TEACHERS CORRECTIVE FEEDBACK ON THE PRONUNCIATION OF ENGLISH FRICATIVES AND AFFRICATES BY NON-ENGLISH MAJOR FRESHMEN

Besides, after theexperimental period, a short questionnaire was distributed to the experimental group to explore their opinions about the use of Teacher’s Explicit Corrective Feedback i

It is undeniable that English has become one of the most popular foreignlanguages in Vietnam However, though Vietnamese learners can speak English, notmany of them have intelligible English pronunciation The fact remains that they areunlikely to perceive their pronunciation mistakes when speaking Suchpronunciation mistakes will be fossilized if they are not corrected by the teacher

Of 24 English consonants, the six consonants /s, z, ʃ, ʒ, ʤ, ʧ/ have beenidentified as the most common pronunciation problems that non-English majorfreshmen at the Diplomatic Academy of Vietnam encounter As a result, the presentstudy aims to help them overcome these problems In order to achieve the aim ofthe study, a quasi-experimental design was conducted on 36 non-English majorfreshmen in class KT40B at the Diplomatic Academy of Vietnam Besides, after theexperimental period, a short questionnaire was distributed to the experimental group

to explore their opinions about the use of Teacher’s Explicit Corrective Feedback inEnglish speaking lessons

The results of the study indicate that students’ pronunciation of the sixEnglish consonants improves significantly when they are provided with Teacher’sExplicit Corrective Feedback Furthermore, the use of this corrective feedback typealso receives support from the experimental students Based on the findings, thethesis provides some recommendations for the application of Teacher’s ExplicitCorrective Feedback, which would be of great assistance for university students inEnglish speaking lessons

LISTS OF TABLES

List of Tables

Table 1: English Consonants

Table 2: Similarities and Differences of English Alveolar Fricatives,

Alveo-Palatal Fricatives and AffricatesTable 3: Percentile Rank of the Pre-test Scores of KT40B Students

Table 4: Placements of 34 Subjects in Pairs for Random Assignment

Table 5: Frequency of the Pre-test Scores

Table 6: The Scoring Process in the Experimental Subgroup 1

Table 7: The Scoring Process in the Experimental Subgroup 2

Table 8: Results of the Pre-test and Post-test of Both Groups

Table 9: Comparison of the Pre-test Scores of Two Groups

Table 9a: Paired-samples Statistics (Pair 1)

Table 9b: Paired-samples T-test (Pair 1)

Table 10: Comparison of the Gain Scores Made by Two Groups after the

Experimental PeriodTable 10a: Paired-samples Statistics (Pair 2)

Table 10b: Paired-samples T-test (Pair 2)

Table 11: Comparison of the Post-test Scores of Two Groups

Table 11a: Paired-samples Statistics (Pair 3)

Table 11b: Pair-samples T-test (Pair 3)

Table 12: The Experimental Group’s Opinions about TECF on Their

Pronunciation of the Six Consonants /s, z, ʃ, ʒ, ʤ, ʧ/

LISTS OF FIGURES AND ABBREVIATIONS

Lists of Figures

Figure 1: The Process of Providing Pronunciation-Focused TECF on Students’

OutputFigure 2: Differences in Gain Scores Obtained by Both groups after the

ExperimentFigure 3: The Experimental Group’s Opinion about TECF on Their

Pronunciation of the Six Consonants /s, z, ʃ, ʒ, ʤ, ʧ/

Lists of Abbreviations

DAV: Diplomatic Academy of Vietnam

TCF: Teacher’s Corrective Feedback

TECF: Teacher’s Explicit Corrective Feedback

L2: Second Language

FL: Foreign Language

TABLE OF CONTENTS

ACKNOWLEDGEMENT i

ABSTRACT ii

LISTS OF TABLES iii

LISTS OF FIGURES AND ABBREVIATIONS iv

PART A: INTRODUCTION 1

I Statement of the Problem and Rationale of the Study 1

II Aim and Objectives of the Study 3

III Research Hypothesis and Research Questions 4

III.1 Research Hypothesis 4

III.2 Research Questions 4

IV Scope of the Study 5

V Methodology 5

VI Significance of the Study 6

VII An Overview of the Rest of the Paper 6

PART B: DEVELOPMENT 8

CHAPTER I: LITERATURE REVIEW 8

I.1 The Importance of Pronunciation Teaching and Learning 8

I.2 Aspects of Pronunciation 9

I.3 The Aim of Teaching Pronunciation: Intelligibility 10

I.4 General Description of Consonants and English Consonants 12

I.5 English Fricative Consonants 14

I.5.1 Production of English Alveolar Fricatives: /s, z/ 14

I.5.2 Production of English Alveo-palatal Fricatives: /ʃ, ʒ/ 15

I.6 English Affricates: /ʤ, ʧ/ 15 I.7 Previous Studies on Vietnamese Learners’ Pronunciation of the Six

English Consonants /s, z, ʃ, ʒ, ʤ, ʧ/ 16

I.8 Teacher’s Corrective Feedback 18

I.8.1 Definition of Teacher’s Corrective Feedback 18

I.8.2 Types of Teacher’s Corrective Feedback 19

I.9 Teacher’s Explicit Corrective Feedback 21

I.10 Theoretical and Empirical Background on TECF 23

I.10.1 Theoretical Background on TECF 23

I.10.2 Empirical Background on TECF 25

I.11 Research Gap 26

CHAPTER II: METHODOLOGY 27

II.1 Context of the Study 27

II.2 Study Design 28

II.3 Selection of the Main Subjects for the Study 31

II.4 Assignment of the Subjects to the Experimental and Control Group 32

II.5 Instruments for Data Collection 35

II.6 Instrument for Data Analysis 39

II.7 Treatment to the Experimental Group and Control Group 40

II.8 Integration of Pronunciation Targets into the English speaking lessons 47

CHAPTER III: RESULTS 51

III.1 Comparison of the Pre-test and Post-test Scores 51

III.1.1 Comparison of the Pre-test Scores of Both Groups 52

III.1.2 Comparison of the Gain Scores of Both Groups after the Experiment 53

III.1.3 Comparison of the Post-test Scores of Both Groups 56

III.2 The Experimental Group’s Opinions about TECF on their Pronunciation of the Six English Consonants /s, z, ʃ, ʒ, ʤ, ʧ/ 57

CHAPTER IV: DISCUSSION 61

IV.1 Discussion of Both Groups’ Pronunciation Gain Scores and Post-test Scores after the Experimental Period 61

IV.2 Discussion of the Experimental Group’s Opinions about TECF

on Their Pronunciation of the Six English Consonants /s, z, ʃ, ʒ, ʤ, ʧ/ 63

IV.3 Recommendations for the Application of TECF in English Speaking Lessons 65

PART C: CONCLUSION 67

I Conclusion 67

II Limitations of the Study 68

III Recommendations for Further Studies 69

REFERENCES 70

APPENDICES I APPENDIX 1: PRE-TEST II APPENDIX 2: POST-TEST V APPENDIX 3: PICTURE DESCRIPTION ACTIVITY VIII APPENDIX 4: QUESTIONNAIRE (ENGLISH) X APPENDIX 5: QUESTIONNAIRE (VIETNAMESE) XI

PART A: INTRODUCTION

This chapter sheds light on the research problem and rationale of the study aswell as its scope and significance More importantly, the aim and objectives of thestudy are emphasized with its research questions Finally, the chapter concludeswith an overview of the rest of the paper, which serves as an orientation for readersthroughout the research

I Statement of the Problem and Rationale of the Study

As a member of World Trade Organization, Vietnam has given foreignlanguages in general and English in particular a greater role than ever before for thepurpose of attracting foreign investment and promoting the economy As English iswidely used in many international settings, the ability to communicate in real-lifesituations is of great importance Therefore, of four English skills, speaking plays anintegral part in every school’s English curriculum throughout the country

In the process of improving speaking skills, many learners face the problem

of pronunciation In fact, according to many foreigners, Vietnamese learners canspeak English; however, not many of them have intelligible English pronunciation

so that they can be understood easily in direct communication with foreigners(Duong, 2009) In her view, the low level of communicative competence of learners

is directly attributed to their deficiencies in pronunciation, not vocabulary andgrammar This is the justification of why teaching pronunciation needs to be givenpriority

For two years working at the Diplomatic Academy of Vietnam (DAV), theresearcher has realized that most of her first-year students show few improvements

in pronunciation after two semesters although they are required to speak English inall English lessons In other words, their frequent mistakes tend to be maintained asthe first days they entered the Academy Through direct observation in manyEnglish speaking lessons at DAV, of 24 English consonants, two alveolar

fricatives /s, z/, two alveo-palatal fricatives /ʃ, ʒ/ and affricates /ʤ, ʧ/ have beenidentified as the most common pronunciation mistakes of the researcher’s students.They have also been more confirmed after considerable discussion with many teachers

at DAV For instance, most of the students tend to pronounce the word social as

/ˈsəʊsl/ instead of /ˈsəʊʃl/, or television as /ˈtelɪvɪzn/ instead of /ˈtelɪvɪʒn/

The fact remains that learners with poor pronunciation at the segmental levelare not always at an advantage In Tench (1981, p.17-20, as cited in Chung),

“spoken language is a social act, and you are expected to fulfill many of thelistener’s expectations and needs.” Therefore, “the more of these are violated, theharder it will be on the listener, and the more “points” will be taken off by listeners

in their judgment of the speaker” (Chung, p.2) This view is echoed by Yates (2002,p.1) who believes, “we often judge people by the way they speak, and so learnerswith poor pronunciation may be judged as incompetent or lacking in knowledge.”That is the reason why the researcher finds it necessary to teach her first-yearstudents how to pronounce individual sounds correctly because it will serve as afoundation for better speaking competence in the next three years In order words,they need to start with individual sounds before moving on other complicatedaspects of speech such as intonation, stress, etc

With regard to the coursebooks used by freshmen in English speaking

lessons at DAV, namely Let’s talk 2 in the first semester and Let’s talk 3 in the

second semester, they do not consist of any sort of pronunciation work onindividual sounds Instead, they contain a variety of speaking activities for thepurpose of developing students’ oral communication skills and fluency Theemployment of these books presumes students to acquire basic knowledge ofpronunciation, including English consonants, to get involved in different Englishconversations However, in many situations, when a student delivers a talk,mistakes like /ˈsəʊsl/ and /ˈtelɪvɪzn/ seem not to be perceived Furthermore, those

who are less likely to notice such pronunciation mistakes are non-English majors

From the researcher’s viewpoint, the problem lies in the fact that students donot receive adequate feedback from the teacher on their pronunciation performance.

In other words, they need to be assisted by the teacher to realize their problems.Kenworthy (1987, p.2) comments:

Learners need to know what to pay attention to and what to work on.Because speaking is for the most part unconsciously controlled, learners maymiss something important For example, they may not realize that when aparticular word is stressed or said in a different way this can affect themessage that is sent to the listener Teachers need to make learners aware of

the potential of sounds (Kenworthy, 1987, p.2)

It is suggested that Teacher’s Corrective Feedback (TCF) can be used to helplearners perceive and discard what is unacceptable or inappropriate from theirinterlanguage In simple words, the students will have a clear picture of what theyare weak at and what they need to improve Considering the benefits TCF can bring

to learners of English, the researcher wants to determine if TCF can solve herstudents’ pronunciation problems in terms of the six English consonants /s, z, ʃ, ʒ,

ʤ, ʧ/ in English speaking lessons where pronunciation work is not included

Last but not least, research into TCF on pronunciation mistakes in Englishspeaking lessons at university level in Vietnam is quite small in number All theaforementioned reasons have motivated the researcher to conduct this quasi-

experimental research on Teacher’s Corrective Feedback on the pronunciation

of English fricative and affricate consonants by non-English major freshmen at the Diplomatic Academy of Vietnam.

II Aim and Objectives of the Study

The study aims at helping non-English major freshmen at DAV improve theirpronunciation of the six consonants /s, z, ʃ, ʒ, ʤ, ʧ/

To be specific, the primary objectives of the study are as follows:

 To examine the effect of TECF on the pronunciation of the six consonants /s, z,

ʃ, ʒ, ʤ, ʧ/ by non-English major freshmen at DAV;

 To investigate the experimental students’ opinions about TECF on theirpronunciation of the six consonants /s, z, ʃ, ʒ, ʤ, ʧ/ after the experimentalperiod.

III Research Hypothesis and Research Questions

III.1 Research Hypothesis

A hypothesis is constructed for the purpose of achieving the aim of the study:

H1: Non-English major freshmen who receive TECF make more significantimprovements in their pronunciation of the six consonants /s, z, ʃ, ʒ, ʤ, ʧ/ than thosewho do not receive TECF

If the above hypothesis is fully accepted, the following null hypothesis will

be obviously rejected or vice versa:

Ho: There is no difference in the pronunciation of the six consonants /s, z, ʃ,

ʒ, ʤ, ʧ/ between non-English major freshmen who receive TECF and those who donot

In order to identify which hypothesis will be accepted, the first researchquestion, posed in Section III.2, needs to be satisfactorily answered

III.2 Research Questions

Based on the aim and objectives of the study, two questions were formulatedand needed to be satisfactorily answered:

(1) What is the difference that Teacher’s Explicit Corrective Feedback brings about

in the pronunciation of the six consonants /s, z, ʃ, ʒ, ʤ, ʧ/ by non-English majorfreshmen at the Diplomatic Academy of Vietnam?

(2) What are the experimental group’s opinions about Teacher’s Explicit CorrectiveFeedback on their pronunciation of the six consonants /s, z, ʃ, ʒ, ʤ, ʧ/ after theexperiment period?

IV Scope of the Study

Initially, corrective feedback as stated in the research title is, in fact, confined

to Explicit Corrective Feedback, one type of TCF in a study of Ellis, Loewen &Erlam (2006, as cited in Fawbush, 2010), for the reasons elaborated on in theLiterature Review Chapter

Regarding English fricative consonants, the study focuses on two Englishalveolar fricatives /s, z/ and two alveo-palatal fricatives /ʃ, ʒ/ When it comes toEnglish affricate consonants, they are /ʤ, ʧ/ Only issues concerning the pronunciation

of these sounds are taken into consideration

Finally, it should be noticed that 36 non-English major freshmen in the classKT40B at DAV were involved in the study but only 34 students were eligible tobecome its main subjects The process of selection will be elaborated in theMethodology Chapter

TECF on their pronunciation mistakes regarding /s, z, ʃ, ʒ, ʤ, ʧ/ after theexperiment.

VI Significance of the Study:

As one of the trail-blazing studies on the impact of TECF on thepronunciation of English consonant sounds by Vietnamese university students, thestudy would be of great benefit for two reasons:

Initially, the research outcome could be used as evidence of the impact ofTECF on Vietnamese university students’ pronunciation of the six consonants /s, z,

ʃ, ʒ, ʤ, ʧ/, which are believed to be one of the biggest problems that most of themencounter As for English teachers in general and English teachers at DAV inparticular, they could base themselves on the results of the paper to make informeddecisions on whether TECF should be encouraged or not In fact, if there exists apositive link between TECF and students’ pronunciation of the six consonants /s, z, ʃ,

ʒ, ʤ, ʧ/, it can bring about a significant change in many English speaking lessons atcolleges and universities, where pronunciation work is not included

In addition, this study can serve as a reliable source of related literature and abasis for other researchers, who share an interest in the topic, to start their future workfrom

VII An Overview of the Rest of the Paper:

The rest of the paper consists of two main parts:

PART B: DEVELOPMENT

This part is divided into three chapters:

CHAPTER I: LITERATURE REVIEW - establishes the solid foundation for the

whole paper Besides clarifying the key terms such as pronunciation, intelligibility,English fricatives and affricates, Teacher’s Explicit Corrective Feedback, the chapteroffers a critical review of related studies

CHAPTER II: METHODOLOGY - clarifies the reason why a quasi-experimental

design is employed in this study Furthermore, it also elaborates on the context of thestudy, the selection of the main subjects, the instruments for data collection andanalysis, and the treatment to the experimental and control group

CHAPTER III: RESULTS - presents and analyzes all the collected data to find out

the answers to the research questions

CHAPTER IV: DISCUSSION - presents the researcher’s discussion based on the

research findings Furthermore, it also offers some recommendations for theapplication of TECF in English speaking classrooms

PART C: CONCLUSION

This part summarizes all the major points raised in the paper Furthermore, it alsoindicates the limitations of the study as well as some suggestions for further studies

PART B: DEVELOPMENT

CHAPTER I: LITERATURE REVIEW

This chapter provides an overview of the literature related to this study,establishing a solid foundation for the whole paper Not only are key terms likepronunciation, intelligibility, English fricatives and affricates, Teacher’s ExplicitCorrective Feedback clarified but background information about the key terms isalso presented to ensure a thorough understanding of the research matters Besides,the research gap is also revealed in this chapter

I.1 The Importance of Pronunciation Teaching and Learning

Gilakjani (2012) considers pronunciation as a set of habits of producingsounds, which is acquired by repeating the sounds over and over again and by beingcorrected when they are pronounced wrongly In other words, when one learns topronounce a second language, he/she is forming new habits of pronunciation and

overcoming the bias of the first language (Cook, 1996) Pronunciation refers to the

production of sounds that are used to make meaning Yates (2002) opines that:

Learners with good pronunciation in English are more likely to beunderstood even if they make errors in other areas, whereas learners whosepronunciation is difficult to understand will not be understood, even if theirgrammar is perfect! (Yates, 2002, p.1)

This opinion is echoed by Hebert (2002, as cited in Shooshtari, Mehrabi, &Mousavinia, 2013), who argues that knowing grammar and vocabulary is importantbut useless if the speaker fails to pronounce those structures and words correctly.When mispronounced, even the simplest words can prevent learners from beingunderstood To sum up, the deficiencies in pronunciation may negatively affectlearners’ communicative competence Without proper pronunciation nobody cansay that he/she knows the language Apparently, this is the approach of scholars and

researchers who consider students’ communicative efficiency as the main purpose

of teaching and learning any foreign language (Hammer, 2001; Yates, 2002; Hebert,2002; Duong, 2009) Under this approach, the important role of pronunciationteaching and learning is emphasized As Harmer (2001) argues, pronunciationteaching not only helps students become aware of different sounds and soundfeatures but also improve their speaking immeasurably:

…concentrating on sounds, showing where they are made in the mouth,making students aware of where words should be stressed – all these thingsgive them extra information about spoken English and help them achieve thegoal of improved comprehension and intelligibility (Harmer, 2001, p.183)

I.2 Aspects of Pronunciation

Basically, pronunciation involves features at segmental (micro) level andsuprasegmental (macro) level The former includes individual vowels andconsonants and the latter involves aspects beyond the level of an individual soundsuch as word, phrase, and sentence stress, intonation, and rhythm (Seferoglu, 2015,

as cited in Tran, 2009)

The issue of teaching segmentals and suprasegmentals is controversial Onone hand, Coniam (2002, as cited in AbuSeileek) opines that segmental aspectsdraw some researchers’ attention because they are more easily explained and taughtthan the suprasegmental ones Burns (2003, as cited in Gilakjani, 2012) alsorecommends giving attention to segmental features because, in his opinion,phonemes are “sounds that, when pronounced incorrectly, can change the meaning

of the word.” On the other hand, as shown in some recent studies, there is a shift tosuprasegmental aspects of the sound system According to Morley (1991),suprasegmental features of pronunciation should be taught because of theirbeneficial functions in interactive discourse This is echoed by Seferoglu (2005, ascited in Gilakjani, 2012) who emphasizes macro features as they help learnersacquire communicative competence However, there are still researchers (Goodwin,

2001; Burn, 2003; Gilakjani, 2012; Moghaddam, Nasiri, Zarea, & Sepehrinia, 2012)

who have a more balanced view that a lack of intelligibility can be attributed to bothmicro and macro features so they are equally important The present study is based

on the belief that both segmental and suprasegmental features of pronunciation play

a significant role in improving communication competence However, at differentstages of learning, attention should be paid to either or both of the two aspects Asdiscussed in Section I in the Introduction part, the focus of the study is on segmentalfeatures instead of suprasegmental features

I.3 The Aim of Teaching Pronunciation: Intelligibility

In the 1970s, proponents of the audio-lingual approach emphasized thenecessity of pronunciation teaching in second language/foreign language (L2/FL)classrooms They are ones who approved of the mastery of native-likepronunciation Later, the audio-lingual approach did not win support of L2/FLspeech research evidence The reason is, according to Moyer (1999), very few adultlearners achieve native-like pronunciation in their L2 Since then, pronunciationbecame an “unteachable” subject due to the inevitability of “foreign accents”(Flege, Munro, & Mackay, 1995; Levis, 2005)

However, these days, the interest in pronunciation teaching has begun torevive It is based on the premise that the ultimate goal of L2/FL speech learning is

to achieve “acceptable” pronunciation According to James (2010, as cited in

Gilakjani, 2012), a learner’s pronunciation has three basic levels:

 Level 1: People often do not understand what the speaker is saying Thespeaker uses the wrong sounds when making English words or uses thewrong prosodic features when making English sentences

 Level 2: People understand what the speaker is saying, but the speaker’spronunciation is not pleasant to listen to because he or she has a distractingand/or heavy accent

 Level 3: People understand the speaker, and the speaker’s English is pleasant

to listen to This is called comfortable intelligibility - the goal of

pronunciation teaching

According to Gilakjani (2012), a speaker has “acceptable” pronunciationwhen other people can understand what he says and the speaker’s English ispleasant to listen to Kenworthy (1987, p.13) provides a more operational definitionthat we can “put to work”: “The more words a listener is able to identify accuratelywhen said by a particular speaker, the more intelligible that speaker is.” Becausewords are made up of sounds, it is necessary to talk about the issue of soundequivalence In practical terms, we are aiming for something “close enough” In

Kenworthy (1987, p.13), an intelligible sound is defined as one that the listener can

match with the sound a native speaker would use without too much difficulty

In fact, learners use a variety of strategies to cope with a new set of sounds.However, some ways of coping may create problems In Kenworthy (1987), anumber of learners’ pronunciation strategies leading to unintelligibility problemsare shown clearly: (as the study only deals with segmental aspects, learners’pronunciation strategies related to suprasegmental aspects will not be included inthis paper)

 Sound substitutions: When a consonant of English does not occur in the

learner’s mother tongue, the “missing” sound is substituted with somethingfrom the speaker’s first language The substitution of one consonant foranother may cause serious confusion for listeners For instance, bysubstituting /z/ for /ʤ/, a Vietnamese learner would produce the word germ

as /zɜːm/ instead of /dʒɜːm/

 Sound deletions: The speaker leaves out a sound In the case of consonants,

either a single consonant at the beginning, middle, or end of a word isdeleted or one of the consonants in a cluster is removed to simplify the

cluster For instance, the word his without the final /s/ would sound like hit.

The word devastating with the sound /s/ deleted would sound /ˈdevəteɪtɪŋ/,which is considered unintelligible

 Sound insertions: Non-native speakers may add sounds To illustrate, the

learner adds the sound /s/ at the end of the words they pronounce, causing theproblem of unintelligibility

Kenworthy (1987) also comments that “intelligibility” is often influenced byhow familiar the interlocutor is to the speaker In her opinion, the teacher himselfcannot make objective judgments of the intelligibility of his students She arguesthat the frequent interactions between the teacher and his students gradually enablehim to “tune in to” their accent In simple words, Vietnamese teachers of Englishmay find it easier to comprehend what Vietnamese learners say Therefore, it isafraid that students’ pronunciation problems that threaten intelligibility may be leftuntreated

I.4 General Description of Consonants and English Consonants

In articulatory phonetics, consonants are articulated in two ways: either it isproduced by a closing movement of one of the vocal organs, forming such a narrowconstriction that the sound of the air passing through can be heard; or the closingmovement is complete, giving a total blockage The closing movement may berelated to the lips, tongue, or throat, but the articulation of consonants is verydifferent from that of vowels, which is relatively open and unimpeded (Roach,1991)

English consonants can be classified according to three main phoneticfeatures: place of articulation, manner of articulation and voicing Manner ofarticulation describes how the tongue, lips and other speech organs are involved inmaking a consonant Six manners of articulation are plosive (a complete obstruction

of the air), fricative (a continuous airflow through the mouth), affricate (a slowrelease of the closure), nasal (the air escaping freely through the nose), lateral (theair escaping laterally over the sides of the tongue) and approximant (vowel-like)

Place of articulation is where, in the vocal tract, the obstruction of a consonantoccurs, and which speech organs are involved Places include bilabial (both lips),labiodental (the upper teeth and inner lower lip), dental (the tongue tip), alveolar(the tongue against the gum ridge), alveo-palatal (the tongue blade), palatal (thetongue), and velar (the tongue against the soft palate) As regards voicing, it refers

to the activity of the vocal cords When the vocal cords are wide apart, consonantsare said to be voiceless (lenis) When the vocal cords are closely together andvibrating, consonants are said to be voiced (fortis) English consonants with theirdifferent features of production are described in Table 1

Table 1: English Consonants

Manner of

Articulation

Place of Articulation Labial Labio-dental Dental Alveolar Alveo-palatal Velar Glottal Plosive p

b

td

kg

Fricative f

v

Θð

sz

I.5 English Fricative Consonants

Fricatives are consonants with the characteristic that when they are produced,air escapes through a small passage and makes a hissing sound, which is sometimes

called “friction” Fricatives are continuant consonants, because you can continuemaking them without interruption as long as you have enough air in yourlungs (Roach, 1991).

The fortis fricatives /z, ʒ/ are said to be articulated with greater force than thelenis fricatives /s, ʃ/ and their friction noise is louder The lenis fricatives have verylittle or no voicing in the initial and final positions, but may be voiced when theyoccur between voiced sounds (Roach, 1991)

I.5.1 Production of English Alveolar Fricatives: /s, z/

These sounds are articulated with the soft palate being raised and the aircannot pass through the nose The tip and blade of the tongue make a light contactwith the upper alveolar ridge, and the side rims of the tongue make a close contactwith the upper side teeth The airstream escapes through the narrow groove in thecenter of the tongue and then causes friction between the tongue and the alveolarridge /s/ differs from /z/ in the fact that while /s/ is a fortis (voiceless), e.g sip, rice,/z/ is a lenis (voiced) with voice from the throat, e.g zip, rise Both /s/ and /z/ can be

at the initial, medial and final positions (Roach, 1991)

I.5.2 Production of English Alveo-palatal Fricatives: /ʃ, ʒ/

With regard to /ʃ/ and /ʒ/, their place of articulation is partly palatal andpartly alveolar The tongue is in contact with an area which is slightly further backthan that for two sounds /s/ and /z/ If you make /s/ and then /ʃ/, you will feel yourtongue move backwards The air escapes through a passage along the center of thetongue as in the case of /s/ and /z/, but the passage is a little wider /ʃ/ is a fortis, e.g.shore, caution, whereas /ʒ/ is a lenis with voice from the throat, e.g garage, vision.Furthermore, in order to produce /ʃ/ and /ʒ/, most speakers have to round their lips,which is an important thing that differentiates them from /s/ and /z/ /ʃ/ and /ʒ/ can

be found in the initial, medial and final positions In the case of /ʒ/, however, thedistribution is much more limited Very few English words begin with /ʒ/ This

sound is commonly found in the medial position, e.g decision, measure, usually(Roach, 1991).

I.6 English Affricates: /ʤ, ʧ/

/ʧ/ and /ʤ/ are the only two alveo-palatal affricates in English (Roach, 1991).These sounds are articulated with the soft palate being raised and the nasal resonator

is shut off A closure made between the tip, blade, and rims of the tongue and theupper alveolar ridge and side teeth creates an obstacle to the air stream At the sametime, the front of the tongue is raised towards the hard palate in readiness for thefricative release The closure is released slowly, the air escaping in a diffuse mannerover the whole of the central surface of the tongue with friction occurring betweenthe blade/front region of the tongue and the alveolar/front palatal section of the roof

of the mouth The vocal cords are wide apart for /ʃ/, but may be vibrating for all orpart of /ʒ/ according to the situation of utterance /ʧ/is voiceless but /ʤ/ is voicedwith voice from the throat

In Duong’s research (2009), the similarities and differences of theaforementioned sounds in terms of voicing, place of articulation and manner ofarticulation are clearly indicated They are shown as follows:

Table 2: Similarities and Differences of English Alveolar Fricatives,

Alveo-Palatal Fricatives and Affricates Sound Voicing

Manner of articulation Place of articulation Same Different Same Different Same Different

A number of studies have been conducted to find out the common problems

of Vietnamese learners regarding English consonants in general and six Englishconsonants /s, z, ʃ, ʒ, ʤ, ʧ/ in particular First, when it comes to the problem ofsound deletion, /s/ in the medial position is often omitted Similarly, the omission ofthe two ending sounds /s/ and /z/ are frequent The reason lies in the fact thatVietnamese speakers, in their language, do not have to pronounce the ending sounds(Duong, 2009; Ha, 2005) In addition, the three sounds /ʒ, ʤ, ʧ/ are not included inthe Vietnamese consonant system; as a result, these sounds are really difficult forVietnamese learners to produce, especially when they occur at the end of the words(Nguyen, 2007; Tran, 2009) Tran (2009) also opines that most Vietnamese learnersare unfamiliar with “the act of holding the tongue against the alveolar ridge for theair to pass through with some friction.” As a result, it is the habit of “swallowing”ending sounds in the mother tongue that “inhibits the pronunciation of ending

sounds in the target language” (Ha, 2005; Nguyen, 2007; Tran, 2009) Second, withregard to sound substitutions, Duong (2009), in the light of Hanoi dialect, finds outfour sound pairs that make learners confused when pronouncing They are shown asfollows:

1 /s/ - /ʃ/

2 /z/ - /ʒ/

3 /ʤ/ - /z/ or /s/

4 /ʧ/ - Vietnamese /ć/

As for the first pair, /ʃ/ is often pronounced as /s/ because either learners fail

to distinguish the difference between the two sounds or they are negatively affected

by their Hanoi dialect (/ʂ/ is pronounced as /s/) Regarding the second pair, withHanoi dialect, many learners pronounce /ʒ/ as /z/ About the next pair, as mentionedabove, the Vietnamese consonant system does not have any affricate consonantsounds; thus, many learners cannot pronounce the sound /ʤ/ They often changethis sound into /z/ or /s/ The word judge /ʤʌʤ/, for instance, is pronounced as /zʌz/

or /zʌs/ When it comes to the last pair, English /ʧ/ is produced as Vietnamese /ć/,which is a voiceless palatal stop and is produced with the blade of the tonguetouching the hard palate To illustrate, English /ʧ/ in child is incorrectly pronounced

as Vietnamese /ć/ in Vietnamese chai

In brief, regarding the six English consonants /s, z, ʃ, ʒ, ʤ, ʧ/, Vietnameselearners, especially those from the North of Vietnam have a tendency: (1) to changethem into familiar sounds existing in their mother tongue (sound substitutions); (2)

to omit the sounds at the medial and final position (sound omission) This maymake their English very Vietnamese but unintelligible, which, subsequently, maylead to many problems in communication with native speakers

I.8 Teacher’s Corrective Feedback

I.8.1 Definition of Teacher’s Corrective Feedback

In the literature of Second/Foreign Language Acquisition, one of the veryfirst definitions of Teacher’s Corrective Feedback is offered by Chaudron (1977),who regards TCF as “any reaction of the teacher which clearly transforms,disapprovingly refers to, or demands improvement of the learner utterance.”Considering this definition, it seems that, according to Chaudron (1977), TCFmeans an evident and direct correction made by the teacher Nevertheless, asHartono (2012) comments, TCF also involves providing learners with some clues toelicit their self-correction It can be seen that in Chaudron (1977) the term TCF isnot be treated properly As a result, it is necessary to seek a more comprehensivedefinition of TCF The definition of TCF by Ellis, Loewen, and Erlam (2006,p.340) really comes in handy:

Corrective feedback takes the form of responses to learner utterances thatcontain error The responses can consist of (a) an indication that an error hasbeen committed, (b) provision of the correct target language form, or (c)metalinguistic information about the nature of the error, or any combination

of these (Ellis, Loewen, & Erlam, 2006, p.340)

Ellis et al (2006) provides a clear picture of TCF that the researcher wants tomention First, it is oral corrective feedback, not written corrective feedback, which

is provided by the teacher Second, TCF can be simple, involving only onecorrective strategy, or complex, involving a number of corrective moves Thefollowing example shows that the teacher uses two corrective feedback strategies,which will be clarified in the next section, in an attempt to make the studentgenerate a repair himself

Student: My mother bought me a new pair of shoes /suː/ on my fifteenth birthday.Teacher: She bought you a new pair of… (Elicitation)

Student: Yes, a new pair of shoes /suː/…

Teacher: A new pair of shoes /ʃuːz/… really? (Recast)

Student: Ah yes, a new pair of shoes /ʃuːz/…

I.8.2 Types of Teacher’s Corrective Feedback

Lyster and Ranta (1997) carried out a study in several French immersion

classrooms in Montreal The subjects were at primary level and their first language

was English The researchers audio-taped four teachers whose lessons weretranscribed These transcriptions provided database for their analyses After that, sixmain corrective feedback types, which later have widely been accepted by manystudies, were explored

 Explicit correction means explicit provision of the correct form At the

same time, the teacher clearly indicates that what the student said wasincorrect and provides the correct form Sometimes the wrong form isidentified along with the provision of the correct form in the teacher’s turn

E.g Student: Her shirt /sɜːt/is very dirty (Phonological error)

Teacher: No, you should say /ʃɜːt/ not /sɜːt/

Student: /ʃɜːt/…

 Recast involves the teacher’s reformulation of all or part of the student’s

utterance minus the error This type of corrective feedback is implicit

because the error is not clearly indicated by phrases like you should say or you mean In other words, the teacher merely provides the correct form

without directly pointing out the student’s error

E.g Student: Her shop /sɒp/is very crowded (Phonological error)

Teacher: Yes, her /ʃɒp/is very crowded Where’s her /ʃɒp/?

Student: Her /sɒp/, ah, /ʃɒp/ is on Hue Street

 Clarification request is used in the form of questions like Pardon? and

Sorry? to indicate that the learner’s utterance is not comprehensible Unlike

explicit correction and recast, this type of corrective feedback refers to theproblems of comprehensibility

E.g Student: They boarded a ship /sɪp/ bound for India (Phonological

error)Teacher: Pardon?

Student: They boarded a /ʃɪp/ bound for India

 Metalinguistic feedback consists of either comments or information about

how the student’s utterance is well-formed but the correct form is notexplicitly provided It focuses on the nature of the error but attempts to elicitthe information from the student

E.g Student: They boarded a ship /sɪp/ bound for India (Phonological

error)Teacher: Do we say /sɪp/?

Student: No, /ʃɪp/

 Elicitation refers to the technique used to directly elicit the correct form

from the student The teacher can strategically pause to allow the student tofill in the blank in his/her utterance with the correct form Moreover, theteacher can directly ask the student to reformulate his/her utterance by saying

Say it again

E.g Student: They boarded a ship /sɪp/ bound for India (Phonological

error)Teacher: They boarded the…

Student: … /ʃɪp/ bound for India

 Repetition refers to the teacher’s repetition of the student’s error In most

cases, the teacher adjusts his/her intonation in order to draw the student’sattention to the error

E.g Student: They boarded a ship /sɪp/ bound for India (Phonological

error)Teacher: /sɪp/? (rising tone)Student: /ʃɪp/

Lyster and Ranta’s research (1997) serves as a fundamental work inidentifying different types of TCF in L2/FL classrooms Based on it, many studieshave been conducted to examine which types of TCF are effective in improving L2/

FL learning This issue is of great importance as it helps researchers answer the

question How should errors be treated? Once certain types of TCF are proved to be

more effective than the others, advices can be given to teachers in order to facilitatethe process of error correction and better L2/FL learning There is also researchwork done in this trend but the six TCF types are grouped under antagonistic

categories: explicit and implicit

I.9 Teacher’s Explicit Corrective Feedback

There is a popular classification of TCF types, which is based on theimplicitness or explicitness of a certain TCF strategy Let’s consider a situation

when a learner says, “Her shirt /sɜːt/ is very dirty” If the response is “No, youshould say /ʃɜːt/, not /sɜːt/”, it is called Explicit Corrective Feedback If the teacher

says, “Yes, her shop /ʃɒp/is very crowded”, her response is called ImplicitCorrective Feedback It can be seen that in the case of Explicit Corrective Feedback,there is an overt indication that an error has been committed whereas in ImplicitCorrective Feedback, the error is not directly shown According to Carroll and

Swain (1993) and Ellis, Loewen, and Erlam (2006), Teacher’s Explicit Corrective Feedback is often operationalized as explicit correction or metalinguistic feedback while Teacher’s Implicit Corrective Feedback (TICF) often takes the form of recasts Clarification request, elicitation and repetition also fall into the category of TICF Metalinguistic feedback facilitates students’ self-repairs while error correction provides learners with correct reformulations and exemplars of the target

features In this study, the focus is on metalinguistic feedback as the researcher

agrees with Swain (1985) who places an importance on students’ modified outputrather than input (his viewpoint will be elaborated in section I.10.1) Furthermore, itshould be noticed that the term TECF and metalinguistic feedback can be usedinterchangeably in this paper

Kenworthy (1987, p.1) comments:

If you’ve never seen a lime before you may think it is an unripe lemonbecause that is the nearest equivalent of the fruits you are familiar with Youmay continue in your misperception until you actually eat one or untilsomeone points out the difference to you (Kenworthy, 1987, p.1)

In his opinion, students’ pronunciation mistakes may be fossilized if they donot receive any corrective feedback from their teacher Therefore, it is suggestedthat TCF plays an important role in L2/FL development Considering thisclassification of TCF, a question arises as to whether or not learners perceive thefunction of a certain TCF type as a correction when it is given explicitly/implicitly(Ding, 2012) In Loewen & Philp (2006, as cited in Ding, 2012), because of itsimplicitness, the corrective intention of recast may not be easily noticed by learners.The following example indicates that the learner is likely to take the teacher’s recast

as a confirmation of what he/she said rather than a correction of his/her erroneousutterance Loewen & Philp (2006, as cited in Ding, 2012) argues that the teacher’srecast is not effective in helping learners notice the gap between their interlanguageand the target form, consequently leading to NO student reformulation not tomention student-generated repair In contrast, the corrective intention of TECF ismore salient as it directly draws the learner’s attention to his pronunciation error

E.g Student: Her shop /sɒp/is near my house on Hue street (Phonological error)

Teacher: I see, her /ʃɒp/ is near your house… (Recast)

Student: Yes, and her /sɒp/ is very crowded

However, there still exists another school of thought about TECF Bearingsome doubt in mind about the effectiveness of TECF, Tornberg (2005, as cited inLange, 2009) opines that TECF not only inhibits the students while communicatingbut also makes them uncomfortable If learners make a lot of pronunciationmistakes, TECF may intimidate their confidence in speaking Furthermore, Lange(2009), who also feels uncertain about the positive impact of TECF, argues that theutilization of this feedback type is time-consuming because the teacher has to waitlong for students’ self-generated output

In fact, despite causing a lot of controversy, this classification of TCF hasbeen of great interest to many researchers who are concerned about the role of inputand output in L2 learning as well as the cognitive roles that TCF plays As the

present study deals with TECF, in the next section, theoretical and empiricalbackground that have boosted the research on TECF will be addressed in detail.

I.10 Theoretical and Empirical Background on TECF

I.10.1 Theoretical Background on TECF

There has aroused much research in the role of input and output in L2/FLacquisition Krashen (1985) proposes the Input Hypothesis, which puts primaryimportance on the comprehensible input that language learners are exposed to.According to him, acquisition occurs when learners understand input containingstructures beyond the students’ current level of competence This means input candirectly affect L2/FL learning Meanwhile, the Output Hypothesis is proposed bySwain (1985) based on her observation of the teaching and learning activities inFrench immersion classrooms The students in these classrooms were observed tohave little difficulty in comprehending the teachers’ instruction; nevertheless, theirproduction was often lack of accuracy As a result, she claims that theComprehensible Input of Krashen (1985) alone was not sufficient for thedevelopment of learners’ language acquisition In her opinion, the production ofoutput in response to input is necessary for further language development In otherwords, Swain’s emphasis is on the role of modified output, which, according to her,

is important and necessary for L2 mastery Swain (1995) also recommended the use

of TCF because it stimulates learners to make more accurate and target-like output.Considering error correction and metalinguistic feedback, it is only metalinguisticfeedback that can stimulate learners to make more accurate and target-like output,

creating opportunities for them to do output practices

Furthermore, the manner in which metalinguistic feedback contributes to thecognitive process of learners can support the theoretical claim that it plays animportant role in facilitating the acquisition of new linguistic features According to

Lyster (2004), the cognitive mechanism of L2/FL learning is described as an

information-processing model, which refers to a gradual change from declarative to

procedural knowledge (Anderson, 1983, as cited in Ding, 2012) Declarativeknowledge refers to the knowledge of the language system while proceduralknowledge refers to the knowledge about how to perform language accuracy,including language comprehension and production As regards languageacquisition, there exist two types: (1) acquisition as the internalization of newforms, and (2) acquisition as an increase in control over forms that have alreadybeen internalized (Ellis, 1997, as cited in Ding, 2012) Therefore, it can beunderstood that the first type of language acquisition means the “acquisition of newdeclarative knowledge” and the second type means “the transition from declarativeknowledge to procedural knowledge” (Ding, 2012) Based on Lyster’s view (2004),

it may be inferred that metalinguistic feedback which elicits target-like output canincrease learners’ control over the already-internalized declarative knowledge,which promotes the transition from declarative knowledge to proceduralknowledge Thus, TECF is supposed to play an important part in the cognitiveprocess of L2/FL learning; however, this notion stills need support from empiricalstudies which examine the effect of TECF in facilitating L2/FL learning process

I.10.2 Empirical Background on TECF

Driven by the theoretical concerns discussed above, the effectiveness ofTECF has become the subject of intensive inquiry The majority of previous studiesaddressing this issue were conducted within an experimental or quasi-experimentalframework, by providing TECF and TICF as different treatments to differentresearch groups, comparing the learning outcomes of these groups on a pre-test-post-test basis and attributing better learning outcomes to either TECF or TICF Theadvantage of this research design lies in the great accuracy it can bring about whenexamining the effects of TCF

There are a number of classroom studies that reported the advantage ofTECF in helping learners achieve better learning outcomes on selected target

linguistic features Spada and Lightbown (1993) demonstrated that TECF increasedlinguistic accuracy Most importantly, their results were maintained in a delay testfive weeks after the treatment Hence, it can be seen that the effect of TECF can still

be present in the long term In White, Spada, Lightbown, and Ranta (1991), theperformance of learners receiving TECF with those who did not receive thistreatment was compared The result showed that the experimental group exposed toTECF showed a higher level of linguistic accuracy than the control group.Furthermore, there were two studies conducted in 2004 showing the same results.One was carried out by Lyster (2004) and the other by Rosa and Leow (2004) Withregard to the design of two studies, the subjects were divided into three groups: (1)TICF in the form of recasts, (2) TECF, and (3) no TCF In Lyster (2004), the groupreceiving TECF outperformed the group exposed to recasts Both of theexperimental groups outperformed the control group In Rosa and Leow (2004), thesame result was produced Hence, the aforementioned studies demonstrate thatTECF may be of pedagogical value (Ellis, Loewen & Erlam, 2006) Additionally,TECF is shown to be more effective than TICF (Carroll & Swan, 1993; Lyster,2004; Ellis, Loewen & Erlam, 2006)

Though the above-mentioned studies do not specifically addresspronunciation, their results are of great importance to pronunciation teaching Theyseem to suggest that the most effective TCF type can facilitate the acquisition oflinguistic features If these results are taken into account in the case ofpronunciation, there are reasons to believe that TECF will be an effective TCF typefor pronunciation too

I.11 Research Gap

As a matter of fact, in Vietnam, a number of studies have been carried out toinvestigate the problems or difficulties that Vietnamese learners have in certainaspects of English pronunciation However, the cause-effect relationship betweenTECF and Vietnamese learners’ pronunciation is still a rather unexplored research

area More specifically, little has been attempted to explore the cause-effectrelationship between TECF and non-English major students’ pronunciation at DAV.Additionally, there is little domestic research on this issue conducted within anexperimental or quasi-experimental framework

Furthermore, according to Ding (2012), human behavior under study insocial research is not “mechanistic” In her opinion, a social phenomenon likeTECF can be understood if the human participants’ opinions, reactions, etc aretaken into account It can be seen that little attention has been paid to experimentallearners’ opinions about TECF they received, leaving a research gap in this area As

a result, further research in this direction is required

To address this major gap, the researcher has carried out a study on theimpact of TECF on non-English major students’ pronunciation at DAV, themethodology of which will be elaborated in the coming chapter

CHAPTER II: METHODOLOGY

The following chapter will depict in detail the methodology of this researchpaper It includes the information about the context of the study, the study design,the selection of main participants, the procedures as well as the instruments for datacollection and analysis

II.1 Context of the Study

The study was conducted at the Diplomatic Academy of Vietnam, which islocated at 69 Chua Lang Street, Dong Da, Hanoi For the 2013-2014 academic year,the academy took in 450 undergraduate students in six disciplines of InternationalRelations, International Law, International Economics, International

Communication, English and French, and 60 college students in InternationalRelations For students of Group A1 and D1, the standard point of enrollment in sixdisciplines ranged from 22 to 23.5 Thus, they can be assumed to have a certainthreshold proficiency of English

On average, freshmen and sophomores spend about nine hours per week onlearning English, which is divided into four sessions Each session lasts for 135minutes and only focuses on one English skill

Generally, students at DAV are divided into two major groups, namely,English majors and non-English majors English majors refer to students who enroll

in the discipline of English and non-English majors are from five disciplines ofInternational Relations, International Laws, International Economics andInternational Communication Both groups have to complete an English foundationcourse in the first three semesters In the fourth semester, there is a difference in thelearning programs of two groups English majors have sessions of Phonetics,Grammar, Pragmatics, etc Meanwhile, non-English majors participate in the ESPprograms

With regard to speaking skills, the teaching materials for non-English majorfreshmen are two books Let’s Talk 2 (used in the first semester) and Let’s Talk 3(used in the second semester) These books include a variety of interesting andinnovative topics that encourage students to develop their oral communicationskills In Let’s Talk 2, an intermediate text, topics include relationships, jobs, sportsand games, travel and transportation, the environment, arts and entertainment, andhumor In Let’s Talk 3, a high-intermediate text, topics are related tocommunication, law, superstition, education, technology, etc There are reviewpuzzles after every four units in order to help students recycle key vocabulary.Activities are accompanied by full-color photographs and illustrations.Nevertheless, it is noticed that there is no room for teaching pronunciation in thetwo books

It cannot be certain that students who sail through the university examinationfor group A1 and D1 are good at English pronunciation because the English testthey take is in the form of a written one Furthermore, as mentioned in the firstchapter, many non-English major freshmen at DAV are observed to make a lot ofpronunciation mistakes when speaking; meanwhile, the teaching materials skip theEnglish pronunciation part Therefore, it is necessary to improve their currentsituation of learning English in general and English pronunciation in particular

II.2 Study Design

To start with, there are some problems that the researcher had to take intoconsideration before choosing an appropriate design for this study:

First, random assignment to different groups is not always possible as it can

be done in a true experimental research In fact, it might be troublesome to ask forthe mixture of different classes for the research purposes in a disciplined context as

my academy In fact, in education settings, the freedom of the researcher tomanipulate and control the conditions under which the research is conducted isoften restricted (Seliger and Shohamy, 1989) The administrators are generallyunwilling to allow classes to be reorganized since it may be unfavorable for theongoing programs Even if the researcher could mix up different classes, the trueexperimental approach would not be a wise choice The learners have the right toknow what is happening with them (i.e., reorganization of classes are necessary forconducting the research) but the findings may be harmed actually (McDonough andMcDonough, 1997) Last but not least, the students are likely to be reluctant tocooperate with the researcher in a true experiment because they may think that itwill interfere in their learning routines and affect their academic outcomes

Of course, the researcher can choose two intact classes for comparison butthis nonequivalent comparison group design appears a threat to the internal validity,especially in the researcher’s situation In that semester, the researcher was assigned

to teach two classes, KT40B and CĐ05B (university level and junior college level

respectively), whose academic achievement levels were assumed not to be equal.Furthermore, junior college students are not the focus of the study As Johnson &Christensen (2007) comments, “any factor that the treatment and control groupsdiffer on are potentially confounding variables that threaten our ability to attributegroup differences after the intervention to a treatment effect” However, the classassignment at DAV was not the researcher’s rights As a result, the researcherdecided not to conduct her study on two intact classes In other words, theresearcher had no alternative but to choose the class KT40B as the sample of thestudy.

To overcome the aforementioned inconvenience, the researcher decided tochoose the following quasi-experimental design:

First, a pre-test was administered to 36 students in the class KT40B for three

purposes:

 To select the main subjects of the study To be specific, the subjects havingmuch lower or higher scores than the rest of subjects (in statistics, they arecalled “outliers”) would be ruled out from the sample In other words, themain subjects, who later were assigned to the experimental and controlgroup, were roughly at the same pronunciation level of English alveolarfricatives, alveo-palatal fricatives and affricates How the main subjects wereselected will be clarified in the Section II.3

 To assign the subjects to the experimental and control group The procedure

of group assignment will be presented in the Section II.4

 To compare with the post-test to find out the answer to the first researchquestion

As regards the written part of the pre-test, the students’ answers were marked

by two Vietnamese teachers When it comes to the sound production part of the test, the performance of each student was recorded and checked by two nativeEnglish teachers to get the final score

pre-After that, in ten weeks, the intervention (TECF) was administered to theexperimental group whereas there was no intervention in the control group How thetwo groups were treated will be explained in the Section II.7

Finally, after ten weeks, both groups sat for a post-test whose structure wassimilar to that in the pre-test The method of scoring is also the same The pre-testand post-test scores of both groups were used for comparison so that the answer tothe first research question can be found

Besides, a short written questionnaire was used as the supplementaryinstrument to explore the experimental students’ evaluative opinions about TECFafter the experiment period

This quasi-experimental design is feasible for the researcher to conduct thepresent study The results of the study were presented in numerical expression (e.g.,test scores) and then analyzed by trustworthy and powerful statistical computersoftware To be specific, in this study, the link between the pre-test and post-testscores of both groups was brought to light by paired-samples t-test with theassistance of the computer software SPSS version 16.0 (Statistical Package forSocial Science), which is widely-used for statistical analysis

II.3 Selection of the Main Subjects for the Study

The sample of the study is 36 members in an intact class KT40B at DAV.They are non-English majors in the discipline of International Economics The classhas 12 male students and 24 female students There are 16 students from the Hanoiand the rest comes from other provinces in the North of Vietnam When it comes tothe university entrance examination, the class has 12 students of group D1, 13students of group A and 11 students of group A1 It can be seen that the students inthe class KT40B are from different geographical locations and educationalbackgrounds

To select the main subjects of the study, a test on the six Englishconsonants /s, z, ʃ, ʒ, ʤ, ʧ/ was administrated to 36 students in the class KT40B To

be specific, the subjects having much lower or higher test scores than the otherswould be ruled out from the sample In other words, the main subjects who laterwere assigned to the experimental and control group were roughly at the samepronunciation level of the above-mentioned sounds

In statistics, outliers are cases that have data values very different from thedata values of the majority of cases in the data set The way a subject having the testscore that differs significantly from the other test scores in the sample are ruled out

is similar to the way an outlier is found First, the lower (first) quartile (Q1) andupper (third) quartile (Q3) were worked out by SPSS Then, the formula Q3-Q1 wasused to calculate the inter-quartile range The next step is to find the “inner fence”and “outer fence” for the data set A point that falls outside the “inner fence” isclassified as a minor outlier, while one that falls outside the “outer fence” is called amajor outlier To find the upper and lower boundaries of the “inner fence”, wemultiply the inter-quartile range by 1.5 and then add the result to Q3 and subtract itfrom Q1 The boundaries of the “outer fence” are worked out in the same way,except that the inter-quartile range is multiplied by 3 instead of 1.5

The pre-test scores of 36 students in the class KT40B can be seen in Table 3.After running SPSS, the researcher has the following table:

Table 3: Percentile Rank of the Pre-test Scores of KT40B Students

Percentiles

5 10 25 50 75 90 95 Weighted Score

Average KT40B

(Definition 1) 63.9750 64.8500 68.0000 69.7500 71.8750 74.0000 75.6000 Tukey’s Hinges Score

KT40B 68.0000 69.7500 71.7500

As indicated in Table 3, the first quartile Q1 is 68 and the third quartile Q3 is71.875 The inter-quartile range is 3.875 Then, it is multiplied by 1.5, which yields5.8125 We add 5.815 to Q3 and subtract 5.815 from Q1 to find the boundaries ofthe inner fence:

 71.875 + 5.8125 = 77.6875

 68 – 5.8125 = 62.1875

Thus, the lower and upper boundaries of the inner fence are 62.1875 and77.6875 respectively There were two cases with the scores falling outside the innerfence They are case 7 scoring 79 points and case 30 scoring 61 points Hence, forthe purpose of the study, the two students were ruled out from the sample The mainsubjects of the study are 34 students in the class KT40B

II.4 Assignment of Subjects to the Experimental and Control Group

After the main subjects were selected, they were assigned to theexperimental and control group based on their pre-test scores The process isclarified as follows:

 The subjects were placed in matched pairs and closely matched pairs forrandom assignment The reason why there exist a number of closely matchedpairs lies in the odd numbers of students gaining certain test scores Forexample, the only subject gaining 67 points (case 2) was paired with theleftover student gaining 68 points (case 11) The following table illustrateshow the subjects were placed in pairs According to the table 4, there are tenmatched pairs and seven closely matched pairs

Table 4: Placement of 34 Subjects in Pairs for Random Assignment

Matched Pairs Closely Matched Pairs

As Slavin (2007, p.29) comments, “whenever it is possible to obtain data oneach subject on variables that could be related to the outcomes we are studying,particularly when the number of students in each group is small, random assignmentshould be stratified on these variables.” When it comes to research on studentachievement, the most important variable that both groups need to be equal on isprior academic achievement level (Slavin, 2007) Since students’ learning ratesdepend to a large degree on how much they learned in the past, group differencesshould be avoided to the maximum so that meaningful interpretation of groupdifferences in the post-test can be made However, based on the frequency of thepre-test scores presented in Table 3, it is impossible to equate the control andexperimental group on their prior achievement level As a matter of fact, there exist

a number of closely matched pairs due to the odd numbers of students gainingcertain test scores Though the assignment of subjects in seven closely matchedpairs to the experimental and control group, decided by flipping a coin, cannotguarantee that both groups are equated on the most important variable – priorachievement level, it is confident to say that both groups are closely matched on thisvariable When “the intervention and comparison groups are very closely matched

in key characteristics”, valid results are likely to be produced (Jon Baron, 2007,p.5) In this way, threat to the internal validity of the study can be reduced.

Table 5: Frequency of the Pre-test Scores

Pre-test Scores Frequency

II.5 Instruments for Data Collection

The three main instruments used to collect the data for the study are as follows:

 Instrument One: Pre-test and Post-test

The pre-test was designed mainly based on Vietnamese learners’ commonpronunciation problems regarding the six consonants /s, z, ʃ, ʒ, ʤ, ʧ/ as mentioned

in the Literature Review chapter It consists of 100 test items and is divided into twomain parts, namely paper-and-pencil and sound production The maximum scorethat a student can get is 100 points

The paper-and-pencil part containing ten questions intends to identify if thestudents could distinguish four consonant pairs: /s/ - /z/, /s/ - /ʃ/, /z/ - /ʒ/, /z/ - /ʤ/ Ineach question, the test takers had to choose one out of four words, which has thesame sound as the underlined part of the given one Harris (1969, p.88-89, as cited