24 THE UNIVERSITY OF DA NANG UNIVERSITY OF FOREIGN LANGUAGE STUDIES LINGUISTIC FEATURES OF ONLINE ENGLISH WRITINGS IN VIETNAMESE HIGHER EDUCATION CONTEXT PHẠM THỊ THU HUONG Major ENGLISH LINGUISTICS C[.]
THE UNIVERSITY OF DA NANG UNIVERSITY OF FOREIGN LANGUAGE STUDIES LINGUISTIC FEATURES OF ONLINE ENGLISH WRITINGS IN VIETNAMESE HIGHER EDUCATION CONTEXT PHẠM THỊ THU HUONG Major: ENGLISH LINGUISTICS Code: 9220201 DOCTORAL THESIS IN SOCIAL SCIENCES AND HUMANITIES (A SUMMARY) Da Nang, 2023 This thesis has been completed at University of Foreign Language Studies, The University of Da Nang Supervisor: Asso Prof Dr Nguyễn Văn Long Examiner 1: ……………………………… Examiner 2: ……………………………… Examiner 3: ……………………………… The thesis was orally defended at the Examining Committee Major: English Linguistics Time: Venue: University of Foreign Language Studies - The University of Da Nang This thesis is available for the purpose of reference at: - Center for Information Technology and Learning Resources, University of Foreign Language Studies - The University of Da Nang - The University of Da Nang – The Center for Learning Information Resources and Communication - National Library of Viet Nam ABSTRACT The rapid development of high technology and the Internet has contributed to English language education for the past few decades Technology provides learners unprecedented opportunities to practice English and involve themselves in authentic language use environments Since the 1980s, many communication and linguistic scholars have done relatively interesting research on computer-mediated communication, which often concerned four aspects: linguistic perspectives, social and ethical perspectives, cross-cultural perspectives, and CMC and group interaction In Vietnam, before the year 2020, conducting blended-learning English courses on online English teaching platforms or software was optional and not as common in colleges or universities in Vietnam as it is today Hence, there has been a limited amount of research on linguistic features or the relationship between linguistic items and computer-mediated discourse/ online English written texts made by EFL students in Viet Nam and abroad The study aims to investigate the linguistic features remarkable in online English writings made by EFL students at some universities in Viet Nam in the last five years Besides, it tempts to determine the relationship between these linguistic features' use and the students' English online writing quality Thereby, it can provide English writing instructors with suggestions for teaching and assessing Vietnamese EFL students' online English writings To conduct the study, the researcher applied descriptive and comparative methods and some instrument tools like the software Antconc, the software Virtual Writing Tutor, and a questionnaire Firstly, it found that these linguistic features were frequently used in the EFL students' online English writings Generally, they appeared more frequently in the final writing version than in the drafts Secondly, most were exploited with correct functions and no grammatical errors, except for the modal verb of ability/possibility and adverbial conjunctions referring to 'concession' Thirdly, in comparing three groups of writings at different English levels, the research revealed a considerable difference in using the linguistic feature regarding their quantity and frequency, which was caused by the different writing genres, the writing topics, and the authors' English levels Finally, the research findings showed that lexical diversity, number of academic words, language accuracy, number of complex sentences, and some cohesive devices (like subordinating and adverbial conjunctions) had considerable and positive influences on the quality of students' online English writings CHAPTER ONE: INTRODUCTION 1.1 Research background The rapid development of high technology and the Internet has contributed to English language education for the past few decades Technology provides learners unprecedented opportunities to practice English and involve themselves in authentic language use environments Since the 1980s, many communication and linguistic scholars have done relatively interesting research on computer-mediated communication, which often concerned four aspects: linguistic perspectives, social and ethical perspectives, cross-cultural perspectives, and CMC and group interaction In Viet Nam, before the year 2020, conducting blended-learning English courses on online English teaching platforms or software was optional and not as common in colleges or universities in Viet Nam as it is today Hence, there has been a limited amount of research on linguistic features or the relationship between linguistic items and computer-mediated discourse/ online English written texts made by EFL students in Viet Nam and abroad All the abovementioned facts have encouraged the researcher to conduct the current study 'Linguistics Features of Online English Writings in Vietnamese Higher Education Context’ 1.2 Aims The study aims to investigate the linguistic features remarkable in online English writings made by EFL students at some universities in Viet Nam in the last five years Besides, it tempts to determine the relationship between these linguistic features' use and the students' English online writing quality 1.3 Objectives The study includes the three following objectives: Investigating lexical features used by Vietnamese EFL students in their online English writings Investigating grammatical features used by Vietnamese EFL students in their online English writings Studying the relationship between the use of lexical and grammatical features and the students' online English writing quality 1.4 Research questions What lexical features are remarkably used in Vietnamese EFL students' online English writing? What grammatical features are remarkably used in Vietnamese EFL students' online English writing? What is the relationship between lexical and grammatical features and the quality of the EFL students' online English writing? CHAPTER TWO: THEORETICAL BACKGROUND AND LITERATURE REVIEW 2.1 Theoretical background 2.1.1 Computer-mediated communication and computer-mediated discourse analysis + Computer-mediated communication (CMC) According to Herring (1996), CMC is communication taking place between human beings via computers In text-based CMC, participants interact by means of the written word, e.g., by typing a message on one computer’s keyboard which is read by others on their computer screens, either immediately (synchronous CMC) or at a later time period (asynchronous CMC) Lauren (2016) defines computer-mediated communication as a broad term that encompasses a variety of semiotic/linguistic modalities, as well as technology interfaces and platforms (computer, mobile phones, tablets, social media, immersive online games, virtual workplace environments, and more) The term refers to communication that takes place through a mediating interface, and these interfaces create levels of structure that require linguistic and social negotiation Wherever there is human interaction, whether it is face-to-face, on a computer in a chat room, or through text messaging on a phone, there is language There are three characteristics in CMC research Firstly, the language of CMC is typed, and hence like writing, but exchanges are often rapid and informal, and hence more like spoken conversation Secondly, because of the fact that participants interact without the benefit of extra-linguistic cues as to the gender, identity, personality, or mood of their interlocutors in CMC, the language of CMC is hypothesized to be impersonal or distancing The last characteristic of CMC relates to the phenomenon of community formation in cyberspace, which generates norms of interaction and conflict resolution procedures in text that can be saved and mined later for insights into the genesis of human social organization (Herring,1996) + Computer-mediated discourse Computer-mediated discourse (CMD) encompasses all kinds of interpersonal communication carried out on the Internet, e.g., by email, instant messaging, web discussion boards, and chat channels (Herring, 2001) In terms of methodology, language-focused CMD research has drawn on methods and key concepts from a variety of linguistic research traditions (including pragmatics, conversation analysis, sociolinguistics, genre analysis, and the ethnography of communication), which have been successfully applied to study how people use linguistic resources to establish contacts, manage interactions, and construct identities within computer networks + Computer-mediated discourse analysis CMDA analyzes computer-mediated communicating products using methods adapted from language-based disciplines such as linguistics, communication, and rhetoric (Herring, 2001) It may be complemented by surveys, interviews, ethnographic observation, or other methods involving qualitative or quantitative analysis However, basically, CMDA is the examination of verbal interaction logs, including characters, words, utterances, messages, exchanges, threads, archives, and so on) In its broadest sense, computer-mediated discourse analysis refers to any study of online behavior based on factual, textual observations (Herring, 1996) CMDA is best considered an approach rather than a "theory" or a single "method." The CMDA method, on the other hand, enables the exploration and testing of alternative theories about discourse and computer-mediated communication According to Herring (2004), there are four domains or levels of language, ranging prototypically from smallest to largest linguistic unit of analysis (such as structure, meaning, interaction, and social behavior) and five discourse analysis paradigms (including text analysis, conversation analysis, pragmatics, interactional sociolinguistics, and critical discourse analysis) commonly invoked in CMDA research Based on the characteristics of online EFL writing instruction context in universities in Vietnam, the subjects, the participants, and the scope of this study, the researcher chooses paradigm 'text analysis' of CMDA to analyze texture or linguistic features employed in the students' online English writing samples when taking various online or blended-learning English courses at the University of Da Nang, Viet Nam 2.1.2 Online English writing The networked computer-mediated writing learning brings learners an online writing environment where students appear to value peer support while composing rather than simply receiving comments on written products (Hyland, 2000) Online writing comprises synchronous and asynchronous writing + Synchronous writing Synchronous writing is written communication in real time via discussion software on LAN or Internet chat sites with all participants at their computers at the same time (Richard, 2003) + Asynchronous writing According to Richard (2003), asynchronous or time-delayed communication using networked computers includes email, newsgroups, and conferencing software Generally, online writing instruction presents stimulating alternatives to traditional paper materials and writing tasks, and many commercial software and Internet sites are easy to learn, straightforward to use, and technically robust However, teachers should carefully consider choosing some software or Internet site most appropriate for the course purposes and know how to integrate them into a coherent writing course so that students can obtain significant benefits from taking part in their online writing courses 2.1.3 Linguistic features in English second language writings In Hinkel’s (2002) study, all textual features the university-level L1&L2 texts for the analysis basing on four criteria: a) Their textual functions, meanings, and the implications of their uses in text; b) The prominence of their uses, meanings, and functions established in corpus analyses of published academic genre in English; c) The prominence of particular linguistic and rhetorical characteristics found in composition instruction and assessment, and d) Whenever possible, consideration of the uses of proximate syntactic, lexical, and rhetorical features in NNSs’ L1s The researcher classified the textual features into two major categories: linguistic features and rhetorical features Generally, identifying the most common features in NS and NNS texts establishes priorities of syntactic, lexical, and rhetorical features in writings made by speakers of different languages Furthermore, determining order ranks of features’ median rates in NNS texts relative to those of NSs can provide valuable insights into L1 and L2 use with practical applications of the analysis applied in curriculum design and L2 teaching (Hinkel, 2002) 2.1.4.The relationship between linguistic features and English writing quality In general, the linguistic features examined by writing researchers fall into three large constructs: lexical, syntactic, and cohesion These features are important elements of discourse structures and have close links to the quality and development of English Second Language (ESL) writings (McNamara et al., 2010) + Lexical features and the English writing quality and development According to Crossley (2020), lexical items are considered the most commonly used linguistic feature to analyze the quality of texts The quality of lexical items can be classified into three categories: lexical diversity, lexical density, and lexical sophistication The first category refers to the number of unique words in a text The second one means the number of contents to function words; and the rest includes words that are more likely found in academic texts or words that are less concrete, imaginable, and familiar, or more specific words, or words that are less diverse based on context + Syntactic complexity and English writing quality and development Lu (2011) defines syntactic complexity as the sophistication of syntactic forms created and the variety of syntactic forms produced The underlying notion is that more complicated syntactic structures signify more advanced writing abilities (Crossley, 2020) Traditionally, standard ways to evaluate syntactic complexity include calculating sentence length with the assumption that longer sentences are more complicated and T-unit counts, where a T-unit is a dominant clause plus all subordinate clauses However, T-units features are problematic because they often report conflicting results across studies and can be challenging to interpret (Norris & Ortega, 2009) + Text cohesion and English writing quality and development Text cohesion is related to the inter-connectivity of text segments of text based on textual features and is an important element of writing because it can indicate lexical, semantic, and argumentative dependencies within a text (Halliday and Hasan, 1976) Text cohesion can occur at the sentence level or across larger segment gaps such as paragraph, chapters, or even texts The most common approach to identifying cohesion within texts is to examine overt connections between text segments including referencing previous elements, repeating lexical items, substituting lexical items and the use of conjunctions to connect ideas If cohesion in a text is not maintained, it may be difficult for readers to evaluate the systematic relationship between shared lexical items, at which point a reader's mental representation of the text may break, affecting comprehension (Crossley, 2020) 2.2 Previous related studies 2.2.1 Previous studies of linguistic features in English writings The reseacher examined some previous studies of linguistic features in English writings such as Biber’s study (1988), Hinlel’s study (2002), and Zein, Sinar, and Nurlela’s study (2017) All these studies focused on examining and comparing the frequency of lexical, grammatical, syntactic, or rhetorical features used in NNS students' texts with NS students' ones Based on the textual functional theory of language features as well as their usage habits in L1 and L2, some studies like Hinkel's analyzed textual contexts and explained the reasons for using each linguistic feature in the essays of different groups of L1 learners Although these studies examine the linguistic features of paper-and-pen English writings, some of their findings (e.g., prominent linguistic features in ESL writing samples with the frequency equal to or even higher than those in NS's ones) will be used as one of the bases for the current research to study linguistic features in online writings of EFL students in the higher Vietnamese education context 2.2.2 Previous studies of CMC and CMDA abroad and in Vietnam The researcher reviewed many previous studies of CMC and CMDA abroad such as Collot and Belmore's (1994) study named "Electronic Language", Yates’ (1996) study titled ‘Oral and written linguistic aspects of computer conferencing,’ Werry’s (1996), Herring’s (1999), Alkhataba's (2018), Thomas Cho's (2010), Fiona and Francesca's study (2013), and Robbert (2008) In addition, some previous studies of CMC and CMDA in Vietnam (such as Nguyen’s (2008, 2011); Pham et al (2014); and Pham’s (2020) were also investigated Generally, previous studies relating to CMC, CMDA, and online English writings focus on the following issues: i) Social, psychological, and pedagogical characteristics of virtual or online environments in English language learning and teaching; ii) Influences of virtual or online English learning environments to learners’ English learning, especially in communicative skill; iii) Students and teachers’ perceptions and acceptance of applying technology or social media in learning English; iv) Students and teachers’ perceptions and reflections on online collaborative writing in ESL classrooms; v) Linguistic features of emails or CM messages in chat groups made by computer users at workplaces Especially, so far, there has been no research on analyzing linguistic features in CMD or online writings made by EFL students in Viet Nam and abroad Therefore, it is highly reasonable and possible to conduct this study in order to bring new contributions to the field of linguistics in CMDA or online English writing learning and teaching in Vietnamese higher education context Furthermore, the researcher studied some previous studies of linguistic features and English writing quality In short, it can consist that there are clear and consistent associations between linguistic features and L2 writing quality and development Many previous studies reported that higher-rated essays include more sophisticated lexical items, complex syntactic features, and cohesiveness Developing writers also move toward more sophisticated words and complicated syntactic structures (Crossley, 2020) These findings will be a reliable reference resource for the current study’s researcher to conduct a similar analysis in chapter 6, finding the answer to this study's third research question CHAPTER THREE: RESEARCH METHODOLOGY 3.1 Research context 3.1.1 Teaching and learning English writing skills in the University of Da Nang The section is a brief description of the English writing curriculum with two main types of English programs: General English and English for specific purposes at several colleges and universities at the University of Da Nang, Viet Nam In terms of writing, there is some description of the purposes and requirements of teaching and learning writing English in each program 3.1.2 Research design The study employed both descriptive and comparative research methods going with quantitative and qualitative approaches to examine linguistic features significantly used by the UDN students in their online English writings The descriptive method was conducted to describe the characteristics of the study subjects, collect quantifiable information or quantitative data for statistical analysis, and hence withdraw the remarkable grammatical, and lexical features with their frequencies in the study samples The result was the answer to the two first research questions The comparative method was used to statistically calculate and compare the frequency of the remarkable grammatical and lexical features in the students' two online English writing versions The findings enabled the researcher to estimate how much the frequency difference of these features was between both corpora The research also uses the qualitative approach in both descriptive and comparative methods to explain the frequency differences in using these linguistic features in the students' online English writings and the impacts of the linguistic features on the quality of the writings 3.1.3 Subject of the Study The subject of this study composes of 135 online writings made by English-major students and Non-English-major students who have taken different English courses at Da Nang University of Technology and Science, Da Nang University of Economics, University of Education, and University of Foreign Language Studies) belonging to the University of Da Nang since the academic year 2017-2018 These writings are papers made in middle-term or ongoing English writing tests, and each one has two versions: Version one (V1) is the students' initial online English writing drafts on some platforms (i.e., Moodle, Google Class) at synchronous online English writing learning sessions without using any online writing aids; Version two (V2) is those edited at asynchronous English writing sessions with computer writing facilities or some online English writing aids 3.2 Data Collection and Analysis 3.2.1 Sample collection All writing samples for this study were written by students from the member colleges/universities of the UDN when taking a progress test or assignment of a blended learning or online GE/ESP course The students were required to complete their work by following a five-step writing process The following Table describes the population and samples of this study Table 3.1 The population and samples University Popul Sample Genres of Level of ation size writings English course Da Nang University of Science and Technology (DUT) Da Nang University of Economics (DUE) University of Education (UED) 135 75 42 27 University of Foreign Language Studies (UFLS) Total 3.2.2 Setting the corpora informal letter essay (descriptive, argumentative) essay (descriptive, argumentative) A1-A2 A2-B1 B1-B2 A1-A2 A2-B1 formal letter A1-A2 102 36 essay (descriptive, argumentative) B1-B2 B2- C1 306 135 11 standard academic writings Secondly, the total number of words in corpus was much more than in corpus As examining the usage of interpretive nouns in the writings by three student groups with different English levels, the study revealed that the first sub-corpora 1.1 &2.1 used the biggest number of interpretive nouns with the highest frequency among three pairs of sub-corpora The second sub-corpora 1.2 &2.2 stood at the second rank, and the rest sub-corpora had the lowest frequency The above statistical results reveal that the higher the English level of the students was, the more interpretive nouns were used in their English writings Besides, the English – major students often used this kind of lexical feature more flexibly and harmoniously than the two other groups of students 4.2 Vague nouns The study found thirteen vague nouns such as boy(s), girl(s), guy, human(s), individual(s), kid(s), man, people/person, society, thing(s), way(s), whatever, and woman in the students' online English writings Regarding frequency, four vague nouns: people, thing, way, and society appeared the most frequently in both corpora Word way and thing stood at the second and the third rank whereas, four nouns boy, guy, man, and woman reached the lowest rank among fourteen vague nouns Moreover, they did not present in the first writing version and appeared only once in the entire corpora In comparing the two writing versions, it found that the writers tended to employ vague nouns increasingly in the later version There were two reasons for this result: Firstly, the mean number of words in each written sample of corpus was considerably more than that of corpus 1; Secondly, the topics of all the writings mostly were about common social issues which usually relate to all kinds of people, general objects or things, and the entire society As examining the usage of vague nous in the writings made by three studentgroups with different English levels, the study revealed that the second sub-corpora 1.2 &2.2 used the biggest number of vague nouns with the highest frequency among three pairs of sub-corpora The first sub-corpora 1.1 &2.1 stood at the second biggest number position, and the rest sub-corpora contained seven vague nouns with the lowest frequency Furthermore, the study revealed that students with English levels B2-C1 did not tend to use many vague nouns in their online English writings The bias in employing the three most vague nouns people, society, and way could be because the topic of two-thirds essays was 'business and corporate social responsibility’ On the other hand, students with English levels A2-B1 and A1-A2 were favour in using people/person, thing, and way because most of the writing topics were about common social issues such as school, studying, or students' life; moreover, their English levels were not high, so their writings often included such basic vague nouns 4.3 Be as main Verb 12 The study found that be as the main verb was the second most favorite linguistic feature used in the students’ online English writings The extremely high usage of the verb be was due to its prevalence in attributive constructions in English texts Especially in the writings made by the student with the lowest levels (A1-A2), this linguistic feature was often used with predicative adjectives in ideationally simple propositions, which typically had a subject and a predicate Besides, it explored that verb be was used in three tenses (including simple present, simple past, present perfect) A partly, verb be was also used with several modal verbs such as can, may, might, should, could, or would to express a hedge or reduce the certainty degree when the authors/students described some events or situations based on their thoughts or opinion As investigating the use of verb be in the three student-group’s writings, the researcher found that it was used rather unequally The unequal usage of be-as main verb in different tenses mainly depended on the topics of the writings/essays For example, the essays in the first subcorpora were about the facts or present social issues, and thus they configurated most of the verbs in present tense and no past tense The writings in the second subcorpora usually described the authors' opinions about studying, work, or living styles, so the writers tended to use more modal verbs going with the verb be Whereas, the last group's writings were about personal life, story, or experiences and hence they had more be-as main verb in the past tenses 4.4 Public verbs There were twenty-five public verbs (including acknowledge, add, admit, agree, argue, assert, confirm, convey, deny, guarantee, insist, maintain, mention, offer, present, promise, protest, report, say, show, speak, state, tell, want, and write) in both versions of the students’ online English writings Of them, five public verbs want, write, show, offer, and present were used the most in both corpora In contrast, acknowledge, admit, assert, confirm, convey, guarantee, report, protest, and state were the nine public verbs with the lowest frequency because each was employed only once or twice in the entire corpora In addition, two-thirds of these public verbs had higher quantity and the frequency in corpus than in corpus There were two reasons for the increase or decrease in using the public verbs in the final versions Firstly, while the drafts were written synchronously on some website or platform at the online learning session, the final writings were composed asynchronously; so the writers had more time and technological assistance to correct grammar or spelling mistakes, choose more appropriate word items, or add some more ideas to make their writings better Secondly, the writers usually used spoken language (e.g., verb say, or speak) as composing initial writings at synchronous section However, they usually tended to employ academic words or written language items when editing their texts at asynchronous sections And as a result, informal words like public verb say, speak were eliminated and the final texts became more perfect 4.5 Private verbs 13 The study found thirty-six private verbs in the students' online English writings Among them, six private verbs: know, find, think, see, feel, and consider appeared the most frequently in both corpora Verb know and see had the highest mean frequency (1.2 tokens know or 1.3 tokens see per sample) Then, five private verbs including believe, hope, study, learn, and mean had the second-highest frequency in the students’ writings Each of them appeared nearly one token in each writing Occasionally, verbs learn, believe, or mean were used twice or three times per sample On the other hand, eleven private verbs (i.e., ascertain, conclude, determine, establish, guess, imagine, notice, prove, reflect, and sense) had the lowest frequency Each of them was employed only once or twice in the whole corpora This fact might be because these verbs belong to vocabulary levels B1-B2, while most of the student participants of this study had levels of A1-B Moreover, the essays or writings' topics played an essential role in impacting the authors' choice and usage of vocabulary for their works In comparing the two writing versions, it discovered that the total frequency of the private verbs in corpus was considerably higher than in corpus Most of these verbs were used more frequently in the final writings, while three verbs (i.e., mean, establish, decide) appeared a bit less in the final writings As investigating three sub-corpora, the researcher found a considerable difference in the use of private verbs in the writings of these three groups There were twentythree private verbs in the writings with English levels of B2-C1 twenty-five other private verbs appeared in the writings at levels A2-B1; And twenty existed in the writings in sub-corpora 1.3 & 2.3 This difference was because the number of samples and the average number of words per writing sample were not the same or similar in each pair of subcorpora In other words, the more writing samples and the higher the mean number of words in each sample there were, the more private verbs appeared in each student group’s writings 4.6 Predicative adjectives The study found two hundred and fifty-eight predicative adjectives in both versions of the students’ writings These adjectives belong to vocabulary with different levels from A1 to C1 They complemented the subject of a sentence or clause in the students' English writings Mostly all the predicative adjectives came after verb be in the sentences or clauses in the students’ English writings (both versions) Besides, there were twenty-two predicative adjectives coming after linking verb become Sixteen adjectives went with verb feel Ten adjectives and six adjectives came after verb look Regarding the frequency, the adjectives coming after verb be had the highest frequency, which was highly outstanding compared with those after the linking verbs This was because nearly all 256 adjectives went with the verb be; and their frequency stood at the second frequency rank in the total corpus and corpus Predicative adjectives going after linking verb become and feel stood at the second and the third 14 frequency ranks in the two corpora Predicative adjectives coming after linking verb look had the lowest frequency in the two corpora Comparing corpus and corpus 2, it revealed that the authors employed considerably more predicative adjectives in the final writings than in the drafts The increase was due to the characteristics of the asynchronous online writing process and the average length of a writing sample in each corpus Investigating three pairs of sub-corpora, it found that the first subcorpora (1.1 &2.1) contained the highest number of predicative adjectives ; The second pair of subcorpora (1.2 & 2.2) stood at the second rank using this kind of linguistic feature The last pair of corpora had the lowest number of predicative adjectives Each pair of subcorpora had the most adjectives going with verb be In the second group, the number of each type of predicative adjectives mostly was at a positive ratio to their frequency: The more adjectives there were, the higher frequency they had in the corpora 4.7 Intensifying adverbs There were twenty-one intensifying adverbs used in the students' online English writings Of them, four intensifying adverbs or amplifiers (such as very, really, extremely, and too) appeared the most frequently in both corpora Especially adverb very had the sharply highest frequency compared to the other adverbs In comparing the two writing versions, it found that the total number of intensifying adverbs in the latter version was slightly less than that in the initial one Particularly, adverb really in the initial writing version had the decreasing frequency compared with those in the last one Some intensifying adverbs (i.e., absolutely, considerably, deeply, fully, and highly) had low frequencies in the drafts but get relatively high ones in the second editions These amplifiers were used to replace tokens very and really that appeared most frequently in the first writings Moreover, in the edited writing samples, the writers used new adjectives to replace the ones going with very or really in the drafts to make the sentences more concise As a result, this substitution also reduced the number of tokens very and really in corpus In comparing the three pairs of sub-corpora, the study found a similarity in the number of this linguistic feature used in the students’ online English writings However, students with English level B2-C1 did not tend to use many intensifying adverbs in their online English writings Besides, there was no bias in employing any adverbs, even though very and really are inherently adverbs most commonly used in daily and computer-mediated communication This could be due to the following reasons: These students' English writings were argumentative essays on business, commerce, science and technology, technology, government, and professions in society Therefore, the writers tended not to use too many lexical items like amplifiers or downtoners (intensifying adverbs) in their essays compared with others with sociocultural topics Moreover, the writing samples' authors were English-major students of the UFLS with pretty high levels (B2-C1), so when composing essays in English, they often avoided repeating vocabulary or sentence structures by using 15 synonyms and antonyms or paraphrasing grammatical structures to make their writings better On the other hand, the authors of the papers in subcorpora 1.3 & 2.3 seemed to be favour of using very and really in their English writings Adverb very occupied the highest frequency of eleven adverbs The adverb really had the secondhighest frequency and was followed by adverbs too and quite The remaining adverbs appeared only a few times in the papers This statistical result is also related to the writers’ English levels, the genres, and the topics of the writings CHAPTER FIVE – RESULTS AND DISCUSSION OF GRAMMATICAL FEATURES 5.1 Modal verbs of ability and possibility There were five modal verbs of ability and possibility (can, could, may, might, and able to) used in one hundred and thirty-five students’ online English writings Among these five modal verbs, can was the most favorite modal verb of the students as they composed the writings, whereas might was the least exploited word in both writing versions Modal verb could, may, and might which are known as hedging, were employed much less than can in the two corpora The writers used these modal verbs in conditional structures to express wishes, predictions, or their individual points of view with their certainty As comparing the two corpora, the study realized a slight increase in using modal verbs in corpus compared with corpus Among the five modal verbs, modal verb can had the highest increase, modal verb could and may appeared three more tokens in the second writing version; Modal verb might increase only one token, and the number of pseudo modal be able to decreased a half in the last writing version When comparing three pairs of subcorpora, the study found a significant similarity in using this type of grammatical feature in three student groups’ writings Besides, the study revealed that the higher English level students had, the more modal verbs of ability or possibility was used evenly Especially, several writings made by the students with English levels A1-A2 contained no modal verbs of possibility in both writing versions This may be because the authors of these writings had the low English level and had not been aware that they need to be cautious in describing their prediction or individual opinions by employing modal verbs instead of using the verb be or ordinary verbs in the simple present or future tense 5.2 Modal verbs of obligation and necessity The study figured out four modal verbs of obligation and necessity (including have to, must, need, and should) in the online English students’ writings However, it revealed that modal verbs of obligation and prohibition such as must, have to, ought to did not appear frequently; whereas, modal verbs should and need were employed much more in the students’ writings and occupied nearly 70% of the total frequency of these four modal verbs In comparing the two versions of the students' writings, the study found that modal verb must had the highest increase in frequency in the final version compared with the 16 initial Besides, although modal have to was another modal verb of obligation with a similar meaning to must, it had the frequency rising in corpus considerably less than must In concerned with the modal verb of necessity, both should and need appeared in corpus less than in corpus Because the writers used should to express more meanings (such as necessity, advice, and recommendation) than need, the number of should was more than need in both versions; and its frequency in version was considerably higher than that of need Examining the three groups of students’ writings, the study found all four modal verbs (should, need, have to, and must) with different frequencies in each group In the first and second groups at levels B2-C1 and B2-B1, most of these modal verbs were distributed relatively evenly and equally except for must Should and need were the students' most favorite modal verbs in all three groups, whereas must was their least favorite In the writings at the lowest levels, there was a considerable bias in using modal verbs expressing advice should and modal verbs expressing obligation have to This could be because of the writers' low English levels, the writing genres, and the writing topics 5.3 The passive voice There were two hundred and four passive patterns with the verb be, and one hundred and nine ones with modal verbs in both corpora Most passive structures were used in three common English tenses: simple present, simple past, and present perfect tenses Generally, the passive voice seemed to be the student participants’ favorite grammar feature employed in their online English writings because the mean number of passive patterns in a sample was relatively high In comparing the two versions of the students’ writings, the study found that the total number of passive patterns in corpus was slightly higher than in corpus Besides, the number of verbs employed in passive voice increased slightly in the final writing version compared with the initial one Sometimes, a few passive sentences with by-phrase used indefinite personal pronouns or general nouns of an agent like everyone, all, people, etc., which are often recommended not to use or limited according to the principle of academic writing However, using the indefinite personal pronoun or general noun of agents could be the authors’ intention to emphasize the actual subject of the activities in the writings In comparing the three pairs of subcorpora, the study found a big difference in using this grammatical feature in the students’ writings at different levels The first subcorpora contained the highest number of passive patterns and ordinary verbs The second subcorpora stood at the second rank regarding the frequency of passive patterns and ordinary verbs used in passive voice in the students’ writings The last subcorpora included the fewest passive patterns and the lowest number of main verbs in passive voice Lastly, it found that the higher the students' English levels were, the more passive structures appeared, and the less bias was in using passive voice with the verb be or modal verbs in their online English writings 17 5.4 Adverb clauses of cause The study figured out forty-three adverb clauses of cause in the studens’ drafts and sixty-two clauses in the final writing version These clauses began with one among four adverbs of causes such as as, because, for, and since Generally, it realized that adverb clauses of cause were not employed frequently in the students’ online English writings In comparing the two versions, the study revealed a slight increase in the frequency of this syntactic feature in the last version compared with the first one The number of adverb clauses beginning with because or as in Corpus was twice as high as in Corpus Contrastly, the number of clauses beginning with since was nearly the same between the two corpora, and the subordinating clauses that started with for appeared only once in each corpus In comparing three groups of the students’ writings, the researcher found thirtythree adverb clauses of cause in sub-corpora 1.1& 2.1, fifty clauses in sub-corpora 1.2& 2.2, and twenty-two clauses in the rest Also, the study discovered a considerable difference in composing adverb clauses of causes in three groups of students' writings The group at levels B2-C1 used all four subordinators as, because, for, and since, to initiate these adverb clauses In the second group, subordinator because was used majorly to start the adverb clauses of cause The students employed as and since to compose this grammar structure, but each of them occupied 10% of the total frequency of subordinators On the other hand, there were two subordinators because and as used to begin the adverb clauses of cause in sub-corpora 1.3&2.3 The writers had much more favor in using because than as to compose this kind of structure in their online English writings This could be because the English levels of this student group were not high (A1-A2), so they tended to overuse familiar conjunctions referring to cause/reason like because in their writings 5.5 Coordinating conjunctions There were five coordinating conjunctions or coordinators (i.e., and, but, for, or, so) used in the students' online English writings considerably Of them, coordinator and and or appeared with the highest and the second highest mean frequency Coordinators but and so stood at the third and the fourth positions with the mean frequency of nearly a third token per sample Coordinator for with the function to explain a reason was the least common coordinating conjunction because it appeared only once in each entire corpus Comparing the frequency of coordinating conjunctions in two corpora, the study found the number of grammatical features in corpus was quite higher than that in corpus Among five coordinators, and and or had the sharp increase in the quantity in the final writing version This is entirely understandable because and is the most basic and common word in any syntactic structure Besides, a writing V2 was usually longer than a V1 and hence the frequency of this coordinator became higher ... 75 42 27 University of Foreign Language Studies (UFLS) Total 3 .2. 2 Setting the corpora informal letter essay (descriptive, argumentative) essay (descriptive, argumentative) A1-A2 A2-B1 B1-B2 A1-A2... 1758 10393 28 8 & DiscusB2-C1 sion essay Argument Group 1379 7 420 146 51 & DiscusB1-B2 sion essay 48 Descriptive Group 1 326 7 125 148 writings, A2-B1 A1-A2 Letters/ Emails Table 3.4 Total Words... tokens sample 36 Argument essay Group B2-C1 1797 1 027 6 28 5 SubCorpus 2. 2 SubCorpus 2. 3 51 Argument essay Group B1-B2 1607 9533 187 48 Letters Group A1-A2 1571 827 9 173 3.3 Selecting linguistic features