Effective collaboration in agile software development hinges on three key elements: communication, cooperation, and coordination. While cooperation can often be ensured through clearly defined roles and levels of expertise, the intricacy of coordination—the manner in which team members interact during collaborative tasks—remains a challenge in objective evaluation. Agile methodologies, renowned for emphasizing teamwork, propose several techniques to enhance coordination, such as planning poker. However, empirical evidence documenting the impact of such techniques on team communication has been limited.
A recent study published on March 22, 2025, leverages Multimodal Analytics (MmA) to bridge this empirical gap by quantitatively analyzing verbal, paraverbal, and non-verbal cues in team interactions. Conducted at the Universidad de Valparaíso, the study evaluated 72 undergraduate students divided into 18 teams of four, engaging them in two software effort estimation tasks. One task employed an ad hoc style without predefined coordination, while the other utilized the planning poker technique. This methodology aimed to measure the impact of these coordination techniques on team communication dynamics.
The research focused on answering three critical questions regarding the effects of these techniques on speaking and attention patterns within the teams. Audiovisual recordings of the sessions were interpreted using MmA, targeting two primary metrics: speaking time and attention time. The findings revealed that although planning poker did not significantly alter total speaking or attention time, it did foster a more equitable distribution of speaking time across participants. This aspect of balanced participation is vital for effective collaboration.
Coordination is paramount in agile environments, where diverse teams must work together to achieve software development goals. As noted in the study, collaboration is not merely about bringing individuals together but requires a structured approach to communication and cooperative tactics. Agile methodologies embody this principle, promoting techniques designed to facilitate interaction among team members. Planning poker serves as an example, outlining a method for team members to express effort estimates collaboratively for various user stories, guiding the team toward efficient project execution.
Despite its theoretical benefits, the concrete benefits of planning poker in real scenarios have been difficult to quantify, particularly in terms of their influence on communication. Previous conceptualizations of collaboration have differentiated between cooperation—where participants contribute based on their roles—and coordination, the procedures governing interactions among team members. This research emphasizes the crucial role of effective communication and coordination to attain successful collaboration in agile software development.
The study meticulously measured communication dynamics using advanced MmA techniques, positioning it as a pioneering effort to objectively assess the influence of planning poker on team interactions. During the experiment, teams first worked without a defined coordination structure, allowing researchers to capture baseline communication behaviors. Upon introducing planning poker in a subsequent session, the resulting data exhibit broader participation across team members, confirming a democratized approach to speaking time within the group.
Several key findings arose from the analysis of speaking and attention time. Although planning poker did not lead to statistically significant changes in the overall attention time—where participants direct their focus during discussions—it provided a noteworthy effect on the variance of speaking time. With a p-value close to statistical significance (p = 0.560153) and a Cohen’s d magnitude indicating practical relevance, this suggests that planning poker creates opportunities for members who may have previously spoken less to contribute more significantly.
Conversely, attention time metrics revealed no substantial differences between the two conditions, implying that the distribution of attention may be shaped by factors extending beyond the coordination method employed, such as pre-existing relationships and group dynamics among participants. This observation echoes prior research findings that indicate individual factors heavily influence attention distribution during collaborative tasks.
The results underscore the importance of fostering equitable dialogue in collaborative contexts, particularly in group settings where imbalances can undermine overall team effectiveness. While planning poker proved effective in leveling the speaking field, the implications go beyond communication patterns—highlighting the need to explore further how various coordination techniques can enhance collaboration in agile software development and similar domains.
As the study concluded, the application of MmA opens new avenues for understanding communication dynamics across varied team settings, not just within agile paradigms. The implications are profound for training environments, where achieving an egalitarian speaking distribution could serve as an indicator of healthy team dynamics.
In addition to planning poker, the study hints at the potential for incorporating additional coordination techniques across different team contexts, including methodologies like daily stand-ups or sprint retrospectives. By systematically evaluating these dynamics through MmA, researchers can gain comprehensive insights on how to optimize team collaboration in ultimately developing successful software solutions.
This research was supported by Grant ANID/FONDECYT/REGULAR/1211905 and represents an important step in understanding communication dynamics within agile software teams. The findings not only contribute empirical evidence but also emphasize the utility of innovative measurement techniques like MmA to analyze subtle communicational aspects that might otherwise remain hidden, thus enhancing the overall effectiveness of collaborative efforts in software engineering.
