Content area
Digital communication has become mainstream in professional as well as in educational collaboration. While it reduces the richness of non-verbal interaction, it also opened up novel, accessible, and low-cost opportunities for professional as well as personal development. This paper presents the findings of applying the framework called digitally supported coaching (DSC) to a self-organized team that partly collaborates via online video conferences. The DSC features a software tool that analyzes and visualizes communication data of several meetings and a team coaching process to use these visualizations for team development. This allows teams to discern helpful and less helpful patterns in their communication and to learn from that. While DSC has already been evaluated in multiple case studies in different contexts, this study explores the use of DSC for a self-organized team for the first time. Applying a mixed method approach, the team communication was analyzed from the subjective perspectives of the team members as well as based on the quantitative data of the meetings and the social network analysis (SNA) derived from it. Compared to hierarchical teams, the results show a much higher variety in team centralities and significant differences in the way meetings were led. Learnings by the team in this case included insights regarding their roleplay, the way they moderated meetings, and their development towards more constructive communication over time. The case indicates the potentials of DCS for self-organized teams particularly regarding their flexibility in taking on roles, tasks, and reflective practice.
Introduction
Through the rise of agile methods, self-organized teams have become a more widespread and appreciated way of collaboration (Komus 2013). The agile manifesto states “The best architectures, requirements, and designs emerge from self-organizing teams” (https://agilemanifesto.org/). Self-organization implies a new way of leadership, as there is no hierarchical leader who tells others how work is distributed, how solutions for problems are found, or how decisions are made.
A special form of self-organizing teams occurs when collaboration is partly or entirely conducted via the Internet. We refer to such teams as self-organized online teams (SOOTs). Virtual collaboration is particularly challenging for SOOTs, as self-organized collaboration in an agile understanding was originally associated with onsite settings. Only then does the density of individual interaction occur that is at the core of the agile principles. The agile manifesto rates “individuals and interactions over process and tools” and further states that “the most efficient and effective method of conveying information to and within a development team is face-to-face conversation” (https://agilemanifesto.org/). By 2001, when the agile manifesto was created, face-to-face conversation typically referred to onsite communication. Therefore, collaboration in an online setting can be particularly challenging for teams that self-organize. In light of this challenge, one may ask how self-organization can be realized in virtual communication settings.
The purpose of this case study is to initially assess the potential of analyzing and visualizing the data generated by SOOTs in online meetings to support the learning process regarding their team collaboration. To this avail, a framework called digitally supported coaching (DSC) is used, an approach that has previously been evaluated in business and learning contexts (Spielhofer and Motschnig 2023; Spielhofer and Haselberger 2023). This framework uses the audio tracks generated in online video calls to visualize their interaction in so-called communication reports. The communication report is then reflected by the team in a coaching session moderated by a professional coach.
A communication report is composed of:
The “airtime”: the percentage of speaking time each speaker had overall
The sequence of speaking events throughout a meeting
The number of times each specific speaker spoke after another speaker (interaction)
Figure 1 shows the communication report for one meeting. The pie chart on the left-hand side depicts the “airtime,” i.e., the share of the overall speaking time, each participant had. The communication sequence in the middle uses one bubble for every point in time (x-axis) a particular participant (y-axis) spoke. The size of the bubbles depicts the speaking time of the respective participant. Thus, the communication sequence reflects the “rhythm” of the conversation, whether, for instance, it is a dense staccato-like interaction of short contributions or a series of long monologues. It can be interesting to reflect on this “conversation rhythm” to assess how well this communication pattern is able to stimulate creativity (Pentland 2014). Moreover, for moderators, the “conversation rhythm” can serve to reflect their communication behavior: is it for instance a series of short interventions stimulating or regulating dialogue of others, or is it rather a mix of moderation and own (longer) contributions on the content level? The latter would result in far larger bubbles than the former on the communication sequence chart. The interaction diagram on the right-hand side shows how many times one person spoke after another person. It can be argued that in a non-pathological conversation this represents interactions between persons (Spielhofer and Motschnig 2023). It is also the basis to conduct a social network analysis (SNA) that would provide additional relevant information. The interaction diagram reflects the communication structure showing aspects such as the strength of the relations between team members or the reciprocity of relationships. It further provides insight on individual team positions by providing metrics to calculate the centrality of team members.
Fig. 1 [Images not available. See PDF.]
Communication report of one meeting
These reports are designed to help teams to discern their communication patterns in online calls and ask themselves which of these patterns are helpful and what patterns they might want to change. A communication pattern in this understanding is a stochastic concept based on Watzlawick’s communication theory (Watzlawick et al. 2011) referring to recurring interactions among human beings that encompass, among other interactions, the way teams make decisions, integrate new team members, or deal with resistance and criticism. To recognize recurring communication structures, reports of several meetings, typically three to five, are inspected and reflected in a DSC session.
Related work
The idea of learning among self-organizing teams and how to foster it has already been examined in the 1970s (Harri-Augstein and Thomas 2008; Thomas and Harri‐Augstein, 2001) and has been further elaborated since (Coombs and Smith 1998). Various practices for coaching self-organized teams were established in the agile community over the past decades (Adkins 2010; Kotrba and Miarka 2019). The so-called Container-Differences-Exchange model (CDE) provides a theoretical framework for understanding the dynamics of self-organized teams (Eoyang 2012). It provides a fundament to understand how self-organized teams can be led and coached from a systemic point of view.
An approach to “automate the identification of salient contributions and patterns in student e-discussions” in self-organized discussions among students (Mclaren et al. 2010) used a machine learning model to classify and tag the content of discussion. This served (among other things) to identify particularly salient contributions by students. Communication patterns in asynchronous online communication have been researched (Massey 2003). Furthermore, Muszyńska and Swacha (2021) analyzed structural aspects of online communication focusing on the reciprocity of interaction.
Several commercial solutions exist that automatically analyze video online communication to provide insights for the meeting participants and their leaders: Microsoft’s online meeting solution MS Teams for instance features a Speaker Coach (https://teams.microsoft.com/). This solution mainly uses content analysis to provide recommendations for meeting participants. Recommendations could for instance include to use a more inclusive language. Cogito (https://cogitocorp.com), a spinoff of the MIT MediaLab, states on its website that it combines the analysis of various aspects of communication to detect emotional intelligence and other relevant factors that determine meeting success and employee satisfaction. This is used to provide recommendations to call center agents to help them in their communication with customers. Due to the confidential nature of these proprietary technologies, no publications on their design could be found in our research.
Using SNA and qualitative methods in mixed methods to understand group dynamics is less widely explored. SNA was used in connection with critical discourse to explore the social dynamics in a classroom (Huber and Froehlich 2020). Chinese organizations were analyzed using a mixed method approach, in which an SNA was derived from the formal organization and compared with the informal organization derived from personal interviews (Fu et al. 2017).
Analyzing interactions with the help of continuously gathered quantitative data to gain social insights is the aim of the so-called social physics. Its founder, Alex Pentland, came to the conclusion that “in studies of more than two dozens of organizations I have found that interaction patterns typically account for almost half of all the performance variation between high- and low-performing groups” (Pentland 2014). Two independent large research projects found that the variance of the so-called conversational turn-taking, the indicator of how many times someone speaks within a meeting, is inversely proportional to group intelligence (Woolley et al. 2010; Duhigg 2016). Group intelligence in both research projects was measured in terms of team performance at certain given tasks like estimation, decision-making, or finding solutions to problems.
Several ways of visualizing online communication have been developed and examined in the past. One research studied the effect of visualizing individual turn-taking and emotional reactions (Samrose et al. 2018). The effect of visualizing the sequence of communication has already been analyzed in 2007 using the so-called FlashMeetingTM videoconferencing system. This application features, among other things, a “linear visualization of the event, in which the horizontal bars represent one turn of the named user” (Scott et al. 2007). Another tool, called MeetCues, provides interactive features for the users, for instance for “tagging points during the meeting with a like/clarify reaction” (Aseniero et al. 2020). It uses this input to create a visualization of the meeting communication as an interactive timeline that enables users to see how the meeting’s mood changes over time. In this interactive approach, the purpose of the tool is to change the course of the meeting as it happens. By allowing the users to give input that is then visualized, the users are having a reciprocal interaction with one another and the tool during their online meetings.
The theoretical foundation of DSC and a multiple case study showing the potential of applying it in a business context have been reported (Spielhofer and Motschnig 2023). Another case study examined the potential benefits of applying DSC in an educational context. Here two university student groups engaged in controversial discussions in online video conferences. The aim was to find out what additional learnings these two student groups could gain through DSC after having reflected their communication in a moderated retrospective featuring external observers beforehand (Spielhofer and Haselberger 2023). In both studies, relevant learnings occurred both on an individual level, such as communication behavior or level of participation, and a structural level, such as the balance of conversational turn-taking or the influence of a moderator on the whole group.
Research questions
While DSC has been studied in various contexts, its value for SOOTs has not yet been investigated. The novelty of this study lies in understanding the crucial social dynamics in SOOTs and how they can be further developed with the help of DSC. Due to the large degree of organizational flexibility and freedom in SOOTs, it is assumed that DSC has the potential to offer useful, easy-to-follow hints for improving communication in SOOTs, in particular. Furthermore, the communication report can be used to compare differences in the communication structures of hierarchical and non-hierarchical teams. Therefore, this research addresses the following questions:
What can SOOTs learn through the reflection of their communication structure during a DSC?
What can SOOTs learn about their roles within a team with the help of DSC?
What can SOOTs learn about their moderation with the help of DSC?
Methods
This research is designed as a case study as described by Yin (2018). The case is bound by encompassing a specific meeting series of a specific team: the regular online coordination meetings a core team of a non-profit organization had. One meeting in 2021 and three meetings in 2023 recorded within 8 weeks were made available by the team for this research. Each of these coordination meetings was set to 1 h. The meeting invitees changed between 2021 and 2023, as the participants of the core team changed. Out of the original eight team members (6 female, 2 male) in 2021, five remained in the core team (4 females, 1 male) while three left to serve in other parts of the organization.
This team was selected, as the first author had previously worked with it as a team coach and the team was open to try out a new method. The research started after all team members were informed about the research and had given their written consent for their anonymized data to be used.
The DSC team coaching session was done in two parts: First, the team was shown the visualizations of the meetings in year 2023, to enable the reflection of their current teamwork. In a second step, the visualizations of the meeting in year 2021 were shown, to allow the team members to discern the development of the communication over a longer span of time and in different team compositions.
A mixed method approach was employed to increase the construct validity of the study. Two sources of evidence were used: quantitative data derived from the analysis of the audio data and qualitative data derived from questionnaires the team members answered at the end of the DSC.
The following questions were asked in the questionnaire:
What did you particularly notice on the communication report?
What did you notice about your roles?
What did you notice about the moderation of your meetings?
What changes have you noticed in your interactions in the online meetings between 2021 and 2023?
In your opinion, what has contributed to constructive communication in your online meetings?
What could lead to even more constructive collaboration?
A reflexive thematic analysis of the answers to the questionnaire, as outlined by Braun and Clarke (Braun and Clarke 2022), was conducted to identify common themes in their findings. Braun and Clarke postulate that the finding of good quality codes and themes “result from dual processes of: (a) immersion and depth of engagement; and (b) giving the developing analysis some distance” (Braun and Clarke 2022). To achieve both immersion and distance seems as paradoxical as accurate. It reflects the need for closeness of the researcher to the social system to discern its core patterns while maintaining the necessary distance to understand it from a new perspective. Embracing this idea, the first author of this paper has coached the team prior to this research and listened to the audio tracks of the meetings to immerge into the group dynamics. He then took some time before elaborating the codes and themes to regain a more distanced perspective. The underlying codes were formed inductively as recommended by Mayring (Mayring 2022). The initial themes were derived by clustering these codes. The result was then reviewed and reflected with a fellow researcher. In this reflection, some underlying assumptions of the researcher about what was relevant in the communication became overt and were discussed. Based on this reflection, the themes were refined, and the final theme names were chosen.
The qualitative data reflect the subjective insights of the participants as expressed in their answers to the questionnaire. The quantitative data reflect the actual “airtimes,” communication sequences, and interactions of the team members in the recorded meetings. It is composed of two steps: First, the audio tracks are used to generate the so-called diarization lists for all meeting participants. One such diarization list holds the point in time and length every time the respective meeting participant spoke. The combined diarization lists of all meeting participants are then used to visualize the team communication. The social network inferred through the interaction was then analyzed and the weight of the relations and the centralities of team members in the meetings were calculated.
The qualitative and quantitative data are then compared to identify commonalities in both views on communication.
Results
Part 1, the meetings in 2023
The themes derived from the qualitative analysis are listed in Table 1. Each of the column headlines represents one question asked in the questionnaire. Underneath, the themes deducted from the answers are listed. Following RQ1, insights of the team regarding their communication structure comprised the distribution of airtime in the meetings and how this varied between meetings. Changes in the balance of speaking times between the 2021 and 2023 meetings were also noticed. Furthermore, the variance in interactions when looking at all meetings was mentioned by the team: “High variance of speaking times and interactions across meetings.”
Table 1. Results of the qualitative analysis
What did you particularly notice on the communication report? | What did you notice about your roles? | What did you notice about the moderation of your meetings? | What changes have you noticed in your interactions in the online meetings between 2021 and 2023? | In your opinion, what has contributed to constructive communication in your online meetings? | What could lead to even more constructive collaboration? |
|---|---|---|---|---|---|
P1 and P3 continuously take large shares of the meeting time. The amount of speaking time also reflects the dominance in the project | P1 talks a lot | By whom and how meetings are moderated changes | Much more balanced speaking times | Respectful listening | Meet at times when people can focus on the meeting |
High variance of speaking times and interactions across meetings | P3 talks a lot | Moderation role is not clearly defined | More constructive communication | Give the floor to hot topics | Rotate the moderation |
Moderation role changes as reflected in the communication report | Only those who want to contribute are at the meeting | Smaller group | Define moderation more clearly | ||
P2 has individual competence, otherwise keeps low profile | P2 automatically took on the role of moderator in 2021 but not in 2023 | Better understanding of roles | |||
P5 speaks when she can, contributions may be on the meta-level |
The airtime diagrams of the three meetings in 2023 are illustrated in Fig. 2. Each of the pie charts shows the relative speaking times of meeting participants 1 to 5 for one meeting. While the airtime of all other participants varied significantly, P1 and P3 together constantly took half of the airtime in all three meetings. Table 2 shows the airtime of all team members and their mean and variance across the recorded meetings. The sum of the airtimes of P1 and P3 ranged in the small span of 48.61 to 50.75% in all three meetings. The variance in airtime of P1 and P3 was only at 0.09 and 0.47, respectively, while the variance of the others was significantly higher, ranging from 10.81 to 56.01. In other words, P1 and P3 always had a constant share of the meeting airtime, even though the topics of the meetings were different. This was partly acknowledged in the DSC: “P1 and P3 continuously take large share of the meeting time,” as the qualitative analysis shows (Table 1). One theme provided an interpretation of this data: “The amount of speaking time also reflects the dominance in the project.” This suggests that, in reflecting on the communication report, the team recognized aspects of their role as asked for in RQ1.1. In particular, they seemed to be interested in the speaking time of team members and whether they remained in balance with each other.
Fig. 2 [Images not available. See PDF.]
The airtime diagrams of the three meetings in 2023
Table 2. Airtimes (in percent) and their variances of the meetings in 2023
P1 | P2 | P3 | P4 | P5 | |
|---|---|---|---|---|---|
Meeting 1 | 25 | 18.04 | 25.75 | 23.44 | 7.77 |
Meeting 2 | 24.96 | 9.07 | 25.69 | 23.16 | 17.13 |
Meeting 3 | 24.34 | 27.4 | 24.27 | 16.33 | 7.66 |
Mean | 24.77 | 18.17 | 25.24 | 20.98 | 10.85 |
Variance | 0.09 | 56.01 | 0.47 | 10.81 | 19.70 |
The interaction diagrams of the three meetings in 2023 are shown in Fig. 3. The structure varies with each meeting, and there is no single person serving as the central communication hub in every meeting. Instead, there is “high variance of speaking times and interactions across meetings” as the participants reflected (see Table 1). In the view of RQ1.1, this indicates that the team reflection concluded that there is no person that continuously dominates the discussion in the team.
Fig. 3 [Images not available. See PDF.]
The interaction diagrams of the meetings in 2023
Figure 4 shows that the centralities of the participants varied from meeting to meeting. No single person continuously held the position of the most central person in the three meetings according to this SNA metric. Instead, the position of highest centrality within the team was one time assumed by person 3, then person 5, and finally person 2. This might correspond with the observations of the team that “moderation role changes as reflected in communication report” and “by whom and how meetings are moderated changes” (see Table 1), showing the teams’ findings about their team moderation in the DSC, as was the focus of RQ1.2. Another aspect in the social network of the team becomes overt when looking at the interaction diagrams of the three meetings: Everyone interacts with everyone else sooner or later, except for participants 1 and 4, who hardly interacted in any of these meetings (the graphs do not show interactions below 14 to keep them easy to read). However, this aspect was not mentioned by the team in the questionnaire after the DSC.
Fig. 4 [Images not available. See PDF.]
The centralities of the five participants (from left to right) in the three meetings
Part 2, comparison of 2021 and 2023
There was only data from one meeting in 2021 available for this research. However, as it is remarkably different from the 2023 meetings, the communication diagrams from this meeting were shown to the team so that they could draw their conclusions. The airtime was even less balanced than in the 2023 meetings, as Fig. 5 shows. There was one speaker who spoke 43% of the time, whereby two others were below 2%. This was noted by the team members in their reflection. When asked about the changes over time, they noted “much more balanced speaking times” in 2023 compared to 2021 as Table 1 shows.
Fig. 5 [Images not available. See PDF.]
The airtime report of the meeting in 2021
This is also reflected by the interaction diagrams (see Fig. 6). The meeting in 2021 shows participant 2 as the central communication hub through which most of the communication ran. On the other hand, the two participants with low airtime had hardly any interactions with anybody in the meeting.
Fig. 6 [Images not available. See PDF.]
The interaction diagrams of the meeting in 2021 and the three meetings in 2023
Discussion
Regarding RQ1.1 “What can SOOTs learn about their roles within a team with the help of DSC,” this case shows significant differences compared to the cases of hierarchically led teams examined in previous research. Looking at the quantitative data for indicators of the team positions, the pattern of the team seems to be that there is no leadership pattern: the centralities of the team members vary significantly from meeting to meeting, as Fig. 4 clearly shows. In this respect, this case of a self-organized team differs fundamentally from the three cases examined in hierarchical organizations (Spielhofer and Motschnig 2023). There, all hierarchically led teams continuously showed the highest centralities for leaders and moderators throughout all meetings. The various interaction diagrams of hierarchical teams also showed a pattern: the same persons serving as a kind of communication hub in the center of the team in all meetings. No such recurring structure can be observed in the interaction diagrams of this self-organized team. If we consider centrality in social network analysis as a measure of the team position, we may come to the hypothesis that positions in this self-organized team are not as fixed as in hierarchically led teams.
However, some continuity in the roleplay was identified in the DSC. The team noticed the remarkably constant airtimes of participants 1 and 3 across the meetings in 2023. As one theme stated, “P1 and P3 continuously take large share of meeting time” (see Table 1). This is coherent with the quantitative data, which shows that P1 and P3 always have quite precisely 25% of the airtime in all 2023 meetings (see Table 2). This is remarkable given that the three meetings had different agendas and it is unlikely that both P1 and P3 had the same amount to contribute regardless of the topics. This gives rise to seek hypotheses for the balance between their airtimes on a social level, for instance in group dynamics. We could for instance assume that the “amount of speaking time also reflects the dominance in the project,” as one theme in the team reflection did (see Table 1). Following this hypothesis, Raul Schindler’s team ranking model (Schindler 1973) can be used to understand the balance of airtime of two players as a leadership dynamic. Accordingly, P1 and P3 would hold or strive for an alpha role, each of them seeking to balance the participation of the other in order to achieve equal presence in the team. Another hypothesis could be derived from a systemic viewpoint, considering P1 and P3 as representatives of different functions or value areas within the team. The belief polarities model for instance states that there are three different value areas in any team: Bhakti, representing compassion, trust, and social competence; Jnana, representing knowledge, insight, and search for truth, and Karma representing order, structure, and the ability to act (Sparrer 2009). As the Hindu names of the three value domains indicate, this model has emerged from a socio-cultural view of organizations and teams, postulating that this triad of value domains occurs in most societies and, consequently, in the social systems within those societies. Following this model, competition and tension between team members could reflect a tension of the different value domains they represent. In their reflection in the DSC, the team in this case study further noticed the strong presence of P1 and P3 and drew assumptions on their position in the team as well as differentiated assessments of the roles of P2 and P5.
Regarding RQ1.2, “what SOOTs can learn about their moderation with the help of DSC,” the team noticed that both “by whom and how meetings are moderated changes” (see Table 1). As a conclusion of the DSC, team members would find it helpful to define the moderation more clearly. A development of the team moderation can also be observed over time. The team members noted that “P2 automatically took on role of moderator in 2021 but not in 2023” (see Table 1). As the airtime and interaction diagrams show, the way moderation was conducted in all three meetings was very different in 2023 compared to 2021. They were moderated in a way that allowed a more balanced involvement of all participants. This stark difference in the communication balance between 2021 and 2023 was also expressed in other answers to the questionnaire, finding “much more balanced speaking times” in 2023 (see Table 1). This is corroborated by the difference in airtimes, as Figs. 2 and 5 show, and the interactions as illustrated in Fig. 6. In the 2021 meeting, two participants were hardly actively participating in the discussion and in the decision-making. This is remarkable, given that this is the core team of an organization, which has the purpose to make day-to-day decisions to keep the organization going. Reflecting this, the team stated that in 2023 “only those who want to contribute are at the meeting.” This is consistent with the interaction diagrams, as the core team reduced to five participants had a much more even level of airtime and interaction in 2023 (see Fig. 6). This conclusion should be treated with caution though, as it is drawn from data of only one meeting in 2021.
Regarding RQ1 “What can SOOTs learn through the reflection of their communication structure during a DSC?”, it could be summarized, that in this case study, learnings occurred regarding the roles within the team, the meeting moderation, and the changes of communication over time.
Limitations
A limitation of this study is that it only covered one case and can therefore only be seen as a basis to form hypotheses regarding the potential outcome of applying the DSC to SOOTs. The authors also need to share that the whole research took place in a western-dominated culture, and we would be excited about broadening our horizon by exploring a culturally more diverse setting and its effects.
Conclusions
This case shows the potential for self-organized teams (SOOTs) to identify and learn about certain aspects in their communication under the right circumstances. One such aspect is team positions and their interaction, which can be well illustrated by the interaction diagram and further supported by quantitative indicators of the SNA. The DSC also proved a fruitful ground to reflect the meeting moderation. The communication report enabled the team to discern how much the moderation role varied, both by whom and how it was done. Examples for improvement included the following: to “define moderation more clearly” and to “rotate the moderation” (see Table 1). These themes formed the responses to the question of what could lead to more constructive collaboration.
When looking at the change of communication over time, the DSC and the communication report can serve as confirmation of the development course a team has chosen. In this case, the team found “much more balanced speaking times” and “more constructive communication” in 2023 compared to 2021 (see Fig. 1). In a team coaching setting, this can be used as positive reinforcement to identify and strengthen the resources that led to the change. It can also be used to identify and curb undesirable tendencies if the report shows otherwise.
On reflection, the visual representation of the communication structure of SOOTs has the added value of explaining a snapshot of the self-organized, real constellation of a team at certain phases of a project. Unlike hierarchically organized teams with pre-defined roles, SOOTs members can use the visualizations to determine their desired structure, so that members can optimally contribute to the goals they have set for themselves in each phase of their project. Conversely, DSC is seen as a framework that is particularly well suited to SOOTs.
Summarizing, this case shows that the DSC can be a valuable, lightweight tool for any self-organized team that partly or wholly communicates through video conferences. Embedded in an appropriately designed team coaching process, the communication report can complement the internal view of team members with a structural view of their communication. A combination that can lead to new insights—insights that can make a difference to the team, as this case shows.
In future work, the current approach could be applied to culturally more diverse team settings. It could also be extended to learn about other aspects of teamwork. One possibility would be to extend the DSC to include contextual information about the cultural backgrounds and language skills of team members, thus supporting studies that aim to determine how people from different backgrounds are integrated into teams. If this approach is applied to many teams within an organization, it could be used to learn about and improve the inclusiveness of the organizational culture. Another research path would be to examine the dynamics of cross-functional teams. The social network analysis part of the DSC could for instance help to find out more about the relationships between software developers and testers in a cross-functional software development team.
Acknowledgements
The authors wish to thank the University of Vienna for the opportunity to conduct this research. Special thanks go to the participating team members for their time and openness to support this research.
Funding
Open access funding provided by University of Vienna. This research received no external funding.
Data Availability
Data of the recorded calls have not been made publicly available due to privacy restrictions.
Declarations
Informed consent
Informed consent was obtained from all subjects involved in the study.
Conflict of interest
The authors declare no competing interests.
Institutional review board statement
This research was conducted in accordance with the ethical guidelines of the University of Vienna and the Vienna Manifesto on Digital Humanism.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
Adkins L (2010) Coaching agile teams: a companion for ScrumMasters, agile coaches, and project managers in transition. Addison-Wesley, Boston
Aseniero BA, Constantinides M, Joglekar S, Zhou K, Quercia D (2020) 2020 IEEE Visualization Conference - short papers: VIS 2020 : proceedings : virtual event, 25–30 October 2020. IEEE VGTC : IEEE Computer Society, Conference Publishing Services
Braun V, Clarke V (2022) Thematic analysis: A practical guide (1st edn.). SAGE Publications Ltd, Thousand Oaks
Coombs, SJ; Smith, ID. Designing a self-organized conversational learning environment. Educ Technol; 1998; 38,
Duhigg, C. What Google learned from its quest to build the perfect team. N Y times Mag; 2016; 26,
Eoyang GH (2012) Human systems dynamics: toward a computational model. 634–637. https://doi.org/10.1063/1.4756214
Fu X, Jar-Der L, Boos M (2017) Social network analysis: interdisciplinary approaches and case studies. CRC Press.https://doi.org/10.1201/9781315369594
Harri-Augstein, S; Thomas, L. A conversational framework for self-organised learning: modelling personal meaning for effective action in the world. Handbook of Conversation Design for Instructional Applications; 2008; 2008, pp. 308-342. [DOI: https://dx.doi.org/10.4018/978-1-59904-597-9.ch019]
Huber M, Froehlich DE (2020) Analyzing group interactions. Routledge, London
Komus, A. Agile Methoden in der Praxis—Studie zur Anwendung und Zufriedenheit. HMD Praxis der Wirtschaftsinformatik; 2013; 50, 91. [DOI: https://dx.doi.org/10.1007/BF03340799]
Kotrba V, Miarka, R (2019) Agile Teams lösungsfokussiert coachen. dpunkt verlag, Heidelberg
Massey, AP. Because time matters: temporal coordination in global virtual project teams. J Manag Inf Syst; 2003; 19,
Mayring P (2022) Qualitative content analysis: a step-by-step guide (1st edition.). SAGE Publications Ltd. https://app.talis.com/textbooks/9781529766738
Mclaren, BM; Scheuer, O; Mikšátko, J. Supporting collaborative learning and e-discussions using artificial intelligence techniques. Int J Artif Intell Educ; 2010; 20,
Muszyńska K, Swacha J (2021) Periodic indicators for graphical monitoring of team communication in information system development projects. 29th International Conference on Information Systems Development, Valencia, Spain
Pentland, A. Social physics: how good ideas spread-the lessons from a new science; 2014; Penguin:
Samrose, S; Zhao, R; White, J; Li, V; Nova, L; Lu, Y; Ali, MR; Hoque, ME. Coco: collaboration coach for understanding team dynamics during video conferencing. Proc ACM Interact Mobile Wearable Ubiquitous Technol; 2018; 1,
Schindler R (1973) Das Verhältnis von Soziometrie und Rangordnungsdynamik. Heigl-Evers, A.(Hg.): Gruppendynamik. Vandenhoeck & Ruprecht, Göttingen
Scott, P; Tomadaki, E; Quick, K. The shape of online meetings. Int J Technol Knowl Soc; 2007; 3,
Sparrer I (2009) Systemische strukturaufstellungen: theorie und Praxis (2., überarb. Aufl). Carl-Auer-Verl
Spielhofer T, Haselberger D (2023) Using digitally supported communication analysis to facilitate constructive discourse in educational Courses. In T. Bastiaens (Ed.), EdMedia + Innovate Learning 2023 (pp. 44–53). Association for the Advancement of Computing in Education (AACE). https://www.learntechlib.org/p/222481
Spielhofer, T; Motschnig, R. Developing teams by visualizing their communication structures in online meetings. Multimodal Technol Interact; 2023; 7,
Thomas, L; Harri-Augstein, S. Conversational science and advanced learning technologies (ALT): tools for conversational pedagogy. Kybernetes; 2001; 30,
Watzlawick P, Bavelas JB, Jackson DD (2011) Pragmatics of human communication: a study of interactional patterns, pathologies and paradoxes. W. W. Norton. https://books.google.at/books?id=YcBUAgAAQBAJ
Woolley, AW; Chabris, CF; Pentland, A; Hashmi, N; Malone, TW. Evidence for a collective intelligence factor in the performance of human groups. Science; 2010; 330,
Yin RK (2018) Case study research and applications: design and methods, Sixth edn. SAGE, Newcastle upon Tyne
Copyright Springer Nature B.V. Dec 2024