in achieving improvements in the targeted linguistic features, and were equal to other, more detailed and explicit, types of feedback. For these reasons, I chose to use a less explicit coded form of metalinguistic feedback in this study. When an error was found, and depending on the student’s assigned group (group one, two, or three), the error was underlined and an abbreviated code was written above: Articles                                                              6 Subject-verb agreement 7 Article      70    (A); Lexical Category (LC) and; Subject-Verb agreement (SV). The three treatment groups varied in terms of the feedback given on their academic writing. The first group received metalinguistic feedback on articles, the second group received feedback on articles and lexical categories, and the third group received feedback on articles, lexical categories and subject-verb agreement. A fourth group, the control group, was included to account for the regular change that might have occurred in these targeted features due to factors external to the treatment. The control group received no form of corrective feedback during the course of this study.  3.4.2 Writing tasks The writing tasks selected for the students were chosen by the participating teachers due to two main reasons: first, the teachers were most familiar with the students’ language level and felt the topics would be well suited for the writers and; second, the writing tasks were chosen so as to maintain the general tone of the classroom. Since the students at this stage of the Writing 420 classes were not yet composing full length academic essays, the tasks chosen by the teachers seemed appropriate to the students’ abilities. The students in this particular course were familiar with having in-class writing assignments of a similar variety so neither the request to write for an hour in class nor the task itself (since they were created by the teachers) were considered onerous or out of the ordinary. The three writing tasks were:  1) If you had to choose between a job you love that only pays $35,000 per year and a job you hate that pays $60,000, which would you choose and why?       2) Think of a holiday in your culture. How do people usually celebrate that holiday?       3) Young couples today are waiting until they are older to marry for the first time. Why?       71    3.5     Data As previously mentioned, grammatical problems still play a very large part in how ESL students’ writing is perceived and graded. Finding efficient ways to target these errors is important, as is deciding which errors to focus on when giving corrective feedback. This is especially the case when giving focused corrective feedback and when faced with the limitation of choosing what to include and what to leave out. My decision to focus on articles, lexical categories, and subject-verb agreement was born out of the frequency of errors in the students’ first composition and reinforced by research outlining the commonality and seriousness of certain ESL errors on students’ grades and writing as well as the treatable nature of these rule- governed errors. Articles and subject-verb agreement have been shown to be problematic grammatical features (Santos, 1988; Tomiyama, 1980; Vann and Meyer, 1984; Sheorey and Ward, 1984).  Lexical category was an error type I decided to focus on during the process of giving feedback on the first composition due to its high occurrence rate. A more detailed explanation of the features is outlined in Table 3.5.              72    Table 3.5  Description of Error Types Targeted features  Description Articles  Including both definite and indefinite articles. An error was counted in the following cases (a) misuse of an article (using a definite article instead of an indefinite article or vice versa); (b) omission of an article.  Examples: a) It is apple. X             It is an apple. b) Alice kicked a ball. A ball was red. X             Alice kicked a ball. The ball was red.  Lexical Category  A lexical category error occurred when the correct root of the lexical item was used but not the correct grammatical category. If, for example, a noun was used in place of an adjective, this was categorized as a lexical category error.  Example: a) I am not passion about the job. X b) I am not passionate about the job.  Participle, gerund, or infinitive errors were not included in this category due to the somewhat ambiguous grammatical classification of these words.  Subject Verb Agreement  A subject-verb error occurred because of subject verb disagreement.  Example:  a) If he discover the problem then he will fix it. X b) If he discovers the problem then he will fix it.   On each of the compositions, the corrective feedback was given directly above where the error occurred (see Appendix G for a sample of the feedback used in this study). If the error was one of omission (in the case of articles), the CF was given above where the item was missing. To explain the feedback criteria and to give an explanation of the error codes, each student was given a handout explaining the grammatical rules and corresponding error code (see figure 3.1).      73    Figure 3.1 Sample Grammatical Codes Handout  The following is a list of the grammatical codes that may or may not appear on your writing assignments. Also, following the codes is a brief description and example of the grammar I will be focusing on for this study.  1) Art = Articles Articles are generally used with common nouns to help show if the noun is general or specific. There are two types of articles in English:  a) Definite articles – the b) Indefinite articles – a, an  • If the noun is specific, then we generally use a definite article – the • If the noun is more general, then we use an indefinite article – a, an  For example: I saw an apple on the ground. The apple was red.  2) SV = Subject-Verb Agreement In English, the verb of a clause must always agree with the subject. For example:  a) She eat an apple. × b) She eats an apple. ☺ Sentence A is not correct because the subject (she), which is singular, requires a singular verb (eats).  3) LC = Lexical Category This is also called word class or parts of speech. In English, there are eight major lexical categories: nouns, verbs, adjectives, adverbs, pronouns, conjunctions, prepositions and interjections.  However, some consider articles to be a separate category as well.  Often times for ESL users, it is easy to mix-up or confuse the lexical category of a word. For example:  a) I am happily. × b) I am happy. ☺ For sentence A, the lexical category of happily is an adverb. However, it should be an adjective (like in sentence B). So sentence A is an example of a word used incorrectly due to its lexical category.       74    3.5.1 Questionnaires There were eight major themes which guided research questions four and five investigating students’ perceptions towards corrective feedback (see appendix C for a table showing the themes and corresponding questions). These themes are: (1) The role of grammar; (2) The role of feedback; (3) Learner history of feedback; (4) General feedback preferences; (5) Specific feedback preferences (what specific types of corrective feedback do the students prefer); (6) Attentiveness towards corrective feedback; (7) Perceptions of efficacy of corrective feedback and; (8) Constructivism. These themes were chosen with the intention of investigating students’ knowledge concerning not only general feedback practices, opinions, and preferences but also as a way for the students to provide feedback on the feedback; or in other words, a way for the students to voice their opinions about the specific CF used in this study. Of the 39 participants in the corrective feedback portion of this study, 37 completed the questionnaires.  3.5.2 Interviews A thematic analysis was performed to interpret the qualitative data gained from the semi- structured interviews in this study. According to Braun and Clarke, “Thematic analysis is a method for identifying, analyzing and reporting patterns (themes) within data. It minimally organizes and describes your data set in (rich) detail” (p. 79). The authors further note that thematic analysis differs from other qualitative analysis (such as grounded theory or thematic discourse analysis) in that it is not “wedded to any pre-existing theoretical framework, and therefore it can be used within different theoretical frameworks...” (p. 81). The following discussion outlines the steps used in the thematic analysis for this research as well as some more general comments regarding the interview process and subsequent analysis of the data.     75    After gaining consent, 11 participants were randomly selected to be interviewed. The pre- and post-study interviews were conducted sporadically over the course of eight days at the research site. The interviews were conducted in a private room and recorded onto a digital recorder. Data were then transcribed onto my computer and entered into a word processor for purposes of comparison and categorization. According to the eight themes which guided the creation of both the interview and questionnaire items, the participants’ responses were analyzed looking for a relationship between their answers and the appropriate theme. This analysis was done by noting key words or phrases which arose in the answers and by the general tone of the response. In short, the data set (the specific data selected to be analyzed depending on the appropriate theme) consisted of all occasions across the larger data corpus which had relevance to the topic being discussed. In some instances, two sources of data (from both the interviews and questionnaires) were combined to form one data set for purpose of analysis. A few factors should be noted about the analysis of the qualitative data in this study. First, in general terms, the interviews were quite short which meant the answers seldom deviated from the theme being targeted in the question. Although there was some fluctuation across the participants, each interview lasted approximately 10 minutes and most students gave very straightforward answers of limited length. Second, the English proficiency of the students most likely had an impact on the level of nuance in their responses and how much they could discuss or expand on a topic. As a result, the responses were generally quite concise and contained little complexity or ambiguity. Third, because the interviews were semi-structured and the interview items were created based on eight previously constructed themes, this tended to have a major influence on the themes which arose from the questions. Or in other words, there seemed to be little deviation in the students’ answers which could not be linked to the targeted theme(s) which     76    the interview items were based on. If, for example, the students were asked a question related to the fifth theme8 of this research, then the students’ responses were almost uniformly dealing with that theme. In turn, I interpreted no new themes as a result of the interviews aside from the ones which were already created. Ely et al. (1997) make an interesting observation regarding the inescapable subjectivity that all qualitative researchers possess and the concept of “emerging themes”. [This concept] can be misinterpreted to mean that themes ‘reside’ in the data, and if we just look hard enough they will ‘emerge’ like Venus on the half shell. If themes ‘reside’ anywhere, they reside in our heads from our thinking about our data and creating links as we understand them. (pp. 205-206) Although I recognize that any attempt to claim with complete accuracy the participants’ true feelings or intentions is problematic, I felt confident that this was kept to a minimum in this research and that the students’ responses were categorized and analyzed accurately.  3.6 Analysis 3.6.1 The effects of tiered CF on grammatical accuracy After word and error counts were completed for each composition, data were first entered into Microsoft Excel and then transferred to SPSS where a range of inferential and descriptive statistics was performed. Error rates (ER) were calculated by the number of errors occurring per every hundred words. Student’s t-tests for paired samples were used to measure the difference in mean error rates of each targeted feature between the first and third compositions. A repeated measures design was used to calculate the between-subject effects of the CF treatment to determine the result of feedback versus no feedback on the different dependent                                                              8  Specific Feedback Preferences      77    variables being investigated in this study (articles, lexical categories, and subject verb agreement). A repeated measures design allows for a smaller number of subjects and was therefore optimal to use in this exploratory study. One of the strengths of this design, and the reason it works well with a small population, is that it reduces problems caused by individual differences (Gravetter & Wallnau, 2008).  3.6.2 Corrective feedback and the learner To answer research questions four and five9, questionnaires (see appendix D) and semi- structured interviews (see appendix E) were used. The questionnaires consisted of 36 items. The interview protocol originally consisted of 10 items but various items were either expanded or omitted depending on the flow of the conversation.     Questionnaires The coded data from the questionnaires were first entered into Microsoft Excel and then transferred into SPSS. The majority of the 36 items in the questionnaire were assigned numerical values. The purpose of the first five items was to determine when the questionnaires were administered and each participant’s ethnicity, age, sex, and level of education. The remaining items were 5-point Likert scales ranging from ‘strongly disagree’ to ‘strongly agree’. These five possible responses were assigned numerical values ranging from ‘strongly disagree,’ one (1) to ‘strongly agree,’ five (5). All of the 5 point scales had a neutral mid-point with a value of three. For purposes of analysis, the filled-out questionnaires were collapsed into three categories: (1) ‘strongly disagree’ and ‘disagree’; (2) ‘neutral’ and; (3) ‘strongly agree’ and ‘agree’. Item 36 was an open-ended item which provided opportunity for students to comment and was                                                              9 4)  What do NNS expect from their instructors as far as feedback on their academic writing is concerned?   5)  How do students perceive different types of corrective feedback?       78    disregarded from analysis due to the very low response rate. Descriptive statistics were calculated to find the frequencies for each of the questionnaire items.     Interviews Two sets of interviews took place preceding and following the CF treatment. The purpose of having two sets of interviews was to help determine whether the particular feedback used in this study impacted the participants’ perceptions and attitudes towards CF. The pre-study interviews took place over the course of four days during the first week of the study, ending the day the first compositions were returned to the students with the assigned corrective feedback. The post-study interviews also spanned four days beginning the day the final compositions were returned. Both the pre- and post-study interview protocols contained 10 items. Depending on the interviewee and the nature of our conversation, certain points were expanded on to continue a lively or informative discussion, or excluded to prevent redundancy and stagnancy in the conversation. Each interview took place at the school where the research was performed and averaged approximately 10 minutes. The sessions were recorded with a digital recorder and the data were transferred onto my computer for transcription and storage purposes. Following this, the interviews were transcribed using Express Scribe, a digital transcription software. After transcription, the interviews were examined using a thematic analysis according to the same set of themes that the questionnaires were based on to provide answers for research questions four and five.         79    3.7 Results 3.7.1 Findings based on Student’s t-tests Paired sample two-tailed t-tests were performed on the three variables for all four groups between the first and third compositions. The t-tests revealed that the corrective feedback resulted in a decrease of mean errors and the reduction of error was statistically significant. For the first group which received feedback on articles, the feedback treatment resulted in a significant decrease of mean articles error rate (t (9) = 2.18, p = .05) from 1.99 errors per hundred words (SD = .86) to .95 per hundred words (SD = .97) (see Figure 3.2). For group two which received feedback on articles and lexical categories, the two-tailed t- tests revealed a significant decrease of mean articles error rate (t (8) = 4.16, p = .00) , from 1.80 (SD = .68) to 0.82 (SD = 0.69)  and mean lexical categories error rate  (t (8) = 3.24, p = .01) from 1.54 (SD = 1.00) to 0.45 (SD = 0.37) (see Figure 3.3). For group three, who received feedback on all three features, the results were mixed. Two of the language features, lexical categories and subject-verb agreement, demonstrated a statistically significant reduction in mean error rates while the third, articles, did not. The decrease of lexical categories mean error rate was significant, t (10) = 5.33, p = .00, with a mean error rate decreasing from 1.46 (SD= .61) to .37 (SD = .57). The decrease of subject-verb agreement mean error rate was also significant, t (10) = 7.20, p = .00, with a mean error rate reduction of 1.30 (SD= .44) to .41 (SD = .52). However, the statistical analysis for articles revealed that the decrease in articles mean error rate was not significant, t (10) = 1.37, p = .20, with a reduction of 1.64/100 words (SD= 1.06) to 0.91/100 words (SD = 1.14) (see Figure 3.4). For the final group, the control group which received no form of written corrective feedback, all features showed varying amounts of increase of mean error rates from composition     80    one to composition three instead of improvement, as is evidenced by the negative t-values. Group four did not demonstrate a reduction of error with articles (t (8) = -1.32, p = .22), lexical categories (t (8) = -.49, p = .64), or subject-verb agreement (t (8) = -1.12, p = .29) (see Figure 3.5).  Figure 3.2: The Mean Error Rate of Articles for Group One       Figure 3.3: The Mean Error Rate of Articles and Lexical Categories for Group Two       81     Figure 3.4: The Mean Error Rate of Articles, Lexical Categories and Subject-Verb Agreement for Group Three      Figure 3.5: The Mean Error Rate of Articles, Lexical Categories and Subject-Verb Agreement for Group Four            82    3.7.2     Repeated measures design In order to address the first research question,10 a repeated measures design was used. The goal of these analyses was to compare the effect of feedback versus no feedback on the different dependent variables, article error rate, lexical category error rate, and subject-verb agreement error rate. The repeated measures design revealed that the feedback had a significant effect in the reduction of targeted error(s) and that there were significant differences between the groups that received feedback and the group that received no feedback on each of these types of errors. The effect of the CF treatment on articles compared with the no feedback control group revealed a statistically significant difference (F (1, 3) = 2.98, p = .044). There was a significant effect of the feedback on lexical categories compared to no feedback (F (1, 2) = 4.44, p = .022). Finally, there was also a significant effect of the feedback on subject-verb agreement compared to no feedback (F (1, 1) = 15.08, p = .00). These findings suggest that in the context of this study the use of corrective feedback was successful in reducing error compared to no feedback.   3.7.3 Findings based on questionnaires and interviews     Questionnaires In general terms, the results of the questionnaires fit well with past research showing that students appreciate and expect corrective feedback on their written work (see appendix C for a percentage distribution of collapsed positive and negative items). Students also appear to have a solid understanding of the importance of grammatical accuracy in the academy. 100% of the students agreed that it is important to have good grammar and 92% agreed that having good grammar is important to academic success. However, fewer appear to believe that grammar is intrinsically linked with academic success. When asked if teachers give higher grades when there                                                              10 What is the effect of focused metalinguistic corrective feedback on the accurate use of targeted linguistic features?     83    are no grammatical errors on their academic writing 61% agreed. This perception may be due the somewhat sheltered existence that ESL students live in terms of language use and expectations. Since language courses are primarily focused on language use (as opposed to content), there may be different expectations and allowances made for error in these classes. Most students (92%) agreed that paying attention to corrective feedback helps improve grammar, although only 72% stated they always do so. This may be related to the prevalence of negative feedback as the main CF treatment on students’ writing and a variety of reasons why students choose to follow or ignore the feedback. When asked whether corrective feedback makes students feel bad about their writing, almost half responded it does (42%) with only 14% stating it does not. The use of negative feedback by teachers on students writing is a perpetual dilemma. It is difficult to say what impact this has on students writing or their attention towards corrective feedback, but perhaps a mixture of negative and positive CF can act as a balance in the larger process of giving and receiving WCF. With respect to specific feedback preferences, most students expressed displeasure with feedback methods that provide only the correct answer with no explanation as to what sort of error occurred. 72% stated they did not like when teachers fix their mistakes with no explanation. Correspondingly, 86% expressed preference when the teacher provides an explanation when fixing their errors. These figures are further corroborated with the mixed responses of students towards the role of constructivist teaching and learning approaches in the classroom. When asked if they prefer when teachers tell them the answers, over half (56%) were either neutral or disagreed. A small majority (53%) felt that self-discovery is better for their learning and a larger 66% preferred when teachers provide guidance but allowed them to discover the answers themselves.     84    Not surprisingly, the vast majority expressed a strong preference towards unfocused CF, greatly favouring to receive feedback on all mistakes instead of none or only one or two. 88% of respondents reported preferring CF on all mistakes while no students reported preferring no feedback on their writing. Only 26% stated that receiving feedback on only one or two types of error was preferable. In general, the overwhelming trend was a preference for direct metalinguistic feedback (86%) in very unfocused ways (88%).     Interviews The interviews provided students an avenue to expand on some of the topics touched on in the questionnaire, and to discuss others not included. For the most part, the results of the interviews served to further confirm those opinions outlined in the questionnaire. One of the widespread opinions to come out of the interviews was the feeling that if a teacher returns a composition with no feedback, then the students assume the teacher does not care, did not pay proper attention to it, or simply did not read it. This is also corroborated in the questionnaires with 94% of respondents disagreeing that they prefer to receive no feedback on their writing. When asked about how they feel if they get an essay back and there is feedback on every mistake, there were mixed responses. One student, Anne11 stated, “maybe I will think my writing is terrible so... so many mistakes have to be changed so I will lose my confidence.” Others also expressed feelings of disappointment (mostly in themselves) when seeing feedback scattered across the page, and another stated that, “I think it’s a little waste of time.” Another had a more optimistic approach while still being realistic. “It's to be honest, it’s difficult to understand everything. But I think it's useful. But to the other hand correct many mistakes is not to help me                                                              11 Names of the participants have been changed for the purpose of anonymity.     85    to remember.” There seems to be a clear message that these students are expressing: extremes in CF practices are neither satisfying nor encouraging. Too much can negatively impact a learners’ confidence and thus lessen his or her motivation, and not enough can make the learner frustrated and confused.  3.8     Discussion As demonstrated by the t-tests and repeated measures analyses, the metalinguistic feedback used in this study appears to have had a positive effect on the targeted grammatical features. Interesting to note are the results of the t-tests in group three. Here, the second and third error categories (lexical categories and subject verb agreement) showed statistically significant reductions in error (p = .00 for both), while the first error group (articles) was not significant as demonstrated by the p-value (p = .20). When students received feedback on one or two types of errors, their number of errors in these categories decreased, However, when they received feedback on three types of errors, the number of their errors decreased significantly on only two of the three types, possibly suggesting they had too much to attend to simultaneously and the effectiveness of the CF lessened. This could provide evidence in support of the argument that reducing the level of focus while giving corrective feedback could result in a decrease of overall efficacy for the CF treatment. Although the results of this study suggest that focused feedback on one or two language features was most effective, the expectations of the students run contradictory to these findings. The students overwhelmingly (88%) reported preferring unfocused feedback (i.e. feedback on every mistake) instead of the limited guidance they received. This contradiction is not surprising despite its perhaps contradictory nature. Adult learners, especially those whom are highly motivated and whom possess the desire to improve quickly, may feel the time needed for focused     86    feedback (to eventually address all the errors in their writing) to be too long and arduous. By receiving feedback on everything (or most things), students might feel more in charge of their own learning and more aware of their own limitations and growth. It is also important to note that unfocused CF, although showing less substantive gains in the short term, may be equally or more effective in the long term. These sorts of questions have not been fully investigated in the literature and may in fact be difficult to accurately assess. It is safe to say, however, that most forms of corrective feedback (including the kind used in this study) appear to be helpful, both in linguistic terms as evidenced by the decrease of error, and in affective terms as students expect and value the feedback they receive. The interviews and questionnaires revealed a general tendency that students want and appreciate corrective feedback. These results are not surprising and fit well with other research that has had similar results (Diab, 2005; Enginarlar, 1993; Ferris, 1995; Radecki & Swales, 1988; Zhang, 1995). The strong preferences of students to receive feedback on all errors were also not a surprising result given the demographic of many of these students. Since the research site is a place of transition for many international students hoping to gain entrance in mainstream university programs, most likely their motivation to succeed is high and the consequences of failure (or underperformance) are great. Regardless of the long process that learning a language requires, many of these students saw this program as a means rather than the end; they had limited time to improve and to do so they viewed focused feedback as too restricted in what it included and what it left out. Somewhat contradictory was the message of students preferring to receive feedback on all mistakes (88%) while admitting that such amounts of feedback on their writing is also a source of stress and disillusionment. Many students reported in the interviews that getting feedback on     87    every error (not limited to grammar) is either not necessary or makes them feel discouraged. However, some did state that despite unfocused CF making them feel “bad” or “disappointed”, it may be more helpful in the end as it serves to illuminate a more comprehensive look into their error frequency.  3.9     Conclusion Most problematic for proponents of focused feedback will be the challenge to get students to invest or believe in the system. Despite research that suggests that focused CF may be more efficient in effecting positive growth in grammar error reduction (Han, 2002; Nicholas, Lightbown & Spada, 2001; Sheen, 2007), the majority of students asked in this study expressed a strong preference towards unfocused feedback over focused feedback. These findings are similar to Leki (1991) who showed students in her study strongly preferred receiving feedback on all errors, both major and minor. If students do not prefer or are not amendable to a focused feedback approach, this could have implications in how they view the teacher and the efficacy of this feedback approach. If students resent the feedback used (or the teacher giving it) this could influence how useful classroom instruction and CF are, regardless of the reported efficacy in reducing error. Both the paired sample t-tests and the repeated measures design provide evidence that the metalinguistic focused feedback used in this study was effective in reducing error and that feedback versus no feedback was preferable in terms of reducing error. All three treatment groups decreased error over the course of the first to the third compositions while the control group showed a minimal increase of errors in all three grammatical categories. The evidence further suggests that the third group, which received CF on all three language features, demonstrated a decreased impact of feedback efficacy. This suggests that feedback on a greater     88    amount of language features may hamper students’ ability to process the feedback and thus lessen its effectiveness. With the small sample size of this exploratory study, it is difficult to make any sort of viable and statistically relevant claims. However, the results do present compelling data in suggesting that as feedback becomes less focused, it also becomes less effective. Further study investigating the effects of focused and unfocused feedback could provide teachers and learners with additional evidence to better refine feedback practices and increase the efficiency of corrective feedback.      89    Chapter 4: Discussion  4.1     Discussion This final chapter will present some concluding remarks regarding the research findings of this study. The following will also discuss the research process and will outline some of the limitations and pedagogical implications that arose in the course of my research and the analysis of the data. The central purpose of this research was to add to the existing evidence investigating the impact of written corrective feedback on the grammatical accuracy of ESL students’ writing. More recent data with a specific L2 focus has shown that CF appears to steadily outperform no CF in terms of improving grammar on both revisions and new pieces of writing (Chandler, 2003; Ferris & Roberts, 2001; Sheen, 2007). The results of this study agree with past research affirming the efficacy of WCF. The following section will refer back to the five research questions which guided this study and will outline how they were addressed in this research. The quantitative findings addressed the first three research questions and dealt with the various effects of the CF treatment on the research participants’ grammatical accuracy in their academic compositions. Q1: What is the effect of focused metalinguistic corrective feedback on the accurate use of the targeted linguistic features? Q2: At what point does focused metalinguistic corrective feedback become unfocused? Q3: Is there a difference in the effect of varying intensities of written corrective feedback (on one, two, or three specified linguistic features) on students’ accuracy in the production of those features?   90    For research question one, the three treatment groups showed a decrease in mean error rates from the first to third compositions. The control group received no form of corrective feedback and demonstrated a slight increase in error frequency. These findings fit well with past studies investigating similar variables, particularly Sheen (2007) who studied the effects of focused CF on articles and the impact of treatment on error reduction and Bitchener (2008) and Chandler (2003) who investigated the use of metalinguistic feedback, including long term effects on reducing error. The second research question is perhaps the most difficult of the three to answer. As Sheen (2007) notes, focused CF may better enable writers to concentrate on problematic features and improve quicker with more sustained effects. It therefore makes sense that there is a threshold between focused and unfocused feedback and that knowing where this line exists could be a useful tool for teachers wishing to improve feedback efficacy. In the context of this study, this threshold, and thus the lessening of efficacy, may begin as the feedback focus reaches approximately three features. For the third group, which received feedback on articles, lexical categories, and subject-verb agreement, although all three features did show a decline in mean error rate, a decrease of efficiency on one feature was evident. A paired sample t-test revealed that the reduction of error in articles was not significant (p = .20). These findings imply that an increase in the number of features for which feedback is given could result in a lessening of effectiveness. Although more research is needed comparing the differences of focused and unfocused feedback, this lessening of efficacy could be a sign that the focused/unfocused line may exist between two and three features. Related to the previous question, the third research question was also based on the results outlined in the paired sample t-tests. The inferential statistics imply that the more linguistic  91    features on which feedback is given, the less effective the treatment is. For group three, the group which received CF on all three grammatical features, only two of the categories showed a statistically significant decrease in error while the third one did not. This may be due to the fact that feedback on three or more types of errors requires a higher degree of cognitive processing. Most likely, this is also closely related with the proficiency level of the students (intermediate English learners) and their developing ability to deal with certain types of error as they continue to progress. Further, there is likely no exact point that exists between effective and ineffective feedback since there is a perpetually moving interaction between students’ proficiency levels and the appropriate types (and number) of feedback that they should receive. For research questions four and five12, the results of the questionnaires and interviews reveal that corrective feedback is a well established tradition that students expect to receive and believe in its value. In particular, most participants in this study were clear in reporting that the more feedback they received, the better they appreciated it. At times, however, these findings were contradictory as many participants also reported feeling disappointed or discouraged by receiving too much feedback on their writing. Still, this contradiction is not very surprising and in many ways is to be expected. Most students, especially at this level, are active learners with their own personal goals, some of them of significant consequence to them and their families’ lives. Engaged students want to improve and thus want to be aware of their mistakes. However, there is a fine line between being aware and being discouraged, especially if the errors (and feedback on the errors) are numerous and potentially overwhelming. This potential discouragement could be countered by balancing the negative feedback students receive with positive feedback. By receiving both kinds of feedback students might feel encouraged by things                                                              12 Q4: What do NNS expect from their instructors as far as feedback on their academic writing is concerned?       Q5: How do students perceive different types of corrective feedback?    92    they have done correctly and therefore, in the big picture, feel less distraught by the potentially large amounts of error correction or negative corrective feedback. Many of the students in the questionnaires reported preferring either direct CF with an explanation (86%) or indirect metalinguistic CF (69%). The similarity of metalinguistic preference indicates that adult learners with presumably high language learning motivation want grammatical knowledge and understand its usefulness in achieving growth. These findings may also be indicative of the language level of the students since indirect-only CF might require knowledge that the learners do not yet possess.  4.2     Limitations In general, I was satisfied with the way this study unfolded and the findings that resulted. Despite these mostly positive results, there are a number of limitations which must be clearly noted. Most noticeably, the small sample size of this research presents a major hindrance regarding any claims that can be made. Because of the different qualitative and quantitative layers of this research and the limitations of both time and resources, I decided that a small study would be most feasible to accomplish. Particularly, with respect to the quantitative findings, having only nine or 10 people per treatment group means the statistical relevance of these results is quite hindered. Despite using a repeated measures design which is typically used in this sort of environment (a low number of participants), future research using larger groups would be beneficial. A potential limitation towards using focused CF is the overwhelming expectation from students to receive more comprehensive feedback on their errors. For intermediate level students, it is still common to see a wide variety and high frequency of grammatical problems in their  93    writing. Getting simultaneous feedback on all of these errors may be overwhelming and discouraging for the writer. However, the students involved in this research did not appear to like or feel satisfied with the focused feedback used. At present, there is little research discussing the discrepancy between the possible efficacy of focused CF and the negative impact on students’ affect as they feel “ripped off” or cheated because of the limited amount of feedback on their writing. Particularly for adult learners who have specific learning goals and targets, they may not be satisfied with focused CF strategies regardless of the potential results. If a structured system were implemented where the focus rotates between various problematic features and where the students are willing participants, then perhaps any dilemma between efficacy and negative perception could be more adequately addressed. Another limitation of this study is the low number of errors that generally occurred in the students’ compositions. This was primarily related to the proficiency of the participants and the length of the compositions, although the genre of the writing prompt might also have played a significant role in guiding the students down a certain grammatical path. On average, the students in this study made a limited number of mistakes on their writing, and the compositions were relatively short in length. This meant that for some students there was little room for growth since the error rates of the targeted features were at or below one mistake per one hundred words. Also, because of limitations of time and the high variability of individual differences with respect to writing speed, the composition length differed greatly. Although some students could write between 200-300 words quite easily over the course of the allotted writing time, others struggled to achieve 100 (or in some cases much less). This low amount of total words meant a low amount of occurring instances of the targeted grammatical features. With little use of the targeted grammar, there were few errors and thus less opportunity for growth. If a  94    lower level of students were used in this sort of study, and given more time to compose, this could provide more opportunities to observe and treat error in a similar context. Changing the writing prompt could potentially achieve the same effect as lengthening the compositions. Having prompts that elicited more difficult responses or which led the students’ writing down different grammatical paths may have allowed more instances of the targeted errors to occur and therefore more opportunities to decrease error in subsequent compositions. I would also conduct a small test study prior to the actual research to make sure I was familiar with the population. I began this study with a set of three grammatical features to investigate (subject-verb agreement, articles, and verb tense) but needed to make last minute adjustments due to the very low number of verb tense problems in the students’ first writing task. I changed my feedback focus to “lexical categories” because of the high rate of error in the first writing task. Had I had the opportunity to look at the some of the students writing before I began, I could have made adjustments earlier. As it was, I found little evidence in the outside literature to reconcile my “lexical categories” with other reports of a high rate of error frequency in ESL/EFL writing. Because the intervention and data collection was on a very tight schedule, I had very little time to give the needed feedback and properly analyze the essays before they needed to be returned. This last minute change of focus was very taxing and led to some late nights and early mornings. Finally, another limitation of this study is the relatively short period, just over three weeks, in which it was done. Undoubtedly, a longitudinal study could have provided evidence whether the feedback was effective in reducing error in the longer term or whether it failed to enact such a substantial, more permanent change. A future analysis of the participants’ writing or a (more) delayed post-test could have provided some further information into this issue and would have  95    enriched this study’s claims. Because of various time limitations it was not possible to have a longitudinal element, but a future study of this kind would benefit greatly from having the ability to gain insight into the CF’s long-term impact.  4.3     Pedagogical implications Corroborating past research, this study found that CF appears to be effective in reducing grammatical errors in ESL students’ writing. There is also evidence that more focused approaches may better enable students to decrease errors on the targeted feature(s) on subsequent pieces of writing. It was clear that the students all had strong opinions regarding their feedback preferences, and generally they had similar things to say. The majority expressed a strong desire to receive feedback on all their mistakes (not just grammar) and preferred to have some sort of metalinguistic explanation instead of just a direct correction or no explanation. These strong opinions led me to question the method of my own feedback practices as well as the kind used in this study. My decision to use different tiers of metalinguistic CF was partially based on my prior practices as a teacher and partially based on information gained from the relevant literature. In retrospect, it would have been interesting to have incorporated the participating students’ views on CF into the actual feedback they received. I believe this has the potential to help better interest and focus students to pay attention to the feedback and to help them believe it is working. In spite of the limitations of this study, the metalinguistic CF used in this research was successful in decreasing error. This provides further evidence in support of the argument to use feedback on L2 students’ academic writing. In particular, it has demonstrated that metalinguistic CF may begin to become less effective as the focus lessens. Knowing this could provide teachers  96    with alternative options to help students decrease grammatical errors in a relatively short period of time. Although students may not feel completely satisfied with receiving feedback on a limited number of errors over long periods of time, a focused approach could be used in certain situations to meet the total learning needs of students. Over the course of a semester (or longer), teachers could rotate focus between various problematic areas which would eventually cover more or all of the needed grammatical features. More specifically, if the focused/unfocused threshold is greater than one (or two) features, this could accelerate the process and hopefully make students more accepting of this CF approach. Finally, there is likely no uniform magical number that will fit every student’s needs. Rather, this line is in constant flux due to changing language proficiency and is dependent on what types of language features are being concentrated on at a certain point in time. The findings in this research are not meant to be definitive, but are only meant to add to the focused/unfocused discussion of written corrective feedback in hopes that this might better aid teachers and students in feedback use.  4.4     Suggestions for further research As Ellis (2009) has stated, there are a lack of empirical studies comparing the differences between focused and unfocused feedback. Sheen (2007) also notes that focused CF may be more effective than unfocused CF for learners who do not possess the linguistic or metalinguistic knowledge needed to process a large amount of errors. A potential problem that advocates of focused feedback must reconcile will be the derision of students who want more comprehensive feedback on their errors. While this is both reasonable and understandable, if more can be learned regarding the potential benefits of focused feedback, CF practices could be refined to increase effectiveness and improve grammatical accuracy.  97    Regarding whether less-focused feedback results in a decrease of efficacy, future research using a larger sample population could provide further evidence investigating if a threshold exists and where, approximately, this may be. This cut-line will never be completely stable and will differ between individuals. However, knowing that a line may exist and that this line divides levels of effectiveness could help writing teachers decide what type or types of CF to give in what frequency and help students learn more efficiently. If future research can provide evidence that unfocused CF is much less effective than focused CF in reducing error, this could have a strong impact on how feedback is given in ESL classrooms. Future longitudinal studies that include a greater number of participants could offer further illumination into the long-term effects of focused metalinguistic feedback CF. A greater number of participants would increase the validity of the quantitative findings and offer greater insight into the true effect that this sort of tiered metalinguistic CF has on a reduction of error. A study which has the ability to use more tiers of feedback (more groups) could better gauge the effect of various levels of feedback and the impact that more to less focused feedback has on the reduction of error. The decision to use three treatment groups and one control group in this current study was primarily due to a limitation of resources. Initially, I had planned to include four treatment groups (with groups one to four receiving CF on one to four grammatical features) but this was reduced to three due to suggestions from my thesis committee regarding the (im)practicality of my initial plan. A study which used four to five treatment groups could better investigate if more feedback groups truly decreases CF’s effectiveness, in both the short and long-term. Other research that incorporates the feedback desires of students before deciding which types of feedback to give could also be enlightening. The socio-affective domains of students  98    play a significant role in language learning, and a study which included students’ feedback preferences would provide an interesting perspective into the impact of student-driven feedback on error reduction. To date, there have been no studies I am aware of that have implicated students as directly in the feedback process as this. Many studies have inquired into students’ perceptions of feedback and their seemingly overwhelming response that they appreciate, expect and believe in its efficacy, but no studies I am aware of have asked students before giving feedback and used the results in deciding precisely which kind of CF to give. Having the students as direct stakeholders could have a significant impact on how much they buy into the system and how much effect the feedback has on their writing.  4.5     Summary The first purpose of this study was to determine the effects of tiered metalinguistic feedback on the grammatical accuracy of English L2 students’ academic writing. Four groups were formed, each receiving either a different tier of WCF or, with the fourth group, receiving no feedback to serve as a control for the study. The participants wrote three separate compositions roughly averaging approximately 200 words. These compositions were then assessed for errors, depending on the students’ assigned groups, and were returned with the appropriate metalinguistic feedback. The following week this process was again repeated until the three compositions were completed. Two-tailed Student’s t-tests for paired samples were conducted to determine what impact the corrective feedback had on the targeted features from the first to the final composition. The results indicate that the corrective feedback was effective in decreasing the errors and that this decrease was statistically significant. Only one language feature in group three, articles, did not  99    display a significant reduction, although a slight decrease in error did occur. This may provide evidence that feedback that is less focused could result in a lessening of feedback effectiveness. A repeated measures design revealed significant between groups effects of the CF treatment of the feedback versus no feedback groups. All three language features showed a statistically significant decrease in error over the course of the treatment as compared to the control group. The repeated measures design provides evidence that the WCF used in this study was effective in reducing errors in the targeted grammatical features. The second purpose of this study was to investigate students’ feelings and perceptions of the feedback process. The qualitative findings revealed similar results to previous studies which demonstrate that students value written corrective feedback and believe that it is useful to help them become better writers. The findings also outlined the students’ general dissatisfaction with focused feedback and their strong preference to receive CF on most or all of their written errors. Despite the potential usefulness of focused CF approaches, teachers will need to be cognizant of these preferences and make appropriate accommodations to combine best practices with the affective demands of the students. As aforementioned, one way to do this is to include students more directly in the feedback process by asking their opinions and incorporating these opinions into the CF process. Despite the various limitations of this exploratory study, the findings can hopefully provide some additional support in favour of written corrective feedback. Although the current scholarship has provided some rich sources of qualitative and quantitative data, there is still much to be done. Giving and receiving feedback is a highly subjective part of the writing process and students and teachers will benefit from continued perseverance to provide additional clues to improve feedback practices. Focused metalinguistic feedback is just one part of a much larger  100    and more complicated process, but perhaps it can provide students with more efficient ways to decrease grammatical errors and improve their writing.  101    References Albrechtsen, D., Henriksen, B., & Faerch, C. (1980). Native speaker reactions to learners’ spoken interlanguage. Language Learning, 30(2), 365-396. Ali, M. S. (1995). The major quantitative findings of a study of the English language- based study skills problems of two groups of foreign students at an American university. ERIC doc. 384 028, p. 26. Allen, P., Swain, M., Harley, B., & Cummins, J. (1990). Aspects of classroom treatment: Toward a more comprehensive view of second language education. In B. Harley et al., (Eds.), The development of second language proficiency (pp. 57-81). Cambridge: Cambridge University Press. Andrade, M. S. (2006). International student persistence: Integration or cultural integrity? Journal of College Student Retention: Research Theory & Practice, 8(1), 57-81. Bitchener, J., Young, S., & Cameron, D. (2005). The effect of different types of corrective feedback on ESL student writing. Journal of Second Language Writing, 9, 227–258. Bitchener, J. (2008). Evidence in support of written corrective feedback. Journal of Second Language Writing, 17(1), 102-118. Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77-101. Carson, J. (2001). Second language writing and second language acquisition. In T. Silva and P.Matsuda (Eds.), On second language writing (pp. 191-200). Mahwah, NJ: Lawrence Erlbaum Associates. Celce-Murcia, M. (1991). Grammar pedagogy in second and foreign language teaching. TESOL Quarterly, 25, 459-480.  102    Chandler, J. (2003). The efficacy of various kinds of error feedback for improvement in the accuracy and fluency of L2 student writing. Journal of Second Language Writing, 2, 267–296. Chomsky, N. (1975). Reflections on language. New York: Pantheon. Cobb, T. (2005). Foundations of linguistics - Approaches and concepts: Constructivism, applied linguistics, and language education. Retrieved September 05, 2009, from Encyclopedia of Language and Linguistics, 2nd. ed.: http://www.lextutor.ca/cv/constructivism_entry.htm DeKeyser, R. M., & Juffs, A. (2005). Cognitive considerations in L2 learning. In E. Hinkel (Ed.), Handbook of Research in Second Language Teaching and Learning (pp. 437-454). Mahwah, NJ: Lawence Erlbaum Associates. Diab, R. L. (2005). Teachers' and students' beliefs about responding to ESL writing: A case study. TESL Canada Journal, 23(1), 28–42. Doughty, C., & Williams, J. (1998). Issues and terminology. In: Doughty, C. and Williams, J. (Eds.), Focus on form in classroom second language acquisition. Cambridge: Cambridge University Press, 1-11. Ellis, R., Sheen, Y., Murakami, M., & Takashima, H. (2008). The effects of focused and unfocused written corrective feedback in an English as a foreign language context. System, 36, 353–371. Ellis, R. (1994). The study of second language acquisition. Oxford: Oxford University Press. Ellis, R. (2001). Introduction: Investigating form-focused instruction. Language Learning, 51, 1– 46. Ellis, R. (2009). A typology of written corrective feedback types. ELT J, 63(2), 97-107.  103    Ely, M., Vinz, R., Downing, M., & Anzul, M. (1997). On writing qualitative research: Living by words. London: The Falmer Press. Enginarlar, H. (1993). Student response to teacher feedback in EFL writing. System, 21(2), 193– 204. Fazio, L. (2001). The effect of corrections and commentaries on the journal writing accuracy of minority- and majority language students. Journal of Second Language Writing, 10(4), 235–249. Felix, U. (2005). E-learning pedagogy in the third millennium: The need for combining social and cognitive constructivist approaches. ReCALL, 17(1), 85-100. Ferris, D. (1995). Student reactions to teacher response in multiple-draft composition classrooms. TESOL Quarterly, 29, 33-53. Ferris, D. (1999). The case for grammar correction in L2 writing classes: A response to Truscott (1996). Journal of Second Language Writing, 8, 1–10. Ferris, D. (2002). Treatment of Error in Second Language Student Writing. Ann Arbor, MI: University of Michigan Press. Ferris, D. (2004). The "grammar correction" debate in L2 writing: Where are we, and where do we go from here? (and what do we do in the meantime...?). Journal of Second Language Writing, 13(1), 49-62. Ferris, D. (2010). Second language writing research and written corrective feedback in SLA: Intersections and practical applications. Studies in Second Language Acquisition, 32, 181– 201 . Ferris, D., & Hedgcock, J. (1998). Teaching ESL composition: Purpose, process, and practice. Mahwah, NJ: Lawrence Erlbaum.  104    Ferris, D., & Roberts, B. (2001). Error feedback in L2 writing classes: How explicit does it need to be? Journal of Second Language Writing, 10(3), 161–184. Firth, A., & Wagner, J. (1997). On discourse, communication, and (some) fundamental concepts in SLA research. Modern Language Journal, 81, 285-300. Fotos, S. (1994). Integrating grammar instruction and communicative language use through grammar consciousness-raising tasks. TESOL Quarterly, 28(2), 323–351. Frodesen, J. (2001). Grammar in writing. In M. Celce-Murcia (Ed.). Teaching English as a Second or Foreign Language. London: Heinle & Heinle: 233–248. Fromkin, V., Rodman, R., Hyams, N., & Hummel, K. (2006). An introduction to language: Third Canadian Edition. Toronto: Nelson. Gaskell, D., & Cobb, T. (2004). Can learners use concordance feedback for writing errors? System, 32, 301-319. Gravetter, F. J., & Wallnau, L. B. (2008). Essentials of Statistics for the Behavioral Sciences (6th Ed.). Belmont, CA: Thomson/Wadsworth. Guenette, D. (2007). Is feedback pedagogically correct? Research design issues in studies of feedback on writing. Journal of Second Language Writing, 16, 40-53. Han, Z. (2002). A study of the impact of recasts on tense consistency in L2 output. TESOL Quarterly, 36(4), 543-572. Hargreaves, L. (2001). Dropout rate among Chinese physics PhD students seems high; community considers why. Physics Today, 54(2), 4. Hedgcock, J., & Lefkowitz, N. (1994). Feedback on feedback: Assessing learner receptivity in second language writing. Journal of Second Language Writing, 3, 141-163.  105    Hillocks, G., Jr. (1986). Research on written composition: New directions for teaching. Urbana, IL: ERIC Clearinghouse on Reading and Communication SkillsiNational Conference on Research in English. Hinkel, E. (2004a). Tense, aspect, and the passive voice in L1 and L2 academic texts.  Language Teaching Research, 8(1), 5-29. Hinkel, E. (2004b). Teaching academic ESL writing: Practical techniques in vocabulary and grammar. Mahwah, NJ: Lawrence Erlbaum Associates. Hinkel, E. (2006). Current perspectives on teaching the four skills. TESOL Quarterly, 40(1), 109-131. Hyland, K., & Hyland, F. (2006). State of the art article: Feedback on second language students' writing. Language Teaching, 39(2), 83-101. Hymes, D. H. (1971). On communicative competence. In J. Pride and J. Holmes (Eds.), Sociolinguistics. Penguin, 1972. (Excerpt from the paper published 1971, Philadelphia, University of Pennsylvania Press.) Johns, A. M. (1995). Teaching classroom and authentic genres: Initiating students into academic cultures and discourses. In D. Belcher & G. Briane (Eds.), Academic writing in a second language: Essays on research and pedagogy (pp. 277-293). Norwood, NJ: Ablex. Johns, A. M. (1981). Necessary English: A faculty survey. TESOL Quarterly, 15, 51-57. Johnson, P. (1988). English language proficiency and academic performance of undergraduate international students. TESOL Quarterly, 22, 164-168. Kassen, M. A. (1995). Responding to foreign-language student writing: Expanding our options. In M. Haggstrom, L. Morgan, & J. Wieczorek (Eds.), The Foreign Language Classroom: Bridging Theory and Practice (pp. 99-117). New York: Garland.  106    Kaufman, D. (2004). Constructivist issues in language learning and teaching. Annual Review of Applied Linguistics, 24, 303-319. Knoblauch, C. H., & Brannon, L. (1981). Teacher commentary on student writing: The state of the art. Freshman English News, 10(2), 1-4. Krashen, S. D. (1981). Second language acquisition and second language learning. New York: Pergamon. Krashen, S. D. (1982). Principles and practice in second language acquisition. New York: Pergamon. Krashen, S.D. (1985). The input hypothesis: issues and implications. New York: Longman. Krashen, S. D., & Terrell, T. (1983). The natural approach: language acquisition in the classroom. New York: Pergamon. Lalande, J. F. (1982). Reducing composition errors: an experiment. Modern Language Journal, 66, 140–149. Lantolf, J. P. (1994). Sociocultural theory and second language learning: Introduction to the special issue. The Modern Language Journal, 78, 418-420. Lantolf, J. P. (2000). Introducing sociocultural theory. In J. P. Lantolf, (Ed.), Sociocultural theory and second language learning (pp. 1-26). Oxford: Oxford University Press. Lantolf, J. P. (2006). Sociocultural theory and second language learning: State of the art. Studies in Second Language Acquisition, 28, 67-109. Lantolf, J. P., & Thorne, S. L. (2006a). Sociocultural theory and the genesis of second language development. Oxford, UK: Oxford University Press.  107    Lantolf, J. P., & Thorne, S. L. (2006b). Sociocultural theory and second language acquisition. In B. van Patten, & J. Williams (Eds.), Theories in second language acquisition (pp. 201- 224). Mahwah, NJ: Erlbaum. Lee, H. K., & Anderson, C. J. (2006). Validity and topic generality of a writing performance test. Language Testing, 24(3), 307-330. Leki, I. (1991). The preferences of ESL students for error correction in college-level writing classes. Foreign Language Annals, 24, 203-218. Leki, I. (2006). "You cannot ignore": L2 graduate students' response to discipline-based written feedback. In K. Hyland & F. Hyland (Eds.), Feedback in Second Language Writing: Contexts and Issues. Cambridge: CUP, pp. 266-285. Leki, I., & Carson, J. (1997). Completely different worlds: EAP and the writing experiences of ESL students in university courses. TESOL Quarterly, 31, 39-70. Light, R., Xu, M., & Mossop, J. (1987). English proficiency and academic performance of international students. TESOL Quarterly, 21(2), 251-255. Lightbown, P. M., & Spada, N. (2006). How languages are learned (3rd ed.). Oxford: Oxford University Press. Long, M. H. (1988). Instructed interlanguage development. In L. M. Beebe (Ed.), Issues in second language acquisition: Multiple perspectives (pp. 115-41). Cambridge, MA: Newbury House/Harper and Row. Long, M. H. (1990). The least a second language acquisition theory needs to explain. TESOL Quarterly, 24(4), 649-666.  108    Long, M. H. (1991). Focus on form: A design feature in language teaching methodology. In de Bot, K., Ginsberg, R. B., & Kramsch, C. (Eds.), Foreign language research in cross- cultural perspective (pp. 39-52). Amsterdam: John Benjamins. Long, M., & Robinson, P. (1998). In C. Doughty & J. Williams. Focus on Form in Classroom Second Language Acquisition (pp. 15-41). Cambridge: Cambridge University Press. Matsuda, P. K. (2003). Process and post-process: A discursive history. Journal of Second Language Writing, 12(1), 65-83. McGirt, J. D. (1984). The effect of morphological and syntactic errors on the holistic scores of native and non-native compositions. Unpublished master's thesis, University of California, Los Angeles. Milton, J. (2006). Resource-rich web-based feedback: Helping learners become independent writers.  In K. Hyland and F. Hyland (Eds.), Feedback in second language writing: Contexts and issues. Cambridge: Cambridge University Press. Nagata, N. (1993). Intelligent computer feedback for second language instruction. The Modern Language Journal, 77(3), 330–339. Nagata, N. (1997). The effectiveness of computer-assisted metalinguistic instruction: A case study in Japanese. Foreign Language Annals, 30(2),187–200. Norris, J. M., & Ortega, L. (2000). Effectiveness of L2 instruction: A research synthesis and quantitative meta-analysis. Language Learning, 50, 417-528. Nicholas, H., Lightbown, P., & Spada, N. (2001). Recasts as feedback to language learners. Language Learning, 51, 719–758. Piaget, J. (1950). The psychology of intelligence. New York: Routledge. Piaget, J. (1953). The origin of intelligence in the child. London: Routledge & Kegan Paul.  109    Polio, C., & Glew, M. (1995). ESL writing assessment prompts: How students choose. Journal of Second Language Writing, 5(1), 35-49. Radecki, P., & Swales, J. (1988). ESL student reaction to written comments on their written work. System, 16, 355-365. Reagan, T. (1999). Constructivist epistemology and second/foreign language pedagogy. Foreign Language Annals, 32, 413-425. Richards, J. C., & Rodgers, T.S. (2001). Approaches and methods in language teaching (2nd ed.). Cambridge: Cambridge University Press. Robb, T., Ross, S., & Shortreed, I. (1986). Salience of feedback on error and its effect on EFL writing quality. TESOL Quarterly, 20, 83–93. Rosenfeld, M., Leung, S., & Oltman, P. (2001). The reading, writing, speaking, and listening tasks important for academic success at the undergraduate and graduate levels. TOEFL Monograph Series 21. Princeton, NJ: Educational Testing Service. Sachs, R., & Polio, C. ( 2007). Learners’ use of two types of written feedback on an L2 writing task. Studies in Second Language Acquisition, 29, 67–100. Santos, T. (1988). Professors' reactions to the academic writing of nonnative-speaking students. TESOL Quarterly, 22, 69-90. Schmidt, R. W. (1990). The role of consciousness in second language learning. Applied Linguistics, 11, 129–158. Schmidt, R. W. (2001). Attention. In P. Robinson (Ed.), Cognition and Second Language Instruction (pp. 3-32). Cambridge: Cambridge University Press. Semke, H. D. (1984). Effects of the red pen. Foreign Language Annuls, 17, 195-202.  110    Sheen, Y. (2007). The effect of focused written corrective feedback and language aptitude on ESL learners' acquisition of articles. TESOL Quarterly, 41(2), 255-283. Silva, T. (1990). Second language composition instruction: Developments, issues, and directions in ESL. In B. Kroll (Ed.), Second language writing: Research insights for the classroom (pp. 11–23). Cambridge: Cambridge University Press. Silva, T. (1993). Toward an understanding of the distinct nature of L2 writing: The ESL research and its implications. TESOL Quarterly, 27, 657–677. Silva, T., & Brice, C. (2004). Research in teaching writing. Annual Review of Applied Linguistics, 24, 70–106. Simina, V., & Hamel, M. J. (2005). CASLA through a social constructivist perspective: WebQuest in project-driven language learning. ReCALL, 17(2), 217–228. Sinor, J., & Huston, M. (2004). The role of ethnography in the post-process writing classroom. Teaching English in the two-year college, 31(4), 369-382. Snow Andrade, M. (2006). International students in English-speaking universities: Adjustment factors. Journal of Research in International Education, 5(1), 131–154. Tashakkori, A., & Teddlie, C. (2003). Handbook of mixed methods in social &. behavioral research. Thousand Oaks: Sage. Trimbur, J. (1994). Taking the social turn: Teaching writing post-process. College Composition and Communication, 45, 108–118. Truscott, J. (1996). The case against grammar correction in L2 writing classes. Language Learning, 46, 327– 369. Truscott, J. (1999). The case for ‘‘the case for grammar correction in L2 writing classes’’: A response to Ferris. Journal of Second Language Writing, 8, 111–122.  111    Truscott, J. (2007). The effect of error correction on learners’ ability to write accurately. Journal   of Second Language Writing, 16, 255–272.  Truscott, J. (2009). Arguments and appearances: A response to Chandler. Journal of Second Language Writing, 18(1), 59-60. Truscott, J., & Hsu, A. Y.-p. (2008). Error correction, revision, and learning. Journal of Second   Language Writing, 17(4), 292-305.  U.S. Department of Commerce, Bureau of the Census. (1995). Dropout rates in the United States. Washington, DC: Author Vann, R.J., Meyer, D.E., & Lorenz, F.O. (1984). Error gravity: A study of faculty opinion of ESL errors. TESOL Quarterly, 18(3), 427-440. VanPatten, B., & Williams, J. (Eds.). (2007). Theories in second language acquisition: An introduction. Mahwah, NJ: Erlbaum. Vygotsky, L. S. (1978). Mind in society. Cambridge, MA: Harvard University Press. Vygotsky, L. S. (1986). Language and thought. Cambridge, MA: MIT Press.  White, L. (1987). Against comprehensible input: The input hypothesis and the development of second language competence. Applied Linguistics, 8, 95-110 Widdowson, H. G. (1988). Grammar, and nonsense, and learning. In W. Rutherford  & M. Sharwood Smith (Eds). Grammar and second language teaching: A book of readings (pp. 146-155). New York: Longman. Williams, J., & Evans, J. (1998). What kind of focus and on which forms? In C. Doughty & J. Williams (Eds.), Focus on form in classroom second language acquisition (pp.139-155). Cambridge: Cambridge University Press. Wood, D., Bruner, J. S., & Ross, G. (1976). The role of tutoring and problem solving. Journal of Child Psychology and Psychiatry, 17, 89-100.  112    Yeh, S. W., &  Lo, J. J. (2009). Using online annotations to support error correction and corrective feedback. Computers and Education: An International Journal, 52(2), 882- 892. Yip, V. (1994). Grammatical consciousness-raising and learnability. In T. Odlin (Ed.) Perspectives on Pedagogical Grammar. Cambridge: Cambridge University Press. Zhang, S. (1995). Re-examining the affective advantages of peer feedback in the ESL writing class. Journal of Second Language Writing, 4(3), 209–222. Zuengler, J., & Miller, E. (2006). Cognitive and sociocultural perspectives: Two parallel SLA worlds? TESOL Quarterly, 40(1), 35–58.                 113    Appendix A: Writing Prompts for Three In-class Compositions    420W Timed Writing #1 Name: ________________________  420W, Fall 2009 Teacher: ______________________  English Language Institute   Instructions:  Develop a paragraph (about 200 words) on the following topic:  If you had to choose between a job you love that only pays $35,000 per year and a job you hate that pays $60,000, which would you choose and why?  Total Time:  50 minutes  Brainstorming:  Brainstorm ideas and examples       Outline Topic Sentence: ______________________________________________________________________________ __________________________________________________________________________  1.   2.   3.  114     420W Timed Writing #2 Name: _______________________  420W, Fall 2009 Teacher: _____________________  English Language Institute   Instructions:  Develop a paragraph (about 200 words) on the following topic:  Think of a holiday in your country.  How do people usually celebrate that holiday?  Total Time:  50 minutes  Brainstorming:  Brainstorm ideas and examples       Outline Topic Sentence: ______________________________________________________________________________ __________________________________________________________________________  1.   2.   3.  115      420W Timed Writing #3 Name: ________________________ 420W, Fall 2009 Teacher: ______________________  English Language Institute  Instructions:  Develop a paragraph (about 200 words) on the following topic:  Young couples today are waiting until they are older to marry for the first time. Why?   Total Time:  50 minutes Brainstorming:  Brainstorm ideas and examples      Outline Topic Sentence: ______________________________________________________________________________ __________________________________________________________________________  1.   2.   3.   116    Appendix B: Form of Grammatical Codes  Grammatical Codes  The following is a list of the grammatical codes that may or may not appear on your writing assignments. Also, following the codes is a brief description and example of the grammar I will be focusing on for this study.  4) Art = Articles Articles are generally used with common nouns to help show if the noun is general or specific. There are two types of articles in English:  c) Definite articles – the d) Indefinite articles – a, an  • If the noun is specific, then we generally use a definite article – the • If the noun is more general, then we use an indefinite article – a, an  For example: I saw an apple on the ground. The apple was red.  5) SV = Subject-Verb Agreement In English, the verb of a clause must always agree with the subject. For example:  c) She eat an apple. × d) She eats an apple. ☺ Sentence A is not correct because the subject (she), which is singular, requires a singular verb (eats).  6) LC = Lexical Category This is also called word class or parts of speech. In English, there are eight major lexical categories: nouns, verbs, adjectives, adverbs, pronouns, conjunctions, prepositions and interjections.  However, some consider articles to be a separate category as well.  Often times for ESL users, it is easy to mix-up or confuse the lexical category of a word. For example:  c) I am happily. × d) I am happy. ☺ For sentence A, the lexical category of happily is an adverb. However, it should be an adjective (like in sentence B). So sentence A is an example of a word used incorrectly due to its lexical category.     117    Appendix C: Table of Questionnaire Themes  Percentage of negatively collapsed: % - Neutral: % = Percentage of positively collapsed: % +  Theme # Question % - % = % + The Role of Grammar  6 7 8 10  14 It is important to have good grammar Having good grammar is important for my academic success Having good grammar is very important in academic writing Teachers give higher grades when there are no grammatical errors on my academic writing Classroom instruction is enough to improve my grammar in academic writing X X 100 X 8 92 X 3 97 3  36 61 58 33 9 The Role of Feedback 22 Corrective feedback is a normal part of learning a language X 8 92 Learner History of Feedback 17 25 My writing teachers provide corrective feedback on my writing assignments I fix the mistakes after the teacher gives me corrective feedback on my grammatical errors 3 22 75 3 28 69 General Feedback Preferences 18  19  20 21 I would like my teachers to provide more feedback on my writing assignments I would like my teachers to provide less feedback on my writing assignments I prefer to receive feedback on every mistake I make I prefer to receive no corrective feedback 6 14 80  91 6 3  11 11 88 94 6 X Specific Feedback Preferences (what types of FB do students prefer: direct, indirect, metalinguistic, etc) 27 28 29 30  31 32 I prefer when teachers fix my mistakes without providing an explanation I prefer when teachers fix the mistake and also give an explanation I prefer when teachers give me a clue and let me fix the mistake on my own I prefer when teachers circle or underline the mistake but do not explain what is wrong I prefer when teachers give corrective feedback on all errors I prefer if teachers give feedback on one or two types of error 72 17 11 8 6 86 14 17 69 58 25 17  8 8 84 37 37 26 Attentiveness towards Corrective Feedback 23  24 I always pay attention to the corrective  feedback on my writing assignments Paying attention to feedback helps my grammar 3 25 72  X 6 94 Perceptions of efficacy of feedback  9  11 12  13 15 16 26 Written corrective feedback (any indication to show that an error occurred) helped improve my grammar. Corrective feedback did not help improve my grammar Corrective feedback alone is enough to improve my grammar in my academic writing Corrective feedback along with regular grammar instruction is enough to improve my grammar in academic writing Corrective feedback makes me feel bad about my writing Corrective feedback increases my confidence about my academic writing Corrective feedback is a waste of time X  14 86 87 11 3 50 39 11  8  42 50 77 17 6 14 44 42 94 6 X Constructivism  33  34 35 I prefer when teachers provide help but allow me to discover answers by myself I prefer when teachers tell me the answer I think it is better for my learning to discover answers on my own 6 28 66  28 28 44 8 39 53  118    Appendix D: Student Questionnaire of Written Corrective Feedback   Questionnaire Evaluation of Written Corrective Feedback    This questionnaire will allow the researcher to find out about attitudes and opinions regarding written corrective feedback. It contains 36 items. At the end of the questionnaire, you are encouraged to add any relevant comments or opinions.  Please respond to these questions in a manner that truly reflects your opinions. If a question does not apply to you, indicate “NA” in the margin. Your answers will be kept confidential.  Please note: if you complete and return this questionnaire, you are thereby consenting to the anonymous use of your responses for research about written corrective feedback in an English as a second language context.  Thank you for your participation.   Part I - General Information (please circle or write your answer) Name:__________________________  1. Today’s date: __________________  3.   Ethnic origin: __________________ 2. Date of birth (dd/mm/yyyy): __________  4.   Sex:    Female      Male  5.   What is the highest level of education that you have completed? (please check the highest level you have completed)   1. primary school  2. some secondary school  3. completed high school  4. some additional training (college or technical school)  5. undergraduate university  5. graduate university  6. postgraduate university       119    Part II – Evaluation of Written Corrective Feedback  6. It is important to have good grammar.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  7. Having good grammar is important for my academic success.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  8. Having good grammar is very important in academic writing.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  9. Written corrective feedback (any written indication to show that an error occurred) helped improve my grammar.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  10. Teachers give higher grades when there are no grammatical errors in my academic writing.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  11. Corrective feedback did not help improve my grammar.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  12. Corrective feedback alone is enough to improve the grammar in my academic writing.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree   120    13. Corrective feedback along with regular grammar instruction is enough to improve the grammar in my academic writing.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  14. Classroom instruction alone is enough to improve the grammar in my academic writing.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  15. Corrective feedback makes me feel bad about my writing.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  16. Corrective feedback increases my confidence about my academic writing.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  17. My writing teachers provide corrective feedback on my writing assignments.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  18. I would like my teachers to provide more feedback on my writing assignments.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  19. I would like my teachers to provide less feedback on my writing assignments.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly Agree     121    20. I prefer to receive feedback on every mistake I make.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  21. I prefer to receive no corrective feedback.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  22. Corrective feedback is a normal part of learning a language.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  23. I always pay attention to the corrective feedback on my writing assignments.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  24. Paying attention to feedback helps improve my grammar.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  25. I fix the mistakes after the teacher gives me corrective feedback on my grammatical errors.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  26. Corrective feedback is a waste of time.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree      122    27. I prefer when teachers fix my mistakes without providing an explanation.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  28. I prefer when teachers fix the mistake and also give an explanation.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  29. I prefer when teachers give me a clue and let me fix the mistake on my own.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  30. I prefer when teachers circle or underline the mistake but do not explain what is wrong.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  31. I prefer when teachers give corrective feedback on all errors.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  32. I prefer if teachers give feedback on one or two types of error.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  33. I prefer when teachers provide help but allow me to discover answers by myself.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree      123    34. I prefer when teachers tell me the answers.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree  35. It is better for my learning to discover answers on my own.  1 Strongly disagree 2 disagree 3 Neutral 4 Agree 5 Strongly agree   36. Please add any comments or opinions related to written corrective feedback here:  ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________        124    Appendix E: Student Interview Protocols  Pre-Study Interviews with the Students The Effects of Tiered Corrective Feedback on L2 Academic Writing   Name: ______________________________________   1. What do you know about corrective feedback?  2. What are your feelings about corrective feedback?  3. Is it helpful or important for you to receive feedback on your writing?  4. Do you prefer lots of feedback, or not much?  5. On what parts of your writing do you prefer to receive feedback? For example, on the grammar or content?  6. If you get an essay back where your teacher has commented on every mistake you’ve made, how does that make you feel?  7. If you get an essay back where there is no corrective feedback (only a checkmark or a grade) how does this make you feel?  8. Do you feel corrective feedback plays an important role in learning grammar? And learning how to be a better writer?  9. Do you think it is important to have good grammar to achieve a high level of English? To succeed in an English academic context/situation?  10. How do you prefer to receive corrective feedback? Do you prefer to receive feedback on all errors or only some? Do you like when the teacher corrects your mistakes, or just circles them? Do you like when teachers use an “error code”, or if they give a clue about what’s wrong?           125    Post-Study Interviews with the Students The Effects of Tiered Corrective Feedback on L2 Academic Writing   Name: ______________________________________   1. Has your opinion of corrective feedback changed? Why or why not?  2. Do you think the corrective feedback you received improved your grammar? Why or why not?  3. Do you think the corrective feedback you received improved your academic writing? Why or why not?  4. What did you like or dislike about the corrective feedback you received?  5. Would you like to see more or less feedback? Why?  6. In the future, will you pay more attention to your corrective feedback?  7. What is the most important part of academic writing, the grammar or the content?  8. Do you think grammar has an important role in your academic success in English?  9. What is better for improving your grammar: when the teacher tells you the answer or when you try to discover the answer by yourself?  10. If you were an English teacher what kind of feedback would you give your students on their writing? What kind of feedback do you like best? What kind do you think works the best?    126  Appendix F: Informed Consent Form  Informed Consent Form                   October 5th, 2009  The Effects of Tiered Corrective Feedback on L2 Academic Writing  Investigator: The principal investigator is Dr. Monique Bournot-Trites, Department of Language and Literacy Education, 604-822-4873. This is a research project for the Master’s degree of Timothy Anderson, Department of Language and Literacy Education, 604-822-4873.  Purpose:  The main purposes of this study are: to determine the effects of focused metalinguistic written corrective feedback on the grammatical accuracy in the academic writing of English as a Second Language students; to investigate the differences between focused feedback (feedback on one linguistic feature) and unfocused feedback (feedback on many features); and what the participating students’ feelings and opinions towards corrective feedback are.  Study Procedures:  There are three parts to this research project.  PART I  is a set of pre-study semi-structured interview questions. The purpose of these interviews is to gauge students’ feelings and opinions about the way corrective feedback works and whether they feel it is effective in improving their grammar and their academic writing.  PART II  is the production of three in-class academic essays. These essays will be given a different tier of corrective feedback depending on the groups participants were randomly selected to be in. Group one will receive feedback on one linguistic feature, group two will receive feedback on two linguistic features, group three will receive feedback on three linguistic features, and group four (the control group) will receive no feedback. Error counts will be done on each essay to determine the total use of the targeted linguistic feature(s) and the times the targeted feature(s) were used both correctly and incorrectly. Over the course of three separate writing tasks, we will investigate the effects that the corrective feedback has on the grammatical accuracy of each subsequent academic writing assignment. The three essays will be part of the regular class requirements, and students will not be required to write anything additional for the purposes of this study.  PART III is comprised of a questionnaire and post-study semi-structured interviews to gain further information on the participating students’ feelings and opinions about the way corrective feedback worked and whether they felt it was effective in improving the production of grammar in their academic writing.  Potential Benefits: From the results of this research, we will be able to provide information to teachers and administrators in ESL settings about the effects of the different tiers of written corrective feedback on the grammatical accuracy of ESL students and further contribute to the understanding of students’ thoughts and feelings about the corrective feedback process.   Confidentiality:  All the data collected in all three parts of the study will be kept absolutely confidential. The participants’ names on the documents will be used initially to group a participant’s information with other information collected from that same participant. For the participants involved in all three parts of the study, the names will be replaced with a random    127  capital letter. All audio recorded data will be transcribed before data analysis and only presented as a written transcription, never as raw data.  Participants will not be identified by name in reports of the completed study.  All information which might directly or indirectly reveal a participant’s identity will be deleted or altered and will not be released or published without specific consent to the disclosure from the participant. All data, including all audio recordings, will be stored in a locked file cabinet at UBC for at least five years. The data will be used for a Master’s thesis as partial requirement of an MA degree and will be shared with my supervisor and the two members of my thesis committee, Dr. Ling Shi and Dr. Steven Talmy. Data analysis and research findings may also be presented at an academic conference and may be published as scholarly work.  Contact for concerns: You may refuse participation in this project or withdraw during the project without any consequence to your position as a student. Your participation is entirely voluntary and you may refuse to participate or withdraw from the study at any time, even after signing this consent form. Refusing to participate or withdrawal will not jeopardize your position as a student. If you have any concerns about your treatment or rights as a research participant, please contact Timothy Anderson or Dr. Monique Bournot-Trites at 604-822-4873 or by e-mail at timanderson306@gmail.com and monique.bournot-trites@ubc.ca respectively, or the Research Subject Information Line in the UBC Office of Research Services at 604-822-8598.  I am writing to request your permission to participate in this study entitled: “The Effects of Tiered Corrective Feedback on L2 Academic Writing.” On the next page, you will find the statement of informed consent to be signed by you and returned to me as soon as possible, whether or not you wish to participate in the project. The first copy is for you to keep.  Thank you for your participation in this study.       Timothy Anderson, BA, B.Ed   Monique Bournot-Trites, PhD MA student      Assistant Professor        Director of Modern Language Education     128  Statement of Informed Consent (copy to keep)  Title of the project: “The Effects of Tiered Corrective Feedback on L2 Academic Writing” Researchers: The principal investigator is Dr. Monique Bournot-Trites, Department of Language and Literacy Education, 604-822-4873. This is a research project for the Master’s degree of Timothy Anderson, Department of Language and Literacy Education, 604-822-4873.  Please fill out the information below. Be sure to keep pages 1-3 for your own records and to return a signed copy of page 4 (Statement of Informed Consent) to me as soon as possible. __________________________________________________________________________  I have read and understand the attached letter regarding the project entitled: “The Effects of Tiered Corrective Feedback on L2 Academic Writing.” I understand that, even if I consent to participate in the study, I can opt out of the study at any time. I have kept a copy of the letter describing the project and a copy of the permission form (Statement of Informed Consent).    ‘I consent / I do not consent (circle one) to participation in PART  I  of this study.’  __________________________  ______________________________________________ Printed name of participant    Participant’s Signature                             Date      ‘I consent / I do not consent (circle one) to participation in PART II of this study.’  __________________________  ______________________________________________ Printed name of participant    Participant’s Signature                             Date      ‘I consent / I do not consent (circle one) to participation in PART III of this study.’  __________________________  ______________________________________________ Printed name of participant    Participant’s Signature                             Date     129  Statement of Informed Consent (copy to return to researcher)  Title of the project: “The Effects of Tiered Corrective Feedback on L2 Academic Writing” Researchers: The principal investigator is Dr. Monique Bournot-Trites, Department of Language and Literacy Education, 604-822-4873. This is a research project for the Master’s degree of Timothy Anderson, Department of Language and Literacy Education, 604-822-4873.  Please fill out the information below. Be sure to keep pages 1-3 for your own records and to return a signed copy of page 4 (Statement of Informed Consent) to me as soon as possible. __________________________________________________________________________  I have read and understand the attached letter regarding the project entitled: “The Effects of Tiered Corrective Feedback on L2 Academic Writing”. I understand that, even if I consent to participate in the study, I can opt out of the study at any time. I have kept a copy of the letter describing the project and a copy of the permission form (Statement of Informed Consent).    ‘I consent / I do not consent (circle one) to participation in PART  I  of this study.’  __________________________  ______________________________________________ Printed name of participant    Participant’s Signature                             Date      ‘I consent / I do not consent (circle one) to participation in PART II of this study.’  __________________________  ______________________________________________ Printed name of participant    Participant’s Signature                             Date      ‘I consent / I do not consent (circle one) to participation in PART III of this study.’  __________________________  ______________________________________________ Printed name of participant    Participant’s Signature                             Date             130  Appendix G: Feedback Sample      131  Appendix H: UBC Research Ethics Board Certificate of Approval    The University of British Columbia Office of Research Services Behavioural Research Ethics Board Suite 102, 6190 Agronomy Road, Vancouver, B.C. V6T 1Z3  CERTIFICATE OF APPROVAL - MINIMAL RISK  PRINCIPAL INVESTIGATOR: INSTITUTION / DEPARTMENT: UBC BREB NUMBER: Monique Bournot-Trites  UBC/Education/Language and Literacy Education H09-02220 INSTITUTION(S) WHERE RESEARCH WILL BE CARRIED OUT: Institution Site N/A N/A Other locations where the research will be conducted: University of British Columbia – English Language Institute (UBC - ELI)  CO-INVESTIGATOR(S): Timothy B. Anderson SPONSORING AGENCIES: N/A PROJECT TITLE: The Effects of Tiered Corrective Feedback on L2 Academic Writing CERTIFICATE EXPIRY DATE:  September 10, 2010 DOCUMENTS INCLUDED IN THIS APPROVAL: DATE APPROVED:   September 10, 2009 Document Name Version Date Questionnaire, Questionnaire Cover Letter, Tests: Tiered Feedback Interview Script N/A August 21, 2009 Tiered Feedback Questionnaire N/A August 21, 2009 Letter of Initial Contact: Tiered Feedback Informed Consent 2 September 2, 2009 Other Documents: UBC-ELI Research Approval  N/A August 21, 2009   The application for ethical review and the document(s) listed above have been reviewed and the procedures were found to be acceptable on ethical grounds for research involving human subjects.   Approval is issued on behalf of the Behavioural Research Ethics Board and signed electronically by one of the following: Dr. M. Judith Lynam, Chair Dr. Ken Craig, Chair Dr. Jim Rupert, Associate Chair Dr. Laurie Ford, Associate Chair Dr. Anita Ho, Associate Chair 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items