Learner agency and potential for creativity in writing task design and Chinese EFL students’ flow experiences in writing: A contextual perspective

Abstract

Ensuring learner agency and potential for creativity in task design are likely to elicit different flow experiences among learners, either facilitating or impeding their task engagement and oral or written performance in foreign language learning contexts. However, so far little attention has been paid to the role of these factors in learners’ flow experiences in task-based second language (L2) writing. This study aimed to investigate the relationships between different task conditions operationalized as a teacher-generated content task (TGT), a learner-generated content task (LGT), and a creative task (CT), and learners’ flow experiences by utilizing Czimmermann and Piniel’s adaptation of Egbert’s flow questionnaire. Moreover, the study also examined the possible role that study contexts might play in students’ flow experiences in English writing. Our investigation involved two groups of Chinese learners studying in Hungary and China, 40 students in total, who produced 120 narratives and filled in a flow questionnaire after each completed task. MANOVA results showed a significant main effect for the study context, suggesting that Chinese learners in Hungary experienced more flow in all of the writing tasks while their peers in China experienced more anti-flow on them. Despite the fact that multivariate analyses did not reveal a main effect for task conditions for either flow or anti-flow experiences or an interaction between study contexts and task conditions, findings from univariate follow-up analyses warrant a closer examination of possible task and interaction effects.

Keywords

creative writing flow task design learner agency study context task condition

I Introduction

Despite being extensively studied in psychology and educational psychology (Csikszentmihalyi, 1975, 1989, 1997), the concept of flow has received little attention in the second language (L2) learning domain. Flow refers to a ‘peculiar dynamic state –the holistic sensation that people feel when they act with total involvement’ (Csikszentmihalyi, 1975, p. 36), a state of feeling interested, being attentive, enjoying what one is doing, and going through ‘self-forgetfulness’ (Maslow, 1971, p. 65). Task-based foreign language studies have suggested that learners’ subjective feelings and cognitions – i.e. interest, attention, manifest learner engagement, and tasks of different characteristics – might bring about different flow experiences for learners, which might either facilitate or impede their task engagement and their subsequent language development (Kormos & Préfontaine, 2017; Lambert & Zhang, 2019).

In the field of second language acquisition (SLA), studies have confirmed the existence of flow experiences among learners of English as a foreign language (EFL) during various classroom activities (e.g. Czimmermann & Piniel, 2016; Egbert, 2003). However, little empirical research has investigated the role of flow in EFL writing (e.g. Liu et al., 2022; Payant & Zuniga, 2022; Zhang, 2021). Still, focusing on writing appears to be important because, on the one hand, writing seems to play a major role in SLA and has been suggested to be effective in promoting SLA (Polio, 2012; Williams, 2012). On the other hand, studies that have compared learners’ performance across different task modalities have shown that in many cases learners’ performance varies according to the written or spoken modality, so they need to be studied separately (Kormos, 2014; Mayo & Agirre, 2019; Zalbidea, 2017). Thus, the present study aims to contribute to research on writing tasks but focuses on task-specific flow, by which term we refer to flow experiences related to a particular task and to attempts to address task features that are associated with flow in writing.

Besides properties of the writing task, learners’ writing performance, writing ability, and competence also seem to be influenced by contextual issues like the broader socio-linguistic environment and the official status of the language being acquired in that particular society although this influence can be conceptualized at several levels (Douglas Fir Group, 2016). At the macro level, research has been conducted on language learning in study abroad (SA) contexts versus formal instruction received at home (EFL context). Specifically, within the SA context, target language contexts of English-speaking countries have been most frequently researched where English is widely used in the society, while there is considerably less research about English as a lingua franca (ELF) contexts, the typical case for immigrants to Europe. Although there are some studies tapping into learners’ affective experiences in different contexts (Dewey et al., 2018; Lee & Kinginger, 2018), it is still unknown whether influences of such contextual differences might be detectable in students’ flow experiences in English writing. Therefore, the current study also attempted to examine the role of study contexts in the flow experiences of Chinese native speakers studying in Hungary and China while performing writing tasks.

II Literature review

1 Flow theory and flow in foreign language learning

Flow is a term originating in psychology, referring to a mental state where the learner is fully involved and engaged with the learning activity or the task (Csikszentmihalyi, 1975, 1989). Csikszentmihalyi (1975, p. 51) proposed the construct of the flow state, claiming that flow is experienced when people perceive opportunities for action as being evenly matched by their capabilities. According to his Quadrant model, maintaining a balance between higher than average skills and challenges results in people being ‘more active, alert, concentrated, happy, satisfied, and creative’ (Csikszentmihalyi & LeFevre, 1989, p. 816). However, if there is a mismatch between the challenges posed by a task and the individual’s skills, anti-flow is experienced. Anti-flow is not simply the opposite of flow which is characterized by full engagement with and deep involvement in what one is doing, but is more associated with emotions, such as anxiety, boredom, and apathy. To illustrate, if skills increase but challenges do not, a state of boredom arises; anxiety appears when the challenges for action greatly exceed the actor’s skills (Csikszentmihalyi,1975; Csikszentmihalyi & LeFevre, 1989); and apathy appears when challenges and skills required are both below the actor’s potentials (Csikszentmihalyi & LeFevre, 1989; Dewaele & MacIntyre, 2014). These anti-flow states can significantly impact learner engagement and performance in language learning tasks (Csikszentmihalyi, 1990; Czimmermann & Piniel, 2016; Dewaele & MacIntyre, 2014). In our study, we have chosen to include anti-flow as a critical variable to provide a more nuanced understanding of the Chinese learners’ experiences while performing writing tasks.

On the phenomenological level, the state of flow can be described by a set of key characteristics (Csikszentmihalyi, 1990). One of them is an intense focus and concentration on the activity that is being performed at that moment. Moreover, the individual perceives total control over their actions, and they feel confident that they are able to cope with anything that comes next. This intense concentration on the current activity results in a merging of action and awareness, which leads to a reduction or total loss of the individual’s reflective self-consciousness. A further characteristic feature is an altered sense of time: The individual’s subjective experience is that time passes faster than normal. The final characteristic is that since these activities are intrinsically motivating, the activity itself is the reward, so in a way the objective outcome of the action appears to be irrelevant (Csikszentmihalyi, 1990).

Recently, foreign language researchers have developed an interest in flow and its role in the L2 learning domain; for example, Egbert (2003) described the hypothetical relationship between flow and foreign language learning on the basis of Csikszentmihalyi’s (1975) flow theory. Egbert attempted to measure the flow her students experienced in connection with different learning tasks, and, in her questionnaire, she divided task-specific flow into four dimensions:

• the balance between challenge and skill that makes the participants willing to explore and complete the task;

• attention from the participants on the task;

• interest from the participants aroused by finding the task intrinsically interesting or authentic;

• a perceived sense of control, mainly referring to the control over an interactive learning situation.

Czimmermann and Piniel (2016) adapted Egbert’s (2003) flow questionnaire and extended it by including the three anti-flow components – namely apathy, anxiety, and boredom (Csikszentmihalyi,1975; Csikszentmihalyi & LeFevre, 1989) – making it a more comprehensive instrument for researching L2 flow experiences.

As regards empirical studies of flow in the L2 learning domain conducted with the help of tasks, the relationship between flow and speaking task modes (individual, pair or group work; Czimmermann & Piniel, 2016) as well as intercultural contact (inter- and intra-cultural communication; Aubrey, 2016, 2017) have been investigated. Moreover, the relationships of flow with task complexity and modality (Cho, 2018; Ghasemi et al., 2021) have also been scrutinized. Since only Aubrey’s studies cover the contextual effects on flow, which is directly relevant to our current topic, a detailed description of them will be given in the section on learning contexts.

There are a number of studies that have been carried out on flow in connection with writing in general, without focusing on any specific writing task. In their study of Chinese EFL learners, Liu et al. (2022) found that intrinsic motivation, perceived competence, and attention control could positively predict writing flow experiences. When researching learners of French, Payant and Zuniga (2022) found that flow occurred in a virtual writing course during the global pandemic thanks to confidence boost resulting from group member discussions. Although the case study conducted by Abbott (2000) involved young native speakers of English, results of this study indicate that flow occurred when students had freedom with regard to important aspects of writing, such as ownership, genre, style, and length. Based on these findings, intrinsic motivation, competence and confidence, attention control, and the freedom of choice all appear relevant with regard to writing flow. However, these studies on writing flow viewed writing activities holistically; thus, their findings might have few direct implications for writing task design and writing instruction. Our study aims to fill this gap by investigating writing flow using specific writing tasks that differ in their affordances for learner agency and creativity.

2 Learner agency and potential for creativity and their hypothesized relationships with flow

‘Agency’ is proposed as a socio-culturally mediated and dialectically enacted construct, which highlights the role of constraints and affordances in a given time and space that determine whether certain actions happen or not (Lantolf & Thorne, 2006). Agency is multidimensional. It does not only concern behavior (Larsen-Freeman, 2019); rather, it entails ‘the ability to assign relevance and significance to things and events’ (Lantolf & Thorne, 2006, p. 143). Agentive language learners take responsibility for their own learning, including setting goals, making choices and decisions, monitoring progress, and evaluating outcomes (Little et al., 2017). Most recent task-based foreign language studies seem to conceptualize learner agency as the freedom assigned to the learner concerning the task content to be produced (Lambert et al., 2017; Lambert & Minn, 2007; Wu & Albert, 2022), which is in accordance with the affordance and the relevance and significance embedded in the tasks proposed by Lantolf and Thorne (2006). Findings tend to suggest that learners performed better on tasks under the learner-generated content task (LGT) condition, where they could create their own ideas and content offering more affordance and less constraint on learners’ part, as opposed to tasks operated under the teacher-generated content task (TGT) condition, where the content was predetermined by the teacher or the task designer (such as picture description). These findings were explained in a way that learners were more likely to find the LGT interesting, meaningful, and relevant to their real life, and, thus, became more engaged by investing time, energy, and effort into solving the task (Maehr, 1984).

Learner agency, often conceptualized as freedom in determining task content whose effects could be made tangible when analysing performance on LGT, has been researched in SLA in recent years (Lambert et al., 2017; Lambert & Minn, 2007; Lo & Hyland, 2007; Poupore, 2014). For example, Lambert and Minn (2007) found that learners experienced more positive affect on speaking tasks implemented in the personal investment condition that offered higher learner agency, and they reported having found the tasks more enjoyable, which slightly increased their confidence in doing them in the future. Moreover, Poupore (2014) found that speaking tasks with personal life themes as content were reported as being more interesting than those whose content was considered irrelevant or of little relevance to their participants’ personal lives. Consequently, the participants were more intrinsically motivated and engaged when doing the tasks with content covering personal experiences. Similarly, Lo and Hyland (2007) also found that the writing activities and topics that were interesting and relevant to their learners’ real personal lives and had genuine audiences enhanced young writers’ writing engagement and built their confidence in producing their texts. Finally, Lambert et al. (2017) found that students reported being more positive about the LGT, indicating that they were more interested in the LGT and enjoyed doing it more. Although none of these studies investigated flow specifically, the interest and enjoyment which was characteristic of learners’ perceptions of tasks that have affordances for exercising learner agency suggests a possible link. Our study follows this line of research by operationalizing learner agency as freedom over what to write, that is the personal narration task in our study, and aimed to investigate its connection to flow in L2 writing.

In addition to addressing learner agency, researchers in applied linguistics also endeavor to design and implement creative tasks in language classrooms (Maley & Kiss, 2018). Creativity associated with language is termed ‘linguistic creativity’ by Tin (2022, p. 127), which has to do with the language used, heard, produced, engaged in, practiced, learned or encountered in daily life. According to Tin (2011), creativity refers to ‘the playful use of language to construct new/unknown meanings, transforming one’s current linguistic and conceptual world and involving several types of creative thinking’ (p. 216). Tin (2012) argues that creative tasks with constraints that require learners to construct new meanings or ‘unknown meaning’ (p. 178) can facilitate creative and complex language use. Therefore, it is apparent that although creative tasks certainly involve the possibility of exercising learner agency by giving freedom to students over specific aspects of task content, this freedom is somewhat limited by task constraints. Studies of classroom writing in particular have suggested that writing tasks with some freedom over content but limited by certain constraints are more likely to enhance task performance and improve writing skills of learners (Abrams, 2019; Tin, 2012; Wu & Albert, 2022). Moreover, recent studies found positive relations between creative EFL writing activities and writing dispositions (i.e. willingness to take risks, motivation, self-efficacy, and interest etc.; Tok & Kandemir, 2015).

Our study follows this line of research by investigating the relation between flow and creative writing tasks. We hypothesized that writing tasks with the potential to be creative with regard to producing content and ideas as well as language use might induce flow in students on the basis of the following grounds. First, creative writing tasks or activities proposed by Tin (2011, 2012) provide both control/freedom over the task and challenges/constraints for learners, which are part of the conditions for flow to occur (Csikszentmihalyi, 1975; Egbert, 2003). Second, as flow has been found to exist in writing in general (e.g. Abbott, 2000; Liu et al., 2022), it is expected to occur in this specific type, namely creative writing. Third, creative writing is considered one type of creative activity, which is where the investigation of flow started in the first place. Researchers found that people experienced flow when they were engaged with creative activities, such as creating works of art (Getzels & Csikszentmihalyi, 1976). Therefore, it seems justifiable to hypothesize that there exists a relation between flow and potential for creativity in writing. Our study attempts to investigate this relationship with the help of the creative task that was operationalized by offering both some degrees of freedom and constraint.

3 English study contexts and writing instruction for Chinese students in Hungary and China

Although emphasizing the conceptual embeddedness of language learning is becoming increasingly popular (Ushioda, 2015), what different authors mean by context exactly is often left unclarified. Thus, the efforts of the Douglas Fir Group (2016) aimed at clarifying the concept of context in a complex framework are laudable. In their seminal paper, they used Bronfenbrenner’s (1979) ecological framework for human development as an inspiration and distinguished micro, meso, and macro levels of context. The micro level focuses on the learner, their interactions with other social actors and the resulting contexts of language use, while the meso level contains socio-cultural institutions and communities which may provide or restrict access to various types of social experiences. The macro level is the broadest; this is the level where large-scale, ideological structures exert their influence also shaping the meso and micro levels. At the same time, the macro level is also influenced by the meso and micro levels. The EFL and ELF contexts investigated in our study can be conceptualized as macro levels, whose effects might be further moderated by factors present at lower systemic levels. Nevertheless, based on the model, characteristic features of the macro level still greatly impact the language learning experiences of learners; therefore, we will describe those macro level characteristics of the Chinese EFL and Hungarian ELF contexts that appear to be relevant to our discussion in this section.

English language learning in China mainly happens in language classrooms since there is limited exposure to English in real-life contexts outside it. Despite this situation, the subject of English is important for Chinese students in China since it accounts for 20% in the college entrance exam in most provinces. English writing is of even more importance, taking up about 17% to 27% of the total score (Yang, 2020). Although its academic status is high, English is not really utilized by learners as a means of communication; it is purely considered as an object of study like any other school subject. It is possible that the rather grim picture suggested by empirical studies of writing which show that students lack interest and experience boredom and anxiety in writing classes (e.g. Jiang, 2022; Liu, 2017; Qin et al., 2015) can be linked to the above described characteristics of the language learning context. These circumstances call for a fine-grained investigation of the situation with the help of specific tasks. Moreover, formal English writing instruction focuses on the genre of argumentation in high schools in China, while other forms of writing are rarely practiced, such as narration about personal experiences and creative writing (Yu, 2021; Zhang, 2016). Finally, in this context where English is seen as an important academic subject rather than a means of communication, the training of writing is often exam-driven with a focus on using writing templates and writing strategies, contributing little to the improvement of students’ writing ability (Liu, 2020).

In contrast, Chinese learners find themselves in a different situation regarding the affordances for learning English in Hungary, a monolingual country in central Europe where English is often used as a lingua franca serving authentic communicative purposes, especially in the case of immigrants who do not speak Hungarian. Chinese immigrant students studying English in Hungary can have more exposure to English input, and they also have more opportunities to use English, either to communicate with people from different countries in social settings or as the medium of instruction in academic settings, than their peers at home (Cogo & Jenkins, 2010; Seidlhofer, 2005). Access to English outside the school setting is much easier, and the opportunities to use English for communication probably shift the focus from its importance as a school subject to being valued as a communicative tool. Additionally, schooling options are varied in Hungary with both state-financed bilingual and self-financed international schools (Öveges & Csizér, 2018); thus, students are likely to have different writing experiences with teachers from different cultural and educational backgrounds. As a result of the differences regarding how English is viewed and taught in the two contexts, we expected differences in flow experiences between the two groups of Chinese students studying in Hungary (CSH) and Chinese students studying in China (CSC).

Existing studies on EFL writing have suggested contextual effects on writing (e.g. Pérez-Vidal & Juan-Garau, 2009, 2011; Sasaki, 2004, 2007). These studies investigated the effects of the study abroad context versus formal instructional contexts in the home country on learners’ writing. They found that learners’ writing interest, motivation, and writing strategies were enhanced after studying for a period of time abroad, as they had more writing experience and training, together with increased exposure to the target language in communication. This, in the long run, improved their writing, possibly due to the transfer from speaking competence (Pérez-Vidal & Juan-Garau, 2011). This explanation for learners’ improvement in English writing resonates with Aubrey’s (2016, 2017) findings in connection with the role of having inter-cultural or intra-cultural interlocutors when performing a task.

In his 2016 study, Aubrey asked EFL learners to perform two tasks with a culturally familiar interlocutor and a culturally unfamiliar interlocutor. He found that inter-cultural contact led to a significant increase in learners’ flow experiences. In a subsequent study, Aubrey (2017) studied flow qualitatively by analysing learner diaries written after each task (five tasks for each group in 5 weeks). There were two groups of participants, one doing intra-cultural oral tasks – since they were paired with partners from the same cultural and first language (L1) backgrounds – and the other inter-cultural oral tasks (since they were paired with partners from different cultural and L1 backgrounds). It was revealed that learners taking part in inter-cultural task-based interactions reported significantly more flow-enhancing experiences and significantly fewer flow-inhibiting experiences than learners who performed the same tasks in the intra-cultural condition. In the inter-cultural task condition, learner pairs probably had less shared background knowledge about the assigned topics, and this might have generated a more genuine need to communicate to develop a shared understanding. This finding is in line with Aubrey’s earlier finding that inter-cultural contact in a task-based context seemed to have increased the likelihood of learners’ achieving a flow state.

A different contextual effect was identified by Lambert and Zhang (2019) in their small-scale qualitative study. They found that Japanese learners of English experienced lower levels of anxiety and felt more motivated than Japanese learners of Chinese when performing an oral task where they were free to determine the content than on a task where the content was determined for them by the teacher. Lambert and Zhang claimed that this was probably due to their previous experiences in formal L2 instruction. For example, learners of English participating in the study had had communicative language learning experience, while learners of Chinese had no exposure to communicative activities in formal instruction, which might have caused them to be anxious while performing the task where they were given freedom.

These studies suggest that, for foreign language learners, being given more opportunities to use and be more exposed to the target language is likely to increase their confidence as well as their competence to perform language-related oral tasks, while having inter-cultural experiences might directly be associated with flow as suggested by Aubrey’s (2016, 2017) studies, which found that the inter-cultural contact had a significant positive relationship with flow. Moreover, being exposed to activities where learners are free to make decisions about the content to be produced probably results in finding those activities more motivating and less anxiety provoking. Since more exposure to the target language and more intercultural experiences are expected to characterize the Hungarian ELF context where it is also more likely that learners are exposed to varied writing tasks, some of which might allow them to create their own content, we expected a positive relationship between the ELF study context (as opposed to the Chinese EFL study context) and writing flow. However, no empirical studies have been conducted on this, so our study aimed to address this research niche.

III Research questions

This research aimed to answer the following research questions:

• Research question 1: How do task conditions influence learners’ flow and anti-flow experiences?

• Research question 2: How does the study context affect learners’ flow and anti-flow experiences?

• Research question 3: Do task conditions and study contexts interact to affect learners’ flow and anti-flow experiences?

IV Methods

1 Research design

This is a quantitative study that utilized writing tasks and a questionnaire to examine the effects of the task conditions and the study contexts on learners’ flow and anti-flow experiences in their English writing. Task condition was a within-participants variable which had three levels: the teacher-generated content task (TGT), the learner-generated content task (LGT), and the creative task (CT). Besides this within-participants factor, we also investigated the effects of study context, a between-participants variable at two levels: Chinese students studying in Hungary (CSH) and Chinese students studying in China (CSC). The effects of these two factors were observed on the four dependent measures of flow and three dependent measures of anti-flow measured by various scales.

Similarly to earlier studies (Lambert et al., 2017; Lambert & Zhang, 2019), we used a TGT in an attempt to provide learners limited opportunities to exercise their agency, and an LGT which had substantial room for agency, as learners were free to determine most aspects of the task content in the LGT condition. The CT condition also gave room for agency, but learners’ freedom was somewhat constrained on this task in order to elicit creative responses, that is, to avoid mundane, everyday solutions and push learners to experiment with language and create new meanings (Tin, 2011, 2012). Study context differences were ensured by using two groups of Chinese students: one studying English in Hungary in an ELF context and the other studying English in China in an EFL context. The association of the above-described independent variables with dependent variables representing different aspects of flow and anti-flow experiences in line with Egbert (2003) and Czimmermann and Piniel (2016) was explored with the help of repeated measures MANOVA tests.

2 Participants

The participants of the study were two groups of Chinese learners of English studying in Hungary (CSH) and China (CSC), and although they were recruited by adopting convenience sampling (Dörnyei, 2007), care was taken that their language proficiency levels should be comparable. This was ensured by asking participants to fill in the Use of English section of the Oxford Placement Test (OPT New 2004 edition), a highly valid and reliable proficiency test, which contains grammatical structures that are regularly covered by most course books and public examinations (Allan, 2004). The Use of English part specifically focuses on grammar that Kormos and Trebits (2012) regard as especially important for writing. The equality of the mean scores of the Use of English section of the two groups was checked by using an independent samples t-test, which did not show significant differences between the group averages (Mean_CHS = 61.15, Mean_CSC = 64.50, p > .05). The proficiency level of each group was around B1–B2 according to the Common European Framework of Reference (CEFR). Eventually 20 participants were recruited for each group, and their age ranged from 12 to 17 years (Mean_CSH = 13.75, SD_CSH = 1.895, Mean_CSC = 15.95, SD_CSC = 0.06). The higher mean age of the group studying in Hungary was caused, on the one hand, by difficulties in recruiting Chinese teenagers in Hungary and, on the other, by the tendency that when teenagers of the same age were compared across the contexts, those studying in China usually had lower proficiency levels. Since we also intended to examine learners’ writing performance (although this analysis is not included in the current article), we opted for matching our groups based on their proficiency levels rather than their age. Besides, it is important to note that an extra requirement for CSH was that they had to have been studying in Hungary for at least 3 years by the time of the data collection. The rationale for this was to examine the role of the Hungarian English study context on their English writing, specifically on their flow experiences in writing, and this period seemed long enough to have a detectable effect on language learning experiences.

3 Instrument

The three writing tasks applied in the present study (see Appendix A) were the TGT, the LGT, and the CT. Though informed by previous similar studies (e.g. Lambert et al., 2017; Lambert & Zhang, 2019), they were designed by the authors themselves to ensure the appropriateness of these activities for the sample population of the current study. The TGT required students to narrate a story depicted by six related pictures in order. With a sequence of pictures provided, students were very restricted in terms of possible content to write; therefore, this task was assumed to allow for the least learner agency. The LGT required students to recall their past experiences and write about a memorable event that had happened in their lives. Students were totally in control and could generate the content and write about whatever they liked. The LGT empowered students with total freedom about what to write within a certain topic and thus was considered as ensuring the most agency for students. The CT was a story-creating task which introduced some constraints while also providing some freedom for students. Students were given 10 unrelated words as a way of introducing restrictions; at the same time, students still had freedom to create their own stories as long as they used at least six of the 10 given words. This task was somewhat constrained and controlled, but it was assumed to have the greatest potential for creativity, according to the arguments proposed by Tin (2011, 2012) as opposed to the other two tasks where content was determined either by the teacher or students’ own experiences. The three writing tasks had the same title ‘A memorable event’ with the same starting sentence as the prompt ‘This is something that I will never forget . . .’ The task instructions were in Chinese to avoid ambiguity for students.

The flow questionnaire applied in the present study was a modified version of Czimmermann and Piniel’s (2016) instrument. The adapted flow questionnaire for the pilot included 35 items, with each construct containing 5 items, where students had to indicate their agreement with the items on a 5-point Likert scale. Then, the modified flow questionnaire was translated into Chinese by one of the researchers, and a back translation to English was also carried out by another expert to ensure the equivalence of the instruments. The discrepancies were settled by negotiation and the adjustment of wordings.

4 Procedures

a Piloting

The three tasks were piloted with three intermediate students studying in Hungary to ensure that the tasks elicited narratives as well as to check the time needed for task completion and minimum length requirements. Due to minor modification to the tasks after piloting (i.e. prolonging the time for task completion), the three students as well as their writings were not included in the final study (for details, see Wu & Albert, 2022).

The adapted 5-point Likert scale flow questionnaire from Czimmermann and Piniel (2016) underwent rigorous piloting as well. As a first step, it was administered to one student (not a participant) to make sure that the content was clear and straightforward. The student was required to write a short piece about anything she liked and completed the flow questionnaire once she finished writing. She was asked to think aloud while completing the questionnaire and was allowed to ask for clarification any time she felt confused about any of the items. All the items were checked thoroughly and corrected if unclear for the student. Then, the flow questionnaire was piloted with 25 Chinese intermediate senior high school students in China who did the questionnaire immediately after finishing writing a short composition on ‘Friendship’.

b Reliability analysis

SPSS 29.0 was used to analyse the reliability of the adapted flow questionnaire. After small modification to the scales, the analysis showed that Cronbach’s alphas for all the scales were above 0.60 which is considered the minimum acceptable (Dörnyei, 2007). However, there was one scale, control, which had only 3 items after the reliability analysis. Since having such a short scale might cause problems in further analyses, we added two new items to the control scale before administering the final version of the questionnaire to our actual participants. The final flow questionnaire (see Appendix B) consisted of 7 scales with 31 items, namely interest (5 items; 1, 5, 11, 16, 27 ), attention (4 items; 3, 8, 14, 28), control (5 items; 2, 6, 12, 17, 30), challenge–skill balance (4 items; 4, 10, 15, 22), boredom (4 items; 7, 18, 25, 29), apathy (4 items; 19, 21, 23, 24), and anxiety (5 items; 9, 13, 20, 26, 31). The reliability figures (Cronbach’s alphas) of this research sample are reported in Table 1 for the flow questionnaire. Cronbach’s alphas for most of the scales are above 0.70 although three figures are only above 0.60, the minimum acceptable (Dörnyei, 2007). Since Cronbach’s alpha is an index of internal consistency and does not provide information regarding the number of factors explaining item correlations (Barbaranelli et al., 2015), principal component analysis was also conducted on the scales to check whether each scale measured one underlying construct. It turned out that although there were minor inconsistencies for the scales of attention and anxiety in the case of two out of three measurements indicating that these scales were potentially problematic, the majority of the scales worked well, showing acceptable reliability.

Table 1.

Reliability analysis and principal component analysis (PCA) of each flow and anti-flow scale.

Measures	N	TGT	PCA	LGT	PCA	CT	PCA
Measures	N	Cronbach’s alpha		Cronbach’s alpha		Cronbach’s alpha
Interest	5	0.80	1	0.85	1	0.86	1
Attention	4	0.72	2	0.75	1	0.61	2
Control	5	0.84	1	0.76	1	0.75	1
CSbalance	4	0.72	1	0.63	1	0.61	1
Boredom	4	0.80	1	0.88	1	0.71	1
Apathy	4	0.74	1	0.73	1	0.73	1
Anxiety	5	0.75	2	0.72	2	0.78	1

Notes. TGT = teacher-generated content task; LGT = learner-generated content task; CT = creative task; CSbalance = challenge–skill balance.

5 Data collection and analysis

The data collection in Hungary and China was administered by one of the researchers. In Hungary, data were collected from Chinese students who went to a private school on weekends in order to learn Chinese, while data collection in China took place within the academic period of high schools. Since recruiting Chinese participants in Hungary turned out to be more challenging, their data were collected first, and then data from students in China at comparable proficiency levels were sought, which meant that we needed to recruit slightly older learners on average from the Chinese context. Once learners with matching proficiency levels were identified, they were asked to perform a writing task by producing a minimum of 100 words within 30 minutes in their regular English classes, similarly to their peers studying in Hungary. Once they finished writing, they immediately completed the flow questionnaire that was attached to the writing task sheet. All essays were hand-written without help from dictionaries or digital tools. Altogether 120 flow questionnaires were collected from the 40 participants (20 in each group), with each student producing three questionnaires for the three tasks they performed. In both contexts, the tasks together with the questionnaires were given to the participants at least one week apart, and the effects of task order on students’ flow experiences were counterbalanced by giving tasks in different orders.

The software SPSS 29.0 for Windows was used to analyse data that were derived from the analytical procedures. To answer the research questions regarding the effects of task condition and study context on students’ flow and anti-flow experiences, repeated measures MANOVAs were conducted, which were followed by post-hoc tests to establish differences between specific groups in the case of significant results. The independent variables (IVs) were the three task conditions (within-participants factor) and the study context (between-participants factor), and the dependent variables (DVs) were the four flow (interest, attention, control, challenge–skill balance) and the three anti-flow components (anxiety, boredom, and apathy). Since the MANOVA procedure is able to test the effect of the independent variables on several dependent variables simultaneously, it is particularly well suited when the dependent variables share a common conceptual meaning (Stevens, 2009) like the underlying components of flow and anti-flow.

Before conducting the repeated measures MANOVA, the data were examined using SPSS Statistics to ensure that its underlying assumptions were met (Bennett et al., 2023). We found no serious violation of the normal distribution of the variables, and the homogeneity of their variance was also established. Neither Mauchly’s tests of sphericity nor Levene’s tests of equality of error variances were significant. Finally, Box’s M was also nonsignificant at a = .001, indicating that homogeneity of variance covariance matrices could be assumed. Because of the small sample size, we set the significance level at p < .05, which we later adjusted in the case of running several univariate analyses. We also calculated the effect size measure of partial eta squared, which gives an indication of the variance explained. Cohen (1988) suggested that partial η² at 0.01, 0.09, and 0.25 should be interpreted as small, medium, and large effect sizes respectively; thus, we followed these guidelines when interpreting our results.

V Results

First, repeated measures MANOVA was performed with task condition as a within-participants factor and study context as a between-participants factor on the four scales of flow as dependent variables. Based on Pillai’s Trace, the repeated measures MANOVA test showed that there was a significant main effect for the study context, F(4, 35) = 3.97, p = .009, partial η² = 0.312, as far as flow was concerned. However, a significant main effect for task condition was not found, F(8, 31) = 1.95, p = .087, partial η² = .335, on flow; also, no significant interaction between study context and task condition was observed, F(8, 31) = 0.99, p = .462, partial η² = .204. The follow-up univariate analyses on the significant main effect for study context revealed statistically significant effects on two measures of flow: with large effect size in the case of control, F(1, 38) = 16.45, p < .001, partial η² = .302, and medium effect size in the case of challenge–skill balance, F(1, 38) = 8.87, p = .005, partial η² = .189. As can be seen in Table 2, which presents means and standard deviations across task conditions and study contexts, Chinese students studying in Hungary reported higher levels of both control (M_TGT = 3.79, M_LGT = 4.13, M_CT = 4.25) and challenge–skill balance (M_TGT = 3.58, M_LGT = 3.84, M_CT = 3.95) than their peers studying in China (for control M_TGT = 3.26, M_LGT = 3.45, M_CT = 3.36, for challenge–skill balance M_TGT = 3.21, M_LGT = 3.36, M_CT = 3.28).

Table 2.

Descriptive statistics of flow scales across study contexts and task conditions.

Measure	Hungary						China						Total
	TGT		LGT		CT		TGT		LGT		CT		TGT		LGT		CT
	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD
Interest	2.99	0.70	3.12	0.82	3.32	0.83	2.88	0.76	3.08	0.73	3.01	0.87	2.94	0.72	3.10	0.77	3.17	0.85
Attention	3.36	0.63	3.53	0.74	3.56	0.69	3.09	0.81	3.13	0.75	3.29	0.68	3.23	0.73	3.33	0.76	3.43	0.68
Control	3.79	0.69	4.13	0.50	4.25	0.48	3.26	0.87	3.45	0.76	3.36	0.56	3.53	0.82	3.79	0.72	3.80	0.68
CSbalance	3.58	0.61	3.84	0.46	3.95	0.48	3.21	0.81	3.36	0.78	3.28	0.60	3.39	0.73	3.60	0.67	3.61	0.64

Notes. N = 20; TGT = teacher-generated content task; LGT = learner-generated content task; CT = creative task; CSbalance = challenge–skill balance.

Although based on the main effects of MANOVA, no association could be detected between task conditions and learners’ flow experiences, we applied univariate ANOVAs to delve into specific aspects of learners’ flow experiences and to explore the nuanced relationships that task conditions have with various components of the flow experience. The follow-up univariate analyses revealed significant independent effects for interest, control, and challenge–skill balance at the p < .05 level. However, in order to control for the inflated family-wise error rates that result from performing multiple tests on the same data, we needed to adjust the significance of the univariate ANOVAs at a Bonferroni adjusted alpha level. To make this adjustment, we divided the family-wise alpha level by the number of dependent variables in the MANOVA, that is by four, setting it at p < .0125. With this stricter alpha level, only one significant univariate effect could be detected for task condition, which was control, F(2) = 4.87, p = .01, partial η² = .114, reflecting a medium effect size. Pairwise comparisons revealed that the statistically significant difference can be detected between the TGT (M_TOTAL = 3.53) and the LGT (M_TOTAL = 3.79). Figures in Table 2 show that learners consistently reported higher levels of control when they had the chance to create the task content themselves (LGT M_CSH = 4.13, M_CSC = 3.45) than when the content was decided for them by their teachers (TGT M_CSH = 3.79, M_CSC = 3.26).

Next, repeated measures MANOVA was performed with task condition as a within-participants factor and study context as a between-participants factor on the three anti-flow scales as dependent variables. Similarly to the first MANOVA, Pillai’s Trace indicated that there was a significant main effect for the study context, F(3, 36) = 8.23, p < .001, partial η² = 0.407, as far as anti-flow was concerned. However, there was no significant main effect for task condition, F(6, 33) = 0.84, p = .549, partial η² = .132 or interaction between the two, F(6, 33) = 1.84, p = .122, partial η² = .250. The follow-up univariate analyses on the significant main effect for study context revealed statistically significant effects on only one measure of anti-flow, anxiety, F = 13.76, p < .001, partial η² = .266. Table 3 reveals that Chinese students studying in China consistently reported higher levels of anxiety (M_TGT =2 .58, M_LGT = 2.54, M_CT = 2.72) than their peers in Hungary (M_TGT = 2.08, M_LGT = 2.05, M_CT = 1.75).

Table 3.

Descriptive statistics of anti-flow scales across study contexts and task conditions.

Measure	Hungary						China						Total
	TGT		LGT		CT		TGT		LGT		CT		TGT		LGT		CT
	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD
Boredom	2.78	0.67	2.45	0.88	2.44	0.69	2.56	0.96	2.56	0.86	2.73	0.83	2.67	0.82	2.51	0.86	2.59	0.77
Apathy	2.28	0.56	2.18	0.68	1.90	0.61	2.49	0.89	2.59	0.74	2.68	0.69	2.38	0.74	2.23	0.74	2.29	0.75
Anxiety	2.08	0.58	2.05	0.67	1.75	0.62	2.58	0.75	2.54	0.63	2.72	0.63	2.33	0.71	2.30	0.69	2.24	0.79

Notes. N = 20; TGT = teacher-generated content task; LGT = learner-generated content task; CT = creative task.

Despite the lack of task condition effects on anti-flow experiences, we decided to explore if any single facet of anti-flow can be associated with different task conditions in order to gain a deeper understanding of their individual contributions, so we performed univariate tests on anxiety, boredom, and apathy. Univariate tests performed to detect the main effect of task condition on the individual anti-flow scales did not identify any independent effects; however, the interaction of task condition and study context was found to have a statistically significant effect on anxiety, F = 4.57, p = .013, partial η² = .107. When checking the profile plot of anxiety (see Figure 1) to aid the interpretation of the interaction effects, we can see that the CT behaved slightly differently from the TGT and the LGT. Chinese learners in Hungary displayed an especially low anxiety level on the CT (M_TGT = 2.08, M_LGT = 2.05, M_CT = 1.75), while their peers in China reported the highest anxiety level on this particular task (M_TGT = 2.58, M_LGT = 2.54, M_CT = 2.72).

Figure 1.

Anxiety profile plot.

VI Discussion

When examining the effects of the factors of task condition and study context on learners’ flow experiences as measured by interest, attention, control, and challenge–skill balance using MANOVA, which treats the dependent scales contributing to flow as one underlying construct, we found no significant effect for task condition, but the effect of the factor of study context was statistically significant. It means that we did not find consistent differences between the three task types – TGT, LGT, and CT – with regard to flow, so we cannot claim that learners’ flow experiences differed significantly when performing these tasks. This finding fails to lend support to Lambert et al.’s (2017) results, who found robust differences between the TGT and LGT tasks concerning engagement. However, our study is different from Lambert et al.’s in two main respects: in the types of writing tasks examined and in the dependent variable used. First, we examined three task types by adding the CT, which allowed us to examine how a combination of learner agency and potential for creativity embedded in tasks might influence flow. This methodological choice was based on the theoretical premise that the interplay of control/freedom and constraint (regarding contents to write and language to use) can shape the cognitive and affective engagement of learners and affect their emotional experiences during task performance. Second, although there are authors who claim that engagement and flow are similar processes, and flow might represent the highest possible level of engagement (Seligman, 2011), according to others, the two constructs appear to be theoretically distinct (Henry et al., 2015; Philp & Duchesne, 2016; Piniel & Albert, 2019), which might also account for the differences in our findings.

When considering the findings of those studies that similarly to our research used the adaptation of the Perceptions questionnaire introduced by Egbert’s (2003) seminal paper, we find that in line with the original study, those investigations tended to measure flow as a sum or average of scores on the different flow scales (Aubrey, 2016, Czimmermann & Piniel, 2016). Therefore, their findings are not entirely comparable with the procedure used in this study, where MANOVA was performed on all four scales of flow simultaneously, treating them as reflecting the same underlying construct via a sophisticated statistical procedure instead of a simple addition of the scores.

We hypothesize that the lack of significant main effect for task condition in the current study might, on the one hand, be ascribed to methodological issues and limitations of the study like the different measurement procedure applied by us, the psychometric characteristics of the flow questionnaire, or the small sample size. On the other hand, it is also possible that the three tasks themselves might not have differed enough in their potential to create flow in learners as this is suggested by the univariate analyses described in the next paragraph, or their differences might have been overshadowed by the fact that they were all low stakes tasks, performed by the students only to comply with the requests of the researchers.

Although follow-up univariate tests suggested that, despite the lack of significant main effect for task condition, the tasks might in fact differ in three out of the four flow dimensions, after adjusting the significance level, only one significant finding remained. Based on the post-hoc tests, it seemed that there was a statistically significant difference between the level of control perceived by the participants on the TGT and LGT tasks: Learners reported higher levels of control on the task where they were free to come up with the content of their story on their own. As regards the CT task, the two learner groups seem to have reacted to it differently; CSH perceived more control while doing it compared to the LGT, while CSC perceived less control compared to the LGT, but both groups reported more control on it than on the TGT. Although the results comparing the CT to the other two task types are not statistically significant, they might still indicate that learners studying in the different contexts could potentially have perceived the CT task differently. In terms of its design features, the CT task gave students freedom to come up with their own content, while also posing some constraint in the form of requiring the use of certain vocabulary items. The simultaneous presence of freedom and constraints might have had different effects on the two learner groups studying in the ELF and EFL contexts. For example, learners in Hungary, the ELF context, might have had more experience in solving problems like the one posed by the mandatory use of certain vocabulary items, while learners studying in China might have had limited experiences in such problem solving both in their classroom and in the wider social context was well (Liu, 2020). However, results of the univariate analyses should be considered preliminary and in need of further validation in future research.

Although initially we mostly expected to find interaction effects when focusing on the study context as a potential factor influencing task performance, meaning that we hypothesized that tasks having different characteristics might result in varied performance depending on the study background of our participants, the MANOVA revealed a significant main effect for the study context itself. This means that Chinese learners studying in Hungary reported higher levels of flow no matter what kind of task they were exposed to than their peers studying in China. Although it is somewhat surprising, this finding seems to lend support to studies that claim that writing is an unpopular language learning activity in China, and learners do not enjoy writing tasks (Yu, 2021; Zhang, 2016). English is a foreign language in China, and it is possible that many learners do not feel the real-life usefulness of practicing writing as the chances of using writing to establish contacts with either native or lingua franca speakers of English are quite slim. It is much easier to use English as a means of communication in Hungary, both in speech and writing, especially in the capital city where our respondents lived. Also, more writing experience and training, together with increased exposure to the target language in communication, in the long run, might improve learners’ writing confidence and competence, possibly due to the transfer from speaking competence (Pérez-Vidal & Juan-Garau, 2011). Moreover, it is possible that learners in Hungary see the practical usefulness of writing and have a more favorable attitude to it, and this is why they report more flow experiences in connection with writing tasks. Finally, the possibility that our results are due to the more favorable attitude of Chinese learners studying in Hungary towards not only writing in English but English in general cannot be ruled out either. Of course, these hypotheses should be investigated in further studies focusing on learners’ attitudes to writing and links between attitudes and actual flow experiences.

The follow-up univariate analyses revealed that when treating the four flow components separately, it was only control and challenge–skill balance that showed significant differences between the two learner groups, in both cases learners in Hungary reporting higher levels. This finding supports, on the one hand, that, when using MANOVA, effects not necessarily significant on their own might combine resulting in a significant main effect (Stevens, 2009) and, on the other, it draws attention to these two dimensions as being particularly significant in causing differences in the flow experiences of the two groups. Perceiving more control and a higher challenge–skill balance in connection with a language task might be hypothesized to be more typical in a lingua franca setting, where learners are regarded as language users (Illés, 2012), which is a more empowering status to hold. Challenge–skill balance, which is the basic condition for experiencing flow (Csikszentmihalyi, 1990, 1997) means that the individual feels that their abilities are in sync with the demands of the task they need to execute. In case the challenge exceeds the skills or skills exceed the challenge posed, anti-flow experiences are expected to be reported according to Csikszentmihalyi. Therefore, the lower level of challenge–skill balance characterizing CSC can be expected to be accompanied by heightened anti-flow experiences, which is exactly in line with the MANOVA results of the anti-flow scales. As regards the non-significant univariate results for interest and attention, a possible explanation could be that all tasks, regardless of the study contexts, may have not been highly motivating or interesting to learners, as they might have considered completing the tasks as doing a favor for the researcher instead of, for example, as an exam situation to get high scores, therefore showing no particularly different level of interest or attention to them (Abbott, 2000; Liu et al., 2022).

The MANOVA conducted on the three anti-flow scales of anxiety, apathy, and boredom resulted in findings very similar to the flow MANOVA, in that the multivariate analysis did not show a significant main effect for task condition, but it did indicate a significant main effect for study context without any significant interaction effects. Although the results were the opposite of what was found for flow, they reinforced the flow findings because they indicated that the CSC group reported higher level anti-flow experiences than their peers in Hungary. The lack of challenge–skill balance, which according to Csikszentmihalyi (1975, 1990) leads to anti-flow experiences, did in fact result in such experiences for students in China. When the unique effect of each scale was examined, only the effect of anxiety was found to be statistically significant at the adjusted alpha level, once again highlighting the potential of MANOVA to detect significant effects that might not be evident at the level of the individual scales (Stevens, 2009). This finding also suggests that not only did students in China experience higher level of anti-flow in general, but they were particularly anxious compared to their peers studying in Hungary. The lack of significant differences between the two groups of learners in the case of boredom and apathy might in fact be rooted in the same causes that led to no significant differences in interest and attention listed in the previous paragraph, as these scales were found to be negatively related in earlier studies (Czimmermann & Piniel, 2016).

Interestingly, it was also the anxiety scale where the univariate analysis revealed significant interaction effects between the study context and the task type, which was judged intriguing enough to report despite the non-significant task effect revealed by the MANOVA. Closer analysis of our data revealed that it was the CT where the two groups of learners behaved very differently. Although the general tendency was that the TGT and LGT resulted in roughly similar anxiety levels, CSC reported higher anxiety levels overall than CSH. Regarding the CT, the two groups of students displayed different tendencies: the CT made learners in Hungary less anxious compared to the other two tasks while it made students in China more anxious compared to the TGT and the LGT. Being given freedom and having to tackle constraints at the same time affected these two learner groups differently. We hypothesize that it might be problem-solving experiences, which are believed to characterize ELF contexts more (Illés, 2012), that helped learners in Hungary to feel at ease when encountering this task. However, the lack of similar experiences in connection with language learning and use might have led to increased levels of anxiety in the case of learners in China (Liu, 2020; Yu, 2021; Zhang, 2016).

Although the lack of significant main effect for task condition in the MANOVA signals the lack of consistent differences that could be linked to task types as far as anti-flow experiences are concerned, the significant interaction effect identified in connection with anxiety might signal that our initial assumption that tasks might work differently in different contexts might not have been entirely incorrect. Nevertheless, further studies involving a larger range of tasks and possibly somewhat different measures would be needed to shed light on such effects. Their existence is also supported by Lambert and Zhang (2019) on Japanese learners of Chinese and English, whose reaction to different task types were most likely influenced by their previous experiences, or lack thereof, with the particular tasks.

VII Conclusions and implications of the study

The study aimed to explore the effects of task conditions and study context on learners’ flow and anti-flow experiences during L2 writing tasks. Research question 1 focused on whether different task conditions – namely TGT, LGT, and CT – would influence the flow and anti-flow experiences of learners. The multivariate analysis of variance (MANOVA) did not reveal significant differences in flow or anti-flow experiences across the three task types. Research question 2 examined the role of study context. The MANOVA results indicated a significant difference in flow and anti-flow experiences between CSH and CSC. CSH reported higher levels of flow experiences, suggesting that the ELF context, with its authentic communicative demands and varied writing experiences, may foster a sense of control and challenge–skill balance necessary for flow. At the same time, Chinese students studying in China reported higher levels of anti-flow experiences consistently across all the writing tasks they were given, manifesting in heightened levels of anxiety in particular. These group-related differences possibly reflect general attitudes that these students have towards writing in general, and we hypothesized that these attitudes might be related to the context in which these learners learn the English language. Research question 3 sought to identify any interactions between task conditions and study contexts, but no significant interaction effects were revealed by the MANOVA tests.

When interpreting the results, certain limitations of the current study should be pointed out and taken into consideration. The first limitation is the sample size. We believe that some tendencies, for example, regarding the effect for task conditions on learners’ flow experiences, might have reached statistical significance if the study had been conducted on a larger sample. In addition, our study would have yielded a deeper understanding of the combined effects of task condition and study context on learners’ flow and anti-flow or certain aspects of them (e.g. interest, attention, boredom, and apathy) if we had also collected qualitative data on learners’ perceptions of the levels of learner agency and potential for creativity embedded in each task with the help of interviews or self-reports, which constitutes another limitation of our study. Moreover, some of the reported results that were obtained using univariate ANOVAs conducted after MANOVA main effects were found to be non-significant should be considered preliminary and exploratory and in need of further investigation.

As for the practical implications of our study regarding writing task design and writing instruction, based on our research, different study contexts having affordances for either broader (English as a lingua franca) or more limited (English as a foreign language) English learning and writing experiences also have an important role in determining learners’ flow experiences in writing. For example, ELF contexts and their communicative requirements of frequent negotiation of meaning and more various writing training and experiences might prepare learners for tasks implemented in various conditions (the teacher-generated, learner-generated, and creative task conditions in our study), as they demonstrated overall greater flow on all the tasks compared to their peers in China who, on the contrary, reported greater anti-flow. Therefore, for students in China, it is imperative, on the one hand, to introduce more writing training and practice opportunities with the help of various tasks so as to raise their interest and build their confidence in writing. On the other hand, placing a bigger emphasis on the communicative function of the English language in society and providing more exposure for learners to use the language might also benefit their writing as a result of possible transfer effects in the long run (Pérez-Vidal & Juan-Garau, 2011).

Footnotes

Appendix A

Appendix B

Acknowledgements

We would like to thank the head teachers who helped contact the participants and organize data collection. We would also like to extend our gratitude to the participants of this study.

Author Contributions

Conceptualization, Q.W. & Á.A; Methodology, Q.W. & Á.A.; Formal Analysis, Q.W. & Á.A.; Investigation, Q.W.; Writing – Original Draft Preparation, Q.W. & Á.A.; Writing – Review & Editing, Q.W. & Á.A.; Supervision, Á.A. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The data from this study can be made available upon request to the corresponding author.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The second author is a member of the MTA-ELTE Foreign Language Teaching Research Group. Her work was funded by the Research Program for Public Education Development of the Hungarian Academy of Sciences.

Institutional Review Board Statement

The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board (or Ethics Committee) of the Doctoral Program of Intercultural Linguistics, ELTE (date of approval: 15 August 2018).

Informed Consent Statement

Informed consent was obtained from all participants involved in the study.

ORCID iDs

Qiaoxia Wu

Ágnes Albert

References

Abbott

(2000). ‘Blinking out’ and ‘having the touch’: Two fifth-grade boys talk about flow experiences in writing. Written Communication, 17, 53–92.

Abrams

Z.I.

(2019). Collaborative writing and text quality in Google Docs. Language Learning and Technology, 23, 22–42.

Allan

(2004). Oxford Placement Test (1 or 2). Oxford University Press.

Aubrey

(2016). Inter-cultural contact and flow in a task-based Japanese EFL classroom. Language Teaching Research, 21, 717–734.

Aubrey

(2017). Measuring flow in the EFL classroom: Learners’ perceptions of inter- and intra-cultural task-based interactions. TESOL Quarterly, 51, 661–692.

Barbaranelli

Lee

C.S.

Vellone

Riegel

(2015). The problem with Cronbach’s alpha: Comment on Sijtsma and van der Ark. Nursing Research, 64, 140–145.

Bennett

Heritage

Allen

(2023). SPSS statistics: A practical guide (5th ed.). Cengage.

Bronfenbrenner

(1979). The ecology of human development: Experiments by nature and design. Harvard University Press.

Cho

(2018). Task complexity and modality: Exploring learners’ experience from the perspective of flow. The Modern Language Journal, 102, 162–180.

10.

Cogo

Jenkins

(2010). English as a lingua franca in Europe: A mismatch between policy and practice. European Journal of Language Policy, 2, 271–294.

11.

Cohen

(1988). Statistical power analysis for the behavioral sciences. Lawrence Erlbaum.

12.

Csikszentmihalyi

(1975). Beyond boredom and anxiety. Jossey-Bass.

13.

Csikszentmihalyi

(1989). The dynamics of intrinsic motivation. In Ames

Ames

(Eds.), Handbook of motivation theory and research (pp. 45–71). Academic Press.

14.

Csikszentmihalyi

(1990). Flow: The psychology optimal experience. Harper Perennial.

15.

Csikszentmihalyi

(1997). Finding flow: The psychology of engagement with everyday life. Basic Books.

16.

Csikszentmihalyi

LeFevre

(1989). The optimal experience in work and leisure. Journal of Personality and Social Psychology, 56, 815–822.

17.

Czimmermann

É.

Piniel

. (2016). Advanced language learners’ experiences of flow in the Hungarian EFL classroom. In Macintyre

P.D.

Gregersen

Mercer

(Eds.), Positive psychology in SLA (pp. 193–214). Multilingual Matters.

18.

Dewaele

J.M.

MacIntyre

(2014). The two faces of Janus? Anxiety and enjoyment in the foreign language classroom. Studies in Second Language Learning and Teaching, 4, 237–274.

19.

Dewey

D.P.

Belnap

R.K.

Steffen

(2018). Anxiety: Stress, foreign language classroom anxiety, and enjoyment during study abroad in Amman, Jordan. Annual Review of Applied Linguistics, 38, 140–161.

20.

Dörnyei

(2007). Research methods in applied linguistics. Oxford University Press.

21.

Douglas Fir Group. (2016). A transdisciplinary framework for SLA in a multilingual world. Modern Language Journal, 100, 19–47.

22.

Egbert

(2003). A study of flow theory in the foreign language classroom. The Modern Language Journal, 87, 499–518.

23.

Getzels

J.W.

Csikszentmihalyi

(1976). The creative vision: A longitudinal study of problem finding in art. Wiley Interscience.

24.

Ghasemi

Rezvani

Namaziandost

(2021). EFL learners’ perception of task experience through flow outlook: Task complexity and modality in focus. Issues in Language Teaching, 10, 301–334.

25.

Henry

Davydenko

Dörnyei

(2015). The anatomy of directed motivational currents: Exploring intense and enduring periods of L2 motivation. Modern Language Journal, 99, 329–345.

26.

Illés

. (2012). Learner autonomy revisited. ELT Journal, 66, 505–513.

27.

Jiang

(2022). A correlation study on topic familiarity and English writing anxiety of senior high school students. Unpublished doctoral dissertation, Jilin International Studies University, Jilin, China.

28.

Kormos

(2014). Differences across modalities of performance. In Byrnes

Manchón

R.M.

(Eds.), Task-based language learning: Insights from and for L2 writing (pp. 193–216). John Benjamins.

29.

Kormos

Préfontaine

(2017). Affective factors and fluent performance: French learners’ appraisals of L2 speech tasks. Language Teaching Research, 21, 699–716.

30.

Kormos

Trebits

(2012). The role of task complexity, modality and aptitude in narrative task performance. Language Learning, 62, 439–472.

31.

Lambert

Minn

(2007). Personal investment in L2 task design and learning: A case study of two Japanese learners of English. Journal of Educational Learning and Innovation, 7, 127–148. https://revistas.uned.es/index.php/ELIA/article/view/18276 (accessed May 2024).

32.

Lambert

Philp

Nakamura

(2017). Learner-generated content and engagement in second language task performance. Language Teaching Research, 21, 665–680.

33.

Lambert

Zhang

G.Q.

(2019). Engagement in the use of English and Chinese as foreign languages: The role of learner-generated content in instructional task design. The Modern Language Journal, 103, 391–411.

34.

Lantolf

Thorne

(2006). Sociocultural theory and the genesis of second language development. Oxford University Press.

35.

Larsen-Freeman

(2019). On language learner agency: A complex dynamic systems theory perspective. The Modern Language Journal, 103, 61–79.

36.

Lee

S.H.

Kinginger

(2018). Narrative remembering of intercultural encounters: A case study of language program reintegration after study abroad. The Modern Language Journal, 102, 578–593.

37.

Little

Dam

Legenhausen

(2017). Language learner autonomy: Theory, practice and research. Multilingual Matters.

38.

Liu

Zhang

Liu

(2022). Flow experience in foreign language writing: Its effect on students’ writing process and writing performance. Frontier in Psychology, 13, 952044

39.

Liu

Y.C.

(2020). Gaozhong yingyu xiezuo jiaoxue gaoxiao celue tanjiu [A study on the senior high school English writing teaching strategies]. Internet Education, 6, 188–189.

40.

Liu

Y.F.

(2017). Gaosan yingyu xiezuo jiaolv de diaocha yu fenxi [A study on writing anxiety in Senior three students]. English Square, 2, 138–139.

41.

Hyland

(2007). Enhancing students’ engagement and motivation in writing: The case of primary students in Hong Kong. Journal of Second Language Writing, 16, 219–237.

42.

Maehr

M.L.

(1984). Meaning and motivation: Toward a theory of personal investment. Research on Motivation in Education, 1, 115–144.

43.

Maley

Kiss

(2018) Creativity and English language teaching: From inspiration to implementation. Palgrave Macmillan.

44.

Maslow

A.H.

(1971). The farther reaches of human nature. Viking Press.

45.

Mayo

M.D.P.G.

Agirre

A.I

. (2019). Task modality and pair formation method: Their impact on patterns of interaction and LREs among EFL primary school children. System, 80, 165–175.

46.

Öveges

Csizér

(Eds.) (2018). Vizsgálat a köznevelésben folyó idegennyelv-oktatás kereteiről és hatékonyságáról [Study on the framework and effectiveness of foreign language education in public education]. Oktatási Hivatal.

47.

Payant

Zuniga

(2022). Learners’ flow experience during peer revision in a virtual writing course during the global pandemic. System, 105, 102715.

48.

Pérez-Vidal

Juan-Garau

(2009). The effect of study abroad (SA) on written performance. EuroSLA Yearbook, 9, 269–295.

49.

Pérez-Vidal

Juan-Garau

(2011). The effect of context and input conditions on oral and written development: A study abroad perspective. International Review of Applied Linguistics in Language Teaching, 49, 157–185.

50.

Philp

Duchesne

(2016). Exploring engagement in tasks in the language classroom. Annual Review of Applied Linguistics, 36, 50–72.

51.

Piniel

Albert

. (2019). Motivation and flow. In Lamb

Csizér

Henry

Ryan

(Eds.), The Palgrave handbook of motivation for language learning (pp. 579–597). Springer.

52.

Polio

(2012). The acquisition of second language writing. In Gass

S.M.

Mackey

(Eds.), The Routledge handbook of second language acquisition (pp. 319–334). Routledge.

53.

Poupore

(2014). The influence of content on adult L2 learners’ task motivation: An interest theory perspective. The Canadian Journal of Applied Linguistics, 17, 69–90.

54.

Qin

X.Q.

Yang

J.J.

(2015). EFL writer’s block and writing apprehension: constructs and source. Foreign Language and Literature Research, 1, 97–106.

55.

Sasaki

(2004). A multiple-data analysis of a 3.5 year development of EFL student writers. Language Learning, 54, 525–582.

56.

Sasaki

(2007). Effects of study-abroad experiences on EFL writers: A multiple-data analysis. The Modern Language Journal, 91, 602–620.

57.

Seidlhofer

(2005). English as a Lingua Franca. Oxford University Press.

58.

Seligman

M.E.

(2011). Flourish: A visionary new understanding of happiness and well-being. Free Press.

59.

Stevens

J.P.

(2009). Applied multivariate statistics for the social sciences (5th ed.). Routledge.

60.

Tin

T.B.

(2011). Language creativity and co-emergence of form and meaning in creative writing tasks. Applied Linguistics, 32, 215–235.

61.

Tin

T.B.

(2012). Freedom, constraints and creativity in language learning tasks: New task features. Innovation in Language Learning and Teaching, 6, 177–186.

62.

Tin

T.B.

(2022). Unpacking creativity for language teaching. Routledge.

63.

Tok

Ş.

Kandemir

. (2015). Effects of creative writing activities on students’ achievement in writing, writing dispositions and attitude to English. Procedia – Social and Behavioral Sciences, 174, 1635–1642.

64.

Ushioda

(2015). Context and complex dynamic systems theory. In Dörnyei

Henry

MacIntyre

P.D.

(Eds.). (2015). Motivational dynamics in language learning (pp. 47–54). Multilingual Matters.

65.

Williams

(2012). The potential role(s) of writing in second language development. Journal of Second Language Writing, 21, 321–331.

66.

Q.X.

Albert

(2022). The role of teacher-generated, learner-generated, and creative content in Chinese EFL students’ narrative writing: A contextual perspective. Languages, 7, 212.

67.

Yang

W.B.

(2020). Gaokao bi shua juan, shinian zhenti (yingyu) [Must-do tests for College Entrance Exam, ten past exam papers (English)]. Kaiming Press.

68.

Q.Q.

(2021). The creative writing to improve students’ English writing skills in high schools: A case study of continuing writing and writing through pictures. Unpublished doctoral dissertation, Southeast University, Changsha, China.

69.

Zalbidea

(2017). ‘One task fits all’? The roles of task complexity, modality, and working memory capacity in L2 performance. Modern Language Journal, 101, 335–352.

70.

Zhang

X.D.

(2016). The composition analysis of college entrance examination between China and America. Unpublished doctoral dissertation, Northeast Normal University, Jilin, China.

71.

Zhang

(2021). A study on the psychological correlates of flow experience in L2 writing. Unpublished doctoral dissertation, Southeast University, Changsha, China.