Sociocultural influence on engineering students’ collaborative design project: an Activity Theory perspective

2.1 Collaborative design

Research in engineering education (e.g., Carter et al. 2007; Gimenez and Thondhlana 2012; Johri and Olds 2011) has illustrated that the process of doing a collaborative task yields opportunities to induct prospective engineers into the norms and practices of the professional engineering community. For example, Carter et al. (2007), who analyzed biology students’ lab reports when conducting experiments in groups of four, found that writing reports helped students engage with the disciplinary expectations of their field. As students attempted to solve problems and write reports collaboratively, as “real scientists” do, they found themselves being trained and becoming members of the scientific community.

Notably, an effective and fair division of work among team members is central to ensuring a successful collaborative experience. In and through collaborative work, each member of the team learns to view the task as a social practice and to take advantage of peer interactions to complete the work. Sometimes one member takes on a dominant role, while her team members contribute less. At other times, the work was distributed more equitably within the team. For example, Volet and Mansfield (2006) who examined why some groups were better at teamwork, reported that the most successful group described multiple goals and focused on group performance and learning as well as teamwork. Thus, while a project can be completed in different ways, it is the challenges of negotiating group dynamics (e.g., Elabdali and Arnold 2020; Gimenez and Thondhlana 2012) that warrant further investigation.

2.2 Storch’s (2002) model of interaction

To better understand individual and collective contributions and the dynamics of collaborative tasks in group work, Storch (2002) introduced a model of dyadic interaction and discussed the implications for group work. She classified the aspects of group work in terms of equality and mutuality. According to Storch, a key part of collaboration is to develop equality and mutuality into one’s academic interaction repertoire in and through collaborative tasks. Guided by the notions of equality and mutuality, she set up four quadrants (Q1 through Q4 in our modified model depicted in Figure 1) in the analytic framework to define distinctive interaction patterns:

1. Q1: collaborative

2. Q2: expert/novice

3. Q3: dominant/passive

4. Q4: dominant/dominant

Figure 1:

Modified version of Storch’s model of interaction (2002, p. 128).

Thus, for example, when group members equally contribute to their work and help each other, their work is considered a collaborative interaction (Q1), as it exhibits a high degree of equality and mutuality in the work. On the other hand, when one group member takes a control for the group work and provides guidance and instructions to others, the group work is considered an expert/novice interaction (Q2) in that it shows less equality and high mutuality in the work.^[1]

2.3 Activity Theory

In the previous section, we looked at Storch’s (2002) writing-specific model. In this section, we briefly discuss the larger learning theory – Activity Theory – that guided our study. Activity Theory (AT) was first proposed by Vygotsky (1978) and later developed by Leont’ev (1978) to explain human cognition in relation to socially motivated activities. More recently, Engeström (1987, 1999) extended the AT idea and, correspondingly, revised his earlier representation of actions that only included a subject, object, mediating artifacts, and outcomes. In his more complex model, Engeström (1987) conceived of the human activity system as including subjects, mediating artifacts, objects (or objectives), rules, communities, division of labor, and outcomes (Figure 2).

Figure 2:

Activity system (Engeström 1987, p. 78).

In his later work (1999, 2001), and drawing on his initial theorization of AT (see Figure 2), Engeström elaborated that an activity is not a discrete individual action intended to accomplish a short-term goal; rather, it is a shared, collective process. Subjects are viewed as the agents of actions. And as agents, subjects make decisions, take control, and regulate their actions as pursuing their objects (Lantolf and Thorne 2006). The object is the goal that the subjects want to achieve. Rules refer to guidelines or regulations the subjects should abide by. The community is any of the social groups that the subjects are involved in during their participation in the activity. The mediating artifacts are the resources that the subjects utilize during the activity. Division of labor refers to how the tasks of an activity are distributed among the subjects in some way. The outcome is a solution or ideas provided by the object. Applying these AT components to ESP classrooms, engineering students are the subjects, and the projects the students want to achieve are the objects. The mediating artifacts are resources such as previous learning experiences or previous assignments that the subjects can rely on, and the rules are the guidelines or syllabus the students follow. The community consists of the groups that the engineering students are involved in. The division of labor is related to how a group of students assigns the projects among themselves.

2.4 Collaborative design projects in Activity Theory (AT)

In ESP setting, AT has proven useful for explaining learner development (Prior 2006; Russell 1997). Because AT considers human cognitive development as a socially mediated process (Lantolf 2000; Vygotsky 1978), it is a useful heuristic to examine student learning in specific educational contexts. Given the centrality of student interaction in AT, several studies that have adopted AT as a framework (e.g., Cho 2017; Fujioka 2014; Son 2022) have focused on such interaction by analyzing student collaborations. For instance, Cho (2017) used both AT and Storch’s (2002) interaction model to examine how interaction patterns were related to individual goals in collaborative writing. She found that her participants’ goals in helping group members influenced the interaction pattern, and the group’s interaction patterns affected the quality of participants’ collaboration. By using AT, Lin (2013) examined the role of agency and group work in writing activities. Lin found that the impact of group work was determined by how individual group members exercised agency and untapped every member’s potential. Simply put, AT provides a useful lens to investigate the dynamic process of student action and the development of collaborative tasks in and out of the classroom. In a more recent, AT-oriented, multi-case study, Lei and Hu (2019) examined the efforts of Chinese nursing doctoral students to publish their research. To overcome conflicts arising from a lack of familiarity with the relevant rules in their activity system, their focal students relied on mediated artifacts, such as journal articles and communities that included fellow students. Their focal students’ mentors and classmates also provided them with the necessary assistance in a socially distributed manner. In short, in keeping with the pivotal role that interaction plays in Engeström’s (1987, 2001 AT system, we join the aforementioned applied linguists who have highlighted the centrality of interaction through collaboration in the learning enterprise.

3.1 Context

The data were collected during the Fall 2015 semester from an introductory engineering design course: EGR 100. EGR 100 is a mandatory introductory course for first-year engineering students to learn the basics of engineering writing.^[2] Throughout the course, the students were expected to learn about an organized methodology to solve new and/or unfamiliar engineering design problems, and subsequently effectively communicate those solutions to scientists, engineers, or the general public. In addition to completing their requisite coursework, the students worked on limited-scope team projects and were expected to learn to collaborate effectively with team members to develop skills in communication, organization, and negotiation as well as to learn to describe the influence of individual personality traits on team dynamics. These core skills were generally discussed in large classes that were comprised of up to 300 students (see Figure 3).

Figure 3:

A screenshot of an EGR 100 course PPT slide.

The course comprised two parts: (1) lectures on values, norms, and practices of the field of engineering from a theoretical perspective; and (2) lab sessions that provided students with hands-on training and guidance to produce marketable products and write the product reports step-by-step. Ben (pseudonym), the graduate student instructor in the lab sessions, noted,

The two projects are designed to allow the students to build something, to develop …. We go through testing or we go through a design framework so they design, build, test, that’s how engineers usually work, design and that’s what we want the students to do.

(Excerpt 1, Ben, interview)

The lab sessions were meant to scaffold students step-by-step in the engineering design process by systematically incorporating tasks such as writing memos, checkpoints (i.e., multiple drafts of the final projects), building prototypes, creating products such as mobile applications, and doing poster presentations (see examples in Appendix A, B, and C). In this study, we focused on the process of the class’s final project (i.e., Project 2; see the assignment schedule for this project in Table 1). Project 2 was the collaborative project that included writing progressive reports (i.e., checkpoints) and making products. After the students submitted their project reports and products, instructors invited the teams with exemplary products to present their projects to a wider audience at an engineering college event.

3.2 Participants and data collection

Ben, the instructor quoted earlier, and two domestic U.S. freshman students (Katherine and Rebecca; all pseudonyms) participated in our study. Katherine’s major was chemical engineering, Rebecca’s major was computer science and engineering. The two student participants who were highly motivated were chosen on a voluntary basis at the beginning of the semester. Katherine and Rebecca were 18 years old at the time of our study. Different from the previous studies (e.g., Cho 2017), our participants’ reflection data constituted our primary data because we also wanted to gain insights into students’ and ESP teachers’ critical reflection.

To triangulate the data from our participants and to enhance the trustworthiness (De Costa et al. 2019) of our results, we also collected multiple secondary data sources which included: two interviews with each participant at different stages of their projects, three checkpoint assignments, the students’ class notes as well as their curriculum vitae, final projects, and poster presentations of the final project (i.e., Project 2; Table 2). From Ben, we collected data from two interviews, lab instructions and slides, writing templates, and grading rubrics for the checkpoints and the final project assignments.

3.3 Analysis

Adopting Activity Theory (AT) allowed us to conduct a multidimensional analysis of Katherine and Rebecca’s participation in Project 2 and capture the complex ecology of their collaborative work (Nishio et al. 2020). Nevertheless, we also remain cautious about applying AT to data analysis in an a priori manner.^[3] Thus, we use AT in the present study in an a posteriori manner, that is, by letting our emergent data – in particular the reflective experiences of our participants – to guide our data analytic process.

We analyzed our reflective interview data, which as mentioned constituted our primary data source, by coding, categorizing, and recoding. In the initial coding, all three authors examined the data closely to become familiar with the data. To approximate inter-rater reliability, Authors 2 and 3 had weekly meetings to examine the data together, code the data, and check the coded data together. In addition to the interview data, for data triangulation purposes, the two students’ assignments, class notes, and Ben’s interviews were also coded on the basis of AT framework. Any disagreement between the two authors’ coding results was resolved through discussion. Because our focus was to examine Katherine and Rebecca’s teamwork into a collaborative design project, the unit of analysis was their respective activity systems.

Our participants were engaged in two core activities during the collaborative design project. The first activity entailed successfully building the final product and writing up each step of that process. The second activity, which often remains unmentioned and is left to students to figure out on their own, is to achieve successful team collaboration through a series of social activities. Katherine and Rebecca’s participation in both activities gave us insight into the dynamics of collaborative design work, which included trust-building among team members, establishing leadership, allocating roles, and developing communication strategies and channels.

An AT lens was used to examine Rebecca’s and Katherine’s learning experiences in these two activity systems (i.e., the task completion activity system, Figure 4; and the collaborative work activity system, Figure 5).

Figure 4:

Task completion activity system-Activity 1.

Figure 5:

Collaborative work activity system-Activity 2.

Our examination focused on the different factors that mediated participation in the two aforementioned activities (i.e., task completion-Activity 1 and team collaboration-Activity 2). We also identified the interactions between these two activities, which in turn contributed to the different collaboration patterns and experiences of our two focal participants.

4.1 Rebecca

Rebecca, a computer engineering major, had taken computer science courses and learned coding since high school. She did not encounter any major problems in the EGR 100 course, disclosing in an interview, “I’m pretty okay at technical writing and writing in general” (interview). Similarly, when asked to describe her first-year learning experience, she described it as “simple”.

4.1.2 Object achieving as mediated by individual agency, division of labor, and community

Unfortunately, the plan to realize the teamwork objective was unsuccessful. Similar issues that led to an unsuccessful teamwork in Project 1 reemerged in Project 2 because there was also limited communication about the division of labor. In fact, Rebecca ended up taking a dominant role and doing most of the work.

Interviewer: So there was no talk about division of labor, “you are going to do this,” and “you are going to do that”?

Rebecca: No, no, I just did it.

(Excerpt 5, Rebecca, interview)

Rebecca’s interaction pattern, analyzed in relation to Storch’s (2002) collaborative interaction patterns, demonstrated both low mutuality and low equality (i.e., Quadrant 3 in Figure 6) by taking charge of the whole project and failing to engage other team members in the teamwork. Such a dominant and passive collaborative relationship (see Figure 6) between her and her team members was further explained by Rebecca in the following two excerpts.

[M]y group shows up, and sometimes they’ll help, but mostly it’s me …. It’s frustrating sometimes, sometimes I understand, that um, my lab partners are busy and other times I feel like it’s better that I do it if I understand what’s happening entirely, and I clearly was a part of every step in the process of the design, so I understand like, the background of what’s happening. I don’t feel entirely comfortable letting my teammates take control of a project that they weren’t very involved in.

(Excerpt 6, Rebecca, interview)

I’m a very driven, motivated person and I wasn’t prepared to sit back and let that kind of conversation happen where nobody does it, and somebody has to at the last minute. And um, so, I guess that’s why I took command of the project.

(Excerpt 7, Rebecca, interview)

Figure 6:

Interaction patterns of the group.

In Excerpt 6, Rebecca identified two factors: (1) different levels of motivation and involvement among team members; and (2) her unwillingness to put “her project” into the hands of the less-engaged team members. Ostensibly in contrast to her team members, she was highly motivated and engaged to work on the project. Therefore, in such a team dynamic, Rebecca was able to achieve agency by taking control of the project and letting her team members take the back seat.

Despite feeling frustrated sometimes, Rebecca was willing to and, more importantly, capable of doing most of the work by herself:

I really enjoy the project I’m working on though. Like, I knew that I was probably going to be the person doing the most work, so I made sure that we picked a project that we really enjoyed. So I don’t, I don’t feel like it’s a burden to do the work because I really enjoy it.

(Excerpt 8, Rebecca, interview)

Excerpt 8 further demonstrates how Rebecca enacted her agency to choose the project that interested her, which in turn helped keep herself engaged. Their project was to create a math tool for visually impaired middle schoolers. Because Rebecca was a computer science major, the project was less of a burden to her. One thing worth mentioning is that Rebecca’s decision with respect to the (non)division of labor and her description of the process of completing Projects 1 and 2 as not posing much difficulty also point to another issue in collaborative work, that is, team members were at different levels of disciplinary literacy and expertise. Moreover, her experience of completing Project 1 also helped Rebecca become more adept at working on Project 2. These past experiences therefore enabled her to complete the project without eliciting much help from her team members.

As they completed Project 2, an unexpected learning opportunity emerged for both Rebecca and her team members when they were selected to present their product to a larger audience at the engineering college. Due to a schedule conflict, Rebecca could not attend the poster presentation and had to allow her teammates to present the project without her. To help prepare her team members for the presentation, Rebecca assigned different parts of the project to her team members based on their familiarity with the project, stating that: “the girl who didn’t show up got the intro, she can do it, and then, people who did show up, I gave the results section and the conclusion” (interview). As a result, the presentation went quite well; Rebecca commented later that she was impressed by how smoothly things went without her. In particular, she commended one of the team members who brought her own expertise to the presentation:

One of the team members, her internship is in marketing, and she did so good at it. She was selling our product, and I was really impressed and pleased with how it turned out …

(Excerpt 9, Rebecca, interview)

This incident can be considered another form of division of labor among Rebecca’s team members, one where Rebecca’s absence (ironically) created an opportunity for her team members to learn about the project more comprehensively. Notably, her team members had to know the project well enough to deliver a good presentation. This opportunity also changed the group dynamics and created a new interaction pattern. That is, the team members’ low mutual engagement was changed to high mutual engagement, and their collaborative relationship took on an expert/novice dimension with Rebecca, the expert, agentively engaging her team members in learning about the project (see the change reflected in Figure 6).

4.2 Katherine

Katherine was a transfer student majoring in chemical engineering during the data collection period. She had spent one year at another university before transferring to the current school. According to Katherine, she had taken a chemical engineering class at her previous university, but it did not require much writing. Similar to Rebecca, Katherine was a diligent and highly motivated student. In the interview, Katherine described her experience of being a female engineering student in a male-dominant discipline:

In the engineering class, I tried to answer the questions that he [the instructor] had. I try to be involved, like stay in the front row or the front three rows …. I feel like I’m a really hard worker and you know whatever needs to be done it’s gonna get done.

(Excerpt 11, Katherine, interview)

Katherine’s strong motivation got turned into actions by enacting her agency when completing Project 2 with her team members. Katherine agentively took different roles, be it the team leader, or the coordinator which required her to accommodate her team members’ needs and keep them engaged and motivated during the Project 2; Such levels of communication and collaboration among team members are considered essential for successful teamwork.

5.1 Participants’ exercise of agency

While Katherine and Rebecca sought to complete Project 2 and engage in teamwork, their achievement of these two goals was greatly affected by how they enacted their agency. Although both identified themselves as highly motivated and driven, their enactment of agency differed in their interactions with their team members. In Rebecca’s case, based on her prior experience working with her team members, she decided to take command of the project because she perceived her team members to be not as involved or engaged in doing the project as she did. In contrast, Katherine exercised her agency, not only working on her assigned parts, but also taking on different roles at different stages to communicate with other team members to balance workload and keep others engaged and motivated, which led to a more successful collaboration experience. This observation corroborates the findings of Lin (2013), who reported that when working in a group, it is important for one to learn to control one’s agency “through interpersonal communication, through accommodating others’ different views, through regulating one’s behavior, and through modifying one’s plan and sharing work with teammates.” (p. 645).

5.2 Collaboration mediated by activity systems

Katherine’s exercise of agency also pointed to another key factor that mediated the group dynamics and shaped the learning outcome: community. Katherine identified each team member as part of the team, with a collective identity in doing different tasks and responding to challenges. Each member’s own skills, talents, and constraints were considered in the division of labor; yet, when individual members were doing their work, the community was there to help and support each other. Furthermore, in Katherine’s account, deliberate consideration was also given to the division of labor. As shown in our data, her team divided the work fairly by taking into consideration team members’ own strengths and constraints. Similarly, in their teamwork, they also created and shared some implicit rules about where, when, and how the assignments should be completed as a team. Therefore, in Katherine’s team, we observed that the components, i.e., community, rules, and division of labor, interacted to contribute to a successful teamwork outcome.

In comparison, in Rebecca’s case, she did not really see her team as a community that could build reciprocal relationships to facilitate project success. Failure to develop a sense of community can be attributed to the following reasons. First, her experience working on collaborative projects revealed some problems with team work, such as different levels of motivation and engagement, difficulty in scheduling team meetings that fit everyone’s schedule, which prompted Rebecca to take charge of the second project. Furthermore, while she was aware of the importance of collaboration, she had no prior successful collaborative experiences and no one to guide her in critically reflecting on unsuccessful collaboration experiences. This also illustrates that one’s previous experience plays an important role in mediating one’s action in the current activity system (Fujioka 2014). Rebecca may have realized that her team was able to complete the project despite limited investment from other team members. Therefore, similar to the participants in Lei and Hu (2019), Rebecca drew from her previous experiences and chose to take on most of the work to ensure a successful completion of the project.

In addition, her team’s collaborative experience was also mediated by the lack of rules to guide their collaboration. Neither Ben (the instructor) nor her team set up explicit or implicit rules to (1) scaffold how to effectively communicate personal goals, (2) decide the division of labor within them, and (3) effectively evaluate team collaboration. As we have discussed, the quality of collaboration can vary depending on the dynamics among group members, which in turn can only be enhanced when students are provided specific guidelines on how to collaborate (Gimenez and Thondhlana 2012). In addition, we maintain that the examination of reflection data – as demonstrated in our study – can illuminate the importance of engaging students in critical collaborative reflective practices.

5.3 Different interaction patterns

Informed by Storch’s (2002) four types of interaction patterns, Rebecca and Katherine’s collaborative interaction patterns can be identified as dominant/passive and collaborative interaction patterns, respectively. More importantly, we observed the interaction patterns are not static or fixed throughout the collaboration. In Rebecca’s case, the apparent constraint, i.e., her absence from the poster presentation, became an unexpected learning opportunity for her teammates. Thus, our findings aligned with those of Cho (2017) and Elabdali and Arnold (2020) who demonstrated that different interaction patterns can lead to different group dynamics and that different group dynamics can facilitate or curtail students’ learning in collaborative work. Hence, we argue that a key part of successful academic collaboration socialization is also learning to incorporate equality and mutuality (Storch 2002) into one’s academic interaction repertoire.