Social Interaction

1. Introduction

Homework support is a growing phenomenon in Sweden, as in many other countries. The term encompasses educational services provided by regular schools, private actors and non-profit organisations, both face-to-face and online. It covers both organised tutoring of students with school-provided assignments or broader preparations for tests and examinations. In this study, we explore the situated and emergent interactional organisation of video-mediated homework support. The analysis builds on video-recordings of a one-on-one tutoring session involving a tutor and an upper-secondary student working together on mathematical assignments. The participants’ interaction is mediated through an online dual-space environment that offers both a video channel and the possibility of writing and sketching on a shared virtual workspace. However, the tasks are completed by the tutee, who writes them with a pen on his notepad while looking at the maths book in front of him. Neither of these resources is visually available to the tutor. Previous studies of help-seeking in face-to-face homework support (Svahn & Melander Bowden, 2019; see also Forsberg, Hallsén, Karlsson, Melander Bowden, Mikhaylova & Svahn, accepted) have shown that tutors and tutees tend to treat the maths book and the notepad as decisive epistemic resources in establishing agreement upon what constitutes the tutee’s problem. Consequently, it is interesting to explore how the participants manage the radically different social and material contexts of video-mediated homework support.

The focus for this study is how the video-mediated environment affords interactional practices that contribute to the specific institutional aims of homework support by exploring the opportunities and constraints on actions that arise and accountably shape these same interactional practices. Drawing on ethnomethodology and multimodal interaction analysis (Goodwin, 2000; Streeck, Goodwin & LeBaron, 2011), we analyse the participants’ ways of dealing with and relating to the conditions and challenges that arise within the video-mediated context. We are particularly interested in how the online platform, comprising a shared virtual worksheet and a video channel, alongside other activity-relevant artefacts, such as maths books and notepads, figure into the organisation of actions in interplay with talk and embodied action. For this reason, the exploration of a specific case promises to contribute further knowledge about coordinated action in video-based online interaction settings more broadly, thereby demonstrating some of the ways in which the medium is procedurally consequential for the participants’ actions.

2. Ecologies of multimodal interaction in video-mediated settings

Video-mediated settings have been characterised as “fractured ecologies”, in which conduct is “fractured from the environment in which it is produced and from the environment in which it is received” (Luff, Heath, Kuzuoka, Hindmarsh & Oyama, 2003, p. 55). In other words, such settings may complicate collaboration, as the participants face the challenge of designing their conduct so that it is sensible and recognisable to co-participants who have only limited access to the environment in which the action is produced. In fact, an important condition for video-mediated settings is that co-participants can only monitor the part of their co-participants’ visual environment that is framed on-screen. This means that mutual access and possibilities for mutual visual monitoring are constrained. A large part of each participant’s immediate surroundings remain off screen and therefore unavailable to remote co-participants. Consequently, within the course of video-mediated interaction, participants have to ensure that they have congruent views of each other’s domains, or that their perspectives are interchangeable (Arminen, Licoppe & Spagnolli, 2016). One important challenge reported in previous research concerns issues of reference. For example, as it is impossible to achieve mutual eye contact with a video image, it is difficult to determine exactly what the individual framed on-screen is looking at or pointing toward (Mlynář, González-Martínez & Lalanne, 2018). Moreover, not only are the domains of the participants and the ecologies around them remote from one another, but the performative significance of gaze, gestures, and bodily movements appears weakened (Heath & Luff, 1993; Luff, Heath, Yamashita, Kuzuoka & Jirotka, 2016). Hence, participants have been shown, for instance, to undertake quite complex activities in order to show an object (Licoppe, 2017).

The concept of fractured ecologies is relevant for an understanding of the interactional specificities of video-mediated homework support in several ways. Firstly, both tutors’ and tutees’ use of gesture and other forms of embodied action has been shown to be crucial for the intelligibility of a help-request (Svahn & Melander Bowden, 2019). This means that the uneven distribution of activity-relevant artefacts, in which rich epistemic resources such as the maths book and notepad are not available to one of the participants, represents an asymmetry that circumscribes interaction. Moreover, this unequal access to objects is not solely a problem of reference, but of more fundamental import relating to the purpose of the activity. With no first-hand knowledge of the assignments or of the teachers’ planning and approach to presenting the mathematical content, the tutors have been found to rely upon the information in the maths book and notepad when designing their instructions and explanations (Forsberg et al., accepted; Svahn & Melander Bowden, 2019). This leads to a second dimension of fragmentation that is actualised in the homework support context, in that the tutors occupy an outsider’s position with respect to the activities in the tutees’ classroom. Consequently, the maths book and notepad play important roles as boundary objects that travel between home and school (Forsberg et al., accepted).

3. Configurations of mediated interaction in educational practices

Although there exists a considerable and growing volume of research concerning interaction in various computer-based learning environments (e.g. Cekaite, 2009; Gardner & Levy, 2010; Melander Bowden, 2019; Musk, 2016), research on interaction in online educational environments is still scarce, with only a small number of studies available. For instance, Duffy and Healey (2014), in a study of video-mediated music lessons, show how the side-by-side arrangement common in co-present lessons (similar to homework support) was replaced by a configuration in which the teacher was looking at the student through the video-frame, while the student was looking away from the screen to the music score. They illustrate how feedback practices were altered, as the student could not monitor the teacher’s non-verbal reactions and feedback while playing. As a consequence, the teacher had to produce extensive verbal instructions in order to achieve, for example, proper bodily positioning by a student. Similarly, Hjulstad (2016) demonstrates how teachers and students of sign language manage the fractured properties of a video-mediated classroom by using pointing for referential mapping in novel ways, to establish reliable addressing practices and next speaker selection.

Collaborative peer group work in chat-based online communities has been a focus of scientific interest, not least with regard to maths education (e.g. Çakir, Zemel & Stahl, 2009; Stahl, 2006 a & b). For example, Zemel and Koschmann (2013) analyse students’ collaborative work on maths problems in an online dual-space environment (i.e. chat and virtual whiteboard), focusing on referential practices and the sequential work of students in order to specify and arrive at agreed-upon representations of, and solutions to, mathematical problems. They conclude that an important constraint of the online environment revealed that participants could only write one at a time, thus imposing a particular kind of sequentiality on the organisation of actions. Moreover, the participants had no visual access to each other, but had to rely solely upon writing actions in the chat or on the shared workspace for the achievement of shared understanding. In another study, Balaman and Sert (2017) explore the multi-layered and laminated co-construction of knowledge in the context of second language learning online. They show how participants coordinate their interactions with their orientations to the task interface in order to enact epistemic progression, which consequently turns the interface into a layer, a semiotic field, and a screen-based resource in the course of achieving intersubjectivity and knowledge co-construction (see also Balaman, 2019; Balaman & Sert, 2012).

In sum, previous research shows some of the affordances and constraints encountered by participants interacting in online educational settings, as well as their consequences. However, hybrid environments consisting of a video channel and a shared online workspace have, to the best of our knowledge, rarely been the focus of previous research. In the following, we will demonstrate how the participants in such a setting collaborate to establish a shared understanding of the problem to be addressed, in ways that prepare the ground for instructional and explanatory trajectories. In particular, the focus will be on scrutinising the participants’ handling of the interactional and epistemic challenges that result from the uneven distribution of activity-relevant materials and artefacts.

4. Data and setting: online homework support

Our selected case is drawn from a larger study exploring homework support across private, municipal, and non-profit organisations in Sweden, where one of the documented settings was a session of online tutoring organised by a private company. The online homework support takes place in one-on-one constellations, in which a tutor and a tutee located in geographically dispersed interactional settings engage in video-mediated homework support activities on a regular basis. The female tutor in our case study was employed by the company on an hourly basis, and performed the tutoring from her apartment. She was an engineering student about to finish her university degree. The tutee was a 16-year-old male, attending his first year in upper-secondary school, and was likewise located in his home during the homework support sessions. Due to them having had weekly tutoring online encounters for slightly more than a year, the participants had a developed relationship. At the time of the video recordings, the tutor was well-acquainted with the tutee’s problem-solving abilities, as well as his general mathematical knowledge. Consequently, the homework support session constitutes a highly routinised activity, in which the participants display a clear and well-established organisation of the activity combined with an accompanying division of labour. Overall, the activity follows a distinct pattern, in which the participants work through a number of maths problems, one at a time, and where the tutee is responsible for deciding what should be worked on and in what order.

As Arminen, Licoppe and Spagnolli (2016) remark, the analysis of a given spate of mediated interaction needs to be contextualised within the ecology of material resources available in the mediated setting at stake. In the present case, the tutor and tutee use an online interface, a multimodal environment consisting of a shared virtual worksheet and a video channel, developed by the private tutoring company. The interface constitutes a resource that forms an assemblage of possibilities and constraints around which actions and practices are shaped. Both participants have equal access to the worksheet, on which they can simultaneously draw and write (cf. Zemel & Koschmann, 2013). Worksheet actions appear on the screen and are indexed by colour-coded cursors (black for tutor and yellow for tutee).

Figure 1. The online platform and user interface

When it comes to the material distribution of significant artefacts, some aspects are particularly relevant for how the activity is organised. While the tutor orients to the shared online worksheet as the main material source for writing actions, the tutee does the majority of his writing on a notepad that he uses for school, which consequently travels between home and school. This is understandable given that the homework is an assignment that is completed at home and then brought back to school. Moreover, in front of the tutee is his maths book, from which he presents the different work tasks. By contrast, the tutor does not have any access to these significant sources of information, except as mediated by the tutee.

The video recordings of the online homework support take the perspective of one of the participants (cf. Olbertz-Siitonen, 2015), here the tutor, with one camera situated next to the tutor capturing her from the side, and another focusing on the screen, recording the tutor’s view. Possible material and contextual constraints – arising from the fact that there is an asymmetric access to the material distribution of visible artefacts between the participants – are thus analysed from the perspective of the tutor. At all times, however, it is the joint participation framework and the various multimodal and semiotic resources upon which the participants draw that are analysed (Goodwin, 2000: Streeck, Goodwin & LeBaron, 2011).

In contrast to many other educational settings, during the homework support sessions, it is the tutee who is responsible for the content and for determining with what he needs help. In our view, this is a constitutive feature of all homework-support settings that we have studied (e.g. Svahn & Melander Bowden, 2019; see also Forsberg et al., accepted). A crucial prerequisite for the activity is that the tutor is not accountable for teaching different mathematical areas, or enhancing the tutee’s overall mathematical knowledge, but rather helps the tutee with specific assignments.

Concerning the overall interactional organisation of the session, larger sequences of interaction are demarcated by a focus on one maths task at a time. Each of these sequences is introduced by the tutee, who in addition to the epistemic conditions mentioned above, by virtue of being in sole possession of the maths book, is responsible for the presentation of the problem to be addressed. The establishment of the particular assignment is followed by a problem-solving sequence that consists of both the tutee’s displays of knowledge (or lack thereof), as well as instructions and evaluations produced by the tutor. As the analysis shows, the participants move on to the next (part of a) problem as soon as they have arrived at an answer that both find satisfactory.

5. Analysis

In this institutional and interactional context, task accomplishment relies upon and is highly structured by the design of the user interface. The video-mediated environment demands coordination in terms of both verbal and embodied interactions and orientations to the interface. The interface may be conceptualised as a semiotic field (Goodwin, 2000) and as a screen-based resource in the formation, ascription and co-construction of action and knowledge (Balaman & Sert, 2017). As will be shown, the participants’ organisation of the activity is also highly dependent upon the unequal distribution of activity-relevant artefacts such as the maths book, which implies a strict division and distribution of labour.

In the following section, we take a closer look at a number of interactional instances that are representative for important and characteristic aspects of the participants’ continuous organisation of the homework-support activity. In the first part of the analytic section, by exploring the initial phases of a typical (in this context) problem-solving activity, we demonstrate (i) the participants’ work to establish a shared point of reference and a ground for understanding the assignment they are working on (Extract 1). In this section, we also examine (ii) the participants’ collaborative work to produce a visual representation in connection with establishing the problem to be solved, and their distribution of labour in this process (Extract 2). In the following section, by exploring the tutor’s interpretative work regarding the details of the problem at hand, we demonstrate (iii) the interactional circumstances through which the activity is able to transition into a more instructional mode (Extract 3). In the final section of the analysis, we focus on (iv) how the participants negotiate a correct way to design a solution to a mathematical problem by drawing upon material objects and the virtual worksheet (see extracts 4a and b).

We begin by showing in detail the collaborative work of the participants in establishing a shared point of reference, and in forming a mutual ground of understanding for the current assignment. As will be demonstrated, this initial part of the problem-solving activity is highly dependent on the tutee (Edward) being able to verbally reconstruct the assignment descriptions from his maths book, as he is the only participant who has access to them and can provide the relevant information, which allows the tutor (Sarah) to interject and offer her guidance. Extract 1 begins with Edward orienting the tutor’s attention to a specific assignment in his maths book.

As has been shown, the tutor’s scope of action during the initial stage of the problem-solving activity is somewhat limited due to the lack of visual availability of the main source, namely the maths book. This means that the collaborative preparations in which the participants engage in the shared virtual workspace in order to proceed to the actual mathematical problem-solving phase are managed through a clear division of labour. This aspect will be further examined in the next extract, which takes place a short while later in the same problem-solving activity as in the previous example. When we join the interaction, Sarah is preoccupied with adding numbers to the baselines of the co-ordinate system drawn by Edward earlier, when he realises that the previously revised line is still not correctly executed, and verbalises this discovery to Sarah (lines 01–03). As will be seen, both participants immediately orient to this as a fact that has to be remedied, but choose different approaches for its execution.

In this next section, we will focus on some other conditions for the successful progression of the activity. Our primary focus is on how Edward is able to display his understanding of the assignments in a manner that is possible to interpret, and how Sarah’s ability to step into a more instructional role is dependent on an invitation from Edward. For this, we will look at a section of the video-mediated interchange in which Edward, in more direct ways, requests Sarah’s help in transforming the descriptions of an assignment into a comprehensible task.

As previously stated, the tutor has only limited access to the tutee’s working materials. Remarkably, this rarely seems to present the participants with any greater difficulties, as they have established a distribution of labour that provides a framework for mutual intelligibility with respect to who is responsible for providing particular information (and knowledge). In this section, we will proceed to examine in more detail two extracts in which the participants negotiate what constitutes a correct way of setting up a solution to a mathematical problem. These examples make visible the consequences of the participants’ unequal access to some objects, as well as the interactional work required to focus attention on relevant parts of the virtual workspace. When we join extract 4a, Edward is providing an answer to the task: “f(x) = 2x + m, determine a value for m so that f(3) = 15”.

6. Discussion

Similar to what has been shown in other online settings (e.g. Heath & Luff, 1993), the performative significance of mutual gaze and bodily gestures for upholding intersubjectivity becomes subordinate to the participants’ verbal contributions in this setting, too. In fact, although part of the screen is allocated to a video channel, the participants’ visual orientation, except when they are talking about off-task topics during the initial and concluding parts of the session (not targeted in this study), is rarely focused on each other, but predominantly towards either the virtual worksheet – or in the tutee’s case, the maths book or notepad. However, bodily gestures, in the form of referential practices, do occur and are, as we have shown, predominantly performed through pointing by use of the online cursors, e.g. in relation to highlighting specific functions of the platform (as in Extract 2) or indicating particularly relevant parts of a drawn or written representation on the worksheet (as in Extract 3). The specific challenges associated with such referential practices are therefore not related to the performing of pointing per se (as in for instance Luff et al., 2003), but in orienting each other’s visual orientation towards the screen at certain times, and in achieving and upholding joint attention. The fact that the tutee has other objects at his disposal creates circumstances in which the tutor at times has to compete for the tutee’s visual attention, and must actively work to redirect it back to the screen (as in Extract 4b, for instance).

Another interesting aspect of the analysed encounters is that they reveal a rather atypical teacher-student relation. Even though a more traditional epistemic asymmetry exists, in terms of the orientation towards the tutor as having not only epistemic access to a mathematical skill set but also epistemic primacy, the unequal access to materials and information concerning the assignments and mathematical problems also transfers the primary responsibility for directing the activity onto the tutee. In relation to the maths book being such a crucial tool, in some sense even a prerequisite for being able to perform the tutoring activity at all, it could be seen as odd that the tutor has not acquired her own copy, but is instead dependent on generating an understanding of the shared task based solely on the tutee’s verbal and visual accounts, in the form of descriptions and presentations. However, after scrutinising the interchanges in more detail, it becomes evident that this particular division of labour leads to some fundamental pedagogical consequences, forcing the tutee into a more active participant role. Another aspect of the tutee having to read out loud from his book (as in Extract 3) or verbally present his executions from his notepad (as in Extract 4a), is that the tutor, in transforming these verbal accounts into writing on the shared workspace, is afforded a possibility for clarification, in terms of visualising the aspects of the information shared by the tutee that are the most essential for solving the problem. In that sense, the tutor becomes a form of interpreter and translator, through the way she is able to model, moment by moment, how the student could most adequately use the information he is sharing, or how he can best and most efficiently approach a solution. Moreover, the analyses show that the tutor is well acquainted with the semiotic structure of the maths book, as she draws conclusions about where to find information without actually seeing the book (as in Extract 3, for instance). The participants’ shared writing on the virtual worksheet is also a product of value, not only in the sense of enabling collaborative problem-solving during the factual session, but as a continuous source of knowledge, in that the tutee can save it and return to it at a later stage, including during future problem-solving activities performed offline.

Overall, the studied setting, in terms of being a “fractured ecology” (Luff et al., 2003), in the sense of there being an unequal distribution of activity-relevant artefacts, still emerges as a smooth-functioning apparatus, free of upgraded gestural practices. In fact, contrary to what has been shown in previous studies (Luff et al., 2003; Luff et al., 2016), the participants continuously manage to collaboratively uphold the tutoring activity without any major obstacles, and although longer pauses in the verbal interchanges do occur, these are predominantly connected to other aspects of the activity (reading, writing, thinking) being in focus, rather than to communicative struggles. It is fair to assume that their respective responsiveness to each other’s online contributions is also connected to a long line of similar encounters, and their previous experience of using the current platform. Another factor relates to them being independent of having to show objects from their respective environments, as the shared worksheet enables them to make referential practices visual on-screen.

Acknowledgement

Financial support from the Swedish Research Council [2015-01754] is gratefully acknowledged.