Using Visualisation to Enhance Intermediate EFL Learners’ Narrative Writing

Studies show that visualisation while reading has a positive impact on both L1 and L2 reader’s comprehension of a text and that visualisation improves L1 learners’ writing, but there are no studies relating to whether visualisation may help L2 learners’ writing. Using visualisation as a learning strategy, this paper reports on how visualisation training might affect a group of Chinese intermediate EFL learners’ narrative writing. Quantitative data from the pre-test and post-test did not show statistical significance, but the analysis of the qualitative data from learners’ written reflections indicated a positive correlation between the use of visualisation and engagement with narrative writing and higher levels of classroom enthusiasm and participation. This study highlights the need for further work to differentiate the most effective aspects of visualisation training for enhancing the quality of narrative writing. It also suggests potential for the integration of visualisation into different writing development curricula and to promote the role of visualisation in the wider process of second or foreign language learning.


Introduction
Visualisation is generally regarded as the process of creating pictures in the mind by deliberate stimulation activities. Majoy (1993) predicts that "…visualisation will become one of the most powerful, effective, and necessary tools for teachers in the years to come. Harnessing inner space will revolutionise teaching and learning" (p.64 ). Majoy's prediction has come to a degree of fruition. For example, Arnold (1999) finds that, "the use of visualisation is becoming more and more common in diverse areas of educational endeavour, whether in science, mathematics, creative writing or second language learning" (p.263). The study by Tomlinson (1998) finds that visualisation is functionally significant in L1 reading. He adds that "L2 learners do not typically visualise when reading in L2", but "those who do so tend to

Rationale of Using Visualisation to Enhance EFL Learners' Narrative Writing
Historically, learning strategies in second language acquisition (SLA) started in the mid-1970s. Early studies (e.g. Naiman et al., 1996;Rubin, 1987) mainly record which strategies were employed by successful learners, as recorded by researchers using their own observation and self-reporting by the learners. Studies that specifically involved SLA learning strategies in China began in the 1990s, with seminal work by Liu (2002) and Wen (2016). Key conclusions of relevance from these studies was that Chinese student writers use their L1 ability and context to inform their L2 writing, particularly in narrative composition writing, observing that, "L1 is more likely to occur in process-controlling, ideagenerating, and idea-organizing activities than in text-generating activities" (Wang & Wen, 2002, p.239).
Visualisation as a potential holistic learning strategy is viable. Cohen (2014, p. 189) reports that test subjects were asked to indicate their use of strategies for writing that included a specific reference to visualisation, which Cohen notes is a technique reflected in general language strategy literature. Earlier in the same book, Cohen (p. 24) has noted that English speakers learning Japanese might need to use a variety of functional visualisation strategies to learn kanji characters, since logographic characters do not relate to the English alphabet. Developing the use of visualisation as a learning strategy, Cohen (2014) notes that experimental findings on a story retelling task seems: to show the positive effects of the treatment in terms of advanced preparation for language tasks. An increase from pre-to post test in "drawing pictures to help remember the story" correlated significantly with increased ability to correctly order the elements of the story. (p199) In the experiment he relates here, using visualisation as a planning strategy, "was reinforced through several different learning activities" (p.199). Cohen notes that visualisation is likely to benefit preparation for writing, being able to "better order the elements of a story", a key element in successful completion of the task. Making the general assumption that one variable will influence another, the influences were also potentially reciprocal. Cohen lists the influences he was referring to then notes that, "use of a mental picture of [a] the favorite city while speaking correlated positively with a higher vocabulary rating (r=.46)" and that "thus, using the strategy of visualization here seems to have helped the students focus on the task at hand" (p.201).
In researching aptitude and motivation in SLA, Dörnyei (2010) notes that what he describes as a cognition-emotion-motivation amalgam featured a salient imagery component. In relating this, Dörnyei (2010) is speaking about the internal construct of possible selves that involve tangible images and senses, as a reality for the individual: people can "see" and "hear" a possible self. He spoke about the possibility for visualisation to translate interior possibilities into intention and thence real action. Understanding visualisation in this context made it a compelling learning strategy to explore and use as a methodological base.

Review of Related Literature
In the literature, the term "visualisation" is not well defined. Psychologists Kolb and Whishaw (1995) call it the ability to conjure up mental images of things that are not physically present. According to them, visualisation is an "ability" which leads to the formation of a mental image product. Tomlinson (1997) defines visualisation as the process of seeing pictures in the mind, where "visualisation" is the process and "pictures" are the product. Later, Tomlinson (1998) further defines visualisation as "the converting of words on the page into pictures in the mind" (p.265). Therefore, visualisation is also a process of "conversion" and pictures in the mind are again the product. Arnold's (1999) definition is that visualisation is the mental images called up for some purposes. "Calling up" indicates a mental process, with images as the product of that process. Nearly a decade later, Arnold et al (2007) take this definition further and define visualisation as the process of forming and using mental images for some purposes, for instance of creating what has been referred to as "movies in the mind". In his renewed definition, visualisation is still a process, with mental imagery as the product, adding the aspect of movement to a static image. Dürsteler (2008) defines visualisation as formation in the mind of the image of an abstract concept, here, visualisation is taken as a process as indicated by the term "formation". Dürsteler does not define whether the images formed are static or moving. Taken together, we can see that the term "visualisation" is not consistently defined. In this study, the term visualisation is used to refer to the formation of mental imagery by means of deliberate stimulation activities. Tomlinson's (1997Tomlinson's ( , 1998 findings significantly correlate higher interest and involvement to reading activities. Tomlinson (1998) also finds that "most L1 readers typically visualise when reading descriptive or narrative texts but do so with differing degrees of vividness" (p.267). Tomlinson (1998) notes L1 readers make much more use of visualisation than L2 readers and comments, "If learners do not see pictures in their minds or the texts they are reading then they will have great difficulty in achieving global understanding and their experience of the texts will be fragmentary and shallow" (p. 227).

Visualisation and its use in language learning
In her research, Masuhara (2000) finds that almost all the L1 proficient readers reported the experience of what she called "multi-dimensional mental representations" such as visual images, audio or kinesthetic representation in the mind, when working with text-based materials. Similarly, Tomlinson (2000) expands on this theme by pointing out that this idea of multi-dimensional representation is vital in all aspects of language production. He sees that when we write or speak, immediately before, during and immediately after communicating we can see that we have mentally represented what we communicated using a combination of internal images, interior speech and representations of our emotional affects. The words we use therefore are basically a means of trying to represent to others what is "see" in our minds. In their study, Islam and Mares (2014) proposed that, Encouraging higher-level cognitive skills means adapting materials in such a way as to require students to hypothesize, predict, infer, make connections and associations and visualize. This type of higher-level cognitive activity engages and motivates students as well as assists in transferring language skills already developed in their first language to the target language. (p.90) Taking Masuhara (2000), Tomlinson (2000), and Islam and Mares (2014) together, it is a reasonable conjecture to conclude that visualisation is related to multi-dimensional representation, which is intrinsic to all aspects of language formation. This clearly requires emphasis on higher-level aspects of cognition, which in turn require higher-level integrative thinking, stimulated by developed materials and significantly enhanced teaching activity that is focused on developing connections between all aspects of human functioning.
Whilst it is true enough to point out that visualisation is a natural enough process, it is also clear that there is a great need to develop whatever innate abilities a student may have and encourage them to expand and apply them to greater effect. The realisation of the potential of visualisation is a worthy goal for learning and development.

Narrative writing by intermediate EFL learners in China
There is no one definition of narrative universally accepted by researchers, partly because "narratives" are almost universal adjuncts to all aspects of human endeavour. They arise in every situation. The importance of narrative has long been recognised. Hardy (1968) places a universal emphasis on narrative in saying that "we live in a world composed of narrative, a world in which we dream in narrative, daydream in narrative, remember, anticipate, hope, despair, plan, revise, criticise, gossip, learn, hate and love by narrative" (p.5). Narrative continues to be recognised as a broad, expanding and important field of study. For instance, Goosseff's (2014) exploration of the differences between objective language and narratives and how differences affect rhetoric demonstrated that only narratives convey the experience of objectivity. This makes "narrative" more effective in persuading people to change than the limiting contribution of "objective" data and explanatory knowledge, that is provided by rhetoric. In other words, narratives contextualise and "humanise" rhetoric, which is a confirmatory, if less prosaic, way of confirming Hardy's emphasis. Thus, at one level, narrative writing is a record of the search for meaning such as lessons learned from experience, but narrative is also the key product of complex, cognitive and intuitive language-related processes. Narrative is written language employed to communicate giving form to ideas and conveying the deep and resonant life-meaning that humans have always derived from the practice of storytelling. Ryan (2007), however, warns against obsession with theory. She says, "Assessing the narrative status of a text is not a cognitive question that we must consciously answer for proper understanding" (p. 33). She warns against inflation of the term narrative into theory that detracts from practical engagement.
In the context of the present research, narrative writing that "reports an event or tells a story of something that happened" by Richards et al. (2002, p.337) is taken as the definition. Narrative writing is a part of the Chinese EFL curriculum in both secondary and higher education. In the context of the present study, narrative writing is one of the four basic genres: narration, description, exposition and argumentation. However, not enough importance has been attached to narrative writing in the curriculum, nor in the actual teaching. This might be the result of the washback effect that operates at two levels: a micro level, in terms of the individuals who are affected by the particular test, and a macro level, in terms of the educational system or society (Bachman & Palmer, 1996). The writing practice by the student participants (in question) who majored in English was very much dominated by the national English examinations such as the Test for English Majors Grade 4 (TEM 4). The analysis of all the writing tasks in TEM 4 from 1995-2019 shows that not a single topic is directly narrative related, although the candidates, in supporting their views in the argumentation might use some episodic narration. In order for the students to perform well in the examinations, both the teachers and the students themselves made every effort to practice argumentative writing, leaving narrative writing unattended. In this research, the author's focus is simply to give the students the tools to explore the relationship between visualisation and narrative.
Based on the above review of literature, and in the context of seeking to engender motivation to learn, three research questions were formulated:  Dai (2017) finds that, "…teachers encouraging students to tell stories in narrative form is often meant to reduce students' fear or aversion to writing, not to systematically improve students' writing ability from the perspective of narrative skills" (p.69). This may be a necessary, but limited perspective that does not explicitly build motivation in the classroom. Dörnyei (2005) process perspective describes the author's motivation as a teacher to design quality learning material. This consists of the four main dimensions: creating the basic motivational conditions, generating initial student motivation, maintaining and protecting motivation, encouraging positive retrospective self-evaluation.

Methodological Approach
The author recognised that these design process dimensions were also related to the outcome the author wished to engender in the students; that is, to use visualisation as a learning strategy for them to take responsibility for their own motivation to learn, following the initial teacher-led impetus, using positive retrospective self-and-peer-evaluation. Dörnyei and Clémant (2001) argue that it is not quantity, but the quality of the selected strategies that mattered, and that teachers cannot be both effective and motivating using only basic techniques. Seliger and Shohamy (1989) assert that "the degree to which the time frame established by the research context can be extended to the real world to which the results of the research will be generalised" (p.110) is relevant to the whole process. This basically means that the closer the research results are to being applicable in real contexts, the more intrinsic value the research may have. To find answers to the research questions, an experimental study was conducted in a university in China. This visualisation training experiment was carried out over five weeks with two-class-hour session per week, the same time frame as in the curriculum. This arrangement was meant to enhance the likelihood that the experiment would have real-world relevance with the research context being closer to a simulation than "simply" transferable or inferential experimental group results. Dörnyei (2005) in writing about linguistic aptitude research notes: Contextually sensitive measures of language aptitude open up brand new possibilities for integrating aptitude research into mainstream SLA studies, and they also allow researchers to link cognitive abilities to instructed SLA and classroom practice in a useful way (p.63). The present research was conceived and implemented in this context, seeking to understand motivation to learn by exploring aptitude through developing self-reliance on the technique of visualisation, which has applications intrinsic to, but not limited to, narrative writing.

Participants
In the present study, fifty seven students and three teachers were involved. The student participants were a cohort of 57 students majoring English language and literature in a Chinese provincial university. They were in two natural classes: with 32 in one class and 25 in the other. The class with 32 students was assigned as the experimental group, with an aim to generate more data, while the class with 25 students was the control group. The mixed male and female participants' ages varied from nineteen to twenty-one. Such a natural group formation was representative of intermediate Chinese EFL learners in mainland China. All of them were studying for a degree of Bachelor of Arts in English Language and Literature; hence it was not a disciplinarily diverse student body. These participants were chosen also because at their learning stage they were beyond simple narrative writing skills but needed assistance with deeper skills in order to become successful, independent student writers. The main benefit of strategic instruction in narrative writing is to encourage students to monitor their own thinking and make connections between writing and their own inner experiences, learning to express their authentic voices in written form. Thus, the more important issue for these students at this stage were to learn, as Hyland (2007) puts it, to produce effective and relevant texts.
Ethical clearance was done before the experiment. All the participants were informed of the major activities they were to be involved in and all of them signed a consent form. To keep the anonymity of the participants, only their initials were used in the present paper.
The students needed stability in their learning routines, and they needed to adhere to the expected curriculum upon which the students would be tested. All the variables were minimised to enhance internal validity for the process.
Teacher participants in this study were the writing teacher (ZL) and two assessors (RQ and LY). The visualization training activities as intervention were carried out by ZL, a colleague of the authors to ensure the objectivity of the resesearch. ZL did the visualization training as part of his teaching routine. ZL had taught English for over ten years by the time of the experiment and he had taught the English writing course for three academic years. As part of his workload, he met all the student participants for a two-hour lesson once every week. All the training activities were designed by the author and were approved by the writing teacher beforehand. The disagreement between the author and the teacher regarding the training activities were settled through discussion beforehand.
The assessors RQ and LY had over fifteen years' intermediate English teaching experience and were experienced in assessing students' writing in English. They graded student participants' written work in both pre-test and post-test. They also participated the process of coding the qualitative data produced by students' reflections. The tests were graded independently by the assessors. Discrepancy up to 3 marks or more were rechecked together to come up with a final score. After the tests were graded, fifteen were randomly selected to check for inter-rater reliability between the two assessors. Inter-rater agreement for scoring was found to be greater than 97% and was judged to be sufficiently high.

Design of the visualisation training activities
The most important tools used in the present study were the visualisation training activities. The main objective of this training was to help and encourage the students to perform visualisation activities before, during and after narrative writing sessions, with the aim of creating enhanced impact. In other words, this research emphasised teaching EFL writing as a real-world model of effective communication.
The researcher realised that this might mean that the activities and their products took precedence over linguistic accuracy in writing narratives, as would be more commonly found in the general practice of teaching narrative writing to the students involved. In designing training activities, three principles were followed: 1) providing interesting materials; 2) designing engaging activities with clear objectives, and 3) writing on personalised topics. The development of the procedure for using the visualisation materials was informed by a) Tomlinson (2003) in developing principled frameworks for material development and b) the visualisation activities in Byers et al. (2008) used as instructional strategies. They maintain that the practice of visualisation is a powerful process that assists learners to construct meaning in listening and reading by means of incorporating their knowledge and previous experiences to form images that extend learners' comprehension, enrich their personal interpretations and stimulate unique ideas for writing. Imaging provides the opportunity for students to experience vicariously what they hear, read and write. Byers et al. (2008) see the process of visualising as powerful in helping students derive meaning by integrating their knowledge and experience into the stimulus information that they receive from reading and listening. They also assert that the images formed as a result of visualisation do not only enhance students' receptive skills, but also stimulate their productive skills such as writing.
Based on the above-mentioned resources and principles, a framework as shown in Table 1 was developed to guide the design the visualisation training activities in the present research.

Table 1 Framework for Designing the Visualisation Training Activities
Since the present research was conducted in sub-optimal, actual classroom time activities, amongst competing priorities for the students, there had to be a level of flexibility built into the activities. The stated aim of affective engagement was in itself a huge multi-variate one that was likely to only have indicative outcomes, at best. However, this framework was the best statement of principle that could be constructed as a guide, both beforehand in preparation and as dynamic adjustments needed to be made in the delivery phase.

Stages
Procedure Principle Objectives Topic Selection Find appropriate topics that are potentially capable of engaging students affectively.
Topics with which learners are able to affectively engage are necessary for effective learning.
To find topics with the potential to engage the participants through visualisation.

Readiness Activities
Introduce topics that will increase learners' readiness to experience familiar situations in multidimensional ways.
Experiencing a familiar topic is an eclectic process involving multi-sensory imaging, establishing affective and cognitive connections To prepare learners to experience topics holistically, in English, in a way that helps learners to relate connections between visualisation and narrative.

Experiential Activities
Devise appropriate activities to help learners engage with the topics experientially.
Holistic comprehension as opposed to linear helps learners to process a text to achieve To enable learners to move away from their tendency to focus on language forms so that they can form images of situations, settings, characters and events.

Response Activities
Devise activities that help learners relate to their experience through imagery.
Learning is facilitated by developing students' mental representations of their experience To encourage learners to visualise their experience before conceptualising text that might represent it.

Activities -Writing
Devise activities that help the learners to put their mental representations into words.
Developing mental representation facilitates richness of narrative writing.
To help learners develop the ability to form mental representations in order to enhance narrative writing intelligently and creatively. Yong Yi Based on the above framework, five two-hour lessons were planned by the author and delivered by the colleague (LC). The lessons would start with generic "instructional" delivery in the common mode for Chinese education; for instance, the importance of writing in one's life and the general possibility of improving L2 writing by using different methods, but the training also focussed on explaining what visualisation was so that the students would become familiar with using this concept of visualisation skill in their language learning and would apply it, especially in their narrative writing. Beyond generic informative tuition, readiness, experiential, intake and input activities were carried out. Examples follow: • Readiness: students were asked to visualise past narrative writing experiences; to see in their mind's eye their experience of writing narratives in the past and imagine they were telling their own students what it was like to write a narrative o After a short discussion, two short stories were introduced. Before reading, students were asked to visualise their own dreams in life, whether they had come true and, if so, how. The purpose of this type of activity was to bring students' past experience to their consciousness so that they could be aware of the difference they might experience in the following lessons. • Experiential: students were asked to read and visualise themselves as the leading character in a story. In one example, they were asked to listen and try to experience how a "dream" was realised; in another, to read and experience how "I" as a stepmother anticipating meeting her adopted daughter for the first time might feel. o After reading each story, students were asked to sketch the images they saw in their minds whilst reading. Students then exchanged their sketches and discussed to what extent they thought they had visualised the stories in interior images. • Intake: students were encouraged to articulate what they know and how they understand a text.
For example, students were asked to talk about the major barriers to realising the "dreams" faced by leading characters in the stories and how they might have removed those barriers, focussing on how "visualisation" plays a part. • Input: here, the activities take the students back to the text to discover how language was actually used in the narrative. For example, they compared two stories and decided which story they liked better, according to the language used in the texts and the way in which preferences were influenced by visualisation imagery. • Output: generally presented as homework. After lessons, students were given a 200-word writing task to finish at home by following a visualisation instruction given to help them practice translating their mental representations into words. Since there were three experienced teachers involved, there was variation in the actual delivery of the lessons -another variable that it was difficult to control in a natural teaching setting. Amendments were not intended or likely to change the overall validity of the plans, as the essence of all the teaching was to train students to use visualisation, but the process showed the complex reality of language teaching where teachers tend to adapt the plans for their specific purposes or to suit their own teaching styles.

Participants' reflections
McDonough and McDonough (1997) note the increasingly persuasive effect of reflection in the narrative form that is of increasing importance in language teaching. They point to the significance of the genre of narrative reflection, perhaps in the form of "diaries" that is used to explore accounts of actual participant experience. In this way, reflection appeared to be an appropriate qualitative research tool for this study. Thus, the participants were required to write a comprehensive reflection on their experience of participating in the visualisation training activities, with no word limit and the language for that writing could be either in English or in Chinese.
By the end of the training sessions, the participants were encouraged to write a refelction on the whole training session. No specific guidelines were given other than writing whatever they wanted to share. It was basically free reflection with an aim to gather a wider range of data. Thirty reflections participants were submitted to the writing teacher ZL. The number of words written varied between about 500 and 1000 words, twenty-five of which were written in English and five written in Chinese because options had been given to them as to what language to use in the reflection.

Pre-test and post-test and the assessment results
Pre-test and post-test were taken respectively before and after the visualisation training. In the pre-test, the participants were allowed 30 minutes in a test mode to write a narrative on a life-event that they had experienced. This event, in one way or another, was to have been seen as "unforgettable". There was no word limit to the writing. The post-test also asked students to write for 30 minutes on another lifeevent whether pleasant or not, but which had perhaps taught a moral lesson. There was no word limit to the writing, either. The topics for both tests were the same in essence, although the wording was slightly different to avoid participants writing the post-test from memory. Both the tests were conducted within class hours. All submitted writing was turned into Word document files for avoiding the influence of handwriting on the assessors. All the Files were named with numbers for anonymity and sent to the assessors who were not aware which was the pre-test or post-test. The assessment was done by following the rubrics of TEM4, Test for English Majors (4) which assesses Chinese students' English proficiency at the end of the foundation stage, the composition part consisting 15 per cent. The average grade of the post-test of both the experimental group and the control group showed minor increase from the pre-test by 0.1(CG) and 0.2(EG) which was of little statistical significance. Therefore, the quantitative data was not analysed in this paper.

Analysis of the Qualitative Data
When all the hand-written reflections from students were collected, they were entered into a document file. Thematic analysis was used analysing the qualitative data in the present study. Thematic analysis is a qualitative, inductive research methodology such as that of Braun and Clarke (2006) use to identify patterns (or "themes"). It involves minimal organisation but clearly reflects the data set in order to capture "something important about the data in relation to the research question and represent[s] some level of patterned response or meaning within the data set" (ibid, p.82).
In coding the data from the participants' reflections, different colour highlighters were used to note where these words and phrases of note appeared, the author and the two colleagues independently identified phrases and words that indicated meaningful and significant comments made by the participants. To avoid the phenomenon known as "confirmation bias", in which any enquirer may be drawn towards what they believe to be significant to them as opposed to what was significant to the student data source (Dahlberg et al. 2007) , the encoders of this study followed the rule "the phenomenon shows itself to us" (ibid, p.236) from the students' perspectives, whilst questioning their "own choice of that particular passage" (Dahlberg et al. 2007, p.238). As in all such coding exercises, it is clear that the deriving and assigning of coding to words and phrases has a degree of arbitrariness about it, but researchers such as Gheradi and Turner (2002) recognise the difficulty, accepting that a researcher is a participant in the process and has to exercise judgment in order to "articulate a theoretical understanding of a non-standardised data" (p.92). Gheradi and Turner conclude that if an emergent pattern is "useful and sensitively constructed… it may turn out to be recognisable, appealing and useful to others as well" (pp.93-4).
Using an iterative process of revision and discussion, each encoder read through the student reflections three times and come up with six themes: a) Newness of the approach b) Positivity in general c) Negativity issues d) Enthusiasm and engagement e) Students' own understanding of "good" writing f) Students' own understanding of visualisation g) The following is a detailed discussion of all the themes with evidence from the student participants' reflections.

The idea of using visualisation as a conscious technique was novel
The majority of the students (93%) expressed curiosity towards the visualisation-training activities from the beginning. They also found most activities were interesting. This was reflected in the reflections. For example, the reflections most often typified was this quote: "The training activities were very interesting". In the reflections, twenty students out of twenty-five mentioned they were interested in the activities in one way or another. There are such comments, as "We have not had such exciting lessons for a long time." Students were used to their course textbooks, in which the teaching of writing follows a pattern of "diction, sentence, paragraph and whole text". In reality however, students don't really engage with "whole text" very much, except perhaps in connection with writing argumentation. Thus, when the students were given encouragement to engage with "whole text" in connection with narrative writing, this was a novel and generative experience for them.

Visualisation training helped students' exposure to a positive new teaching method
All the student participants expressed surprise at this teaching method. Student XQ wrote three diary entries. She was the only one who mentioned reading about visualisation before. She wrote that long ago she read an article about a teacher asking the students to form a picture in their minds and then describe the picture. "I did not expect to be learning this method in my own class." To Student XQ, forming these pictures in the mind was like watching a film that she was directing by herself. She wrote: "…the pictures are rich in details. The leading characters smiling, frowning, even an eye contact can be the focus of my writing. I find that the big and empty words I used to use very funny and it seemed that I was talking nonsense(sic)." She reflected that before she received the visualisation training, she was not able to come up with details resulting from visualisation, and that was one of the reasons why she only used "big, empty" words to describe the vague and general sketches without many details. Although Student XQ found that the details were very useful, she sometimes felt that it was hard to decide what specific details should be chosen for the writing. "When there are many details occurred to my mind, I find it hard to get the focus and hard to organise the whole structure so that the whole composition may be coherent." She also asked herself, "Just as in a film the shots are given one after another and there are inevitably leaps in between. I was wondering that there might be disconnection when I chose one detail after another. Can this be difficult for the reader to understand?" FX reflected in her diary that this way of learning to write encouraged creativity and the expression of true feelings, which was not what they usually do. She wrote: We always write composition in a model, not many beautiful words, not many true feelings. By this means, we can develop our imagination, of course, readers could like to read this kind of articles that are full of thoughts. This participant spoke highly of the visualisation method as she thought that this approach helped to create diversity, which was lacking in students' usual writing. The visualisation activity also stimulated students' imagination and helped them to see the world differently. She wrote: I like to read beautiful articles of true stories. I really practice myself to pay attention to these when writing compositions. We don't need similar compositions with similar structure and similar sentences. What we need is to create! So I think this kind of new teaching method is very useful and necessary. It helps us to broad our imagination, see something different from others and use the most vivid words to express. We will pay more attention to our expressions and how to arouse readers' interest. FX commented that, "We don't need similar compositions with similar structure and similar sentences. What we need is to create!" This comment was a significant response. FX was expressing an integrating personal observation that indicated the potential for the visualisation exercises to bridge and develop the concepts that the present study was attempting to realise. The move to creation is a strong human drive speaking to motivation, meaning and purpose and this spontaneous comment was indicative of the presence of this drive.

There was a high level of engagement, enthusiasm and classroom participation
Most students (85%) were ready to participate in the class activities according to their own comments on personal and class participation levels. This enthusiastic participation and the energetic volunteering of responses was significant evidence of students' engagement. The reticence of students to participate for fear of giving "wrong" answers was lessened by the realisation that giving their opinion, or experience could not be "wrong" and these engendered high levels of dialogic engagement.
Among all the materials used, the participants found "The Beautiful Girl in the Mirror" used in Lesson 3 was both the most interesting and the easiest to understand. For instance, XW wrote: "I found the activities were so exciting, especially those for the Beautiful Girl in the Mirror. None of my visualisation was the same as what was in the story. Neither of my classmates.
Everybody had a different version. We were just fascinated." In the reflection, XW actually related a lived experience of genuine excitement. She expressed wonder at the differences that arose between the students completing the same exercise. The differences engendered were noteworthy. Although it is very natural for people to develop different imagery after reading the same text, students who had not experienced specific training before might have found this surprising. The tutor's comments and the students' reports indicated a high level of discussion in the EG, which supports the idea that the visualisation training was engaging. Chinese EFL students will have certain expectations of their work regarding its presentation, content and the learning process. Difference "excited" them and there was a positive response to the discoveries made within themselves and between the classmates; there appeared to be increased arousal and engagement in the teaching and learning process. In XW's response of "wonder", increased potential could be seen to have opened up minds to new possibilities.

There was evidence of students' preoccupation with "long and complicated sentences" as features of native speakers' writing
Forty-four per cent of the participants expressed their preoccupation with technical aspects of writing, as this is the examination focus of language teaching in China. Participant LM expressed dissatisfaction with some of the samples used in the visualisation training exercises. The author was not sure of the exact meaning of the original comments and required clarification, which was given. The participant had cited two examples used to show poor pieces of writing. Unfortunately, LM had taken the samples given as "poor" to be "good" ones. Her explanation as to why they were not appropriate examples was typical of participants who regard poor writing to be that which is not like that of native writers, according to their own understanding. She wrote: First, most of them are composed of short and simple sentences. It seems that their writers failed to make full use of language devices to make their work appear more native, like Conjunctions, Attributive Clauses, Appositive Clauses and so on. What's more, I also find little logic in the structure of some samples. In other words, some of the samples didn't follow the basic principles of writing strictly. …And as for the conclusion, most of them are made up of no more than three simple sentences. Compared with those articles of native people, Chinese students' work are obviously less cohesive, logical and natural. LM's comments in this regard were a fair and appropriate critique of the samples -and part of the reason they were included to illustrate "poor" writing. Interestingly, this description of "good writing" in English included words like, "native, cohesive, logical"; by native, she meant "make full use of language devices… like Conjunctions, Attributive Clauses, Appositive Clauses and so on", as well as being "cohesive, logical and natural". So LM (wrongly) took the complexity of sentences as signs of text being written by a native English speaker, because Chinese English language teaching values these aspects of writing, however erroneously. Another case in point was from participant PF's reflection. PF also mentioned her frustration by not being able to write the long beautiful sentences she hoped to express. She wrote "sometimes when I read what I write, I find the sentences are so simple. Almost all sentences are simple sentences. I don't know how to write very beautiful sentences." The understanding of complex sentences as "desirable" and simplicity as being "undesirable" appears to be induced by the nature and focus of the Chinese EFL teaching system.
Despite the fact that visualisation-training activities engendered great positivity amongst many students, significant resistance was seen with those students whose preoccupation only with what will help them pass examinations most effectively. This is a valid perspective, given the Chinese emphasis on high grades and what kind of writing is marked highly. It was clear that visualisation training needed to address these fears and concerns and not be content with "simply" being seen as more enjoyable learning sessions or even engendering improvement in writing just for its own sake. Bereiter and Scardamalia (1987) distinguish two levels of thinking processes involved in written communication: a lower level and a higher level. The former refers to the mechanical demands of producing written language, including handwriting or other means of transcription, spelling, capitalisation, punctuation, and formatting, while the latter refers to all those thinking processes which focus on the meaning a writer intends to communicate to readers. In this study, 83% of the students expressed the idea that in their writing practice using the visualisation concepts, they did in fact pay more attention to how to express their meaning clearly than how to use a certain sentence structure correctly or spell words correctly, which was a sign of higher level thinking process; but they noticed too that this brought with it a realisation of the need to be competent in the lower level process and that the visualisation training highlighted deficiencies in this level of their competence, which was significant.

Visualisation training benefited participants' higher-level thinking processes
Reflections on the points above tend to show the necessity of using visualisation training in conjunction with development of all aspects of writing in balance, so that visualisation is seen as a holistic development tool for writing that improves and expands competence in communication across the whole spectrum of need, specifically addressing the technical competence that is perceived to lead to higher marks in tests, as well as enhancing creativity and imagination, viewed as more optional by most students.

There was a tendency to feel the frustration with using the same training pattern
All the above themes were in one way or another positive regarding the experiment and the training activities. But twenty-eight percent of the student participants expressed frustration with the training activities because, throughout the five lessons, they were frequently asked to engage with the admittedly repetitive process of "read, visualise, discuss and visualise again." For example, CQ's problem was that "after the first few activities, my visualisation reached the end of the story and in the later exercises I could not stop going along my own line instead of following what was actually in the story." This phenomenon concorded with Tomlinson's findings in his studies in 1997 where even after the readers was given the text regarding what happened in the story the participants could not change the original image formed in their mind before they read the story further. This response would need to be examined further in the future development of effective learning materials; at this point it can be seen that the common process of "stop-start" feeding a story to students, which is a common teaching device in language training, may well not work when using visualisation. This indicates the necessity for optimising visualisation-training materials for more positive outcomes that balance technical and creativity needs.

Conclusion
This learning strategy-based, real-time, real-situational exploratory study answered its research questions to a certain degree: before the training, no intentional visualisation had been used by the students; but after visualisation training, the great majority of students were convinced of the potential value of using this visualisation methodology, not only for writing, but for wider aspects of learning. Statistically, there was no significant improvement in the quality of their narrative writing per se. However, the study showed that there was a definite place for visualisation training activities in narrative writing development, subject to being able to address important aspects arising from the exercise, including recognising hindrances intrinsic within the present system of learning. Important points noted for development were: • The training activities used in this study were based exclusively upon providing written guidance narratives; this prevented flexibility in presentation and use. Repetition of the same format was found to be disengaging for the students, as revealed in the reflection. In future study, it would be necessary to use more diverse and stimulating audio-visual material to develop visualisation training. • All the participants' reflections were only capable of inferring that the students had little natural propensity to use visualisation prior to the training exercises, as the evidence was not clear. It was, therefore, difficult to establish the "baseline" for conscious prior use of visualisation and thus to be sure about improvement as a result of the visualisation training intervention. The evidence for positive development was drawn principally from the students' responses to the activities and comments about how this approach was novel and (usually) helpful. Since there is more emphasis on argumentative writing in the Chinese language curriculum and this aspect of writing is, therefore, what is tested -all participants in this study noted that the visualisation training exercises had the potential to be effective in developing argumentative writing as well and the application of visualisation training to argumentation might be positive and worthy of exploration in later studies.
Although this study is an enquiry into the effect of visualisation on narrative writing in China, there is a significant possibility that visualisation as a technique would be effective for enhancing all aspects of Chinese EFL students' language. It was interesting to note that Liu and McCabe (2018) conclude from their extensive research, that student "novice writers" need, "a deeper understanding of the effect of the choices they make in writing, an understanding which goes well beyond lists of metadiscoursal lexical items, and into discourse semantics and their realisations in texts across contexts and situations" (p. 97). This is one of the potentials that visualisation training offers and in a small, but indicative way; this study promotes through the technique of visualisation the goal to, "help students become more aware of the possible effects of their linguistic choices" and thus "choose their expression to match their intentions in meaning and in discourse goals" (ibid).
Summarising, visualisation can be developed as a tool for language teaching across all subjects. In this way visualisation would become a common method for extending the creative and authorial higherlevel faculties required for artistic expression alongside the lower order, but still vital, competence in structure and vocabulary. This links with the argument that visualisation is a powerful tool for education in its broadest sense, in that it is a facility of the brain that enhances the mind's capabilities -and that this potential for enhancement is by no means restricted to language learning. There is necessity, however, to urge that further account be taken of new research into the phenomenon of aphantasia. How and in what ways students will need consideration in visualisation-oriented development if they do not possess the visualisation faculty requires careful consideration. It may be that, as with dyslexia or any other communicative difference, different approaches will need to be considered. Nevertheless, this research also has potentially significant value in enhancing the communicative aspects of narrative writing and expression, as it will enable real-life engagement with the English-speaking people of the world, most of whom will have been brought up in systems that value communicative creativity and expression over pure technical competence. Tomlinson, B. (2003). Developing materials for language teaching. London, UK: Continuum. Tomlinson, B. (2011). Seeing what they mean: Helping l2 readers to visualise. In B.