A study on 11th form students’ acquisition of english relative clauses = nghiên cứu về độ tiếp thu mệnh đề quan hệ tiếng anh của học sinh lớp 11

ABSTRACT This study investigated the processing difficulty of relative clauses in two groups of Vietnamese students with high and low level of English proficiency.. This study aimed to e

VIETNAM NATIONAL UNIVERSITY, HANOI

UNIVERSITY OF LANGUAGES AND INTERNATIONAL STUDIES

FACULTY OF POST-GRADUATE STUDIES

-oOo -

NGUYỄN THỊ HOA

A STUDY ON 11TH FORM STUDENTS’ ACQUISITION

OF ENGLISH RELATIVE CLAUSES

(NGHIÊN CỨU VỀ ĐỘ TIẾP THU MỆNH ĐỀ QUAN HỆ

TIẾNG ANH CỦA HỌC SINH LỚP 11)

M.A MINOR PROGRAMME THESIS

Field: English Teaching Methodology Code: 60140111

HANOI - 2017

VIETNAM NATIONAL UNIVERSITY, HANOI UNIVERSITY OF LANGUAGES AND INTERNATIONAL STUDIES

FACULTY OF POST-GRADUATE STUDIES

-oOo -

NGUYỄN THỊ HOA

A STUDY ON 11TH FORM STUDENTS’ ACQUISITION

OF ENGLISH RELATIVE CLAUSES

(NGHIÊN CỨU VỀ ĐỘ TIẾP THU MỆNH ĐỀ QUAN HỆ

TIẾNG ANH CỦA HỌC SINH LỚP 11)

M.A MINOR PROGRAMME THESIS

Field: English Teaching Methodology Code: 60140111

S T T T P , Ph.D

HANOI - 2017

DECLARATION

Title: “A STUDY ON 11TH FORM STUDENTS‟ ACQUISITION OF

ENGLISH RELATIVE CLAUSES”

I certify that no part of the thesis has been copied or reproduced by me from any other works without acknowledgement and that the thesis is originally written

by me under strict guidance of my supervisor

Hanoi, 2017 Student‟s signature

Nguyễn Thị Hoa

ACKNOWLEDGEMENTS

First of all, I would like to express my deepest gratitude to my supervisor, Doctor Tran Thi Thanh Phuc, for her careful guidance, critical feedback, and enormous enthusiasm Without her support, this thesis cannot be completed

I am grateful to all the lecturers of the Faculty of Post-graduate Studies, University of Languages and International Studies, Vietnam National University, Hanoi for their valuable teaching and assistance during my study at the college

I would like to express my thanks to all my friends who were willing to help me in sharing materials and ideas when this thesis is worked on

My sincere thanks also go to all my colleagues and students for their assistance during the process of data collection

Particularly, I owe my indebtedness to my husband for his support, encouragement and tolerance when I was in the process of writing this thesis

ABSTRACT

This study investigated the processing difficulty of relative clauses in two groups of Vietnamese students with high and low level of English proficiency It also aimed at exploring errors on relative clauses made by 11th form Vietnamese students at a high school in Hanoi 65 students from two classes were invited to join

in the research They answered a test including 30 questions about different types of English relative clauses, namely, SS, SO, OS, and OO The results indicate that participants in both two groups faced more processing difficulty in OS and OO type than the others, providing weak evidence for any of the four hypotheses: SOHH (Subject-Object Hierarchy Hypothesis); NPAH (Noun Phrase Accessibility Hierarchy Hypothesis); PFH (Parallel Function Hypothesis; PDH (Perceptual Difficulty Hypothesis) Moreover, although students in the high proficiency group performed better than those in low proficiency group, they all made a number of mistakes in doing relative clause exercises

TABLE OF CONTENTS

DECLARATION i

ACKNOWLEDGEMENTS ii

ABSTRACT iii

LIST OF ABBREVIATIONS vi

LIST OF TABLES vii

CHAPTER I: INTRODUCTION 1

CHAPTER II: LITERATURE REVIEW 4

2.1 Overview of relative clause 4

2.1.1 Definition 4

2.1.2 Classification 4

2.2 Predictor hypotheses of relative clause acquisition 5

2.2.1 The parallel function hypothesis (PFH) 6

2.2.2 The perceptual Difficulty Hypothesis (PDH) 6

2.2.3 Noun Phrase Accessibility Hierarchy (NPAH) 7

2.2.4 Subject - Object Hierarchy Hypothesis (SOHH) 8

2.3 Previous studies 9

2.3.1 Previous studies on the PFH 9

2.3.2 Previous studies on the PDH 9

2.3.3 Previous studies on the NPAH 10

2.3.4 Previous studies on the SOHH 12

2.4 Error analysis 13

2.5 Summary 14

CHAPTER III: METHODOLOGY 15

3.1 Setting and participants 15

3.2 Methods of collecting data 15

3.2.1 Relative Clause test 16

3.2.2 Interview 17

3.3 Data analysis 17

CHAPTER IV: FINDINGS AND DISCUSSION 19

4.1 Research question 1 19

4.2 Research question 2 24

4.3 Other findings 30

4.4 Summary 32

CHAPTER V: CONCLUSION AND IMPLICATIONS 34

5.1 Implication 34

5.2 Conclusion 36

5.3 Suggestion for further research 36

REFERENCES 37 APPENDIX I

LIST OF ABBREVIATIONS

LIST OF TABLES

Table 1: Item purpose 16 Table 2: Case Processing Summary 19 Table 3: Results of t-test of difference between high and low proficiency groups 19 Table 4: Results of statistical analyses basing on Test of Homogeneity of Variances 24 Table 5: Results of statistical analyses using ANOVA 25 Table 6: Results of statistical analyses basing on Robust Test of Equality of Means 26

CHAPTER I: INTRODUCTION

Learning English as a foreign language is a dynamic process in which grammar plays an important role Among different grammatical aspects, relative clauses have been showed to be a great obstacle for students This study aimed to explore the validity of different hypotheses on the acquisition of relative clauses among Vietnamese high school students

of students are confused about the forms and functions of the English relative clauses They often make mistakes in combining sentences containing relative clauses and fail to distinguish between restrictive and non- restrictive relatives Therefore, I would like to know and understand more about this issue

Research on RCs has been conducted in a number of countries The majority of these studies focus on testing the validity of different hypotheses related to the acquisition of RCs For example, Alotaibi (2016) identifies that the most difficult type of RC for Kuwaiti EFL learners is OO (object-object), while Abdolmanafi and Rahmani (2012) points out that students encounter more difficulty in SO (subject-object) In Vietnam, research following this approach has been quite limited One study by Lan (2009) does not focus on testing RC hypothesis, but aims to point out typical mistakes committed by Vietnamese students She identifies these as the wrong use of relativizers, the misuse of the relative clauses, and the misuse between restrictive and non-restrictive relatives

This study aims at investigating Vietnamese high school learners‟ acquisition of RCs Among the various types of relative clauses, this study only investigate those RCs that the role of the two head nouns and two relativized noun phrases are either subject or object Basing on four basic types of relative clauses, namely, SO, SS,

OS, and OO, this study explored whether four hypotheses (SOHH, PDH, PFH and NPAH) could predict Vietnamese learners‟ acquisition of relative clauses

1.2 Aims of the study

This study is aimed at exploring errors on relative clauses made by 11th form Vietnamese students at a high school in Hanoi More specifically, it investigates which hypothesis is the most appropriate and best predicts the hierarchy of difficulty or the acquisition order of English relative clauses by two groups of students: high proficiency and low proficiency In addition, it identifies whether students of high proficiency group outperformed those of low proficiency group in the RC test

1.3 Research questions

The study addresses the following research questions:

1 Which among the four hypotheses - SOHH, PDH, PFH and NPAH – best predicts Vietnamese high school learner‟s acquisition of relative clauses, regarding high and low proficiency learners?

2 To what extent does the high proficiency group outperform the low proficiency one in the relative clause test?

3 What are possible reasons for students‟ performance on the re lative clause test?

1.4 Methods of the study

The study is primarily a quantitative one A test on RCs is delivered to student participants The test result is analyzed by SPSS to yield conclusion Based on the outcome of the test, a number of students are invited to be interviewed on how they completed the test

1.4 Design of the study

The study is organized into five chapters

Chapter I – Introduction: This chapter describes the rationale, aims, research questions, methods and design of the study

Chapter II – Literature review: This chapter presents theoretical foundations for the research and summarizes previous studies on the same topic

Chapter III – Methodology: This chapter provides information regarding how the study is conducted

Chapter IV – Findings and Discussion: This chapter details the findings of the study and compares the findings with previous studies

Chapter V – Conclusion: This chapter concludes the study and suggests further investigation of the topic

CHAPTER II: LITERATURE REVIEW

2.1 Overview of relative clause

2.1.1 Definition

A relative clause, also called an adjective clause, is a dependent clause that modifies a noun or a pronoun by making it more specific or adding additional information about it Relative clauses are typically found after a noun phrase and provide some information about the person or thing indicated by that noun phrase They are sometimes called „adjective clauses' because, like many adjectives, they often describe and help to identify the person or thing being talked about (Yule,

1998, p 240) According to Cowan (2008), relative clauses are defined as “one kind

of dependent clause introduced by a special set of relative pronouns” Huddleston and Pullum (2005, p.183) consider a relative clause as “a special kind of subordinate clause whose primary function is as modifier to a noun or nominal”

An RC is introduced by either relative pronouns (who-, whom-, which-, that-, whose) or relative adverbs (when-, where-, why-) The omission of relative pronouns can be seen as the zero relative pronouns “Where” and “when” replace a prepositional phrase of place and of time respectively, to function as adverbial of the relative clause In this case, the “preposition + which” can be replaced by particular relative adverbs which are “where” and “when”

2.1.2 Classification

McArthur (1992, p.859) considers a restrictive relative clause (also defining relative clause) “a relative clause with the semantic function of defining more closely what the noun modified by the clause is referring to.” As defined by Yule (1998, p.248), restrictive relative clauses are ones that “define” or “restrict” the reference of the antecedent noun They help to identify or classify the person or thing being talked about

This is an example of a restrictive relative clause:

The woman whom John is talking to is a teacher

In this sentence, the relative clause “whom John is talking to” restricts the reference of “the woman” by specifying “John is talking to” The restrictive modification is necessary to identify “The woman” that the speaker mentions and distinguish the woman from any others in this context

It can be seen from the above definitions that restrictive relative clauses define and identify the antecedent This helps to limit possible references of the antecedent and the interpretation of the sentence Restrictive relative clauses can be introduced

by all wh-relative pronouns and adverbs and also “that” or “zero” relative pronoun

A non-restrictive relative clause (also non-defining relative clause) adds information not needed for identifying what a modified noun is referring to (McArthur, 1992) Yule (1998, p.248) defines that “a non-restrictive relative clause gives extra information about an antecedent It provides additional information, not identifying information.” This is an example of a nonrestrictive clause:

My father, who is working in a hotel, will retire next year

In this sentence, there exists a nonrestrictive relative clause who is working in a

hotel that provides additional information about the father However, his identity is

already clear and identified thanks to the possessive determiner my Any

modification is extra explanation and if we leave out the adding clause, the sentence still makes sense Therefore, non-restrictive relative clauses add information or describe the antecedent rather than define it as restrictive relative clauses do Thus, they are not essential for identification and can be left out

2.2 Predictor hypotheses of relative clause acquisition

In order to evaluate the acquisition of English relative clauses, a number of hypotheses have been proposed They are supposed to demonstrate the relative ease and difficulty in acquiring different types of RC sentences, namely Hamilton‟s (1994) Subject-Object Hierarchy Hypothesis (SOHH), Keenan and Comrie‟s (1977) Noun Phrase Accessibility Hierarchy Hypothesis (NPAH), Sheldon (1974) Parallel

Function Hypothesis (PFH), and Kuno‟s (1974) Perceptual Difficulty Hypothesis (PDH) This section presents an overview of these hypotheses

2.2.1 The parallel function hypothesis (PFH)

The PFH, which was first suggested by Sheldon (1974), shows the order of difficulty in the comprehension of relative clauses by children learning English as their native language The PFH identifies the acquisition order of relative clauses in terms of the identification of grammatical functions between the head noun in the matrix clause and a relative pronoun in the embedded clause When the grammatical function of the head noun is the identical with that of the relative pronoun, the hypothesis predicts difficulty, whereas ease of acquisition is anticipated when the function of the head noun is parallel with its co-referential relative pronoun Therefore, the predicted order of acquisition of relative clause is represented as follows (Sheldon, 1974)

Subject – Subject (SS) & Object – Object (OO)

> Subject – Object (SO) & Object – Subject (OS) (“>” means “easier than”; “&” means “as difficult as”)

Based on this order of acquisition, it can be predicted that learners of English acquire the sentences in (a) and (d) prior to those in (b) and (c)

a SS: The girl who lives next door is beautiful

The head nouns “the girl” in the sentence of (a) and (d), as a subject and an object of each matrix clause, have identical functions with relative pronouns of each sentence respectively On the contrary, the head nouns “the girl” in (b) and (c), as a subject and an object of each matrix clause noun are not parallel with their co-referential relative pronoun

2.2.2 The perceptual Difficulty Hypothesis (PDH)

Another hypothesis, the PDH, has come from universal constraints on the process of embedding across languages as a basis for prediction The human cognitive processing in the matrix clause can be interrupted when the embedding of

a restrictive relative clause occurs in the middle of a relative clause (Kuno, 1974) Therefore the location of the relative clause in the matrix clause is critical in the PDH In other words, based on the limitation of the human memory system, it is predicted that owing to short term memory limitations, center embedding is perceptually more difficult to acquire than right embedding This is because center embedding interrupts the processing of the matrix sentence while right embedding does not Therefore the sentence in (b) is considered to be harder to acquire than sentence in (a)

(a) center embedding

The boy that the woman that the man loved scolded was intelligent

(b) right embedding

The man loved the woman that scolded the boy that was intelligent

Regarding to RCs, the PDH predicts that RCs formed on the subject are more difficult than those formed on the object In details, “object embedded, subject focus” (OS) and “object embedded, object focus” (OO) types of RCs should be easier than “subject embedded, subject focus” (SS) and “subject embedded, object focus” (SO) types, as illustrated here:

Object - Subject (OS) & Object – Object (OO)

> subject – Subject (SS) & Subject – Object (SO)

2.2.3 Noun Phrase Accessibility Hierarchy (NPAH)

Keenan and Comrie‟s (1977) Noun Phrase Accessibility Hierarchy (NPAH) has attracted a number of researchers in terms of studying the acquisition orders of the different types of RCs Keenan and Comrie carried out a detailed comparative study

of RC structures in different languages before proposing the NPAH, which suggests the subject is placed higher than object and other grammatical functions

Subject > Direct object > Indirect Object > Object of preposition

> Genitive > Object of comparison (The symbol „>‟ here means “is more accessible than”)

Example of each RC type adopted Hamiton (1994) is as follows:

(b) Object RC: the man [that the woman knows ]

Keenan (1975) hypothesized that relativized subjects are more accessible than relativized objects Therefore in terms of the functions of head noun and relative pronoun, SS and OS types are predicted to be eaiser than SO and OO

(g) SS & OS > SO & OO

(“>” means “easier than”; “&” means “as difficult as”)

2.2.4 Subject - Object Hierarchy Hypothesis (SOHH)

The fourth hypothesis that has gained attention is Hamilton‟s (1994) SOHH The SOHH recommended an implicational relationship between four types of RC clauses, namely, OS, OO, SS and SO The specific order of difficulty is predicted based on the number of discontinuities in the structure, as: OS> OO/ SS> SO According to the SOHH, the number of gaps in the structure determines the difficulty Thus, OS types should be the easiest to acquire because they contain only one gap within the RC With two gaps OO types are more difficult than OS Similarly the SS types contain two discontinuities and are as difficult as the OO types With three discontinuities the SO types are considered to be the most difficult

2.3 Previous studies

A number of studies have been conducted to test the predictions of different hypotheses of RC acquisition Four hypotheses have been tested the most, namely PFH, PDH, NPAH and SOHH, which are represented as below

2.3.1 Previous studies on the PFH

Among rather limited studies on RC related to PFH hypothesis, Sheldon (1974) identifies that children could better understand sentences in which the head noun and the relative pronoun served the same function in their respective clause (e.g., SS and OO relativization) However, this finding has not been duplicated by other EFL researchers Doughty (1991) concludes that the PFH has not been supported by the studies of L2 relativization as it was predicted In addition, different studies find inconsistent results with the PFH (Bowerman, 1979; Gass and Ard, 1980; Ozcan, 1997; Ozge, Marinis, and Zeyrek, 2010; Sadighi, 1994 For example, Rahmany and Haghpour (2015) investigates the difficulty order of RC types and concludes that the most demanding RC type for Persian speaking EFL learners was the SS type due to the difficulty in identifying SSMS, MS referring to the Matrix Subject position (Mean=3.88) It seems that due to the embedding of RC within the matrix clause, participants misunderstand the object of the RC as the subject of the matrix clause This result is inconsistent with predictions of Sheldon (1974) Another research by Özge, Marinis, and Zeyrek (2010) investigating the production of RCs

in Turkish children and adults shows that both children and adults used more subject than object RCs and children were less accurate in the production of object compared to subject RCs

2.3.2 Previous studies on the PDH

Wong (1991) collects 170 English essays written by four ESL learners from five classes in a secondary school in Hong Kong and attempts to check if the NPAH

is valid in the order of RC acquisition By conducting conducted pre- and error-frequency counts of the learners‟ production of six types of relative clauses, he argues that analyses were more in line with Kuno‟s (1975) hypothesis Therefore,

post-the order of post-the acquisition of relative clauses by post-these learners was: OS> OO> SS>

SO Similarly, another study is conducted by S J Abdolmanafi (Rokni) & Z Rahmani (2012) to identify the rank order of mastery of the four types of relative clauses - SS, SO, OS, and OO - in Iranian EFL learners It is found that the rank order for mastery was OS (80%) > OO (52.5%) > SS (20%) > SO (2.5%) Therefore, the findings obtained from this study (OS + OO = 294 + 282 = 576, while SS + SO = 291 + 261= 552) are consistent with the results of a number of previous researchers (Ioup& Kruse, 1977; Kuno, 1974; Wong, 1991 and Ramin Rahmany and Mina Haghpour, 2015) that OS and OO relative clause types would

be easier to acquire than SS and SO types

Different results which show the invalidity of the PDH was achieved through the study conducted by Song Tiehua (2016) Through examining college EFL learners‟ comprehension and production tasks of RC, the acquisition of English relative clauses in the context of foreign language learning in China has been shed light on It identifies that the difficulty order of the production task for Chinese participants was O-DO > O-IO (indirect object) > S-DO > SS & S-IO > O-S > S-OPREP, whereas that in the comprehension task was O-DO > O-OPREP > S-OPREP> O-IO> S-S> O-S> S-DO> S-IO The results generally contradict the prediction of the PDH

2.3.3 Previous studies on the NPAH

A significant number studies supports the NPAH, showing that it is likely a valid prediction about relative clauses To be specific, Gass (1979) investigates the acquisition of relative clauses by 17 high-intermediate and advanced learners who belonged to different linguistic backgrounds e.g Korean, Thai, Italian, French, Japanese, Portuguese, Persian, etc A sentence combination task was used to examine directly the difficulty order of RC types The participants‟ answers revealed that NPAH was manifest in terms of universal order on the hierarchy For instance, the subject position was the easiest one to relativize being the highest position on the hierarchy, whilst the most problematic one was no other than the

object of comparison Jong- Bai Hwang (2003) investigates the relevance of the three hypotheses (PFH, PDH, and NPAH) for predicting Korean learners‟ acquisition of English relative clauses Fifty-nine learners from a high class took three tasks (grammaticality judgment, production and gap filling task) included four types of clauses, and the results showed that the NPAH appeared to be the most relevant hypothesis among the three In grammaticality judgment task, the subjects received the highest mean score in their judgement of OS-type relative clauses (M=65.76), and the second highest score on the SS type relative clauses (M= 60.52) The mean scores of subjects‟ judgment on the SO and OO type relative clauses were lower than the others (M= 55.56, and 48.68, respectively) Therefore,

in this task, the difficulty orders exactly the same as that predicted in the NPAH: SS

& OS > SO & OO

While a number of studies found their results in line with the NPAH, few other studies identify that the NPAH was not able to account for the learner‟s RC difficulty order (Gibson, 2003; Hsiu-chuan, 2000; Ozeki and Shirai, 2007; Ramin Rahmany and Mina Haghpour, 2015) For example, with an aim to investigate the difficulty order of RC types, the results of the study conducted by Rahmany and Haghpour (2015) shows the most demanding RC type for Persian speaking EFL learners was the SS type due to the difficulty in identifying SSMS (Mean=3.88) It was concluded that the results had conflicting results with the NPAH Another example was the study conducted by Jong- Bai Hwang (2003) which investigated the relevance of the three hypotheses (the PFH, the PDH, the NPAH) in predicting the acquisition order of English restrictive relative clauses by Korean EFL learners The result showed that although NPAH appeared to be the most relevant hypothesis among the three, it seemed to be weak for predicting the difficulty order in the gap – filling tasks The difficulty order was SS > OS > OO > SO compared to SS & OS >

SO & OO predicted by the NPAH According to the researcher, the surface order of the difficulty in the gap-filling task is the same as the NPAH predicts, but the order

is not sustained by the statistical analyses since no significant difference is found between different types of relative clauses

2.3.4 Previous studies on the SOHH

The SOHH, which is based on the notion of processing discontinuity in both RC and matrix clause, has also been proved to be valid For example, Izumi (2003) analysed data from 61 learners of English as the second language in three different tasks and concluded that the results of all tests supported the predictions by the SOHH As noted by the researcher, the data were consistent with the prediction of the SOHH but differed with regard to the two OPREP relatives (OO Prep and SO Prep)

partially supports SOHH The processing difficulty of relative clauses in two groups

of Persian-speaking students with high and low level of English proficiency is investigated 165 (92 female) university students aged 18 to 30 did sentence comprehension exercises consisted of four types of restrictive English relative clauses, namely, SS, SO, OS, and OO ANOVA measurement was conducted for

RC types (SS, SO, OS, OO) for the within-group variable, and proficiency level (high, low) for the between-group variable The results showed that there was no significant interaction effect between the types of RCs and the proficiency level The results of pairwise comparisons showed that participants‟ performance did not differ in OS (Mean= 4.58) and OO type (Mean= 4.47), but their performance in OS and OO type was better than SO (Mean= 4.16) and SS type (Mean= 3.78) Therefore the order of processing different RC types is OS & OO > SO > SS The participants encountered more processing difficulty in SS type than in OS and OO ones According to the researchers, although SO type was demanding for learners in comparison with OS and OO type, SS was the most difficult which was contrary to the SOHH (OS > OO & SS > SO) That explains why the study reveals a partial support for the SOHH

Another research conducted by Gao (2014) has different findings With 40 Chinese English learner participants, she cannot verify the SOHH In her study, the difficulty order is OO > OS >OO prep> SO > SS > SO prep in sentence combination

In conclusion, although there have been multiple researches on different RC hypotheses, the findings across those findings are far from consistent This study, therefore, aims to contribute to this literature review by reporting findings from Vietnamese high school learners

2.4 Error analysis

Besides testing different hypotheses, a number of studies aim at identifying common errors made by students when completing RC tasks Alotaibi (2016) conducts a study to examine the extent to which 120 Kuwaiti EFL learners are aware of the structure of relative clauses in English A sentence combination task was used to measure the participants‟ ability to produce correct relative clauses in English The results indicate that Kuwaiti EFL learners may not be fully aware of the formation rules of relative clauses in English (total mean=60.4%) Moreover, statistically significant differences are identified between the answers of the advanced (76.3%) and intermediate learners (44.4%) The participants made a number of errors, such as deletion of the relative pronoun, wrong relative pronoun, repetitive use of pronouns, passivisation of the relative clause and problems with the indirect object and genitive relative clauses

Yee (2005) shows another problem associated with RC acquisition, which is the unnecessary repetition of a pronoun in a position which formed a wrong relative structure, such as “The girl who was sick she went home.” A potential reason for this would be explained by the difference between Chinese and English In Chinese,

a pronoun is required when the head noun is not in the subject or direct object positions This may influence Chinese learners of English in their formation of English relative clauses In addition, there are other factors such as the difficulty in resetting the wh-movement parameter and the specifier-head agreement parameter, the irrelevance of the universal principle constraining the occurrence of wh –movement or the licensing wh-traces, the occurrence of negative L1 transfer and the influence of the processing load

In Vietnam, the setting of this study, research on the acquisition of RC has been rather limited Lan (2009) conducts a research on sematic and syntactic problems in using relative clauses in Vietnamese high school students Data collection is administered to 40 high school students to identify their errors related to RC in English The most noticeable errors made by the students are the wrong use of relativizers, the misuse of the relative clauses, and the misuse between restrictive and non-restrictive relatives For example, the result shows that the most mistaken

uses of relativizers was that, which and whose with 77.5%, 75% and 70 % of wrong

answers respectively In terms of the use of the relative clauses, students did not know which part of the sentence to make a relative clause, and what head noun to

be modified by a relative clause Lan (2009) assumes that this is because the students lack back ground knowledge as well as understanding of syntactic features (especially the position) of relative clauses

2.5 Summary

This section has presented theoretical issues related to relative clauses It introduces definitions of RC, types of RC, four hypotheses that are often investigated in relation to language learners‟ acquisition of RC, and previous studies The following section will present the methodology of the research

CHAPTER III: METHODOLOGY

This chapter presents the methodology of the research with detailed description

of the participants, methods of collecting data, and data analysis It is primarily a quantitative study, with findings and conclusions resulting from the analyzing of a test However, it also contains some qualitative aspects, which is data obtained from interviews

3.1 Setting and participants

The study was conducted at a high school in Hanoi, Vietnam Ninety eight students in two classes were invited to do the RC test All the informants started learning English when they were at primary school At the time of collecting data, they had completed three quarters of a school year They all studied the book Tiếng Anh 11 – the official English textbook for Vietnamese high school students By this time, they had been introduced and taught about English relative clauses in previous lessons

Among 98 students who did the RC test, only 65 of them were selected to be participants of the study Because the study aims to identify differences between high proficiency students and low proficiency students in their acquisition of RC in English, I examined the students‟ academic record From the academic record, I picked up 30 students whose average score of English was 8.0 or above and 35 students whose average scores of English was 6.5 or lower The average score was calculated based on the students‟ oral and written English test from the beginning of the school year up to the moment of test delivery The 30 students with higher average scores formed the high proficiency group, and the other 35 students formed the low proficiency group

3.2 Methods of collecting data

A number of methods were used to collect the data for the study, including a test and subsequent interviews with some students

3.2.1 Relative Clause test

According to Hughes (1989), a test refers to any structured attempt to measure language abilities; and „abilities‟ refers simply to what people can do in, or with, a language When investigating learners‟ acquisition of RC, a proficiency test is often implemented (e.g., Alotabi (2016), Lan (2009), Rahmany & Haghpour (2015)

In this study, the RC test aims at testing the four hypotheses of RC acquisition The test consists of 30 sentences, classified into three tasks: a multiple choice task,

an error correction task and a sentence combination task The format of the test in terms of specific sentence structures resembled those in English grammar books, while the exact words used in the test was chosen so as not to contain new words other than those students had encountered previously This was to ensure that students‟ performance on the test would not be influenced by their unfamiliarity with the words The test consists of eight items of OS structure, eight items of SO structure, and seven items of SS and OO structures respectively

The test was piloted with ten students of the same cohort but in different classes than those who did the actual test After obtaining the outcome of the pilot test, some changes were made so that the test would be more reliable The actual test was administered to students of two classes at the same time, and was collected after 30 minutes

All test papers were marked by me, and only answers that satisfactorily completed all the requirements of the test were marked as correct For example, in the error correction task, an answer was marked as correct when the student can both identify the error and propose the correct alternative

The format of the test and the purpose of each test item are represented as follows

Table 1 Item purpose

3.2.2 Interview

This study is primarily quantitative However, I would like to identify possible reasons for the students‟ selection of options and explanation for their answer Therefore, I decided to conduct further interview with some students An interview

is a conversation with a purpose (Berg & Lune, 2012) It enables research participants to share their opinions about the subject matter (Richards, 2005) and allows the researcher to gain a deep understanding of the issue (Stake, 1995) All interviews were conducted in Vietnamese so that the students fully understand the questions of the interviewer and could use their mother tongue – the language they were best at – to give answers

Of the 65 participants, I selected five students with the highest score on the RC test (28/30 correct answers) and five students with the lowest score on the RC test (under 10/30 correct answers) to interview Each interview lasted approximately 10 minutes I first asked them general questions related to their experience of learning English, such as their feeling and difficulties when learning English Then, I moved down to their test I selected up to 15 items, including those that students answered correctly and incorrectly respectively and elicited what drove them to such answers Some questions that I asked were:

2 You have said that you chose option B randomly for the first question in task 2 So, are there any clues for your choice?

3 Why did not you pick any options for the question six in task 2? Was this because you did not understand the meaning or because of any other reasons?

4 In error correction task, you underlined this word but you did not give the correction for it Can you complete it now?

3.3 Data analysis

* Quantitative analysis

The test results were analysed by me using SPSS software, version (64 Bit)

20 SPSS is software that allows comparison of data across different groups SPSS can handle data flexibly and perform statistic procedures accurately The data

obtained from the two groups were entered into SPSS The low proficiency and high proficiency factor were entered as independent variables, and students‟ score were entered as dependent variables A total score of each type of clause tested (SO,

OS, SS, OO) was calculated as dependable variables

After all the scores were entered, I conducted some formulas to obtain the information needed Firstly, I calculated the means obtained from the two groups of students to identify the means for each type of clauses for each group respectively This helps to determine which hypothesis best predict the acquisition of RC in each group

Then, the one-way analysis of variance (ANOVA) was used to determine whether there are any statistically significant differences between the means of the two groups This would help to identify whether the high proficiency group performed better than the low proficiency group

* Qualitative analysis

Qualitative analysis was conducted with the transcript of the interviews with the students Based on their answers, I worked out whether there were common patterns related to how they selected particular answers and difficulties when they did the test

CHAPTER IV: FINDINGS AND DISCUSSION

In order to measure the ability of Vietnamese learners to produce correct relative clauses in English, an SPSS analysis was used, including the calculations of percentages, means and standard deviations of the participants‟ answers on the test These calculations were shown according to the four types of relative clauses

Table 2 Case Processing Summary

Cases

4.1 Research question 1

Table 3 shows the answers of 65 informants of the study to the 30 questions in the test In the first ten questions, participants are requested to select an appropriate relativizer for each blank The next ten questions are designed as an error correction task, which inquires students to find out the mistakes and provide the correct alternatives In the last ten questions, students have to combine sentences using relative clauses

Table 3 Results of t-test of difference between high and low proficiency groups

It is clear from the chart that thirty – five students in low proficiency group received the highest mean score in their result of SO type RCs (M = 3.8571), and the second highest score on the SS - type RCs (M=3.8) The mean scores on the two other types, OO and OS were lower (M= 2.9143, and 1.8 respectively) Besides, the highest standard deviation belonged to SO RCs with 1.97250 It means the analysis

of the participants‟ results illustrated the informants in low proficiency group of this study found SO and SS - type RCs easier than the other types and OS was considered the most difficult one Therefore, to low proficiency group, the RC acquisition order is SO & SS > OO > OS, which supports none of the hypotheses Similarly, students in high proficiency group performed SO (M=7.5667) and SS-type RCs (M= 6.8667) better than the other two types, OO (M = 5.5) and OS (M

= 6.1333) Among four types of RCs, OO – type RCs showed the highest 1.04221 standard deviation There was a slight difference in the order of RCs acquisition between two groups While weak students had a higher mean score of OO than OS, students in high proficiency group got adverse result (OS > OO).Therefore, the difficulty order of RCs that the students in high proficiency group gained is SO >

SS > OS > OO, indicating that all hypotheses also fail to be proven

According to the results, SO relative clauses were classified into the easiest type However, it seems that they are still difficult to students in low proficiency group Four out of five weak students made errors on both two questions and could not explain the reason for their choices For example, in question 5, the informants

were asked to select one relativizer out of the four relativizers who, whose, whom or

which to fill in the blank of the following sentence: “The teacher with………… we

studied last year no longer teaches in our school For this question, most informants selected “who” while some chose “whose” When being interviewed for the reason

of choosing the option “who”, one student in low proficiency said: “The teacher” is

a person and that is why I do not chose “which” Instead, I chose “who” to indicate

a person I do not know how to use “whose” or “whom” Another example is that the students were required to combine two following sentences: “The bed has no