“I Became Self-Motivated”: Investigating Chinese Students’ Motivation in Assessed Collaborative Writing Tasks

Abstract

While the existing literature on collaborative writing has implied the benefits of its implementation in second language (L2) classrooms for students’ language learning, the assessment of collaborative writing has received scant attention in both language assessment literature and L2 writing literature. This study proposes an assessment scheme considering both products and processes when assessing online writing undertaken collaboratively. It employs a case study approach to explore two pairs of Chinese students’ motivation toward assessed collaborative writing tasks. Drawing on data from participants’ reflections, pair interviews, stimulated recall interviews, online pair-talks, and revision histories, this study reveals that students with similar L2 proficiency may hold different motivations during product-graded, web-based collaborative writing assignments compared to product & process-graded assignments of the same nature. Additionally, this study finds that assessment can lead to changes in students’ motivations, which in turn affect learner participation in collaborative writing. The results can enhance our comprehension of the impact of assessment on learners’ motivation in computer-mediated collaborative writing tasks, offering fresh insights into strategies for strengthening learner motivation during the implementation of such tasks. The study suggests teachers should prioritize assessment as a pivotal element when incorporating computer-assisted collaborative writing, conscientiously consider students’ motivations in activity design, and recognize the growing significance of computer technologies in L2 learning.

Plain language summary

Investigating Chinese students’ motivation in assessed collaborative writing tasks

This study focuses on how collaborative writing tasks in second language (L2) classrooms can benefit students’ learning, but it notes that assessing these tasks has not been well-studied. The researchers propose a way to assess collaborative writing that looks at both the final writing and how students work together. They studied two pairs of Chinese students and found that students with similar language skills can have different motivations when the task is graded only on the final product compared to when it is graded on both the product and the process. They also found that how students are assessed can change their motivation, which affects how they participate in collaborative writing. The study suggests that teachers should consider assessment when designing collaborative writing tasks, think about students’ motivations, and recognize the importance of using technology in language learning.

Introduction

Collaborative writing, recognized as an effective and viable instructional approach, has found widespread adoption in second language (L2) education over recent decades (Fernández Dobao, 2012; Wigglesworth & Storch, 2012; Zhang & Plonsky, 2020). Collaborative writing involves the collaborative effort of two or more learners to jointly create a single text (Storch, 2013). Prior research has demonstrated numerous advantages of collaborative writing, such as providing learners with opportunities to collaborate in resolving language-related challenges and jointly constructing new language knowledge (Fernández Dobao, 2012; Wigglesworth & Storch, 2012). It also enhances linguistic accuracy (Wigglesworth & Storch, 2012) and promotes attention to language structures and discourse (Swain & Lapkin, 1998). The application of collaborative writing is rooted in Vygotsky’s (1978) sociocultural theory, which asserts that interaction not only facilitates language learning but also acts as a driving force in language acquisition. Moreover, this theory views all learning as inherently social, deeply embedded in sociocultural contexts. In essence, sociocultural theory underscores the significance of social interaction and language as pivotal tools in the language learning process. Research within the realm of L2 learning from this perspective has demonstrated that engaging with others provides learners with opportunities to pool knowledge and formulate language concepts, thereby fostering language learning (Storch, 2002; Swain & Lapkin, 1998). For instance, Fernández Dobao (2012) investigated the benefits of collaborative writing tasks. When working in pairs, learners have opportunities to collaboratively tackle language-related issues, resulting in the production of linguistically more precise written and oral texts through interaction.

Moreover, an extensive body of research on L2 learning has delved into learners’ interaction patterns during collaborative writing (Li & Zhu, 2013; Storch, 2002; Watanabe & Swain, 2007). Storch (2002) conducted a longitudinal, classroom-based study, examining the nature of interaction based on criteria such as equality and mutuality. Equality denotes the extent of control or authority that group members wield over the task, while mutuality reflects the level of engagement group members have with one another’s contributions. Four distinct interaction patterns emerged: collaborative, dominant/dominant, dominant/passive, and expert/novice (Storch, 2002). These four patterns can be visually represented by the quadrants formed through the intersection of two axes, as illustrated in Figure 1.

Figure 1.

A model of dyadic interaction (Storch, 2002).

Quadrant 1 represents the collaborative pattern of interaction, where participants demonstrate relatively high equality and mutuality. Quadrant 2 illustrates the dominant/dominant pattern of interaction, characterized by high equality but relatively low mutuality. This quadrant also encompasses the cooperative pattern of interaction, in which learners may contribute equally to the task but engage minimally with each other’s input. Quadrant 3 displays the dominant/passive pattern of interaction, marked by low equality and mutuality. In this case, one member takes a controlling role while the other assumes a passive role. Quadrant 4 depicts the expert/novice pattern of interaction, featuring high mutuality but low equality. Here, an individual acts as an expert, encouraging a novice participant to participate in the task. Researchers have applied this model across diverse contexts (Li & Kim, 2016; Li & Zhu, 2013; Storch & Aldosari, 2013; Watanabe & Swain, 2007). Among these patterns, the collaborative pattern is most conducive to language learning (Watanabe & Swain, 2007), with both collaborative and expert/novice patterns also fostering language acquisition (Storch, 2002).

Patterns of interaction offer advantages to L2 learning by providing learners with opportunities for languaging (Swain, 2006) and generating more language-related episodes (LREs). LREs refer to instances where learners self-correct or deliberate on linguistic elements, grammatical choices, spelling, and pronunciation (Swain & Lapkin, 2001). LREs represent ongoing L2 learning (Swain & Lapkin, 1998; Swain & Watanabe, 2012), positively impacting L2 development (Leeser, 2004). Prior research suggests that the occurrence of LREs during collaborative tasks is influenced by factors such as proficiency level (Kim & McDonough, 2008; Leeser, 2004; Watanabe & Swain, 2007), interaction patterns (Storch, 2002; Watanabe & Swain, 2007), level of pair engagement (McDonough, 2004; Storch, 2008), task type (Suzuki & Itagaki, 2007; Swain & Lapkin, 2001), and participant count (Fernández Dobao, 2012). For instance, Leeser (2004) showed how learner proficiency affects the quantity and types of LREs. Williams (1999, 2001) found that LRE numbers increased with higher course levels, with more LREs generated by proficient learners. Fernández Dobao (2012) compared group, pair, and individual work, revealing that participant count did not heavily influence LRE focus but positively impacted their frequency and resolution. Groups produced more LREs, and these were mostly resolved correctly (Fernández Dobao, 2012). Regarding the effects of pair interaction patterns on LREs, a study by Watanabe and Swain (2007) demonstrated that participants with a collaborative orientation (collaborative and expert/novice) produced more LREs than those with a non-collaborative orientation (dominant/passive and expert/passive)

As demonstrated by the aforementioned studies, not all types of interaction prove conducive to collaborative learning, underscoring the significance of exploring interaction patterns in collaborative writing. Numerous factors have been identified as influencers of pair interaction patterns, resulting in varied LRE generation. These factors encompass personality differences (Villamil & de Guerrero, 1996), grouping strategies (Dobao, 2014; Mozaffari, 2017), proficiency level (Kowal & Swain, 1994; Leeser, 2004; Williams, 2001; Yule & Macdonald, 1990), and attitude (Chen & Yu, 2019). For instance, Chen and Yu (2019) discovered that learners’ attitudes toward collaborative writing may shift during activities, with positive attitudes yielding more learning opportunities. While these studies offer insights into factors influencing interaction patterns, what remains underexplored in the current body of L2 collaborative writing research is the nexus between motivation and peer interaction patterns.

Motivation, which can be broadly defined as the driving force behind an individual’s choices, actions, and the effort they invest in pursuing their goals (Dörnyei & Ushioda, 2011, p. 3), is a central theme in our manuscript. Within the realm of motivation, it is crucial to distinguish between intrinsic and extrinsic forms of motivation due to their distinct impacts on outcomes. The Self-Determination Theory (SDT), a comprehensive theory of human personality and motivation, is particularly relevant in this context. SDT elucidates the various forms of motivation, including intrinsic and different types of extrinsic motivation, and how they influence responses across different situations, as well as their effects on social and cognitive development and personality (Deci & Ryan, 1985). SDT emphasizes the critical influence of the social and cultural context in either supporting or hindering individuals’ basic psychological needs, their sense of self-direction, performance, and overall well-being (Ahmadi et al., 2023).

In our study, we explore motivation within the context of L2 (second language) learning. L2 motivation encompasses a range of factors, as identified by Csizer and Dornyei (2005), including Integrativeness, Instrumentality, Vitality of the L2 Community, Attitudes toward the L2 Speakers/Community, Cultural Interest, Linguistic Self-Confidence, and Milieu. While there is limited existing research addressing participant motivation in the collaborative writing context, it is widely acknowledged that motivation plays a pivotal role in determining the success of collaborative learning (Dörnyei, 2001). For example, Storch (2004) found that learners’ task goals and orientations impact the formation of interaction patterns. Moreover, Chen (2020) demonstrated that students with comparable proficiency levels might exhibit varying levels of motivation, potentially influencing their participation. Despite these insights, it remains unclear how motivation specifically shapes the formation of interaction patterns within collaborative writing tasks. With the growing development and importance of Web 2.0 technologies, such as wikis, Google Docs, blogs, RSS feeds, and Shimo, for negotiation and collaboration, there has been an upsurge in the use of these web-based collaboration tools during the asynchronous collaborative writing. Such web-based collaboration tools have been considered to be convenient and friendly, and to improve online collaboration and interactive learning (Doering et al., 2007; Kai-Wah Chu, 2008; Kim et al., 2022). Previous studies have investigated a few advantages of web-based asynchronous collaborative writing, including facilitating learners’ expression of their thoughts and group interaction, improving their L2 writing ability and organization skills, and fostering learners’ collaborative learning ability (Kessler, 2009; Kessler & Bikowski, 2010; Kost, 2011; Lee, 2010; Leight & Leight, 2008; & Li & Zhang, 2021). A body of research has explored the patterns of interaction in web-based asynchronous collaborative writing. For instance, Li and Kim (2016) examined the dynamics of peer interaction during two collaborative writing tasks that used a Wikispaces site, and the study found that the nature of tasks has effects on students’ wiki writing interaction.

However, existing studies offer limited insights into how motivation impacts learners’ interactions in the context of web-based collaborative writing tasks. Prior research has predominantly focused on product-graded collaborative writing assignments, where the assessment primarily considers the final written work, such as content, organization, and language usage. To address this gap, there is a need for further investigation that centers on product & process-graded collaborative writing tasks. These assignments assess not only the resulting text but also the entire writing process, including peer interactions and the quality of LREs.

However, these studies offer limited, if any, insights into the manner in which motivation influences learners’ interaction during web-based collaborative writing tasks. Prior research has often centered on product-graded collaborative writing assignments, wherein participants’ writing performance is evaluated in terms of content, organization, and language usage. There exists a need for further investigation focusing on product & process-graded collaborative writing tasks, wherein learners’ writing performance is assessed not only with respect to the resulting text but also the writing process itself—encompassing peer interaction and the quality of LREs. With these goals in mind, this study aims to address the following questions:

• RQ1: What motivates students to participate in collaborative writing, whether for product-graded or product & process-graded assignments?

• RQ2: In what ways, if any, does the introduction of product & process-graded assessment impact students’ motivation to engage in collaborative writing throughout the writing process?

Methodology

This study employs a qualitative case study approach to explore the motivation of two pairs. This approach facilitates an examination of motivational shifts across tasks and their impacts on interaction patterns and the emergence of LREs.

Context and Participants

This study was conducted within a 16-week composition class at a university in mainland China, where English proficiency has become a pivotal criterion for achieving success. The class consisted of 42 intermediate-level students who had undertaken a proficiency test upon their initial enrollment in the language institute. Given their placement at the intermediate level, the students displayed relatively similar L2 writing proficiencies. The course was led by the first author, serving as the instructor. While these students shared a fairly consistent level of second language (L2) proficiency, we anticipated significant individual variations in their computer-assisted language learning experiences and their beliefs related to the assessment of collaborative work. Pairs were initially formed through the students’ free choice, with occasional adjustments made by the instructor to ensure a mix of participants with varying levels of language proficiency (Li & Kim, 2016; Storch, 2013). During the class sessions, students were tasked with composing two argumentative essays, with topics of their choice (centered around education and technology), working in pairs and utilizing the Shimo tool. The selection of argumentative essay writing was motivated by its importance in academic writing. Essay topics were drawn from the College English Test Band 4 (CET4), a national English proficiency examination. The participants’ scores on the proficiency test varied between 55 and 65 out of a possible 100. Out of the 42 individuals in the study, 30 were female, and 12 were male. Their ages fell within the range of 19 to 20 years (with an average age of 19.3), and they had, on average, 9.74 years of formal English learning. Weeks 1 to 10 comprised the standard teaching weeks, focusing on intermediate reading and writing skills. At the onset of the research, students were furnished with writing guidelines detailing how to collaboratively write and revise their essays using the Shimo tool. Additionally, they received training before embarking on their initial writing task.

Out of 21 pairs, two pairs were purposively selected for participation in this study based on their distinct and contrasting interaction patterns observed at the study’s commencement and conclusion. This selection was guided by the research focus on examining the evolution of collaborative writing dynamics. Pair 8 (consisting of a male and a female student) transitioned from a dominant/dominant interaction pattern to a collaborative pattern by the study’s conclusion. Similarly, Pair 10 (comprising two female students) evolved from a dominant/passive pattern to a collaborative pattern. These pairs were chosen because their interactional trajectories highlight significant shifts that align closely with the study’s objectives. All the participants had previous exposure to the concept of collaborative writing, having engaged in it during the English course they took in the previous semester with a different instructor (not the first author of this study). Importantly, none of the students were familiar with Shimo, the specific computer-mediated environment used for this study, and they had no prior experience with collaborative writing on such platforms. To ensure anonymity, all data were assigned pseudonyms for analysis, with the master list of real names and pseudonyms subsequently discarded.

Data Collection Procedure

The data for this research were drawn from various sources, including audio-recordings of semi-structured interviews and stimulated interviews, pair chats, revision histories, written texts, and reflection journals. Table 1 provides an overview of the data collection timeline and procedures. In contrast to the traditional rubric, the new rubric (see Appendix 1) incorporated three conventional parameters for assessing writing product—content, organization, and language use—along with two parameters for evaluating the collaborative process—mutuality and equality. The inclusion of the latter dimensions draws on recent literature on collaboration dynamics (e.g., Li & Zhu, 2013, 2017; Storch, 2013). The mutuality scale, on a five-point continuum, gaged the extent of active response and affective involvement between pairs, while the equality descriptors assessed joint contributions and control over the task.

• Stage One: In week 11, all pairs completed a pre-task survey to explore their previous collaborative writing experiences.

• Stage Two: In week 12, students were introduced to the collaborative writing tool, Shimo, and its concept. They were given the autonomy to select pair members of their choice and collaboratively complete a writing task using Shimo.

• Stage Three: In week 13, students reviewed the traditional rubric and were allocated 2 hr to write an essay on either “education” or “technology.” Post-assignment, students composed reflections on pair dynamics, individual contributions, and motivations related to the writing activity.

• Stage Four: The new rubric (Appendix 1) was introduced in week 14. The second web-based asynchronous collaborative writing activity centered on the chosen topic of “education” or “technology.” Reflection journals akin to those in Stage Three were collected.

• Stage Five: In week 15, post-task interviews were conducted with each pair, aiming to refresh their memory of collaborative writing experiences and elucidate specific behaviors and ideas during the two activities. Each interview, audio-recorded, lasted 15 to 20 min.

• Stage Six: In week 16, two pairs underwent stimulated interviews to delve into their motivations for both the product-graded and product & process-graded writing tasks, as well as to identify influencing factors. These interviews, audio-recorded, spanned 20 to 30 min.

Table 1.

Data Collection Procedures and Timeframe.

Stage	Time	Tasks
1	Week 11	Pre-task survey
2	Week 12	Pre-task training
		Pair formation
3	Week 13	Traditional rubric was reviewed
		Writing task 1
		Reflection journal 1
4	Week 14	New rubric was introduced
		Writing task 2
		Reflection journal 2
5	Week 15	Post-task interviews
6	Week 16	Stimulated interviews

Note. Weeks 1 to 10 were regular teaching weeks.

Data Analysis

The pair chats, along with the recordings of semi-structured interviews and stimulated interviews, were meticulously translated and transcribed in their entirety by the researchers. The pair chats were subjected to analysis, focusing on interaction patterns using Storch’s (2002) model of dyadic interaction. Additionally, they were scrutinized for instances of language-related episodes (LREs), as outlined by Swain and Lapkin (2001), in order to delve into learners’ deliberate attention to language use during their interactions

LREs: Building upon Swain and Lapkin (1998, 2001), LREs encapsulate learners’ explicit language attention and signify their progress in learning. The initial coding of the pair chats centered on LRE identification. Subsequently, these episodes were further categorized into three groups: grammar-based (F-LREs), lexis-based (L-LREs), and mechanics-based (M-LREs), based on the specific linguistic aspect each LRE pertained to. F-LREs involved concerns like verb tense selection and article usage; L-LREs concerned word meanings and word choice, including preposition usage; and M-LREs focused on spelling, punctuation, and pronunciation (Storch & Aldosari, 2013). Furthermore, each LRE’s resolution was designated as either correctly resolved, incorrectly resolved, or unresolved, adhering to the framework established by Leeser (2004)

With the exception of excerpt 1, derived from the data of Qi and Long (all names used in this article are pseudonyms), an illustrative instance highlights the coding of an L-LRE that was correctly resolved. In this excerpt, Long and Qi deliberate on enhancing their collaborative writing through word choice revisions

Excerpt 1: L-LRE resolved correctly

Long:

Let’s check the following things: vocabulary, grammar and sentence pattern. Ok?

Qi:

Ok. I think maybe we should change the word “however”.

Long:

Why? “however” means grammatical turn, it’s suitable here.

Qi:

Yes, but we need to end this paragraph here because we wont continue this paragraph. So we should choose another suitable word here.

Long:

You are right. Which word you think is better?

Qi:

How about “fortunately”? It can be helpful to end this paragraph.

Long:

Good, I agree.

Excerpt 2 shows an example of when Tian and He took responsibility to find a correct word to resolve a language problem. The LRE was resolved correctly.

Excerpt 2: L- LRE- Resolved correctly

He:

“They should not be made”……Um, how to speak qiangzhide in English?

Tian:

Um…I don’t know.

He:

How about finding it from a dictionary?

Tian:

Ok, let me find for you. [looks up in a dictionary] Look, we can use the verb “obligatory” because it means “ that you must do because of the law, rule, etc.)

He:

Great!

Patterns of interactions: The pair chats conducted during the collaborative writing activity were also subjected to an analysis of interaction patterns, utilizing Storch’s (2002) model, which categorizes interactions into four patterns: collaborative, dominant-dominant, dominant-passive, and expert-novice. These patterns are classified based on Damon and Phelps’s (1989) concepts of equality and mutuality. Building on the insights from Mozaffari (2017), interaction patterns were considered predominant if they were present in more than 70% of the episodes.

Excerpt 3, derived from the pair chats between Tian and He, offers an illustration of how an episode of pair chats was categorized according to its dynamic interaction type

Excerpt 3: Dominant/passive

Tian:

There are three points in our written texts. Firstly, it’s convenient because we can search vocabulary at any times. Secondly, it can provide us accurate words because it incorporates different dictionaries. At the same time, we also can get the word’s pronunciation. Thirdly, there are some new ways for students to search, such as voice input and picture input. It’s enough to put these three points in this paragraph.

He:

Why not use the sentence of “in my opinion” in the beginning of the first paragraph?

Excerpt 3 furnishes an instance of a dominant/passive interaction pattern, wherein Tian assumed task control and provided the bulk of the ideas. In contrast, He exhibited limited engagement and minimal contribution to the task, aside from initiating a query about the text’s content.

To ensure reliability of the results, two additional raters who were not the authors independently analyzed all the interview transcripts, patterns of pair interaction, and LREs. The level of agreement between these raters for the interview transcripts was 89.2%, for classifying LREs was 95.3%, and for resolving LRE solutions was 93.9%. Any disagreements were resolved through thorough discussion.

Results

In this section, we present the findings from two pairs: Pair 8 and Pair 10. We delve into the participants’ motivations for two computer-based collaborative writing tasks and analyze how the product & process-graded assessment influenced participants’ collaborative writing endeavors throughout the writing process

Case 1: Qi and Long:“We Learned to Respect Each Other.!” (Task 1: Dominant/Dominant; Task 2: Collaborative)

At the outset, both participants’ motivations for collaborative writing were primarily fueled by the prospect of product assessment. They exhibited similar motivations toward the product-graded writing task. Qi, for instance, elucidated that his impetus was to complete the assignment and attain a favorable grade during Task 1:

Our teacher assigned us this collaborative writing task using the Shimo tool. This task is part of our assignment, so completing it is a necessity. We were informed that our text quality would be evaluated. Naturally, we aimed for high grades, hence putting in our best writing efforts. Throughout the writing process, I didn’t focus much on collaboration; my main aim was to swiftly finish the task. (Qi, Pre-task interview and Reflection journal 1)

Similarly, Long conveyed that the task requirements and the potential for a high grade constituted his primary motivation to engage in collaborative writing with his partner:

In my perspective, nothing held greater importance than task completion. Our teacher would evaluate our collaborative output. I aspired for our writing to earn substantial points, which prompted active collaboration with my partner. Both of us contributed ideas for our assignment, yet there were instances when we couldn’t persuade each other. (Long, Pre-interview and Reflection journal 1)

During Task 1, where the teacher solely evaluated the pair’s work based on text quality, both participants were primarily driven by the desire for a good grade and compliance with the teacher’s task stipulations. Qi and Long were motivated by the grade and task requirements, leading them to adopt a dominant/dominant interaction pattern. However, they placed less emphasis on true collaboration. Qi acknowledged that, “I wished for our produced text to secure a high grade, so I employed sophisticated sentence structures and eloquent vocabulary. However, I don’t believe my partner (Long) fully comprehended my ideas.” (Post-task interview) Similarly, Long noted, “There were moments when I disagreed with Qi’s ideas and expressed my viewpoint, but she didn’t seem receptive. I’d rather have worked on it alone.” (Post-task interview)

Hence, their motivation driven by grades and task specifications fostered a dominant/dominant interaction pattern between Qi and Long. Excerpt 4 from the pair chats of Qi and Long underscores the dominance of this pattern (high equality and low mutuality), where both actively voiced opinions but struggled to persuade each other:

Excerpt 4: Dominant/dominant -L- LRE- Unresolved

Qi:

Online dictionaries are easy to take.

Long:

We should change this sentence slightly. Online dictionaries are easy to carry and read. We don’t use the word “take ”.

Qi:

In my opinion, it’s unnecessary to use the verb “read ”.

Long:

Um…I can’t agree with you.

Qi:

Um…

In Excerpt 4, Long initiated a suggestion about a sentence. However, Qi dismissed Long’s idea and asserted his own interpretation. Pair members displayed sporadic positive attitudes, occasionally encouraging each other to collaboratively complete tasks. Yet, they frequently fell short of establishing a consistently positive and supportive environment. This context may have contributed to the pair generating a mere 34 LREs and resolving only 21 of them correctly (62% of the total LREs) during Task 1.

Intriguingly, participants exhibited diverse motivations in Task 2, where the teacher evaluated their collaborative work considering both product and process (mutuality and equality). This dual assessment structure significantly influenced participants’ motivations for collaborative writing. Apart from task requisites and grades, other motivations surfaced in response to the influence of process assessment. Long’s motivation centered around self-improvement and learning from his partner during Task 2, as revealed in his reflection journal:

For Task 2, our teacher indicated that she would assess both our collaborative writing process and our product. Hence, during the writing process, I paid heightened attention to collaborating effectively with my partner. In this task, I grasped the importance of respecting my partner’s opinions. I made a conscious choice to listen attentively to his ideas, particularly when our viewpoints differed. We engaged in discussions when grappling with language challenges and collaborated to surmount difficulties. Furthermore, I observed that my partner demonstrated genuine respect for my viewpoints and provided substantial assistance. (Long, Post-task interview and Stimulated interview)

At this juncture, Long’s statements underscored how the product & process-graded assessment acted as a catalyst, driving him to illuminate the collaborative aspect of his partnership. This influence propelled him to hold his partner in higher regard. As a response to this assessment framework, pair members engaged actively with each other’s queries and responded cohesively to task-related challenges.

Qi echoed similar sentiments, indicating the profound impact of the product & process-graded assessment on her motivation. This assessment structure prompted her to be highly motivated to enhance her learning through collaboration with her partner. Qi articulated her motivation, stating, “I felt a sense of excitement to learn from him; it presented an opportunity for personal enrichment.” (Reflection journal 2) Reflecting on Task 2, Qi further expounded:

During Task 1, I was skeptical about receiving accurate feedback or assistance from Long, resulting in my frequent disregard of his points. However, in Task 2, our teacher informed us of the assessment of our collaboration process, prompting me to earnestly collaborate with my partner. Through this collaborative journey, we learned to respect each other’s efforts. We encouraged and shared our ideas extensively. Astonishingly, I found myself gaining far more knowledge than I had anticipated. (Qi, Post-task interview and Stimulated interview)

As we can see, Qi’s motivation also changed in Task 2 after knowing their collaboration process would be assessed, and she had begun to respect her partner’s assistance. Pair members discussed the direction of the writing and collectively made decisions. Both of them made equally tremendous efforts to contribute to the task.

With the shift in the pair’s motivation, they prominently displayed a collaborative interaction pattern characterized by elevated equality and mutuality in their contributions. This collaborative dynamic was evident as they collaborated on Task 2. Remarkably, the number of LREs during Task 2 surged to 65, with a corresponding increase in correctly resolved LREs to 53 (82%). These outcomes highlighted a marked enhancement in both the quality and quantity of LREs during Task 2 in comparison to Task 1. This transformation coincided with the shift in the pair’s motivation, triggered by the product & process-graded assessment structure. Excerpt 5 extracted from the pair chats serves as a prime illustration of the collaborative pattern, characterized by heightened participation and the successful resolution of LREs:

Excerpt 5: Collaborative- -L- LRE- Resolved correctly

Long:

I think that art classes are so important for students’ development that should be compulsory in high school.

Qi:

I think this sentence can be described in another way, that would be better.

Long:

Um…How?

Qi:

So art classes are important for students’ development and should be

compulsory in high school.

Long:

It is simpler than the older one.

Qi:

Of course.

Long:

Ok, we use this new version.

Qi:

Wait, we have used many times of “so”, can we choose another one?

Long:

Sure, how about “therefore”?

Qi:

It performs the same grammatical function as “so”.

Long:

Yes, let’s change.

In summation, the data portrayed a distinct evolution in this pair’s motivation toward collaborative writing across two tasks, instigated by the implementation of product & process-graded assessment. This transformative influence in motivation subsequently reverberated across multiple facets of their collaborative endeavor, including patterns of interaction and the abundance and quality of LREs. Hence, a meticulous scrutiny of the collected data unveiled that the introduction of product & process-graded assessment significantly impacted the pair’s motivation toward collaborative writing. Beyond the intrinsic factors of task requisites and grades, their newfound motivation encompassed a genuine desire to foster mutual learning and self-improvement. This motivation, in turn, galvanized their concentrated focus on the collaborative process throughout the activity

Case 2: Tian and He: “Assisting Others is Another Way to Learn” (Task 1: Dominant/Passive; Task 2 Collaborative)

Parallel to Case 1, Case 2 depicted a congruent pattern of motivation concerning collaborative writing, propelled by the product assessment aspect of Task One. Tian characterized collaborative writing as a task to be completed and pinpointed the considerable influence of both the teacher’s assignment and the prospective grade on his motivation to partake in collaborative writing. In his perspective:

As per the teacher’s instructions, our writing work would be subject to evaluation. Personally, I view writing as an individual endeavor, and perhaps I could have written more effectively if I had worked on it independently. Nonetheless, it was our responsibility to collaboratively fulfill the teacher’s task, isn’t it? Naturally, I aspired for a high grade. I held the belief that I contributed more than my partner, who, in my view, simply followed my lead. (Tian, Post-task interview and Reflection journal 1)

Similarly, He expressed, “Writing in tandem with another person proved to be a challenging endeavor.” (Stimulated interview) He aligned his motivations for collaboration with the task requirements and the promise of a favorable grade. His sentiments about collaborative writing were articulated as follows:

For me, it boiled down to a mere writing task. If we could execute it adeptly, a commendable grade would be the outcome. My partner adopted his own approach to writing without considering whether I grasped the content or not. Eventually, I relinquished my involvement and permitted him to proceed. This marked my inaugural experience with collaborative writing alongside my partner, an encounter I found to be less than satisfactory. (He, Post-task interview and Reflection journal 1)

In Task 1, Tian exhibited a lack of full acknowledgment for his partner’s contributions, coupled with minimal participation in discussions. This apprehension on Tian’s part was not lost on He during the task. He decided to retract his active participation and regarded Tian as an overbearing partner. Labeling Tian as a “dominating and bossy partner,” He opted for a passive role. At this juncture, the pair’s interaction pattern materialized as dominant/passive, characterized by diminished equality and mutuality. This dynamic yielded a total of 56 LREs, of which only 22 (39%) were resolved correctly. This excerpt illustrates this pattern:

Excerpt 6: Dominant/passive -L- LRE- Resolved correctly

Tian:

Let’s check our grammar and vocabulary

He:

Okay.

Tian:

Is there any problem?.

He:

Um…

Tian:

Here, “online dictionaries is bring much more convenience for our study and work”, maybe “online dictionaries is more convenient for us to study,” this new sentence is simple and more suitable.

He:

Fine.

Tian:

We delete some words and choose the other words to replace, so now the sentence is well-organized.

He:

Yes.

In excerpt 6, Tian asserted a significant degree of control over the task’s direction, resulting in a disparity in contribution between him and He. The engagement between the pair was minimal, and the allocation of effort was not equally balanced, leading to a dominant/passive interaction pattern.

During Task 1, where only the product of the writing text was assessed and the collaboration process was not under scrutiny, the primary motivations of the participants were driven by the task’s requirements and the associated grade. However, Task 2 introduced a shift in motivations due to the revelation that the teacher would assess the collaboration process.

In Task 2, Tian explained, “I focused more on collaborating with my partner because the teacher would evaluate our collaboration process.” (Stimulated interview) Reflecting on Task 2, he elaborated on this shift:

As I turned my attention to the collaboration process, I found the value in offering assistance to my partner. This change in perspective altered my perception of my partner, convincing me that aiding others could simultaneously enhance my own abilities. (Tian, Post-task interview and Reflection journal 2)

This transformation in motivation was characterized by Tian’s aspiration to learn from his partner and enhance his own skills. He too underwent a similar transition, shifting his motivation away from the grade toward a focus on collaborating with his partner, influenced by the product & process-graded assessment. He articulated:

I initially held back from active involvement due to my unfavorable experience in Task 1. However, as I recognized the potential benefits of collaboration with my partner, driven by the understanding that the teacher would assess our collaborative process, my perspective changed. I needed to appreciate Tian’s efforts, especially considering my lower English proficiency, and his significant assistance. (He, Post-task interview and Reflection journal 2)

The data illustrated that in Task 2, He altered his motivation toward collaborative writing, driven by the knowledge that their collaboration process would be evaluated. This motivational shift exerted an influence on their interaction patterns and the generation of LREs. Task 2 witnessed an increase in both the quantity and quality of LREs, with the total number reaching 102 and the correctly resolved LREs rising to 88 (86.3%). The enhanced LRE production stemmed from their altered motivation, which led to heightened engagement in the collaboration process. Throughout the product & process-graded assignment, participants engaged in more extensive discussions and collaborations, thus exhibiting a collaborative interaction pattern marked by high equality and mutuality. Excerpt 7 from the pair chats illustrated this pattern, featuring resolved LREs:

Excerpt 7: Collaborative -L- LRE- Resolved correctly

Tian:

The addition of painting class may aggravate this…this…this…

He:

May I help you?

Tian:

Yes, I don’t know how to spell xianxiang in English.

He:

Oh, I don’t know either. Don’t worry, let me google.

Tian:

Good idea.

He:

Oh, here, it should be phenomenon.

Tian:

Can you spell it please?

He:

Sure. P-H-E-N-O-M-E-N-O-N.

Tian:

Alright.

With the change of the pairs’ motivation, the generation of LREs and patterns of interaction also underwent corresponding changes. In Task 1, the teacher only assessed the collaborative work’s product, so regardless of whether it was Case 1 or Case 2, the students’ motivation for computer-based collaborative writing was centered around task requirements and grades. In Task 2, however, the teacher not only evaluated the collaborative work’s product but also assessed the collaboration process. As a result, the motivation of both pairs changed. They became motivated to learn from each other and improve themselves, and this shift in motivation significantly influenced the pairs’ computer-based collaborative writing process.

Discussion

The present study has demonstrated that students with similar L2 proficiency may hold different motivations during the product-graded collaborative writing assignment and the product & process-graded collaborative writing assignment using the Shimo technology. Our research builds upon earlier studies (e.g., Arnold et al., 2012 ; Hsu, 2020 ; Li & Zhu, 2013) that examined the effects of various factors (e.g., computer-assisted language contexts, task complexity, task form, group size) on L2 students’ participation and learning during collaborative writing. We have added process-graded assessment as a mediating factor to this equation. As a result of the product & process-graded assessment, the variations in motivations toward collaborative writing between Case 1 and Case 2 across the two tasks led to significant differences in LRE generation and interaction patterns among pairs. This finding offers empirical evidence of how product & process-assessment impacts learners’ motivations during web-based collaborative writing tasks. It also underscores the crucial role that product & process assessment plays in collaborative writing assignments.

Importantly, a majority of prior research has concentrated on exploring the correlation between motivations, pairs’ interaction patterns, and LREs in face-to-face interactions (Chen, 2020; Storch, 2013). This has led to a lack of research regarding how motivation influences learners’ collaborative writing in computer-mediated environments. As a result, this study furnishes empirical evidence that the evaluation of the collaborative writing process has the potential to influence students’ motivation within computer-assisted language contexts.

Furthermore, this case study illustrates how learners’ motivations toward computer-based collaborative writing undergo changes due to the influence of product & process assessment. Past research on students’ motivations in collaborative writing has predominantly focused on evaluating the written product (Cho, 2020; Li & Zhu, 2017; Zhai, 2021). This study advances the existing research by elucidating that learners’ motivations play a pivotal role in computer-assisted collaborative writing activities. Moreover, it underscores that alterations in motivation toward collaborative writing exert an influence on collaboration patterns and the generation of LREs.

Moreover, building upon prior research that highlights how learners’ objectives for collaborative tasks significantly influence the dynamics of collaboration (Cho, 2020) and how these goals can shape their interactions throughout the task (Chen & Yu, 2019), this study advances existing knowledge by proposing that assigning importance to the collaboration process in the assessment could have contributed to heightening learners’ awareness of the pivotal role of collective decision-making and problem-solving. Such awareness, in turn, may have facilitated learners in setting a goal of achieving a high level of collaboration

The current study is accompanied by certain limitations that warrant acknowledgment. Foremost and of paramount importance, the study was conducted within a specific instructional context—a college-level writing class. Consequently, the findings might not be readily transferable to alternate learning settings. Secondly, although the inclusion of reflection journals, post-task interviews, and stimulated interviews facilitated the researchers in tracing the shifts in motivations within the two pairs, the present study did not encompass an exploration of other potential variables that might interplay when motivation becomes intertwined with collaborative writing. Thirdly, due to the adoption of a case study approach in this research, focusing solely on two pairs, there arises a potential oversight of other pairs and certain behavioral factors that could have been generalized as influencers of motivations.

Despite the aforementioned limitations, this study yields numerous implications for classroom application. Primarily, assessment emerges as a pivotal element that commands significant attention from teachers when incorporating computer-assisted collaborative writing into their pedagogical approach. As demonstrated by the data, the impact of product & process-graded assessment on students’ motivations toward collaborative writing becomes evident in their heightened engagement within discussions and an increased perceived value of peer assistance. Hence, the assessment criteria should effectively recognize and reward both the writing process and the co-created text. This approach could potentially prove advantageous for learners, enhancing their writing outcomes while also exemplifying collaborative patterns.

Secondly, another instructional implication for educators is to conscientiously consider students’ motivations while designing collaborative writing activities within the classroom. The findings of this study underscore the significance of learners’ motivation in shaping effective teamwork dynamics. The study indicates that when students are motivated to foster respect and learning from one another, it contributes to the cultivation of collaborative interaction patterns, subsequently enhancing both the quality and quantity of LREs. Teachers are encouraged to engage in discussions with students about the potential advantages and merits of collaborative writing. Furthermore, offering evidence and examples of successful collaborative writing experiences before initiating such tasks can aid in instilling robust motivations for students to actively engage in collaborative endeavors with their partners.

The third implication arising from this study highlights the growing significance of computer technologies in L2 learning. It becomes imperative for educators to enhance their comprehension of how technology can effectively facilitate collaborative writing and language development. Teachers must cultivate familiarity with the inherent advantages of different technologies, such as providing learners with enhanced opportunities for communication and collaborative writing, unrestricted by temporal and spatial constraints. To this end, it is crucial to furnish teachers with ample and well-structured technological knowledge. By doing so, educators can harness the potential of up-to-date technology to implement L2 collaborative writing within their instructional approaches.

Conclusion

In conclusion, this study delved into the impact of process and product-based assessment on learners’ motivation and subsequently explored how these motivational shifts influenced interaction patterns and language learning opportunities. This exploration was conducted by investigating students’ motivations across two collaborative writing tasks: a product-graded collaborative writing assignment and a product & process-graded collaborative writing assignment, both situated in a computer-mediated context. While this study meticulously examined interaction patterns, as well as the quality and quantity of LREs, there remains a recommendation for future research to employ a mixed-method approach. This approach would encompass an analysis of students’ motivations coupled with a quantitative evaluation of their collaborative written texts. By doing so, a more holistic understanding could be derived, emphasizing the pivotal role that product and process-based assessment plays in reaping the potential benefits of collaborative writing within computer-mediated environments. Due to the limitation of a small sample size, future research could extend its focus by exploring motivational changes across multiple pairs and various writing tasks, resulting from product and process-based assessment.

An exciting avenue for future research involves a comprehensive examination of motivational behaviors within the context of collaborative writing, drawing from the insights provided by advanced classification systems, such as the one devised by Ahmadi et al. (2023). In their study, Ahmadi and colleagues employed a Delphi method to develop a classification system for teacher behaviors, grounded in the SDT. This classification system offers a structured approach to categorizing and understanding the motivational behaviors that influence students’ motivation. By adapting this classification system to the collaborative writing context, researchers can systematically explore the motivational behaviors that emerge during collaborative writing tasks. These behaviors may include aspects like fostering autonomy, competence, and relatedness, all of which are central to the principles of SDT.

This expansion would provide a deeper insight into the indispensable role that product and process-based assessment plays in the collaborative efforts of pairs. Moreover, further research is warranted to investigate whether product & process-graded assessment extends its influence to other closely interconnected factors in collaborative writing, such as learners’ attitudes toward collaborative writing. This exploration is especially relevant given that previous researchers often correlate motivation with attitude, particularly within collaborative writing activities.