the use of lexical collocations in writing by vietnamese learners of english a corpus based study

To explore Vietnamese EFL learners’ correct collocation use and common errors of lexical collocations, as well as the reasons why they fail to produce correct lexical collocations in the

INTRODUCTION

Background to the study

Vocabulary instruction has been considered to be of great importance to the second language acquisition of EFL learners Hence, lists of core vocabulary are created and proven to be useful for teaching and testing purposes as those provide a good basis or lexical repertoire for learning English as a foreign language New core vocabulary lists have been compiled as the interest in this particular field has not lessened (Brezina & Gablasova, 2015) However, one weakness of these core vocabulary lists has been identified; it is the fact that they only include single words, underestimating the multi-word units which are generated by the high- frequency words

These multi-word units are sometimes considered to be more frequent than the single items, and they must be taught formally (O’Keeffe et al., 2007) Excluding multi-word units from vocabulary lists may negatively affect lexical assessment as the number of words categorised as highly-frequent will increase, which may not correspond to the whole expression’s frequency in the target language input (Lindqvist et al., 2012) Furthermore, much of the language itself comprises prefabricated chunks or multi-word units, and collocations, and thus, learning these is considered important (Hoey, 2005) It is also believed that collocations are easier to learn compared with isolated words (Lewis, 2000) Regarding this, collocations play a crucial role in EFL learners’ language productive skills, assisting their mastery of the target language

Collocations play a pivotal role in language proficiency, as research has established a strong correlation between knowledge of collocations and overall language ability (Cantos-Gomez & Sánchez, 2001; Nesselhauf, 2005) This association is particularly evident in English language learning, where lexical collocations are found to be crucial for learners' proficiency (Al-Zahrani, 1998; Sung, 2003) This suggests that learners with a robust understanding of collocations possess a significant advantage in their language acquisition journey.

2 acquire, the more competent the learners’ communicative skills become Nevertheless, few studies have investigated this issue in the context of Vietnam, especially high-school EFL learners in Vietnam (Bui, 2021), which stems from the reality that grammar receives more attention and isolated words are more likely to be taught, not collocations The consequences can be significant, as Vietnamese EFL high-school learners can encounter problems or hindrance related to English proficiency

Collocation difficulties faced by non-native learners stem from various causes Language transfer, an interlingual factor, significantly contributes to collocation misuses (Bahns, 1993) Additional interlingual causes include the learner's native language structure and semantic differences between the two languages Intralingual causes, arising within the target language, also play a role These include insufficient exposure to authentic language, lack of language awareness, and learners' reliance on dictionaries.

EFL learners frequently exhibit linguistic influence from their native tongues when producing the target language, often employing translation strategies (Eldaw, 1993; Biskup, 1992; Yamashita & Jiang, 2010) Intralingual factors, such as overgeneralization and rule restrictions ignorance, also contribute to erroneous collocations (Richards, 2015) Furthermore, insufficient exposure and knowledge of collocations in materials and textbooks can hinder acquisition and impact production (Koya, 2003; Laufer & Waldman, 2011; Liu, 1999; Nesselhauf, 2003).

By and large, collocations are believed to be essentially important to the development of English proficiency Thus, an investigation into learners’ lexical collocation use in writing and causes of lexical collocation errors can provide a valuable insight into EFL learners’ acquisition and production of collocations This study is conducted to investigate the use of lexical collocations of Vietnamese EFL learners, the differences in the collocation use between higher-proficiency learners and lower-proficiency learners,and the factors influencing learners’ collocation use

Aims of the study

Although collocations play a crucial role in assisting EFL learners to achieve mastery in the target language, and numerous studies investigating EFL learners’ use of collocations in general have been conducted (Biskup, 1992; Cao, 2018; Dang, 2014; Do & Le, 2022; Durrant & Schmitt, 2009; Harta et al., 2021; Koya, 2003; Mahmoud, 2005; Nesselhauf, 2003), little is known about how Vietnamese EFL high school learners produce lexical collocations in their writing, the types of errors they have, the types of lexical collocations that cause the most challenge to them and the underlying causes of such erroneous use Therefore, the present study, employing the mixed-methods approach, has been carried out

The purpose of this study is to investigate how Vietnamese EFL high school learners produce lexical collocations in their writing and the causes of their erroneous collocation use The aims of this study are given below:

(1) To investigate how Vietnamese EFL high school learners use lexical collocations in their writing

(1a) To explore the extent to which those lexical collocations are produced accurately

(1b) To investigate the types of lexical collocation errors Vietnamese EFL high school learners tend to produce

(2) To explore the possible factors contributing to those lexical collocation errors, as perceived by learners and teachers

To align with the aims introduced above, the following research questions are formulated to guide the study:

1 How do Vietnamese EFL high school learners use lexical collocations in their writing?

1a To what extent are the lexical collocations used accurately by higher-proficiency learners and lower-proficiency learners?

1b What lexical collocation errors are made by higher-proficiency learners and lower-proficiency learners?

2 What are the factors contributing to Vietnamese EFL high school learners’ lexical collocation errors, as perceived by learners and teachers?

Significance of the study

This research aims to highlight the significance of lexical collocations in English language teaching and learning, particularly within Vietnamese high schools By investigating the production of collocations by learners at different proficiency levels, the study identifies common errors and explores the factors influencing their usage These findings contribute to the growing body of research on collocation instruction, providing valuable insights for educators both within Vietnam and globally, and fostering informed pedagogical practices.

With regards to the practical aspect, the findings of this study can provide teachers and educators with an insightful understanding of learners’ common collocation errors so that appropriate teaching methodology and techniques can be applied to assist learners Moreover, learners can pay more attention to common collocation errors and frequently used lexical collocations so that frequency and accuracy of such can be increased The implications can help teachers to have more awareness concerning the approaches, techniques and strategies used in the classroom to teach lexical collocations to EFL learners Furthermore, with the knowledge of the types of lexical collocations that learners find most difficult to acquire, teachers can take proactive action by providing exercises, extra materials and activities that can raise awareness of the importance of collocations, draw learners’ attention to the target collocations in the lesson and improve autonomy to acquire collocations better.

Scope of the study

The present study focuses on lexical collocations, types of lexical collocation errors and the possible underlying causes of the errors Idioms are not considered in this study Within this focus, a number of factors need clarifying Firstly, the sample of participants included 54 tenth graders of a class in a public high school in Ho Chi Minh City, whose level of English proficiency ranged from

5 lower-intermediate to higher-intermediate These students had been learning English for at least seven years Secondly, this study only explores how these learners produce lexical collocations in their writing, which means that language proficiency reflected in the other language skills, namely reading, listening and speaking, is not examined Finally, the present study only investigates the production of lexical collocations, and collocation knowledge is outside the scope of this study.

Organization of thesis chapters

This thesis is divided into five main chapters

Chapter 1 consists of the background to this present study, aims and research questions, significance and the scope of the study

Chapter 2 reviews, analyzes and synthesizes the body of literature related to the topic under investigation Research on identification and classification of collocations, collocation errors, causes of those errors and the process of teaching collocations is reviewed

Chapter 3 introduces the research design, participants and research instruments of the present study The process of designing the Collocation Test and Writing Test, as well as the procedure of collecting and analyzing both the quantitative and qualitative data, is explained

Chapter 4 reports and discusses the major findings of the present study It also addresses the research questions and provides the foundation for the implications and suggestions in the next chapter

Chapter 5 concludes the study by summarizing the main findings Pedagogical implications are provided to make suggestions to different stakeholders Limitations of the present study, as well as suggestions for future research, are also provided

In the next chapter, a review of the existing body of literature regarding the topic under investigation is provided

LITERATURE REVIEW

Collocation identification and classification

Many researchers have attempted to propose a definition for the term collocation However, most of the attempts are in reference to the researchers’ specific research interest, resulting in the lack of a universal definition Why can curiosity, criticism or concern be aroused, but attention cannot? Instead, we can draw, attract or catch somebody’s attention The government can fine people heavily for speeding, but not strongly These are considered the typical combinations of words, which are naturally understood and utilized by native speakers According to Lewis (1997), collocations are the backbone of his Lexical Approach, which stresses the notion of chunking and that ready-made chunks are said to be more easily retrieved by students as they are stored in their long-term memory Similarly, Thornbury (2002) refers to the behavior of words combining with a number of other particular words as hunting in packs To comprehensively recognize and process these chunks, different approaches and viewpoints have been proposed

Collocation is generally defined as the grouping together of words However, how words are grouped together is not the same to different researchers According to Benson et al (1986), free combinations are combinations of words that are not united closely with each other, whereas idioms are frozen expressions whose meanings do not reflect the meaning of each component of the combination Benson et al (1986) uses combinations with the noun murder to prove these two points Regarding free combinations, to condemn a murder and to discuss a murder are two of many acceptable combinations of this noun Meanwhile, to scream blue

7 murder is an idiom including this node noun, and the meaning of this whole expression (to scream loudly) does not reflect the meaning of each constituent part in the expression, making it a frozen one Existing between the two ends of this spectrum are loosely fixed combinations, or also known as collocations whose (1) meanings reflect the meaning of each component in the combination This characteristic sets it apart from idioms Other points worth noting are that collocations are (2) frequently used combinations and (3) psychologically salient, or in other words, spring up to one’s mind immediately, which sets collocations apart from free combinations By and large, Benson et al (1986) concludes that fixed combinations, recurrent combinations or collocations are fixed, identifiable, and non-idiomatic phrases and chunks

With regards to the length of a combination, while Sinclair (1991) believes that in typical investigations of collocation, the maximum range is 4-span including the node word, Kjellmer (1994) is more interested in strictly adjacent word co-occurrence More interestingly, the span proposed by Sinclair (1991) was later viewed as unnecessary because the lexical hierarchy formed by collocates within linguistic units such as sentences and phrases truly matters (Cantos-Gomez

& Sánchez, 2001) Regarding the number of words in a collocation, there have been different views upon how many words and what types of word included, which can form a collocation Whereas Partington (1998) claims that traditionally collocations are two-word combinations, there are combinations of more than two words that are still deemed collocations For example, according to Kennedy (1991), combinations of more than two words in length are considered collocations, such as flash through my mind Besides, other views such as collocational framework by Renouf and Sinclair (1991), slot filling (Nattinger &

Collocations, combinations of words that co-occur and exhibit restrictedness, are a crucial component of language These pairings can include lexical bundles (DeCarrico, 1992) and phrasal templates (Smadja, 1993) The identification and classification of collocations serve as important linguistic tools.

Collocations can be identified depending on whether they are viewed as phraseological or statistical phenomenon, which are the two major approaches

8 (Cao, Pho & Dangnguyen, 2021) The phraseological approach (Cowie, 1994) identifies a particular collocation using its restricted suitability, while the frequency-based approach (Sinclair, 1991) identifies a collocation based on the frequency of a word combination’s co-occurrence rather than by chance

2.1.1.1 The phraseological approach in collocation identification

In traditional phraseology research, collocation is identified in a number of ways and is generally deemed as “words co-occurring within a short space of each other” with varying degrees of arbitrary restriction from free combinations (Sinclair, 1991, p.170) There are also restrictive collocations (e.g conduct research, do an experiment) and frozen expressions (e.g generally speaking)

Collocations are traditionally considered to be restricted by their semantic or syntactic environment because of the association with semantic opaqueness and substitutability of arbitrary restriction Due to the blurred boundaries between free combinations and restrictive collocations and the various extents to which each of them entails arbitrary restriction, the act of distinguishing them can be challenging

According to Yamashita and Jiang (2010), collocations can be defined differently from formulaic sequences in a number of important ways Firstly, collocations are looser combinations of words, compared to formulaic sequences because a component word in a collocation can collocate with other words to form other collocations For instance, strong can collocate with tea and man, while on the other hand together makes a unit whose constituent words cannot be substituted by another However, the flexibility of component words when recombining with one another to form other collocations is limited Secondly, while formulaic sequences tend to be language specific, which means that the existence of its counterpart or equivalent in another language is not certain, collocations are cross-linguistic and usually have their counterparts in another language, excluding cases in which culture-specific concepts are involved For example, both English and Japanese languages have the collocation hot tea, but what English people refer to as strong tea is called dark tea in Japanese This

9 results in the distinction between congruent collocations and incongruent collocations

Meanwhile, many studies adopting the phraseological approach have indicated that collocation is a word combination possessing a particular grammatical pattern in the first place (Cowie, 1994; Nesselhauf, 2005), which means that a collocation contains a base The base is always the noun head in patterns that contain nouns such as verb + noun (e.g., pursue a career) and adjective + noun (e.g., strong tea) and a collocate (e.g., pursue and strong) Based on the transparency in the combination’s meaning, which can be identified by the restricted substitutability (Cowie, 1994; Nesselhauf, 2003), a free combination and an idiom can be distinguished

In addition, in the traditional approach, unlike grammatical collocations which usually consist of one closed-class word such as prepositions or determiners and one open-class word such as verbs or nouns, lexical collocations consist of two open-class components, which are the phraseological units investigated in many previous studies (Granger, 1998; Laufer & Waldman, 2011; Nesselhauf, 2003) In the present study, in order to identify a collocation, the grammatical pattern of a collocation, which consists of a base, or the node word, and a collocate, is considered as a criterion

2.1.1.2 The statistical approach in collocation identification

According to the statistical approach, collocations are identified as the co- occurrence of words with a level of frequency (Durrant, 2009; Nation, 2001; Sinclair, 1991) According to Nation (2001), frequency of co-occurrence, opaqueness and uniqueness of meaning, adjacency, fossilization and collocational specialization are important factors to consider when identifying collocations For a simpler approach, the two major factors that need taking into consideration when identifying a collocation are the strength of the collocation and the frequency of the collocation (the degree of usage)

In previous studies, frequency has been used in order to provide a reasonable cutoff range to identify meaningful collocations, or collocations that

10 have pedagogical value For example, in an attempt to compile a list of academic formulas, Simpson-Vlach and Ellis (2010) used the cutoff ranges of 10 instances per million word, as proposed by Biber et al (1999), although in other studies, ranges between 10 and 40 instances per million were adopted The reason is because Simpson-Vlach and Ellis (2010) took advantage of other statistical measures such as mutual information (MI) score to select and rank the formulas, so a “less restricted data set” (p 492) was preferred Meanwhile, in a similar attempt to compile the Academic Collocation List (ACL), Ackermann and Chen (2013) decided upon the normalized frequency of ≥1 instance per million words so as to filter the candidate collocations A normalized frequency of ≥ 1 instance per million words would equal 96 times of occurrences in the British National Corpus (BNC) and 1000 times in the Corpus of Contemporary American English (COCA), which is relatively high frequency because this would result in pedagogically valuable academic collocations However, for this present study, the ultimate goal is to explore the collocation production of EFL learners, regardless of the academic characteristic of the collocations Therefore, this study adopted a lower frequency for a combination to be considered a collocation

Regarding the statistical measures available in native speaker corpora such as MI score, t-score or Log Dice, these have been adopted in many studies to identify collocations For example, the span of 4 words and a MI score of 3.0, proposed by Jones and Sinclair (1974), are adopted in a study conducted by Durrant and Schmitt (2009) so as to calculate frequency Meanwhile, in a study carried out by Tsai (2015) in Taiwan, an additional criterion, which is a t-score larger than 2.0, was used to ensure a combination’s frequency so that it is deemed a collocation However, in the present study, the researcher did not take statistical measures other than frequency of occurrences into consideration because of the following reasons

Collocations and language proficiency

Collocations constitute a crucial part of idiomaticity, which means that collocational knowledge is undoubtedly necessary for non-native learners to produce appropriate language and develop their language proficiency The general volume of collocational knowledge has been proven to be of great importance to EFL learners’ language mastery, and whether there is a difference in EFL learners’ use of collocations, regarding their proficiency levels, has been explored in a lot of previous studies These studies revolve around determining a relationship between EFL learners’ levels of proficiency, either determined by a test or their academic curriculum level, and their production of collocations

To begin with, some studies have discovered that EFL learners who are more proficient in the target language would be more likely to produce multiword units such as collocations in their writing production Read and Nation (2006) revealed that higher proficiency level test-takers produced more formulaic multiword strings than other band test-takers Similarly, in a study to investigate learner lexical development in the written discourse of Cambridge ESOL’s Skills for Life test, Vidakovic and Barker (2010) suggested that lexical bundles were rarely used in the written discourse of low proficiency learners, whereas the type and token of lexical bundles increased in the discourse of intermediate and advanced proficiency learners In Laufer and Waldman’s (2011) study to investigate use of English verb-noun collocations of Hebrew native speakers in argumentative and descriptive writing, a statistically significant relationship was determined between these EFL learners’ proficiency levels and the quantities of verb-noun collocations they produced in their essays To be specific, among three

15 groups of EFL learners, advanced, intermediate and basic, advanced EFL learners produced significantly more collocations than basic learners

Nevertheless, conflicting results were demonstrated in Staples et al.’s (2013) study, which investigates learners’ use of formulaic sequences such as lexical bundles in the Test Of English as a Foreign Language’s (TOEFL) writing section The findings suggest that more lexical bundle tokens were observed in low-level TOEFL iBT written responses than in higher-level responses because of the prompts This means that lower-level learners have a tendency to make use of the suggested lexical bundles, whereas higher-level learners showed less reliance on prompted bundles but a higher rate of non-prompt-based bundles This indicates that learners who are not yet proficient in the target language are more likely to repeat the unanalyzed language provided to them, while learners of higher proficiency tend to produce their own recurrent sequences instead of relying solely on the given bundles

Although frequency counts from corpus data are undoubtedly the most important indicator for phraseological awareness, and L2 learners are capable of frequently using collocations or other forms of phraseological items, many of these multi-word sequences are actually used inaccurately, as “this quantitative gain is not paralleled by a qualitative gain” (Paquot & Granger, 2012, p 137) To investigate whether the quantitative gain of collocations is directly proportional with the qualitative gain, a number of empirical studies have been conducted to compare the level of accuracy in EFL learners’ collocation production and learners’ proficiency levels In Shaw and Weir’s (2007) study, although advanced EFL learners attempted to produce more collocation tokens of various types in their written output for the Main Suite exams, which are Cambridge ESOL’s core General English exams, than their lower-proficiency counterparts, it is inevitable that a great number of collocation errors are produced by advanced learners despite the lack of grammatical errors In Nesselhauf’s (2005) study to investigate advanced EFL learners’ production of verb-noun collocations, 2,000 verb-noun collocations were extracted and a quarter of these were erroneous collocations This means that even advanced EFL learners may still have difficulty producing

16 collocations accurately, in spite of the significant quantities of the produced collocations

In the study conducted by Al-Zahrani (1998), 81 Saudi EFL university learners of four different academic levels sat a blank-filling test of verb-noun collocations and a writing test, as well as a TOEFL test Al-Zahrani (1998) determined a statistically significant difference in the collocation knowledge of first and second academic levels, second and third academic levels, as well as third and fourth academic levels This means that there is a strong relationship between EFL learners’ collocation knowledge and their overall language proficiency This finding is positively supported by findings in Sung’s (2003) study to investigate the connection between lexical collocations and speaking fluency In this study, the relationship between non-natives’ lexical collocation knowledge and their speaking proficiency, as well as their collocation knowledge and their use of lexical collocations was investigated Sung (2003) concluded that there was a strong relationship between non-natives’ lexical collocation knowledge and their use of those in their speaking, and between their speaking proficiency and their knowledge of lexical collocations, as well as their use of lexical collocations As a result, lexical collocation use can be a reliable indicator of language proficiency in productive skills

However, in Laufer and Waldman’s (2011) study, in terms of the production of erroneous collocations, learners of all three proficiency levels produced a large number of unusual collocations, accounting for approximately 30% of the total number of collocations produced When comparing the numbers of erroneous collocations and the numbers of correct collocations produced by learners of different levels of proficiency, Laufer and Waldman (2011) did not identify any statistically significant relationship between learners’ proficiency and their erroneous collocation use However, when comparing the quantities of erroneous collocations and the total number of words produced in all texts, a significant relationship was identified between advanced learners and basic learners, as well as between intermediate learners and basic learners This means

17 that advanced learners and intermediate learners have the tendency to produce more erroneous collocations than basic learners

Overall, there is a strong relationship between EFL learners’ language proficiency and their collocation use Hence, it is evident that lexical collocations play an important role in an EFL learner’s journey to language mastery With a view to improving EFL learners’ English proficiency, learners’ use of collocations should be improved initially To achieve this ultimate goal, different levels of proficiency of EFL learners should receive more attention so that appropriate pedagogical methodologies can be adopted to assist learners of all proficiency levels.

Collocation errors by EFL learners

With regards to collocational misuses, collocations often contain inflective or positional variations (e.g results obtained, broader contexts, achieving objectives), posing the great challenge of how to collect and combine these relevant forms and present them in a consistent way, assisting the process of teaching and learning these in the EFL context Studies have been conducted to investigate the challenges collocations pose to EFL learners of various nationalities (Dang, 2014; Do & Le, 2022; Harta et al., 2021; Nesselhauf, 2005; Quping & Pramoolsook, 2014) According to Nation (2001), because collocations contain a number of elements of grammatical or lexical unpredictability, it is believed to be challenging to L2 learners’ acquisition process of such Moreover, Laufer and Waldman (2011) sum up the findings from studies of error analyses and corpus analyses and claim that irrespective of years of instruction received in L2 or task types L2 learners are required to carry out, collocation use can cause problems for them

Studies have explored collocational difficulties encountered by learners, particularly advanced learners who face challenges with frequent and versatile dynamic verbs in verb-noun collocations (Dang, 2014; Do & Le, 2022; Harta et al., 2021; Nesselhauf, 2005) For instance, verbs like "get" and "make" possess multiple meanings and can be used in diverse contexts, presenting a complexity that learners must navigate within verb-noun collocations.

18 collocate with different nouns makes it difficult for advanced learners to produce such collocations accurately (Ringbom, 1998) In addition, in a study conducted by Nesselhauf (2005), verb-noun collocations that are frequent and appear in everyday expressions pose the most problems to German learners of English This is also explained by Howarth (1996) that almost every sentence structure in English includes the structure of a verb preceding an object, or a noun The extracted collocations were then classified according to their degree of restriction to determine whether each of them is a free combination (F), a restricted combination (RC) or an idiom (I), whose acceptability in English was later evaluated A quarter of the combinations produced by the learners in the study contain one or a number of mistakes, distributed evenly over the essays Moreover, as well as the verb and the noun of a combination, non-lexical elements belonging to the combination were also judged based on acceptability Nesselhauf (2003, p 232) identified common errors in these learners’ collocational use: a wrong choice of verb or non-existent verb (e.g., carry out races is wrong, and was corrected to hold races) b wrong choice of noun or non-existent noun (e.g., close lacks is wrong, and was corrected to close gaps) c combination exists but is not used correctly (e.g., take notice is inappropriate, and was corrected to to notice) d combination does not exist and cannot be corrected by exchanging single elements (e.g., hold childen within bounds is wrong, and was corrected to show children where the boundaries lie) e preposition of a prepositional verb missing, unacceptable or wrong (e.g., fail in one’s exam is wrong, and was corrected to fail one’s exams) f preposition of a noun missing, unacceptable or wrong (e.g., raise the question about is wrong, and was corrected to raise the question of) g article or pronoun missing, unacceptable or wrong (e.g., get the permission is wrong, and was corrected to get permission) h noun used in singular instead of plural or vice versa (e.g., pass one’s judgements is wrong, and was corrected to pass judgement)

19 i wrong syntactic structure (e.g., make somebody friends is wrong, and was corrected to make friends with somebody)

Mahmoud's (2005) study identified three main types of collocation errors made by Arabic learners: word choice errors, word form errors, and contextual errors Word choice errors involve using one or both words in a combination incorrectly, such as "do inspection" (incorrect) instead of "conduct inspection" (correct) Word form errors occur when the form of a word is incorrect, like "free-flow of corruption" (incorrect) instead of "corruption-free" (correct) Contextual errors, while grammatically correct, are used inappropriately in a given context, for instance, "plagued by mass destruction" (incorrect) instead of "thrown in turmoil" (correct).

The model by Mahmoud (2005) provides a general perspective of collocation errors made by EFL learners without particular erroneous patterns for each category of lexical collocations Therefore, in this present study, the researcher decided to adapt the model of Nesselhauf (2003, p 232) and Mahmoud (2005) and add error codes with specific error descriptions for each lexical collocation category (see Appendix B) Any combinations belonging to any of these seven types of lexical collocations in learners’ essays that are erroneous based on these seven subtypes are deemed errors, with inflectional morphological errors, article errors (e.g attend course instead of attend a course), and minor spelling errors being disregarded Inflectional morphological errors include errors related to inflectional morphemes such as the plural marker (-s/-es), the possessive marker (’s), suffixes indicating comparative or superlative adjectives (-er/-est), third person singular present tense, past tense, passive or perfect participle, and progressive participle (Carstairs-McCarthy, 2002, pp 28-42) The reason why

20 grammatical errors and minor spelling errors are not considered in this study is because this study only examines EFL learners’ collocation use, excluding their knowledge of grammatical features, so only errors related to collocation use are investigated.

Factors influencing EFL learners’ collocation use

Laufer and Waldman (2011) claim that collocation use poses great difficulty to L2 learners regardless of the length of exposure to the target language and the types of tasks learners practice with In this section, factors that influence EFL learners’ use of lexical collocations are reviewed

Many studies in second language acquisition of collocations have discovered that learners’ mother tongue influences their L2 collocation learning (Bahns, 1993; Bahns & Eldaws, 1993; Bisk-up, 1992; Koya, 2003; Laufer & Waldman, 2011; Nesselhauf, 2003, 2005) One major problem that learners have when acquiring collocations is language transfer, which takes place when learners find something in the L1 similar to the target language, which is considered negative transfer

A number of studies have attempted to prove that language transfer is the main cause of difficulties in learners’ collocation acquisition In a study conducted by Bahns (1993), it is concluded that inappropriate collocational usages may result from L1 transfer, especially when learners’ L1 is closely related to English This finding corroborates the findings in a study carried out by Biskup (1992) to investigate L1 transfer observed in German and Polish learners of English When EFL learners produce collocations, most of the errors are the use of inappropriate synonyms (Biskup, 1992), and L2 learners show too much reliance on a small number of some particular collocations that they are able to produce correctly

Learners of English as a foreign language (EFL) may make errors in collocation usage due to similarities between their native language (L1) and English Research by Wang and Shaw (2008) found that EFL learners whose L1 (Swedish) was closer to English made more collocation errors compared to learners whose L1 (Chinese) was less similar This suggests that the degree of congruence between L1 and the target language significantly impacts learners' collocation accuracy.

21 to English; therefore, they are more likely to attempt to create collocations, resulting in errors Meanwhile, the significant incongruence between Chinese and English suggests that they are less likely to try creating collocations, resulting in fewer collocation errors Similarly, the findings of the study conducted by Nesselhauf (2003) suggest that L1 influence on the production of English collocations by German speakers is remarkably high, because the degree of congruence of the collocations between their L1 and L2 is associated with the likelihood of learners producing incorrect collocations, which also means that congruent collocations, which sound natural in both their L1 and L2, cause less difficulty to learners than non-congruent collocations

These findings are consolidated in Ying’s (2009) study, which explores the collocation use of Chinese-English majors and non-English majors It is concluded from the study that collocations which are congruents and translatable from L1 to L2 are easier for learners to acquire, and result in fewer collocation errors This study also discovers that overgeneralization is a major cause as learners tend to look for the equivalent collocation in L1 without the realization that incongruity exists between their L1 and L2 Similarly, it is discovered that EFL learners tend to overuse highly-frequent adjectives to form collocations, which may lead to the formation of odd collocations or ones with vague meanings (Jullian, 2000; Pavičić Takač & Lukač, 2013)

Collocations often contain inflective or positional variations (e.g results obtained, broader contexts, achieving objectives) and a number of elements of grammatical or lexical unpredictability, so they are believed to pose a great challenge to EFL learners regarding how to present these forms in a consistent way Likewise, Laufer and Waldman (2011) discovered that EFL learners have the tendency to depend on their L1 without considering the restriction of word combinations EFL learners’ writing in English was examined in Laufer and Waldman’s (2011) study to identify collocational misuses related to verb + noun collocations and the results suggested that half of the total number of erroneous collocations identified in the study were caused by L1 interference and this influence did not decrease over time during the span of the study The learners

22 have the tendency to ignore restrictions on word combinations and construct the meaning of a combination from individual words, instead of noticing the prefabricated patterns

Similarly, Liu (1999) claims that learners do not know the collocational restrictions which are based wholly on the meaning of the word and range The underlying reason is that learners rely on individual lexical items instead of prefabricated chunks to construct a message When learners encounter difficulty producing the desired collocations, they have the tendency to avoid using those, which is referred to as the avoidance strategy This is discovered in a study carried out by Koya (2003), in which learners of low level of proficiency refrained from producing collocations that are not translatable from Japanese This can be understood that the insufficiency of collocational knowledge or lexical items may hinder the acquisition and production of certain collocations Liu (1999) also finds out that the lack of collocational concept leads to learners’ inability to know which word collocates with the node word although they know the basic meaning of the node word

Besides, coinage and approximation are contributing factors in the production of collocational misuses Word coinage means that learners make up new words or combinations in order to get their message through Liu (1999) gave an example of one of the collocational misuses identified in Taiwanese learners’ blank-filling tests, which was fell the exam, instead of failed the exam Meanwhile, approximation is a strategy used by the learners to produce incorrect words or phrases that bear similar semantic features with the correct ones For example, learners produced attend my goal, instead of attain my goal (Liu, 1999) The inability to produce collocations correctly or use appropriate collocations can make non-native speakers sound unnatural and even unintelligible to speakers of the target language Thus, a high level of collocational competence is crucial if learners aim for advanced proficiency According to Lewis (1997) and Thornbury (2002), the Lexical Approach should be implemented in the EFL classroom and the centrality on lexis can be included in the curriculum to effectively provide learners with sufficient knowledge and skills related to vocabulary in general, and

23 collocations more specifically Another point to take into account is the use of authentic language obtained from concordances, authentic materials and exercises to improve lexis skills This can result in learners getting to know a lexical item more intimately through acquiring the collocates of the item

2.4.2 Textbook input and recycling factors

Textbooks are a valuable resource for learners, providing essential input for language acquisition The frequency and variety of collocations within a textbook can greatly impact its effectiveness Therefore, it is crucial to evaluate textbooks based on the usefulness of the collocations they present, ensuring that these collocations are commonly used by native speakers and are adequately recycled throughout the material By investigating these factors, educators can select textbooks that provide optimal support for students' English language mastery.

Research suggests that ELT textbooks' provision of lexical phrases, including collocations, idioms, and phrasal verbs, is subpar Koprowski (2005) found that useful collocations were often omitted, while Vilkaite (2016) showed that formulaic sequences account for only 41% of English, with collocations being underrepresented Similarly, Koya (2004) reported limited coverage of collocations in Japanese ELT materials These findings indicate a lack of emphasis on collocations in textbooks, hindering the development of native-like proficiency in non-native learners Therefore, collocations merit attention, and their inclusion in textbooks is crucial for learners to acquire and strengthen their collocational knowledge.

Not only is the collocation use in the materials important, but the extent to which each collocation occurs in the material also needs paying attention to, since

24 in many foreign language teaching and learning contexts, textbook is the primary source of target language exposure, which means that the collocational knowledge contained in the textbook can be crucial to non-native learners If students are sensitive to the frequency of occurrence of collocations in their textbooks, it is possible to expect them to respond to collocations derived from their textbooks faster and more accurately than novel sequences of words which do not appear in their books Empirical studies suggest that the recycling of each lexical item is more important to EFL learners than to ESL learners (Kojic-Sabo & Lightbrown, 1999) and the positive impact of various encounters of those items on learners’ vocabulary acquisition has been proven (Webb et al., 2013) In the study conducted by Kojic-Sabo and Lightbrown (1999) to investigate vocabulary learning strategies, the results suggest that EFL learners need to make more efforts than ESL learners to practice new English words because of the limited exposure to English outside the classroom

For effective language acquisition, EFL learners must engage in regular vocabulary activities to reinforce word retention and accurate usage The concept of lexical item recycling is crucial in this regard, as it emphasizes the need for repeated exposure to vocabulary According to certain models, only frequently encountered items are stored in long-term memory (Wray, 2002), while alternative theories suggest that each exposure to a word contributes to its entrenchment and processing (Bybee) Thus, ongoing vocabulary review is essential for EFL learners to develop a robust and usable vocabulary.

& Hopper, 2001) Therefore, it is necessary for a learner to encounter one lexical item from five to sixteen times or even more in order to acquire it (Nation, 1990) Meanwhile, according to Peters (2016), a collocation should be recycled at least three times so that learners can begin acquiring it However, in the study conducted by Koya (2004) to investigate the collocation input provided in some Japanese ELT textbooks, it was found that more than half of the collocations introduced in the textbooks are repeated only once This may raise a question about whether such low frequency of occurrences in the textbooks can successfully facilitate learners’ acquisition of the collocations, because the frequency with which units of language occur is a driving force behind chunking processes and, all else being equal, each

25 exposure to a sequence of words or sounds will affect its subsequent processing (Bybee & Hopper, 2001)

The conceptual framework of the present study

This study employs key theories, exploring seven types of lexical collocations with defined error codes Correct and erroneous collocation usage is analyzed to determine the relationship between learners' collocation use and their proficiency levels (Figure 2.1).

The conceptual framework of the present study

29 This chapter discussed the approaches based on which lexical collocations are identified and classified into seven types Then, the relationship of collocation production and language proficiency, along with collocation errors and possible causes of those errors, was reviewed In the next chapter, the research design, participants, research instruments used in the present study, and the procedures of collecting and analyzing data will be reported

METHODOLOGY

Research questions

As presented in Chapter 1, this study aims to explore (1) how Vietnamese EFL high school learners use lexical collocations in their writing regarding the correct use, common errors and (2) causes of their erroneous use of lexical collocations The following research questions are formulated to guide the study:

1 How do Vietnamese EFL high school learners use lexical collocations in their writing?

1a To what extent are the lexical collocations used accurately by higher-proficiency learners and lower-proficiency learners?

1b What lexical collocation errors are made by higher-proficiency learners and lower-proficiency learners?

2 What are the factors contributing to Vietnamese EFL high school learners’ lexical collocation errors, as perceived by learners and teachers?

This research sought to investigate Vietnamese EFL learners' lexical collocation use, including accuracy, common errors, and influencing factors By comparing learners at different proficiency levels, the study aimed to establish a connection between proficiency and lexical collocation usage The findings provide guidance for teaching and learning collocations in Vietnamese EFL contexts, contributing to improved understanding of vocabulary acquisition and language proficiency.

Research design

This study is a corpus-based investigation in which learner’s written essays were collected so that their collocation use would be analyzed Besides, so as to gain an even deeper understanding of learners’ and teachers’ perceptions about the

31 matter under enquiry, semi-structured interviews were conducted This study, hence, adopted the explanatory mixed-methods approach so that a wider picture of the topic under investigation would be painted According to Riazi and Candlin (2014), the primary purpose of adopting the mixed-methods approach is to provide triangulation which seeks corroboration among the results obtained using different methods, leading to the elaboration of the results This is the complementary nature of mixed-methods research

There are two types of data collected in this present study The correct answers, incorrect answers in the Collocation Test, and the correct and incorrect collocations identified in the learners’ essays made up the quantitative data about learners’ use of lexical collocations After the quantitative data had been collected, interviews were conducted and qualitative data were obtained One important feature of this study is that the learners were categorized into two level groups of lower proficiency and higher proficiency for comparisons The following subsections demonstrate information about the participants, the research instruments and procedures of data collection and data analysis.

Participants

One class of tenth graders in a public high school in Ho Chi Minh City consisting of 55 students was chosen as this class satisfied a number of criteria for the study These criteria include the number of participants, the level of language proficiency, the use of the new EFL textbook, and the exposure to argumentative essay writing Learners in this class are currently using the new English textbook titled Friends Global 10 (Vu et al., 2022), published under the obligation of the new National English Curriculum issued by the Ministry of Education and Training in Vietnam (MOET, 2018) These learners’ exposure to English is approximately seven to nine years, and their level of language proficiency ranges from lower intermediate to higher intermediate This class has seven periods of English (out of 40 periods) per week, with two extra periods with a foreign teacher, about writing and speaking skills for the International English Language Testing System (IELTS)

To conduct the study, the researchers sought approval from the English teacher and school principal The teacher's agreement allowed the researchers to proceed Before data collection, the researchers visited the site and provided learners with a detailed explanation of the study The teacher offered incentives for participation, such as bonus points for completing both tests with satisfactory performance The researchers informed the teacher about bonus points after grading the papers.

Obtaining informed consent is crucial in research involving human subjects Prior to data collection, all 55 students in the designated class received consent forms outlining the study's purpose and procedures Students thoroughly reviewed and signed these forms, indicating their voluntary participation in the research By collecting these signed consent forms before initiating data collection, researchers ensured compliance with ethical guidelines and protected participants' rights.

54 students were present on that day, and hence, only 54 consent forms were signed and collected This also means that the number of participants in this study are 54 learners The English teacher participated as the supervisor observing the learners while they completed the two tests

During the semi-structured interview phase, nine learners agreed to participate To gather additional perspectives, the researcher contacted high school English teachers using convenience sampling As the initial teacher declined due to scheduling conflicts, three teachers from three distinct high schools in Ho Chi Minh City ultimately agreed to participate in the interviews (details are provided in Table 3.1).

Demographics of the teachers participating in the interviews

Teacher #1  3 years of teaching experience

 using Friends Global textbook students of grade 10 and 11 lower-intermediate to higher- intermediate

Teacher #2  2 years of teaching experience

 using Friends Global textbook students of grade 10 lower- to higher-intermediate

Teacher #3  3 years of teaching experience

 using Friends Global textbook students of grade 10 and 11 lower-intermediate to intermediate

All of the interviews with the students and teachers were conducted via Google Meet and the procedure is described in Section 3.5.

Research instruments

There are three main research instruments in this study, which are the Collocation Test, the Writing Test and semi-structured interviews The Collocation Test includes a translation task and a sentence transformation task The Writing Test requires the participants to write an argumentative essay The semi-interviews were conducted to collect qualitative data Below is the description of each instrument and the modifications made to each after they were piloted

In the present study, the first research instrument, called the Collocation Test, is the test of controlled production of lexical collocations The use of tasks with two levels of controlled production is required to help the researcher explore whether learners’ L1 influences their production of collocations The first task in this instrument is a translation task with 28 items For this task, learners are required to read a sentence written in English and the Vietnamese translation of that sentence In the English version, the target collocation is hidden and the learners are asked to read the bold and underlined part in the Vietnamese translation to find the lexical collocation that is suitable to fill in the blank The

34 frequency of each collocation in the Collocation Test was checked against COCA to ensure its conventionality The Collocation Test is presented in Appendix D

Before the actual study was conducted, the researcher had carried out a pilot study In the pilot study, the places for participant’s answers were designed to be located below the questions This became a challenge for the researcher while grading the answers because the blank space was quite narrow and insufficient for the participants to write in, or to rewrite another answer if any To enhance the ease of data collection, a separate answer sheet for each test was provided so as to avoid the similar problem as in the pilot one

This study investigates seven types of lexical collocations, each represented by four items To assess the memorability of collocations used in the Friends Global 10 textbook, the researcher selected two items per type from the textbook Additionally, one item was taken from the English 9 textbook, previously used by participants, and the final item was derived from the Academic Collocation List, which contains collocations commonly used by native speakers.

Demographics of collocations in the translation task

Type of collocations Friends Global 10 English 9 ACL Total

35 The second task in the Collocation Test is a sentence transformation task Learners are required to read a sentence and use the suggested word between the parentheses to form an appropriate lexical collocation so that the meaning of the new sentence is similar to the first sentence There are seven items in this task and each item tests one type of lexical collocation All of the target lexical collocations appearing in this task were taken from the book Friends Global 10, as one of the aims of the study is to confirm whether the frequency of occurrences of the collocations in the textbook contributes to the acquisition and production of those collocations After 40 minutes, the researcher asked the teacher to collect all of the answer sheets of the Collocation Test before distributing the answer sheets for the Writing Test to ensure that their use of lexical collocations in the Writing Test would not be influenced by the collocations they encountered previously

The underlying reason for the choice of the translation task and sentence transformation task is that in previous interlanguage studies, in order to determine whether learners can successfully produce the target collocations whose meanings can be understood from the translation in their mother tongue, translation tasks and cloze tasks were given In Bahns and Eldaw’s (1993) study to investigate German learners’ collocation acquisition, a translation task and a cloze task were provided for the participants The chosen collocations were translated into German and embedded in fifteen German sentences Learners were supposed to provide the English translations of those German sentences, and the expected English collocations would be contained in those translations Regarding the cloze task, in an English sentence, the verb collocate was missing and only the noun node was presented The learners were supposed to comprehend the context embedded in the sentence to figure out the missing verb collocate and fill in the blank By doing this, the researchers were able to investigate those participants’ use of collocations Also, this would test the freedom to express the meaning of the collocational combinations in English given the fact that they did not know the collocation In terms of the scoring criteria, the collocations are considered acceptable if they are semantically and idiomatically correct, and unacceptable when they are semantically and idiomatically incorrect Moreover, places that were left blank

36 were deemed unacceptable Similarly, in the present study, if learners do not attempt to provide an answer, that is regarded as unacceptable

In order to build a corpus of learners’ production, the researcher designed a test of free production and learners are required to write an argumentative essay, from 180 to 250 words, to one of the two task questions (see Appendix E) The researcher carefully examined all of the possible topics for the writing test so that the collocations that appeared in the Collocation Test cannot be taken advantage of for the essay, which increased the validity of this test

The standardized marking criteria for IELTS writing, or the IELTS Writing Band Descriptors (see Appendix F), was utilized, because the majority of the participants are taking IELTS classes outside the classroom, and every week, at the public high school, where the study was conducted, all classes have two periods with a foreign teacher who teaches IELTS writing and speaking Each of the essays was graded by the researcher and an IELTS trainer who has taught IELTS for three years Neither the researcher nor the IELTS trainer teaches at the same school where the learners currently study Although writing skill is only one of the four language skills, it is usually considered the most challenging skill to master for EFL learners Yet, it contributes greatly to not only learners’ academic journey but also their future careers In addition, this study only focuses on how EFL learners produce lexical collocations in their written productive performance, so the researcher only considered the participants’ proficiency level, referred to as either higher-proficiency (HP) or lower-proficiency (LP) levels in the present study, based on their written productive performance

Students with essays that received a score of 6.0 and higher were categorized into the higher-proficiency level group, while students with writing works lower than 6.0 belonged to the lower-proficiency level group According to www.ielts.org, an IELTS score of 5.5-6.5 is equivalent to B2 level of the Common

European Framework of Reference for Languages (CEFR), which can be interpreted that when a learner receives a score between 5.5 and 6.0, this learner’s level of proficiency belongs to the lower tier of B2, and from 6.0 to 6.5 to the

37 higher tier of B2 As a result, the score of 6.0 is an ideal cut-off point for this study All essays were anonymous and coded with a number The feedback related to content, cohesion and coherence, vocabulary and grammar was provided in each text (see Appendix G) Afterwards, the scores graded by the IELTS teacher and the researcher were compared and discussed There were no essays graded lower than 5.5 or higher than 7.0, and there was no difference in the decisions of the researcher and the IELTS trainer during the grouping process (see Appendix H)

The two topics for the Writing Test were inspired by IELTS Writing Task

2 questions Learners were supposed to write this essay in forty minutes (within a class period) without assistance from the teacher In order to give learners some ideas for the essays, a suggested outline for each task question was provided in the test paper as a reference However, learners were not obliged to strictly follow the suggested outline In addition, learners were encouraged to use as many lexical collocations in their essays as possible The two task questions have neither appeared in their English textbook nor been taught in the IELTS periods with the foreign teacher This ensured that learners would not write the essays from their memory The English teacher in charge of this class promised to give the students bonus points for their participation in the study

3.4.3 Questions for semi-structured interviews

After the quantitative data were collected and analyzed, the researcher invited nine students and three teachers to twelve semi-structured interviews When the analysis of the quantitative data was completed, the researcher was able to identify possible areas, from which questions could be formulated The researcher prepared a list of ten questions for the learners and ten questions for the teachers Probing questions were required during the interviews to elicit all aspects of the participants’ answers, providing a deep understanding of the matter

To prepare for the interviews, the researcher established an initial framework of concepts based on themes identified during the pilot study This framework was further refined by incorporating insights from Nguyen's (2024) research on teachers' perspectives on teaching collocations in the Vietnamese context This cross-referencing of findings ensured that the interview questions effectively addressed the key themes and nuances related to the study's objectives.

38 study and Nguyen’s (2024) study, and the matter under inquiry are similar to the present study

Maps of concepts for question formulation

The lists of questions asked in the semi-interviews in the present study can be found in Appendix I, and an extract of an interview with a learner and a teacher is presented in Appendix J.

Data collection procedure

Before conducting the study, the researcher presented all of the key information of the study to the English teacher of the participants The files included the title of the study both in English and Vietnamese, the researcher’s personal information and contact details, the timeline and detailed procedures of the study, along with guidelines for distributing the test papers, collecting the answer sheets and informing students about the study’s purposes After the teacher raised a number of questions for clarification, the researcher arranged a time for the participants to sit the two tests 54 answer sheets of the Collocation Test and

54 essays were collected within two periods in one day

39 After collecting all of the answer sheets from 54 participants, the researcher scanned all of the pages of the answer sheets for storage For the Collocation Test, the researcher annotated directly on the answer sheet In order to grade learners’ Collocation Test answer sheets, the researcher developed a two-point scale If the learner provides a correct collocation for each test item, the answer receives one point, regardless of grammatical errors and minor spelling errors If the learner provides a combination that is contextually and semantically incorrect according to the codes for identifying collocation errors in this study, or does not match the Vietnamese translation (in Task 1), or fails to form a meaningful and similar sentence to the provided sentence (in Task 2), the answer receives zero point If no collocation is provided in the blank, no point is given

The researcher manually graded answer sheets and recorded scores in Microsoft Excel for a Cronbach's alpha test to assess internal data consistency This test, crucial in research and education (Cortina, 1993), evaluates not only attitudes and affections but also learners' knowledge and understanding (Taber, 2018) After grading all 54 answer sheets, the researcher conducted the Cronbach's alpha test using IBM SPSS Statistics (Version 20) As the Collocation Test assessed participants' knowledge of seven lexical collocation categories, all 35 test items tested the same construct The alpha value for all 35 items was 720 (N = 35), indicating high reliability and internal consistency according to Cortina (1993).

For the Writing Test, all of the essays were typed into txt files, so that the researcher could use software AntConc 3.5.9 (Anthony, 2020) to extract the concordances containing the node word and the collocates with greater convenience Although the participants provided their real names on the answer

40 sheets, their essays remained anonymous when digitized The researcher coded each essay using a combination of letters and numbers For example, for the code

HP_S22, HP means that the learner belongs to the higher-proficiency level group and S22 means that this is the twenty-second learner in this group The procedure of counting the correct collocations and erroneous collocations to obtain scores for a comparison between two learner groups is described in Section 3.6.1

Regarding the semi-structured interviews, nine learners that agreed to take part in the interviews gave their telephone numbers to their English teacher The researcher contacted the teacher and set up an interview schedule for them The researcher texted the English teacher to inform the time and the Google Meet link to the interview for each participant Three hours before the interview, the researcher texted the English teacher to forward the schedule message to each participant again to ensure that the participant was well aware of the schedule Another consent form for their participation in the semi-structured interview was sent to the English teacher’s mail, and the form was printed out, signed by the participants and collected by the teacher, then collected by the researcher Each interview took place via Google Meet, and lasted approximately 30 minutes All interviews took place within one day Overall, throughout the process of collecting both the quantitative data and data from the interviews, the English teacher’s supervision was maintained to ensure the participants’ rights and the research study’s ethics About the interviews with the three English teachers, the interviews took place via Google Meet and the schedule was informed to each teacher a day earlier.

Data analysis procedure

In this study, besides an analysis of correct collocations, a five-stage error analysis model adapted from Gass and Selinker (2008) was followed to analyze erroneous collocations According to Gass and Linker (2008), a systematic error analysis model consists of six main steps, namely, (1) data collection, (2) error identification, (3) error classification, (4) error quantification, (5) error source analysis, and (6) remediation In the present study, the researcher adapted Gass and

41 Linker’s (2008) model with a minor adjustment The step of remediating the errors would be combined with the step of classifying the error as this would be more time-saving The procedure to identify erroneous collocations was adapted from Cao (2018) To check for the frequency of a collocation, COCA is used as the main reference corpus

For the Collocation Test, the researcher directly annotated on participants’ answer sheets to grade their answers The researcher had prepared an answer key for the Collocation Test If the learner provides the correct lexical collocation as the answer key, the researcher graded one point, regardless of any grammatical errors or minor spelling errors There was the possibility of alternative answers besides the target collocations in the answer key It was possible that the learners may produce other collocations which may be both correct and contextually appropriate For alternative collocations which were not in the original answer key, the researcher first checked their meanings using the online Cambridge Learner’s Dictionary, before checking against COCA for the frequency of occurrences and concordances If the alternative collocation or the combination extracted from the essays does not meet the normalized frequency threshold, the researcher put it into a list for native speakers’ consultancy

For the Writing Test, after the essays were collected, they were stored as txt files so that they could be analyzed using the software AntConc 3.5.9 (Anthony, 2020) In this software, the files of both groups were opened The researcher retrieved the list of all the words that have appeared in all essays, along with their frequencies, using the tool Word List in AntConc 3.5.9 (Anthony, 2020) (see Figure 3.2)

The interface of the tool Word List in AntConc 3.5.9

The researcher chose to look into nouns, verbs, adjectives and adverbs in the list, as the researcher believed that this approach saved time and ensured that all potential combinations would be “sifted” and found Unlike in the study conducted by Durrant and Schmitt (2009), which attempted to extract combinations that do not contain an element such as (1) proper nouns (names of objects or people, identified by capitalization), (2) acronyms or initialisms, (3) semi-determiners (such as same, other, last, next, etc.) and (4) numbers/ordinals, in this study, combinations that contain an acronym, or a proper noun were still identified For example, the combination high GDP would not be identified in

Durrant and Schmitt’s (2009) study because it contains an initialism, but in this study, this combination would be identified and extracted because the raw frequency of GDP in COCA is 10,557, which means that GDP is frequently used

Moreover, pronouns and possessives found in the combinations would make the combinations not considered for the study by Durrant and Schmitt (2009) The reason was because they investigated only adjective + noun and noun

+ noun collocations, and to keep the calculation of association measures straightforward, only directly adjacent pairs of words were identified However, in this present study, the researcher also looked at verb + noun, noun + verb (action), adverb + verb or verb + adverb collocations, so the appearance of pronouns or possessives within the combinations is understandable Therefore, in this study,

43 combinations which contained pronouns and possessives were extracted and analyzed

At this stage, the extracted strings of words were considered combinations Regarding the span of the combinations, only directly adjacent pairs or combinations within the span of four to five words are considered In the previous body of research into this issue, spans of ±2 (Clear, 1993), ±3 (Ackermann & Chen, 2013), ±4 (Sinclair et al., 2003) and ±5 (Stuart & Trellis, 2006) have been utilized but the span of ±4 is claimed to be the most ideal for identifying collocations (Jones

In the study, different spans were applied to various lexical collocation categories to ensure the inclusion of relevant data For six specific categories, a span of ±4 was implemented However, for noun 1 of noun 2 collocations, known as "positionally variable collocations," a wider span of ±5 was deemed more appropriate This modified span ensured that valuable data was not overlooked in this particular category.

The interface of the tool Collocates

To initiate error analysis, data collection is crucial This involves identifying word combinations, classified as candidate collocations, from learners' essays The methodology employed in this process is adapted from Cao's (2018) study, which meticulously categorizes lexical collocations into seven distinct types.

In cases of more than one adjective modifying a noun, combinations will be extracted separately Gerunds and present participles acting as adjectives need distinguishing, and only combinations that contain present participles acting as adjectives modifying nouns will be extracted

There are many other structures that may be regarded as the (LC2) pattern, rather than solely verb + object structure, according to Nesselhauf (2005) Other structures such as verb + complement (e.g go to school) and verb + object + complement (e.g take something into consideration) are some syntactic structures of verb-noun collocations Regarding cases of finite or non-finite clauses acting as the object of a verb, as in example (1), the verb-noun combinations (suggested as the verb, and taking photos on the beach as the noun) will not be extracted

(1) Julie suggested taking photos on the beach

However, any verb + noun combinations lying in the finite or non-finite clauses will be extracted, for example, taking photos as in example (1) In the case of a non-finite clause acting as the object in a sentence, as in example (2), besides the verb-noun combinations being extracted, any noun (implicit subject) + verb combination may be extracted as well As in example (2), children is the implicit subject of the verb break, so the combination children + break and the combination break + window will also be extracted

(2) The children denied breaking the window

As in example (3), more than one verb refers to a noun, so separate combinations will be extracted prepare + meal and cook + meal are extracted as separate combinations, as both verbs prepare and cook are coordinated with the noun meal

(3) Her mother quickly prepared and cooked the meal

FINDINGS AND DISCUSSION

Learners’ correct use of lexical collocations

The results of the analysis of correct answers in the Collocation Test, including the translation task (28 items) and the sentence transformation task (7 items), and the correct collocations extracted from 54 argumentative essays in the Writing Test are reported in the following subsections

In the translation task, there are four collocation items for each collocation category Because the higher-proficiency level group has 29 learners, the total number of collocation items in the translation task is 116 items per collocation category (NHP = 116), whereas for the lower-proficiency group, that is 100 items per category (NLP = 100) The total number of items in the Collocation Test is 216 items (NT = 216) The figures for correct (LC3) (86.1%), (LC5) (83.8%) and (LC7) collocations (83.8%) were the highest in both groups The numbers of correct collocations of all seven categories produced by higher-proficiency learners were higher than those produced by lower-proficiency learners (see Table 4.1)

Correct collocations (CC) in the translation task

In terms of the sentence transformation task, there is one collocation item per category For the higher-proficiency level group, the total number of collocation items in the sentence transformation task is 29 items per collocation category (NHP = 29), whereas for the lower-proficiency group, that is 25 items per category (NLP = 25) Higher-proficiency learners produced a greater number of correct collocations of six lexical collocation categories than lower-proficiency learners, except for the (LC6) category Among seven types of lexical collocations, learners of both groups produced the most significant figures for correct collocations of (LC1) (55.6%), (LC3) (83.3%) and (LC6) collocations (66.7%) The number of correct (LC7) collocations in this task was the lowest (18.5%), which means this category caused all learners most difficulties (see Table 4.2)

Correct collocations (CC) in the sentence transformation task

CC % CC No of CC % CC No of

The results of Mann-Whitney U tests indicate that 29 learners in the higher- proficiency level group produced significantly more accurate collocations in the Collocation Test (U = 525.00, P = 005), compared to 25 learners in the lower- proficiency level group Regarding each lexical collocation category, 29 higher- proficiency learners significantly produced more correct (LC2) collocations (U =

220.500, P = 010), (LC4) collocations (U = 243.000, P = 032) and (LC7) collocations (U = 256.500, P = 049) in the Collocation Test, compared to 25 lower-proficiency learners (see Table 4.3)

Correct use of lexical collocations in the Collocation Test

Mean Rank Sum of Ranks U value P value

This also means that for the four categories (LC1), (LC3), (LC5) and (LC6), higher- proficiency learners did not significantly produce more accurate collocations, compared to lower-proficiency learners Following this, how accurately learners produced lexical collocations in the essays is reported in the next subsection

Higher-proficiency learners demonstrated superior writing skills, producing essays with significantly more words (281.3 vs 240.4) and a richer vocabulary, evidenced by a higher frequency of lexical collocations (35.4 vs 18.6).

Description of essays in the Writing Test

Average tokens per essay Higher-proficiency learners’ argumentative essays 29 8158 281.3 1027 35.4

Lower-proficiency learners’ argumentative essays 25 6010 240.4 465 18.6

In total, learners from both groups produced 1135 correct lexical collocations out of 1492 lexical collocations extracted from 54 essays Furthermore, 29 higher-proficiency learners (M = 35.414, SD = 6.827) compared to 25 lower-proficiency learners (M = 18.440, SD = 3.513) produced significantly more collocation tokens in the Writing Test (t(52) = 11.208, p = 000) (see Table 4.5)

Total collocation tokens in learners’ essays

Error Mean t value df Sig

The learners produced great numbers of correct collocations of all categories, compared with the total numbers of collocations of each category Higher-proficiency learners produced large numbers of correct (LC1) (84.9%), (LC4) (88.2%) and (LC6) collocations (84.8%) Meanwhile, for lower-proficiency learners, the figure for correct (LC3) collocations (93.5%) was the most noticeable (see Table 4.6)

Correct collocations (CC) from learners’ essays

Higher proficiency Lower proficiency Total

(LC4) adv + adj 34 30 88.2 16 14 87.5 50 44 88.0 (LC5) adv + v 78 51 65.4 60 49 81.7 138 100 72.5

While analyzing the extracted combinations by checking for their frequency using COCA, the researcher found 141 combinations whose frequency was below

50 occurrences Six combinations were considered Questionable (?), eight combinations were regarded as Unacceptable (-), and the remaining 126 combinations were considered correct collocations (see Appendix M)

The results of Mann-Whitney U tests are that higher-proficiency learners produced significantly more correct collocations (U = 3.000, P = 000) than lower- proficiency learners In addition, regarding the accuracy rate in each essay (AR), or the number of correct collocations out of the total collocation tokens produced in each essay, higher-proficiency learners had significantly higher accuracy rates of lexical collocation production in essays (U = 192.000, P = 003) than lower- proficiency learners’ Regarding each lexical collocation category, the result is that higher-proficiency learners significantly produced more correct (LC1) (U = 325.000, P = 000), (LC2) (U = 376.500, P = 000) and (LC4) collocations (U = 550.000, P = 005) than lower-proficiency learners (see Table 4.7)

Correct use of lexical collocations in the Writing Test

Mean Rank Sum of Ranks U value P value

Among the combinations whose frequency of occurrences in COCA did not meet the threshold of 50 occurrences, although a number of combinations were ultimately regarded as Acceptable (+), the informants also gave some suggestions for improvement Although these combinations could be considered collocations, there are other ways of generating the same idea to make it sound more natural and native-like (see Appendix N)

To conclude, of the total 1512 items in the translation task and 378 items in the sentence transformation task in the Collocation Test, there were 1348 correct answers, accounting for 71.3% Of the total 1492 collocation tokens extracted from

54 essays, there were 1135 correct lexical collocations, accounting for 76.1% (see Table 4.8)

Correct collocations collected from two research instruments

Note Percentages out of the total number of collocation items are presented in the parentheses

Higher-proficiency learners produced significantly more accurate lexical collocations than lower-proficiency learners in both the Collocation Test and the Writing Test For the Collocation Test, learners from both groups produced large numbers of correct (LC3), (LC5) and (LC7) collocations in the translation task, (LC1) (LC3) and (LC6) in the sentence transformation task, suggesting that these five lexical collocation categories did not pose a great challenge to the learners in controlled production Regarding each separate category of lexical collocations in this test, higher-proficiency learners only produced a significantly higher number of correct (LC2), (LC4) and (LC7) collocations than lower-proficiency learners

For the Writing Test, learners from both groups attempted to produce great numbers of correct (LC1) and (LC2) collocations in their essays Higher- proficiency learners produced significantly more correct collocations, even when compared with the total collocation tokens in each essay, than lower-proficiency learners Moreover, correct (LC1), (LC2) and (LC4) collocations produced by higher-proficiency learners were significantly higher than those produced by lower-proficiency learners

Learners’ common collocation errors

The results of the analysis of incorrect answers in the Collocation Test and the erroneous lexical collocations extracted from 54 argumentative essays are reported in the following subsections

In the translation task, incorrect (LC2) (27.3%), (LC4) (42.6%) and (LC6) collocations (25.9%) accounted for the highest figures, while incorrect (LC3) collocations (13.9%) represented the lowest figure (see Table 4.9)

Erroneous collocations (EC) in the translation task

This indicates that (LC2), (LC4) and (LC6) collocations were the most problematic to all learners, whereas they had least difficulty finding the most appropriate (LC1),

(LC3), (LC5) and (LC7) collocations to match the Vietnamese translation

Regarding the sentence transformation task, learners from both groups produced the least numbers of erroneous (LC6) (33.3%) and (LC3) (16.7%) collocations Interestingly, learners from the higher-proficiency group had the tendency to produce more erroneous (LC6) collocations than learners from the lower-proficiency group (see Table 4.10)

Erroneous collocations (EC) in the sentence transformation task

Overall, learners with higher proficiency (n=29) demonstrated significantly greater accuracy in collocation usage on the Collocation Test compared to their lower-proficiency counterparts (n=25) Conversely, the lower-proficiency group exhibited a significantly higher incidence of erroneous collocations.

(LC2), (LC4) and (LC7) collocations in the Collocation Test, compared to 25 lower-proficiency learners

Although the error codes are presented in Appendix B, to clarify the collocation errors of each category, the researcher briefly described the types of errors in the result tables The first error type is the wrong choice of either component in a collocation (w n/v/adj/adv/n1/n2) For example, in an (LC1) adjective + noun collocation, the wrong choice of the adjective is briefly described as w adj, next to the code LC1.1 The second error type is that the existent collocation is used contextually inappropriate (inappr.) The third error type is the non-existent collocation (n-ex.)

Higher-proficiency learners made 27 erroneous (LC1) collocations in the

Collocation Test (NHP = 27), while thre were 35 erroneous (LC1) collocations produced by lower-proficiency learners (NLP = 35) Among a total of 62 erroneous

(LC1) collocations (NT = 62), 22 erroneous collocations consist of a wrong noun (35.5%), and 28 consist of a wrong adjective (45.2%) In addition, while lower- proficiency learners had the tendency to provide wrong nouns, and non-existent combinations, higher-proficiency learners were more likely to struggle with providing appropriate adjectives (see Table 4.11)

(LC1) collocation errors in the Collocation Test

For example, instead of writing natural habitat for the translation of môi trường sống tự nhiên, learners wrote natural environment One learner provided the answer heredity disorder for the translation rối loạn di truyền heredity is a noun, but the part of speech of this component should be an adjective Therefore, this combination was ruled as incorrect Also, combination wrong genetic, which is a non-existent combination, was also provided, but it was identified as an erroneous collocation In the sentence transformation task, since the noun base impact was provided, learners only made the wrong choice of the adjective, provided a collocation that is contextually inappropriate, or wrote a non-existent combination To match the negative connotation of the verb damages, learners were supposed to provide an adjective with a negative connotation However, some learners wrote great impact, big impact, or strong impact, producing a sentence whose meaning is not similar to the given sentence Consequently, these collocations were regarded as existent collocations but not contextually appropriate

Higher-proficiency learners made 35 erroneous (LC2) collocations in the

Collocation Test (NHP = 35), while there were 47 erroneous (LC2) collocations produced by lower-proficiency learners (NLP = 47) The most common errors were the wrong choice of the verb (53.6%) and wrong choice of the noun (25.6%) While higher-proficiency learners were more likely to provide wrong verbs, lower- proficiency learners had more problems with nouns (see Table 4.12)

(LC2) collocation errors in the Collocation Test

For example, some learners wrote give an impression, cause an impression, or have an impression for the translation gây ấn tượng, while the correct answer is make an impression For the translation phạm tội, few learners made the wrong choice of the verb and wrote cause crimes or do crimes, while the correct collocation should be commit crimes Also, few learners made the wrong choice of the noun and wrote commit a sin While commit a sin (freq 189) is considered a collocation, sin is not contextually appropriate because sin means an offence that breaks the religious or moral laws, but the item refers to a general context of an illegal activity according to legislation, so it was regarded as incorrect

In the sentence transformation task, a number of learners of both groups provided the wrong verb To give the sentence the meaning of having the expected outcome, learners were required to write match the expectation or meet the expectation, but the majority of learners wrote have expectation Although have an expectation is a collocation (freq 655), it is not contextually appropriate

Higher-proficiency learners produced 19 erroneous (LC3) collocations in the Collocation Test (NHP = 19), while there were 30 erroneous (LC3) collocations produced by lower-proficiency learners (NLP = 30) The most common errors were the wrong choice of the second noun (44.9%) and the non-existent combination (26.5%) Higher-proficiency learners were more likely to provide inappropriate first nouns, whereas lower-proficiency learners had the tendency to provide inappropriate second nouns, or non-existent combinations (see Table 4.13)

(LC3) collocation errors in the Collocation Test

For instance, for the translation phòng chờ khởi hành, which is the departure lounge, few learners wrote departure room, which does not exist, while some others wrote waiting room (freq 2869), which is a frequently used collocation but not contextually appropriate for the translation, and passenger room (freq 12), which is existent, but not semantically appropriate

Higher proficiency learners produced a significant number of erroneous collocations in the Collocation Test, with 59 errors recorded In contrast, lower proficiency learners made 30 similar errors The majority of these errors involved the incorrect use of adverbs (37.5%) and non-existent collocations (50.8%).

(LC4) collocation errors in the Collocation Test

For instance, for the translation thể chất tốt, whose English collocation should be physically fit, few lower-proficiency learners provided well physical, or physically conditional, while some higher-proficiency learners wrote strongly physical In the sentence transformation task, for the item physically demanding, although some learners wrote highly demanding, which is a correct collocation

(freq 53), it does not demonstrate the requirement for physical strength, so it was not considered appropriate

Higher-proficiency learners produced 33 erroneous (LC5) collocations in the Collocation Test (NHP = 35), while there were 37 erroneous (LC5) collocations produced by lower-proficiency learners (NLP = 37) The most common error is the wrong choice of the adverb (82.9%) Lower-proficiency learners had difficulty providing appropriate verbs, while higher-proficiency learners had more difficulty finding appropriate adverbs (see Table 4.15)

(LC5) collocation errors in the Collocation Test

For the translation biến mất bất thình lình, some lower-proficiency learners wrote disappear unfortunately, or disappear immediately, while disappear suddenly (freq 110) is the correct collocation For learners from the higher- proficiency group, the (LC5) collocation in the sentence transformation task caused them problems Few learners from this group wrote well-protected, which is an adjective modifying a noun, so this combination is not a (LC5) collocation, whereas some other learners wrote well protected (freq 279) without the hyphen symbol, which is regarded as a (LC5) collocation Some interesting answers of learners from both groups to this item were found, such as legally protected (freq 38), which is not contextually appropriate because no element of laws or legislation is mentioned in the original sentence

4.2.1.6 (LC6) noun (subject) + verb (action)

Higher-proficiency learners produced 41 erroneous (LC6) collocations in the Collocation Test (NHP = 41), while there were 33 erroneous (LC6) collocations produced by lower-proficiency learners (NLP = 33) The most common error is the wrong choice of the noun (subject) (55.4%) (see Table 4.16)

(LC6) collocation errors in the Collocation Test

Causes of learners’ collocation errors

4.3.1 Perceptions of lexical collocation challenges

When analyzing the interviews of learners and teachers, the researcher identified two themes about the difficulties learners encountered: (1) learner’s reluctance to use collocations in writing, and (2) categories of lexical collocations that learners found most problematic Understanding how learners perceive the challenges related to lexical collocations and problems teachers may encounter can result in valuable insights into the causes of learners’ erroneous collocation use

Most lower-proficiency student interviewees claimed that they do not often use collocations in their essays, because they are not confident with their collocation knowledge Specifically, they believe that they do not have the habit of generating an idea using a rather fixed combination of words When asked the same question, most of the higher-proficiency student interviewees claimed that they often attempt to use collocations in their essays because they consider them natural language used by native speakers However, sometimes learners write without paying attention to collocations, because they forget

Sometimes, I cannot recall a certain collocation for an idea, so I generate the idea into a sentence without using an appropriate collocation There are days on which I try to use some good collocations that I have learnt in my essays (Student #HP4)

Moreover, they have the tendency to rely on familiar collocations, because they are comfortable using those without pressuring themselves to use an unfamiliar one

Limited vocabulary often leads students to rely on familiar collocations, despite having learned numerous effective expressions Students may resort to using collocations they recall or have used frequently, as unfamiliar ones pose a challenge.

When asked about which type of collocations causing the most trouble, most of the student interviewees shared that they have had difficulty using (LC2) verb + noun, (LC7) noun 1 of noun 2 and (LC5) adverb + verb collocations For (LC2) collocations, some interviewees thought that some common verbs are the

76 collocates of many other nouns, which makes it hard for them to remember all of them, and they may make mistakes

Understanding verb + noun collocations can be challenging for learners, especially when the reasoning behind the combinations is unclear For instance, the collocation "take medicine" is commonly used even though medicine is typically consumed with water Additionally, verbs such as "make," "take," and "do" have numerous collocations, making it difficult to memorize and recall the correct pairings.

There are so many adverbs and sometimes I fail to distinguish some of them because they may seem familiar Also, similar adverbs are not used similarly, and that’s what causes difficulty to me For instance, strongly and heavily may bear some similar senses, and they are the collocates of different verbs (Student #LP6)

For (LC7) collocations, 33% of the student interviewees told the researcher that they rarely encountered collocations of this category Collective nouns, which indicate groups or organizations of items, people or animals, are considered problematic Quantifiers or nouns that indicate quantities are sometimes confusing to the student interviewees

Collocations, or fixed word combinations, can pose challenges for learners due to their idiosyncratic nature A learner expresses their difficulty in recalling and comprehending why certain collocations are used, such as "a school of fish" and "a swarm of bees." They speculate that these expressions are simply conventions used by native speakers.

Teachers gave relatively different answers, when asked about the most challenging type of lexical collocations to learners Teacher #3 believed that (LC6) collocations are the most problematic to her students

My students simply think that as long as the verb carries a meaning, regardless of the action, it can go with any subject They do not have a concept that combinations in the form of (LC6) noun (subject) + verb

(action) are considered collocations too, and they should learn about those, instead of trying to match the subject and a random verb (Teacher #3)

However, Teacher #1 thought that (LC1) adjective + noun, and (LC2) verb + noun collocations pose the most challenge to her pupils, as she shared, “Their mother tongue influences their collocation processing a lot, in my opinion They always have problems with (LC1) and (LC2) collocations because they think these collocations are the same as in Vietnamese.” Agreeing with the other two teachers,

Teacher #2 emphasized that learners encounter problems with all seven types of lexical collocations, to a certain extent

Regarding teachers’ own difficulties related to the process of teaching collocations, Teacher #2 claimed that in EFL classrooms in Vietnam currently, collocations have not received sufficient attention, as single lexical items are still more focused on Meanwhile, Teacher #1 argued that EFL teachers in high schools may still have difficulty adapting the textbook into their own classroom

The nature of the General Education Program issued by the Ministry of Education in 2018 is an open approach, which means that teachers have the freedom to adapt the textbook to their classroom as long as they meet the requirements in the curriculum This has been mentioned and clarified in textbook introduction sessions and training sessions before an academic year, but it seems that some teachers still find it difficult Therefore, more in-depth training sessions related to assessment, class management, teaching methodology innovations … are truly needed (Teacher #1)

The introduction of new ELT textbooks in Vietnamese high schools has presented a significant challenge for EFL teachers Despite introductory sessions and training courses, many teachers continue to face obstacles in effectively utilizing the textbooks This has resulted in difficulties creating engaging lessons and meeting student performance expectations, highlighting the need for further support to ensure the successful implementation of the new materials.

According to the analysis of the erroneous collocations from the Collocation Test, which is controlled production, learners from both groups found

(LC4) collocations most challenging, besides (LC2), (LC5), (LC6) and (LC7) collocations Regarding free production in the Writing Test, both higher- proficiency and lower-proficiency learners struggled with (LC1) and (LC2) collocations This corroborated with what the teacher interviewees shared According to the teachers, all lexical collocation categories posed challenges, especially (LC1) and (LC2) In addition, learners tend to be reluctant to produce lexical collocations in their writing because they are not yet familiar with the concept of collocations, or they tend to rely on familiar ones

Discussion

From the analysis of quantitative data, it was found that higher-proficiency learners produced significantly more correct lexical collocations than lower- proficiency learners in both the Collocation Test and the Writing Test Learners of both levels of proficiency have more problems with (LC1), (LC2) and (LC6) collocations than with the other four categories The negative influence of L1 demonstrated through the strategy of word-by-word translation and the use of synonymy and approximation, the process of teaching collocations, as well as the recycling of collocations in the ELT textbook were found to be the factors influencing learners’ collocation use In this section, two key points that emerged from the findings will be discussed These two points are (1) Vietnamese EFL learners’ use of lexical collocations in writing, and (2) factors influencing Vietnamese EFL learners’ use of lexical collocations

4.4.1 Vietnamese EFL learners’ use of lexical collocations in writing

Results from the Collocation Test data demonstrate that higher-proficiency learners produced significantly more accurate lexical collocations than lower- proficiency learners With regards to the Writing Test, the accuracy rate or the number of correct collocations out of the total collocation tokens produced in each essay of the higher-proficiency group was significantly higher than that of the lower-proficiency group This also means that regarding the total collocations produced in an essay, lower-proficiency learners had lower accuracy rates than higher-proficiency learners This finding corroborates the results in many previous studies in which a strong relationship between learners’ proficiency levels and their production of collocations is determined (Al-Zahrani, 1998; Read & Nation, 2006; Shaw & Weir, 2007; Yan-Ren, 2005) In other words, the more proficient EFL learners are at the target language, the more likely they are to produce accurate lexical collocations

The present study also discovered that higher-proficiency learners produced a significantly higher number of collocation tokens than lower-proficiency learners in their essays In addition, when comparing only the numbers of erroneous collocations, not regarding the total collocation tokens produced in each essay, it was found that higher-proficiency learners produced a significantly higher number of errors than lower-proficiency learners This echoes a finding in Laufer and Waldman’s (2011) study, which indicates that advanced learners and intermediate learners in their study produced significantly more unusual collocations than the basic ones Laufer and Waldman (2011) argue that learners with a higher proficiency level tend to be more confident with their language use, and so they attempt to produce more collocations, which results in the larger number o erroneous collocations

Another interesting finding in the present study is that higher-proficiency and lower-proficiency learners produced large numbers of correct (LC1) adjective

+ noun and (LC2) verb + noun collocations in their essays Simultaneously, these two collocation categories were also the most problematic to learners of both groups For (LC1) collocations, learners tended to provide an inappropriate

86 adjective, while for (LC2) collocations, they were more likely to use an inappropriate verb This corroborates the findings in many previous studies (Bahns

& Eldaw, 1993; Biskup, 1992; Cao, 2018; Nesselhauf, 2003, 2005), which demonstrate that EFL learners have the most problems with (LC1) and (LC2) collocations

(LC2) collocations are believed to pose the greatest challenge to EFL learners because of many reasons The first reason may be the prevalence of this collocation category In any sentence, the pattern of a verb preceding a noun, which is the object, is utilized (Howarth, 1996) This is why EFL learners encounter this category often and they are more likely to make errors Nesselhauf (2005) also discovered that German EFL learners had a lot of difficulty with verb-noun collocations because of their exposure to these on a daily basis Another compelling reason is that there are a number of core verbs, such as get, make, have, do, and take, which can collocate with many different nouns, making it difficult for EFL learners to successfully memorize and acquire In the study by Ringbom (1998), because of the polysemous nature of some high-frequency verbs such as get and make, EFL learners, except for Swedish-speaking learners, underused collocations of the verb make, when compared with the native control corpus

Core verbs collocate with various nouns, but this doesn't imply multiple collocations for a single noun Limited options exist for learners, as exemplified by the verb "take." Despite its wide range of collocations, it offers only "take the medicine" or "swallow the medicine" to express taking a pill for medicinal purposes This contrasts with other collocation categories like adverb + adjective or adverb + verb, where learners have more options The difficulty with core verb collocations stems from synonymy, as similar-meaning verbs may not collocate with the same nouns.

Avoiding collocation mistakes is crucial for EFL learners; however, it's important to focus on correct usage rather than complete avoidance Interviews with students and teachers reveal that similar words can have distinct collocations, such as "contain" and "include." To use verb-noun collocations accurately, learners must understand the context and semantic sets of the verbs.

Besides (LC2) collocations, (LC1) adjective + noun collocations are found to be problematic to Vietnamese EFL learners, which is consistent with findings in previous studies (Dang, 2014; Do & Le, 2022) This category of lexical collocations does not only pose a challenge to Vietnamese EFL learners, but also to EFL learners of other nationalities such as Indonesian (Harta et al., 2021) and Chinese (Quping & Pramoolsook, 2014) The most possible reason is the arbitrariness of the collocations, which means that the restriction between the choice of the adjective and the choice of the noun is arbitrary, rather than

“semantically motivated” (Harta et al., 2021, p 20) Harta et al (2021) gave an example of tight condition, which is considered an error, versus strict condition, which is considered a correct collocation (p 20) The researchers concluded that the reason why strict collocates with condition, but tight does not, is the result of arbitrary language convention, instead of the semantic features Another possible explanation is that EFL learners have the tendency to rely on highly-frequent adjectives such as good and bad to form collocations with vagueness in meaning, while more specific meanings could be embedded with the use of other adjectives (Jullian, 2000; Pavičić Takač & Lukač, 2013) In other words, EFL learners tend to resort to the L1 synonyms of an L2 expression, without attempting to use other possible synonyms, and overuse some adjectives, which may lead to the formation of odd collocations

While (LC1) and (LC2) collocations have been found to cause EFL learners a lot of difficulty, (LC6) noun (subject) + verb (action) collocations have not been referred to by previous studies as a problematic lexical collocation category In Do and Le’s (2023) study, the proportion of (LC6) erroneous collocations accounted for 1% of all errors, which is a very insignificant figure, compared with (LC1) and

(LC2) collocations Likewise, in Cao’s (2018) study, erroneous noun + verb

In the Collocation Test, learners often provided incorrect nouns within collocations During the Writing Test, misidentifying either collocate was the most common error Learners may struggle with collocations due to a lack of comprehensive understanding of the concept They often assume that the subject or noun can collocate with any verb This suggests the need for educators to emphasize these collocations with learners.

4.4.2 Factors influencing EFL learners’ use of collocations

It has been found in the present study that Vietnamese EFL learners’ use of lexical collocations is adversely influenced by a number of factors, the most important of which are interlingual and intralingual factors, resulting in learners’ word-by-word translation strategy, synonymy and approximation Besides, ELT textbooks and classroom activities also contribute to learners’ acquisition and production of lexical collocations

The phenomenon of EFL learners’ L1 negatively influencing their acquisition and production of L2 has been investigated in many previous studies with EFL learners of different nationalities such as Polish (Biskup, 1992), German (Bahns & Eldaw, 1993; Nesselhauf, 2005), Chinese (Quping & Pramoolsook, 2014), and Indonesian (Harta et al., 2021) In the present study, it was found that learners sometimes resort to the word-by-word translation strategy to obtain the collocation Although some students reported that the translation strategy often resulted in odd and incorrect collocations, they still subconsciously adopt this strategy This finding is consistent with the results from the study conducted by Nguyen (2024) to investigate teachers’ perceptions of teaching and learning collocations in Vietnamese context It is confirmed by all four teacher interviewees, three of whom work with high school learners, that Vietnamese EFL learners of different levels of education have the tendency to make “weird

89 combinations” due to the habit of translating word by word from Vietnamese into English (p 79)

Besides the strategy of translating directly from their L1, the strategy of using a synonym to produce a collocation with a level of approximation has been identified in this study This echoes the findings in the study conducted by Biskup (1992) in which most of the errors found in the study resulted from the use of inappropriate synonyms This strategy is similar to overgeneralization found in Ying’s (2009) study in which Chinese university students tended to overgeneralize in their production of collocations in L2 without realizing that there is incongruity existing between their mother tongue and L2 This was confirmed by a finding in Nesselhauf’s (2003) study in which the lower the degree of congruence of a lexical item between the learners’ L1 and L2, the more likely that these learners produce erroneous collocations What can be deduced from this is that if EFL learners’ mother tongue is not significantly incongruent compared with the target language, these learners will be less likely to produce erroneous collocations

However, Wang and Shaw (2008) discovered that Chinese learners in their study produced fewer collocation errors than Swedish counterparts The L1-L2 distance between Chinese and English is larger than that between Swedish and English, so Chinese EFL learners were less likely to create collocations, resulting in a lower rate of errors The most reasonable explanation for the likelihood of Vietnamese EFL learners to produce collocation errors, although the L1-L2 distance is not significantly large, is because they are used to learning single words, rather than collocations or multi-word units Nguyen and Webb (2017) discovered that Vietnamese EFL learners have limited collocation knowledge because there is a lack of attention to teaching formulaic sequences and collocations Because of this vocabulary learning strategy, Vietnamese EFL learners remember single words without actually understanding how the word is used and its collocates Also, when learners are more comfortable using single words, combining with a limited collocation repertoire, learners may conveniently substitute words with similar meanings in a combination to form collocations, resulting in unusual and erroneous ones For instance, in Do and Le’s (2023) study, a learner wrote allocate

CONCLUSION

Summary of findings

This study has explored that higher-proficiency learners produced significantly more correct lexical collocations in both the Collocation Test and the Writing Test than lower-proficiency learners This also means that learners of a lower proficiency level have the tendency to produce more erroneous lexical collocations

A finding that corroborates the results of a number of previous studies is the large quantity of erroneous (LC1) adjective + noun and (LC2) verb + noun collocations produced by learners of both groups Another interesting finding from the semi-interviews with the learners and the teachers is that (LC6) noun (subject)

+ verb (action), and (LC7) noun 1 of noun 2 also caused learners great difficulty because of their rare exposure to these collocations Students tend to make wrong choices of the verb in (LC2) and (LC6) collocations, wrong choices of the adjective in (LC1) collocations, and wrong choices of the first noun which can be a collective noun or a quantifier in (LC7) collocations Moreover, learners tend to provide wrong choices of the adverb in (LC4) adverb + adjective and (LC5) adverb + verb collocations

With regards to the possible causes of the learners’ erroneous use of collocations, learners have the tendency to translate from Vietnamese to obtain the English collocations they need This strategy results in strange and incorrect combinations, adversely affecting the cohesion and coherence of their writing The low frequency of occurrences of some particular collocations in the currently in- use textbook was found to be one of the underlying causes, which was agreed by the teachers It has been proven in previous studies that only lexical items that are recycled sufficiently, approximately from three to sixteen times (Nation, 1990; Peters, 2016), can be stored in their long-term memory and successfully acquired

94 Meanwhile, the limited chance of production in speaking and writing was claimed to exert a more significant impact than the limited chance of exposure.

Pedagogical implications

After examining the potential factors contributing to the incorrect usage of collocations in non-native English learners' writing, the researcher proposed recommendations for EFL instructors and ELT textbook authors.

Firstly, it is important that teachers emphasize the importance of collocations in mastering the English language Without sufficient emphasis from teachers, learners may fail to pay enough attention to this area, hindering the acquisition process To successfully assist learners during their process of collocation acquisition, teachers can provide learners with examples and clear explanations of how a collocation is formed, and in what context it is used It is important that when teachers introduce a collocation whether it is from the textbook or it is a part of incidental teaching, example sentences be provided so that learners can comprehend how the collocation is used in a specific context This can be done through the use of example sentences in the textbooks, concordances from online corpora and dictionaries

To enhance collocation acquisition, educators should provide ample practice opportunities Extra exercises in workbooks or classroom settings, such as multiple choice questions, word formation tasks, and writing assignments, can effectively support this process Expanding learners' collocation repertoire involves introducing useful phrases beyond textbooks, including collocates and synonyms Additionally, teachers can curate reliable online resources and dictionaries to facilitate collocation learning both inside and outside the classroom, serving as valuable references for students.

95 as well as enhance their autonomy in language learning To higher-proficiency learners, this would become the driving force for them to further advance their English language skills

In the classroom, a number of classroom activities can be of great assistance, such as mini games, occasional debate activities and group projects to encourage learners to use collocations in their spoken or written productive performance Interesting learning activities such as games give them more opportunities to revise and use the collocations The games also attract learners’ attention, creating a competitive environment, thanks to which they would actively learn the collocations Meanwhile, debates and presentations can be great speaking activities in which learners are required to utilize their productive skills such as public speaking, note taking, synthesizing and presenting information These activities not only offer them chances to use the collocations and structures they have been taught, but these also allow them to practice a variety of skills which can benefit them academically and professionally in the future

In textbook development, collocations merit increased emphasis Common and topic-specific collocations should be consistently recycled and highlighted, leveraging the efficacy of textual input enhancement A Collocation List at the textbook's end fosters student attention and retention, potentially including meaning translations for lower-proficiency learners Moreover, supplemental exercises dedicated to collocation usage provide valuable reinforcement in both the Student Book and Workbook.

For education administrators, it is worth noting that Vietnamese EFL learners have limited vocabulary knowledge, especially the meaning and form of

96 formulaic sequences (Nguyen & Webb, 2017), and that the common procedure of vocabulary teaching in a lot of EFL classrooms in Vietnam still maintains the focus on single words (Vu & Peters, 2021) Regarding the advent of the new English textbooks in high schools in Vietnam, it has posed a great challenge to English teachers A number of English teachers in high schools may have difficulty adapting the textbook into their own classroom Despite the provision of introductory sessions and training courses, many teachers are still struggling to effectively make use of the new ELT textbook, successfully adopt the communicative approach and ensure learners’ qualities and capacities, according to the 2018 General Education Program issued by the Ministry of Education and Training (see Section 4.3.1) Therefore, it is essential that the significance of collocations in English mastery should be further emphasized in ELT textbooks and training sessions for teachers.

Limitations of the study

This study has been conducted with care and ethics However, a number of limitations were inevitable

Firstly, the sample size of this study is not large enough for a reliable generalization of the findings, as the results may be only valuable to subjects having the same background or in the same setting The participants in this study come from a relatively high-ranking high school in Ho Chi Minh City, so they are not representative of all Vietnamese EFL high school learners Involving high school learners from different high schools may result in a more diverse collection of collocation errors, painting a more detailed picture of this phenomenon

Secondly, although the researcher attempted to invite the English teacher of the chosen class for an interview, the invitation was declined due to objective reasons, which was disappointing as this teacher’s perceptions and sharings can more effectively triangulate her students’ words Although the involvement of three teachers in this present study resulted in an insightful exploration of the phenomenon and thought-provoking opinions, their teaching background may not be representative of all teachers of English in high schools in Vietnam

97 Thirdly, learners’ production of lexical collocations may not be thoroughly reflected in one collection of essays There are seven categories of lexical collocations identified in this study and the topics of the writing tasks may not be large enough to elicit all patterns of lexical collocations

Last but not least, the native informants’ subjective intuition may not have been the most ideal instrument to determine whether a combination is a collocation or not Fully aware of this, the researcher attempted to invite four native speakers to check the conventionality of the combinations Despite that, this limitation is inevitable.

Recommendations for further research

For further research, researchers can take the limitations of this study into consideration for better results

First and foremost, a larger sample would result in more significant generalizability of the findings Moreover, the students chosen can be more representative of Vietnamese students, resulting in more valuable insights into how Vietnamese EFL learners produce collocations Moreover, more than one set of essays should be collected, with a variety of writing topics, so that more patterns of collocation errors can be more effectively elicited

As Vietnamese EFL learners increasingly pursue the IELTS exam for education exemptions, their speaking proficiency in collocations becomes crucial The high percentage of IELTS test-takers between 16 and 22 demonstrates the need for high school learners to master this skill Collocations play a significant role in their overall language proficiency and success in achieving high IELTS scores.

Collocation acquisition, the process of learning word combinations, is imperative for language proficiency However, cultural differences can impact this acquisition A study investigated techniques and activities that aid Vietnamese learners in collocation acquisition, considering the cultural differences between Vietnam and English-speaking countries This research contributes to a better understanding of collocation acquisition and cultural influences, informing language teaching practices and materials development.

98 difficulties while learning collocations will add more valuable findings into the existing body of literature of this very matter Most importantly, adopting experimental designs to investigate the cause-and-effect relationship between a varible and the use of lexical collocations by learners

Ackermann, K., & Chen, Y H (2013) Developing the Academic Collocation

List (ACL) - A corpus-driven and expert-judged approach Journal of

English for Academic Purposes, 12(4), 235–247 https://doi.org/10.1016/j.jeap.2013.08.002

Alhaisoni, E (2016) EFL teachers’ and students’ perceptions of dictionary use and preferences International Journal of Linguistics, 8(6), 31 https://doi.org/10.5296/ijl.v8i6.10267

Al-Zahrani, M S (1998) Knowledge of English lexical collocations among male

Saudi college students majoring in English at a Saudi University

[Doctoral dissertation, Indiana University of Pennsylvania]

Anthony, L (2020) AntConc (Version 3.5.9) [Computer Software] Tokyo,

Japan: Waseda University Available from https://www.laurenceanthony.net/software

Bahns, J (1993) Lexical collocations: A contrastive view ELT Journal, 47(1),

Bahns, J., & Eldaw, M (1993) Should we teach EFL students collocations?

Benson, M., Benson, E., & Ilson, R F (1986) The BBI Combinatory Dictionary of English: A guide to word combinations John Benjamins Publishing

Benson, M., Benson, E., & Elson, R (1997) The BBI dictionary of English word combinations, Amsterdam: John Benjamins

Bhalla, V., & Klimcikova, K (2019) Evaluation of automatic collocation extraction methods for language learning In ACL 2019 - Innovative Use of NLP for Building Educational Applications, BEA 2019 - Proceedings of the 14th Workshop (pp 264–274) Association for Computational

Linguistics (ACL) https://doi.org/10.18653/v1/w19-4428

Biber, D., Johansson, S., Leech, G., Conrad, S., Finegan, E., & Quirk, R (1999)

Longman Grammar of Spoken and Written English Longman

100 Biskup, D (1992) L1 influence on learners’ renderings of English collocations:

A Polish/German Empirical Study Vocabulary and Applied Linguistics, 85–93 https://doi.org/10.1007/978-1-349-12396-4_8

Brezina, V., & Gablasova, D (2015) Is there a core general vocabulary?

Introducing the new general service list Applied Linguistics, 36(1), 1–22 https://doi.org/10.1093/applin/amt018

Bruton, A (2007) Vocabulary learning from dictionary referencing and language feedback in EFL translational writing Language Teaching

Bui, T., Boers, F., & Coxhead, A (2020) Extracting multiword expressions from texts with the aid of online resources ITL - International Journal of

Applied Linguistics, 171(2), 221–252 https://doi.org/10.1075/itl.18033.bui

Bui, T L (2021) The role of collocations in the English teaching and learning

International Jourrnal of TESOL & Education, 1(2), pp 99-109 EOI: http://eoi.citefactor.org/10.11250/ijte.01.02.006

Bybee, J L., & Hopper, P J (2001) Frequency and the emergence of linguistic structure John Benjamins Pub Co https://doi.org/10.1075/tsl.45

Cantos-Gomez, P., & Sánchez, A (2001) Lexical constellations: What collocates fail to tell International Journal of Corpus Linguistics, 6(2), 199–228 https://doi.org/10.1075/ijcl.6.2.02can

Cao, D T P (2018) The effect of an online collocation dictionary on advanced learners’ use of collocations in L2 writing [Doctoral dissertation,

University of Leeds] https://etheses.whiterose.ac.uk/22829/

Cao, D T P., Pho, P D., & Dangnguyen, A C (2021) Profiling collocation use in English textbooks for Vietnamese students Journal of English

Education, 6(2), 82-91 http://doi.org/10.31327/jee.v6i2.1579

Carstairs-Mccarthy, A (2002) An introduction to English morphology: Words and their structure Linguistics, 160

101 Clear, J (1993) From Firth principles: Computational tools for the study of collocation In M Baker, G Francis, & E Tognini-Bolleni (Eds.), Text and technology (pp 271-292) Amsterdam/Philadenphia: John Benjamins

Cortina, J M (1993) What is coefficient alpha? An examination of theory and applications Journal of Applied Psychology, 78(1), 98–104 https://doi.org/10.1037/0021-9010.78.1.98

Cowie, A P (1994) Phraseology In E A Ronald (Ed.), The Encyclopedia of

Language and Linguistics (pp 3168–3171) Oxford: Pergamon

Dang, D (2014) Errors in lexical collocations by academic writing students at

Faculty of English Linguistics and Literature In English Language

Teaching in ASEAN Countries, Ho Chi Minh (pp 41-46) University of

Do, N H M., & Le, Q T (2022) Lexical collocation errors in essay writing: a study into Vietnamese EFL students and their perceptions International

Journal of Language Instruction, 2(2), 1-20 DOI: https://doi.org/10.54855/ijli.23221

Duong, P T., Perez, M M., Desmet, P., & Peters, E (2021) Differential effects of input-based and output-based tasks on L2 vocabulary learning

Canadian Journal of Applied Linguistics, 24(3), 120–144 https://doi.org/10.37213/cjal.2021.31183

Durrant, P (2009) Investigating the viability of a collocation list for students of

English for academic purposes English for Specific Purposes, 28(3), 157–

Durrant, P., & Schmitt, N (2009) To what extent do native and non-native writers make use of collocations? IRAL - International Review of Applied

Linguistics in Language Teaching, 47(2) https://doi.org/10.1515/iral.2009.007

Evert, S (2008) Corpora and collocations In A Ludeling & M Kyto (Eds.),

Corpus linguistics: An international handbook (pp 1212–1248) Berlin,

Germany: Mouton de Gruyter doi:10.1515/9783110213881.2.1212

102 Gass, S M & Selinker, L (2008) Second language acquisition (3rd ed.)

Gopika, J S., & Rekha, R V (2023) Awareness and use of digital learning before and during COVID-19 International Journal of Educational

Granger, S (1998) Prefabricated patterns in advanced EFL writing: Collocations and lexical phrases In A Cowie (Ed.), Phraseology: Theory, analysis and applications (pp 145–160) Oxford, UK: Oxford University Press

Harta, I G W., Bay, I W., & Ali, S W (2021) An analysis of lexical collocation errors in students’ writing TRANS-KATA: Journal of

Language, Literature, Culture and Education, 2(1), 15-25

Hausmann, F J (1989) Le dictionnaire de collocations In F J Hausmann, O

Reichmann, H E Wiegand, & L Zgusta (Eds.), Wửrterbỹcher: ein internationales Handbuch zur Lexikographie Dictionaries Dictionnaires

(pp 1010–1019) Berlin: Mouton De Gruyter

Howarth, P A (1996) Phraseology in English Academic Writing Phraseology in English Academic Writing De Gruyter https://doi.org/10.1515/9783110937923

Hoey, M (2005) Lexical priming: A new theory of words and language

Routledge http://dx.doi.org/10.4324/9780203327630

Jones, S., & Sinclair, J M (1974) English lexical collocations - A study in computational linguistics Cahiers de lexicologie, 24, 15-61

Jullian, P (2000) Creating word-meaning awareness ELT Journal, 54(1), 37–

Kennedy, G (1991) Between and through: The company they keep and the functions they serve In K Aijmer & B Altenberg (Eds.), English corpus linguistics (pp 95-110) Essex: Longman

Kjellmer, G 1994 Dictionary of collocations Oxford: Clarendon Press

Kojic‐Sabo, I., & Lightbown, P M (1999) Students’ approaches to vocabulary learning and their relationship to success The Modern Language Journal,

103 Koprowski, M (2005) Investigating the usefulness of lexical phrases in contemporary coursebooks ELT Journal, 59(4), 322–332 https://doi.org/10.1093/elt/cci061

Research on collocations, or word combinations, has been conducted in various linguistic contexts T Koya investigated English and Japanese noun-verb collocations in secondary school textbooks (2003, 2004) Additionally, B Laufer and T Waldman analyzed verb-noun collocations in second language English writing (2011).

Le-Thi, D., Rodgers, M., & Pellicer‐Sánchez, A (2018) Teaching formulaic sequences in an English- language class: The effects of explicit instruction versus coursebook instruction TESL Canada Journal, 34(3), 111–139 https://doi.org/10.18806/tesl.v34i3.1276

Lew, R., & Doroszewska, J (2009) Electronic dictionary entries with animated pictures: Lookup preferences and word retention International Journal of

Lexicography, 22(3), 239–257 https://doi.org/10.1093/ijl/ecp022

Lewis, M (1997) Implementing the Lexical Approach: Putting theory into practice http://ci.nii.ac.jp/ncid/BA35070833

Lewis, M (2000) Materials and resources for teaching collocation In M Lewis

(Ed.) Teaching collocations: Further development in the lexical approach (pp 186-204) London: Language Teaching Publications

Lindqvist, C., Gudmundson, A., & Bardel, C (2012) A new approach to measuring lexical sophistication in L2 oral production Eurosla

Liu, C.P (1999) An analysis of collocation errors in EFL writing The

Proceedings of the English International Symposium on English Teaching

(pp 483-494) Taipei: Crane Publishing Limited

Mahmoud, A (2005) Collocation errors made by Arab learners of English

Asian EFL Journal, 5(2), 117–126 http://www.asian-efl

104 journal.com/monthly-editions-new/collocation-errors-made-by-arab- learners-of-english/

McIntosh, C., Francis, B., & Poole, R (2009) Oxford collocations dictionary for students of English

MOET Chuong Trinh Giao duc pho Thong: Chuong Trinh mon Tieng Anh

[English Curriculum for the Secondary School]; Education Publishing

Nation, I S P (1990) Teaching & learning vocabulary Heinle ELT

Nation, I S P (2001) Learning vocabulary in another language Cambridge:

Nattinger, J R., & DeCarrico, J S (1992) Lexical phrases and language teaching Oxford: Oxford University Press

Nesselhauf, N (2003) The use of collocations by advanced learners of English and some implications for teaching Applied Linguistics, 24(2), 223–242 https://doi.org/10.1093/applin/24.2.223

Nesselhauf, N (2005) Collocations in a learner corpus John Benjamins

Nguyen, C., & Boers, F (2018) The effect of content retelling on vocabulary uptake from a TED talk TESOL Quarterly, 53(1), 5–29 https://doi.org/10.1002/tesq.441

Nguyen, C D (2016) Creating spaces for constructing practice and identity: innovations of teachers of English language to young learners in Vietnam

Research Papers in Education, 32(1), 56–70 https://doi.org/10.1080/02671522.2015.1129644

Nguyen, H X C (2023) Students’ perceptions of the debate activity in grade 9

Intensive English program VietTESOL International Convention

Proceedings Retrieved from https://proceedings.viettesol.org.vn/index.php/vic/article/view/66

Nguyen, H X C (2024) Teaching and learning collocations in the Vietnamese context: Teachers’ voices AsiaCALL Online Journal, 15(1), 71-89 DOI: https://doi.org/10.54855/acoj.241515

105 Nguyen, T M H., & Webb, S (2017) Examining second language receptive knowledge of collocation and factors that affect learning Language

Teaching Research, 21(3), 298-320 https://doi.org/10.1177/1362168816639619

Nguyen, V L., (Ed.) (2016) English 9 Vietnam Education Publishing House. Nunan, D (2003) The impact of English as a global language on educational policies and practices in the Asia-Pacific region TESOL Quarterly, 37(4),

O’Flynn, J (2018) Developing an Academic Collocation List for Arts and

Humanities (pp 238–245) [Master’s thesis, University of Warwick] https://dialnet.unirioja.es/servlet/articulo?codigo"85393

O’Keeffe, A., McCarthy, M and Carter, R (2007) From corpus to classroom

Othman, H G., & Murad, I (2015) A study on Kurdish students’ attitudes to group work in the EFL classroom European Scientific Journal, 11(11) https://eujournal.org/index.php/esj/article/download/5454/5243

Paquot, M., & Granger, S (2012) Formulaic language in learner corpora Annual

Review of Applied Linguistics, 32, 130–149 https://doi.org/10.1017/S0267190512000098

Partington, A (1998) Patterns and meanings: Using corpora for English language research and teaching Amsterdam: John Benjamins

Pavičić Takač, V., & Lukač, M (2013) How word choice matters: An analysis of adjective-noun collocations in a corpus of learner essays Jezikoslovlje,

Peters, E (2012) Learning German formulaic sequences: The effect of two attention-drawing techniques The Language Learning Journal, 40(1), 65-

Peters, E (2016) The learning burden of collocations: The role of interlexical and intralexical factors Language Teaching Research, 20(1), 113–138 https://doi.org/10.1177/1362168814568131

106 Quping, H., & Pramoolsook, I (2014) Non-English major EFL learners’ lexical collocation errors in a Chinese context Suranaree Journal of Social

Rahimi, M., & Momeni, G (2012) The effect of teaching collocations on

English language proficiency Procedia - Social and Behavioral Sciences,

Rasti-Behbahani, A (2021) Why digital games can be advantageous in vocabulary learning Theory and practice in language studies, 11(2), 111–

Read, J., & Nation, P (2006) An investigation of the lexical dimension of the

IELTS speaking test IELTS research reports 6 IELTS Australia/British Council

Renouf, A., & Sinclair, J.M (1991) Collocational frameworks in English In K

Aijmer & B Altenberg (Eds.), English corpus linguistics (pp 128-143) Harlow: Longman

Riazi, A M., & Candlin, C (2014) Mixed methods research in language teaching and learning: Opportunities, issues and challenges Language

Richards, J C (2015) Error analysis: Perspectives on second language acquisition Taylor and Francis https://doi.org/10.4324/9781315836003

Ringbom, H (1998) High-frequency verbs in the ICLE corpus In A Renouf

(Ed.), Exploration in Corpus Linguistics (pp 191–200) Amsterdam and Atlanta: Rodopi

Simpson-Vlach, R., & Ellis, N C (2010) An academic formulas list: New methods in phraseology research Applied Linguistics, 31(4), 487–512 https://doi.org/10.1093/applin/amp058

Sinclair, J M (1991) Corpus, concordance, collocation Oxford, UK: Oxford

Sinclair, J M., Daley, R., & Jones, S (2003) English Collocation Studies: The

OSTI Report http://ci.nii.ac.jp/ncid/BA70095683

107 Shaw, S D., & Weir, C J (2007) Examining Writing: Research and practice in assessing second language writing http://ci.nii.ac.jp/ncid/BA82799424

Smadja, F (1993) Retrieving collocations from text: Xtract Computational

Smith, J A & Osborn, M (2015) Interpretative phenomenological analysis In

J A Smith (Ed.), Qualitative psychology: A practical guide to research methods (pp 53-80) Los Angeles/London/New Delhi/Singapore: Sage

Staples, S., Egbert, J., Biber, D., & McClair, A (2013) Formulaic sequences and

EAP writing development: Lexical bundles in the TOEFL iBT writing section Journal of English for Academic Purposes, 12(3), 214–225 https://doi.org/10.1016/j.jeap.2013.05.002

Sung, J (2003) English lexical collocations and their relation to spoken fluency of adult non-native speakers [Unpublished doctoral dissertation, Indiana

Taber, K S (2018) The use of Cronbach’s alpha when developing and reporting research instruments in science education Research in Science Education,

Thanh, H (2023, December 29) Graduation exemptions for IELTS entice more

Vietnamese youths VnExpress Retrieved from: https://e.vnexpress.net/news/news/education/graduation-exemptions-for- ielts-entice-more-vietnamese-youths-4694569.html

Thornbury, S (2002) How to teach vocabulary Essex: Pearson Education Ltd http://archives.umc.edu.dz/handle/123456789/111117

Trinh, T L A., Tran, T K N., Vo, T B N., & Huynh, T T S (2021) The difference effects of paper dictionaries vs online dictionaries AsiaCALL

Trinh, T H., Nguyen, M N., & Tran, T T H (2022) Teachers and students' perceptions of using digital games in improving vocabulary at non-

English-majored class AsiaCALL Online Journal, 13(5), 112-131 https://doi.org/10.54855/acoj.221358

108 Tsai, K.-J (2015) Profiling the collocation use in ELT textbooks and learner writing Language Teaching Research, 19, 723–740

Vidakovic, I., & Barker, F (2010) Use of words and multi-word units in Skills for Life writing examinations IELTS research reports 41 IELTS

Vilkaitė, L (2016) Formulaic language is not all the same: comparing the frequency of idiomatic phrases, collocations, lexical bundles, and phrasal verbs Taikomoji Kalbotyra, 8, 28–54 https://doi.org/10.15388/tk.2016.17505

Vu, M.L., Huynh, D.H., Nguyen, T.L., Huynh, N.T.T, & Tran, T.T.T (2022)

Tieng Anh 10 Friends Global Student Book [English 10 Friends Global Student Book] Vietnam Education Publishing House

Vu, D.V., & Peters E (2021) Incidental learning of collocations from meaning input: a longitudinal study into three reading modes and factors that affect learning Studies in Second Language Acquisition 44(3), 685-707 doi:10.1017/S0272263121000462

Vu, D V., & Peters, E (2022) Learning vocabulary from reading-only, reading- while-listening, and reading with textual input enhancement: Insights from Vietnamese EFL learners RELC Journal, 53(1), 85-100 https://doi.org/10.1177/0033688220911485

Webb, S., Newton, J., & Chang, A (2013) Incidental learning of collocation

Language Learning, 63(1), 91–120 https://doi.org/10.1111/j.1467-

Wray, A (2002) Formulaic squences and the lexicon Cambridge: Cambridge

Yamashita, J., & Jiang, N (2010) L1 influence on the acquisition of L2 collocations: Japanese ESL users and EFL learners acquiring English collocations TESOL Quarterly, 44(4), 647–668 https://doi.org/10.5054/tq.2010.235998

109 Yan-Ren, D (2005) Use of formulaic language as a predictor of L2 oral and written performance Journal of PLA University of Foreign Languages https://en.cnki.com.cn/Article_en/CJFDTOTAL-JFJW20050300A.htm Ying, S (2009) Study on collocations in English writing by Chinese students

Appendices Appendix A Verb patterns of grammatical collocations (Benson et al., 1986)

(P1) An indirect object can be moved to a position before the direct object of transitive verbs For example: ã He gave her some flowers (or He gave some flowers to her.)

(P2) An indirect object cannot be moved to a position before the direct object by deleting to Common verbs that fit this category are mention, return, scream, etc For example: ã They described the book to her

(P3) The deletion of for that goes with a transitive verb is allowed and the indirect object can be moved to a position before the direct object For example: ã She bought a scarf for her daughter (or She bought her daughter a scarf.)

(P4) A collocation of a verb with a specific preposition and an object is formed For example: ã This plan contributed to the success of the company

(P5) Verbs are followed by to + infinitive For example: ã They start to shout loudly

(P6) Verbs are followed by infinitives without to For example: ã We must leave now

(P7) Verbs are followed by the second verb in –ing For example: ã We enjoyed playing badminton

(P8) Transitive verbs are followed by an object and to + infinitive For example: ã They asked the students to participate in discussion

(P9) Transitive verbs are followed by a direct object and an infinitive without to For example: ã We let them use the car

(P10) Verbs are followed by an object and a verb in –ing For example: ã He found the children reading in the library

(P11) Verbs can be followed by a noun or pronoun and gerund For example:

111 ã I cannot imagine her shouting at the children

(P12) Verbs are followed by a noun clause beginning with conjunction that For example: ã He admitted that he was late for the conference

(P13) Transitive verb can be followed by a direct object, an infinitive to be and adjective, past participle, noun or pronoun For example: ã They considered him to be a rising star

(P14) Transitive verbs are followed by a direct object and adjective, past participle or noun or pronoun For example: ã He finds watching horror movies interesting

(P15) Transitive verbs are followed by two objects For example: ã The teacher asked the students questions

(P16) Intransitive, reflexive or transitive verbs must be followed by an adverbial (an adverb, a prepositional phrase, a noun phrase or a clause) For example: ã The meeting will last three hours

(P17) Verbs can be followed by an interrogative word, such as how, what, when, etc For example: ã She knows when to say a few words of encouragement

(P18) Dummy ‘it’ is followed by transitive verbs (often expressing emotions) and by to + infinitive or by that + clause or by either For example: ã It shocked me to hear the news

(P19) Some intransitive verbs are followed by a predicate noun or a predicate adjective For example: ã He will make an excellent yoga instructor

Interview questions in the present study

Student makes the wrong choice of adjective, uses a non-existent adjective, or provides the adjective with a level of approximation, but the part of speech is incorrect

In recent society, air pollution is a serious problem

Students may make several common errors when using nouns in their writing These errors include selecting the wrong noun, choosing a non-existent noun, omitting a necessary noun, or providing a noun with an incorrect approximation or level of specificity Each of these errors can detract from the clarity and effectiveness of the writing.

Smartphones are one of the most popular applicants used by young people

Suggested improvement: popular devices/appliances

LC1.3 The combination exists but it is contextually incorrect

We should reduce the use of public transport and switch to toxic-free transportation such as bicycles

The combination does not exist The combination does not bear a semantic meaning, and therefore, does not fit semantically and contextually

In growing societies, a lot of investment goes into developing technology

Suggested improvement: developing / fast-growing countries

Student makes the wrong choice of verb, uses a non-existent verb, or provides a word with a level of approximation, but the part of speech is incorrect

People living in big cities have to face many diseases Suggested improvement: suffer from many diseases

Student makes the wrong choice of noun, uses a non-existent noun, does not provide a noun, or provides a word with a level of approximation, but the part of speech is incorrect

Smartphones can cause their health

Suggested improvement: cause health problems

LC2.3 The combination exists but it is contextually incorrect

If we do not have a way now, the problem is going to become worse

Suggested improvement: take action / take measures

The combination does not exist The combination does not bear a semantic meaning, and therefore, does not fit semantically, and contextually

The discovery of the mammal graveyard in the Gobi Desert unfolds many mysteries

Suggested improvement: reveals many secrets

Student makes the wrong choice of adjective, uses a non-existent adjective, or provides a word with a level of approximation, but the part of speech is incorrect

Residents in large cities can have a better living quality Suggested improvement: life quality

Student makes the wrong choice of noun, uses a non-existent noun, does not provide a noun, or provides a word

Energy-saving appliances are said to have a high energy effect

114 with a level of approximation, but the part of speech is incorrect

LC3.3 The combination exists but it is contextually incorrect

Close to 1,000 surveys were handed out to travellers waiting in the waiting room Suggested improvement: departure lounge

The combination does not exist The combination does not bear a semantic meaning, and therefore, does not fit semantically, and contextually

Air pollution is caused by the smoke coming from factories and working constructions Suggested improvement: construction sites

Student makes the wrong choice of adverb, or uses a non-existent adverb, or provides a word with a level of approximation, but the part of speech is incorrect

The soldiers stand extremely still in front of the mausoleum

Student makes the wrong choice of adjective, uses a non-existent adjective, does not provide an adjective, or provides a word with a level of approximation, but the part of speech is incorrect

Energy-saving appliances are said to have high energy effect

LC4.3 The combination exists but it is contextually incorrect

In order to become good physical condition, we need to exercise regularly

The combination does not exist The combination does not bear a semantic meaning, and therefore, does not fit semantically, and contextually

In order to become well physical, we need to exercise regularly

Student makes the wrong choice of adverb, uses a non-existent adverb, or provides a word with a level of approximation, but the part of speech is incorrect

My kittens sometimes gone suddenly when it is sunny outside

Student makes the wrong choice of verb, uses a non-existent verb, does not provide a verb, or provides a word with a level of approximation, but the part of speech is incorrect

Smartphones are completely allowed in many high schools

LC5.3 The combination exists but it is contextually incorrect

In many big cities, air pollution is growing faster Suggested improvement: becoming more serious

The combination does not exist The combination does not bear a semantic meaning, and therefore, does not fit semantically, and contextually

Videos on tiktok are becoming more and more unregulatedly famous

Student makes the wrong choice of noun, or uses a non-existent noun, provides a word with a level of approximation, but the part of speech is incorrect

The bom explodes in the water and fish are killed Suggested improvement: bomb explodes

Student makes the wrong choice of verb, uses a non-existent verb, does not provide a verb, or provides a word with a level of approximation, but the part of speech is incorrect

This problem happens because students depend on smart phones too much

Suggested improvement: problem arises / problem occurs

LC6.3 The combination exists but it is contextually incorrect

A lot of vehicles operate by fuel, which causes pollution

Suggested improvement: vehicles run on fuel

The combination does not exist The combination does not bear a semantic meaning, and therefore, does not fit semantically, and contextually

Trafic on the road wastes dirty air every day

Vehicles on the road emit

Student makes the wrong choice of first noun, uses a non-existent noun, or provides a word with a level of approximation, but the part of speech is incorrect

There is a swarm of fish along the warm currents

Suggested improvement: shoal of fish

Students frequently encounter errors in their noun usage, such as selecting incorrect nouns, employing non-existent nouns, omitting nouns altogether, or using words that are not nouns These errors can compromise the clarity and coherence of their writing To improve their noun usage, students should ensure they choose appropriate nouns, verify the existence of nouns, avoid leaving nouns out, and select words that accurately represent nouns.

Maria is well-known for her jokes and a sense of fun

Suggested improvement: sense of humor

LC7.3 The combination exists but it is contextually incorrect

Reading negative posts can affect their wealth of happiness

Suggested improvement: a level of happiness

The combination does not exist The combination does not bear a semantic meaning, and therefore, does not fit semantically, and contextually

Starting the lesson with some mini online games can give students an emotion of utopia Suggested improvement: a sense of euphoria

Appendix C Collocation activities in EFL classrooms (Lewis, 1997)

Table C.1 A matching exercise to teach collocations (p 88)

8 to set a a balanced diet b your weight c an exercise program d routine e weight-loss f targets g food h benefits

Table C.2 A deleting exercise to teach collocations (p 94)

One word in each group does not make a strong word partnership with the word in capitals Which is the odd one?

BRIGHT idea green smell child day room

HIGH season price opinion spirits house time

STRONG possibility doubt smell influence views coffee

Table C.3 A story-telling activity to teach collocations of the verb have (p 150)

Write a story using these collocations with have had a bit of good/bad luck had no alternative but to had an accident had no doubt that had a feeling that had a talk to had no effect had a think

Table C.4 A matching exercise to teach collocations in business English (p 92) bill presentation costs invoice discount debt lunch message expenses de al service calculation mistake money promise pay make give

Task 1 – Read the sentences and their translation Fill in the blank with a collocation (kết hợp từ) that fits the meaning of the underlined phrase in the translation

Bring your raincoat because there is going to be _ this afternoon!

(Mang theo áo mưa đi, vì chiều nay sẽ có mưa lớn!) Answer: heavy rain

For question 1-4, you need to fill in the blank with a verb + adverb or adverb + verb collocation (e.g., rain heavily, smile brightly, firmly believe)

1 The government has begun to _ the drivers who park in the bus lanes

(Chính phủ bắt đầu phạt tiền nặng những tài xế đỗ xe trong làn ưu tiên dành cho xe buýt)

2 Expensive products are believed to _

(Những sản phẩm đắt tiền thì được cho là hoạt động một cách hiệu quả)

3 The most devastating tsunami can _ all houses and road systems

(Trận sóng thần kinh hoàng nhất có thể hoàn toàn phá hủy nhà cửa và hệ thống đường xá)

4 My kittens sometimes _ when it is sunny outside

(Những con mèo con của tôi đôi lúc biến mất bất thình lình khi ngoài trời nắng đẹp)

For question 5-8, you need to fill in the blank with a verb + noun collocation (e.g., make a mistake, have a party)

5 Warhol’s pop-art works still _ on art lovers around the world

(Các tác phẩm pop-art của Warhol vẫn gây ấn tượng với những người yêu nghệ thuật trên thế giới)

6 Humans need to _ their _ in order to save the planet

(Con người cần phải thay đổi cách cư xử của họ để cứu hành tinh này)

7 Many people were found to _ after the terrible earthquake struck the town

121 (Nhiều người được tìm thấy đã mất ý thức sau khi trận động đật ập vào thị trấn)

8 Families have to bear the responsibility when their relatives _

(Các gia đình phải chịu trách nhiệm khi người thân của họ phạm tội)

For question 9-12, you need to fill in the blank with a noun (subject) + verb (action) collocation (e.g., a bomb explodes, bees buzz)

9 Global warming is worse as _ at a more rapid rate every year

(Sự nóng lên toàn cầu đang trở nên tệ hơn vì mực nước biển dâng cao với một tốc độ nhanh chóng hơn)

10 Warmer temperatures are making the _ faster

(Nhiệt độ ấm hơn đang làm các chỏm băng tan nhanh chóng hơn)

11 When an earthquake shakes a large city, _ quickly

(Khi một trận động đất rung chuyển một thành phố lớn, các tòa nhà sụp đổ nhanh chóng)

12 Whenever I call her and the _ twice before heading into voicemail, I know she is busy

(Bất cứ khi nào tôi gọi cô ấy và điện thoại reo hai lần trước khi vào hộp thoại thì tôi biết cô ấy đang bận)

For question 11-16, you need to fill in the blank with an adjective + noun collocation (e.g., heavy rain)

13 People should be encouraged to use _ more in big cities to reduce traffic jams

(Người dân nên được khuyến khích sử dụng phương tiện giao thông công cộng nhiều hơn trong các thành phố lớn để giảm kẹt xe

14 Children with a rare _ are more likely to suffer from mental health problems (Những trẻ em với một rối loạn di truyền hiếm thường dễ bị ảnh hưởng bởi các vấn đề về sức khỏe tinh thần)

15 The _ of some wild animals in the USA is being polluted by chemicals from nearby factories

(Môi trường sống tự nhiên của một số loài động vật hoang dã ở Mỹ đang bị ô nhiễm bởi các chất hóa học từ các nhà máy lân cận)

16 This book provides the readers with a _ of royal families in some European countries

(Quyển sách này cung cấp cho người đọc một sự mô tả ngắn gọn về các gia đình hoàng gia ở một số quốc gia ở châu Âu)

For question 17-20, you need to fill in the blank with an adverb + adjective collocation (e.g., bitterly disappointed)

17 Artificial intelligence (AI) is now one of the most _ industries

(Ngành trí tuệ nhân tạo hiện nay là một trong những ngành công nghiệp cạnh tranh cao nhất)

18 In order to become _, we need to exercise regularly and eat healthily

(Để có thể chất tốt, chúng ta cần tập thể dục thường xuyên và ăn uống lành mạnh)

19 Although Switzerland remains an ideal tourist destination, smaller Mediterranean islands are becoming _

Mặc dù Thụy Sĩ vẫn là điểm đến du lịch hấp dẫn, nhưng những hòn đảo nhỏ hơn ở khu vực Địa Trung Hải đang trở nên ngày càng được yêu thích hơn.

20 The soldiers guarding at the Ho Chi Minh Mausoleum have to stand _ during their shift

(Những người lính canh gác ở Lăng Hồ Chí Minh phải đứng yên hoàn toàn trong suốt ca trực)

For question 21-24, you need to fill in the blank with a noun + noun collocation (e.g., university degree)

21 According to the _, countries with a tropical climate will experience heavy rain next week

(Theo dự báo thời tiết, các quốc gia có khí hậu nhiệt đới sẽ có mưa to vào tuần tới)

22 You can find the new light bulb that tells you about _ in the local store (Bạn có thể tìm thấy loại bóng đèn điện mới mà báo cho bạn biết hiệu suất năng lượng trong cửa hàng địa phương)

23 Before boarding, passengers are instructed to wait and relax in the _

(Trước khi lên máy bay, hành khách được hướng dẫn đợi và thư giãn trong phòng đợi khởi hành)

24 is the foundation for a peaceful and prosperous world because females and males have the same rights

(Bình đẳng giới là nền tảng của một thế giới bình yên và thịnh vượng vì nam và nữ có những quyền giống nhau)

For question 25-28, you need to fill in the blank with a noun (1) of noun (2) collocation (e.g., a loaf of bread)

25 Anna is one of the most popular students because of her great _

(Anna là một trong những học sinh nổi bật nhất nhờ khiếu hài hước tuyệt vời của cô ấy)

26 Her mom asked her to buy a _ on the way home from work

(Mẹ cô ấy nhờ cô ấy mua một thanh xà phòng trên đường đi làm về nhà)

During our diving expedition on Phu Quy Island, we encountered a colossal school of fish that left us in awe The sheer size and abundance of these creatures created an unforgettable experience, showcasing the vibrant marine life that inhabits the pristine waters surrounding the island.

28 A study of lottery winners has discovered a normal _ half a year after their win

(Một nghiên cứu về những người thắng vé số đã tìm thấy một mức độ hạnh phúc bình thường nửa năm sau khi họ trúng số)

Task 2 – Rewrite the following sentences so that the new sentence has the same meaning as the given sentence Remember to form a collocation using the provided word in the brackets

This institute is famous for its well-qualified tutors (reputation)

➡ This institute has a _ for its well-qualified tutors

Answer: This institute has a good reputation for its well-qualified tutors

1 The performance was not as good as expected (expectations)

2 Greenhouse gas emissions damage the ecosystem (impact)

➡ Greenhouse gas emissions have a(n) on the ecosystem

3 Farm workers’ job is often extremely tiring and requires a lot of strength (demanding)

➡ Farm workers tend to have a(n) job

4 In the Ring of Fire, before a volcanic eruption, people are evacuated (volcano)

➡ In the Ring of Fire, before a _, people are evacuated

5 The guide took us to famous sightseeing areas (tourist)

➡ The tour guide took us to famous

6 The government allows nobody to destroy nature on this island (protected)

➡ Nature on this island is by the government

7 After the camping trip, everyone felt pleased and satisfied (pleasure)

➡ After the camping trip, everyone had a(n)

Question no Answer key Acceptable answers Freq

6 alter (their) behavior change (their) behavior

27 school of fish shoals of fish 26

Question no Answer key Acceptable answers Freq

(LC2) v + n 1 meet (the / my) expectations match expectations satisfy expectation

(LC1) adj + n 2 negative impact bad impact detrimental impact

88 (LC4) adv + adj 3 physically demanding

(LC3) n + n 5 tourist attractions tourist areas tourist destinations

(LC5) v + adv 6 strictly protected heavily protected well protected

445 (LC7) n1 of n2 7 a feeling of pleasure

Write an argumentative essay to one of the following questions Write at least 250 words Try to use as many collocations as possible The suggested outline is for reference only

A Online education is becoming more and more popular Some people claim that e-learning has so many benefits that it will replace face-to-face education soon Others say that traditional education is irreplaceable Discuss both views and give your opinion

B Many governments think that economic progress is their most important goal Some people, however, think that other types of progress are equally important for a country Discuss both views and give your own opinion

Suggested outline for writing task A Structure of paragraphs and suggested ideas

You need to summarize the two viewpoints and state your opinion ã agree with which viewpoint?

Main idea 1: First viewpoint discussion ã distance learning? ã resources?

Main idea 2: Second viewpoint discussion ã interactions with teachers and friends? ã inspirations?

You should summarize key ideas in your article and restate your opinion You can also add some suggestions for the readers at the end ã what are the key points that you have discussed? ã what is your opinion again? ã any further suggestions?

Suggested outline for writing task B Structure of paragraphs and suggested ideas

You need to summarize the two viewpoints and state your opinion ã agree with which viewpoint?

Main idea 1: First viewpoint discussion ã economy? ã people’s feeling?

Main idea 2: Second viewpoint discussion ã medical? ã education?

You should summarize key ideas in your article and restate your opinion You can also add some suggestions for the readers at the end ã what are the key points that you have discussed? ã what is your opinion again? ã any further suggestions?

Appendix F IELTS Writing Band Descriptors (retrieved from www.ielts.org )

Appendix G Samples of learners’ essays

Appendix H Scores of participants’ essays and grouping decisions

Essay code Score by researcher Score by IELTS trainer Decision

Essay code Score by researcher Score by IELTS trainer Decision

Appendix I Interview questions in the present study Questions for student interviewees

1 What is a collocation? Can you give an example?

2 Does your English teacher give you the definition of collocations and emphasize the importance of collocations? Can you describe an occasion on which a collocation was taught?

 Does the teacher encourage you to use the collocation?

3 Where do you usually learn about collocations from? Do you have the habit of using reference materials or tools to learn collocations?

4 How do you write a collocation that you want to use? For example, you want to write gây nhiều tranh cãi, but you do not know the collocation provoke/create controversies?

5 Have you ever translated from Vietnamese to English to create a collocation? Is that collocation correct or incorrect?

6 Have you ever created a collocation based on a knowledge of a collocation you have learned? (For example, "failed the test" but a student writes "fell the test")

7 Do you feel comfortable using a collocation that you are very familiar with?

8 Have you ever avoided using a certain collocation due to its difficulty, or have you encountered difficulties in remembering it?

9 In the Collocation Test, there are 14 collocations taken from the Friends Global 10 book and 7 collocations taken from the English program for grades 6-9 Do you feel familiar with the collocations taken from the book? When studying from textbooks, do you notice the collocations appearing in the texts?

10 Do you have any suggestions for teaching and learning collocations for high school students?

 more relevant and useful collocations for certificate exams

 additional exercises to reinforce the knowledge

 the presentation of the textbook

1 Do you introduce the definition of collocations to students? Do you provide examples of collocations to students?

2 Can you provide an example of a time when you taught collocations to students (teaching steps, example sentences, etc.)?

3 Do you introduce additional collocates of a word that appears in the textbook (Friends Global 10) when you teach?

4 Do you provide different exercises for students to memorize and use collocations? What types of exercises are they? Can you provide examples?

5 In your opinion, which type of collocations do students struggle with (adjective + noun, verb + noun, noun + noun, adverb + verb, subject + verb, noun1 of noun2)? Why do you think so?

6 Do your students often make mistakes with collocations? If so, what types of mistakes are they (wrong word, grammar errors, spelling mistakes, incorrect collocations)?

7 In your opinion, why do students make those mistakes? Is it due to objective or subjective reasons (strategies that students use to write

137 collocations: using synonyms, using words with similar or related meanings, creating new words or translating from Vietnamese)? Can you provide examples?

8 Some collocations only appear 1-3 times in the students' textbook In your opinion, is the limited appearance of collocations a reason why students have difficulty memorizing them and make mistakes?

9 In your opinion, what tools can support the learning and teaching of collocations (dictionaries, AI-powered tools, reference books, etc.)? Can you suggest some ways to support students in learning collocations?