Search and Compare course prices, ratings, and reviews. Over +350 Design and Technology courses in one place!

How Personality and Communication Patterns Affect Online ad-hoc Teams Under Pressure

marvin meyer SYTO3xs06fU unsplash1 | How Personality and Communication Patterns Affect Online ad-hoc Teams Under Pressure | Coletividad
Photo by Marvin Meyer on Unsplash

The Task

To study participant interactions in ad-hoc teams of strangers under pressure, we turn to crowdsourcing, and a custom-made task. Our task is inspired by the “Keep Talking Nobody Explodes” (Knuth, 2021) puzzle video game. Participants work in dyads, and their common mission is to defuse a bomb that is placed within a maze, by combining information that is unique to each one of them. One participant is assigned the role of the “Defuser”: they can “walk” inside the maze toward the bomb and defuse it, but they do not know where the maze walls are. The other participant is assigned the role of the “Lead Expert”: they have the map of the maze but they cannot walk in it. The Defuser and the Lead Expert must exchange information and actions, to defuse the bomb within a limited amount of time. The task has been designed to have the same critical characteristics as actual emergency response tasks, namely a high-demanding environment, enforced role division, performance pressure and stress.

Teams in Classical High-Demand, Time-Pressing Settings

Critical, time-bounded, and high-stress tasks, like incident response, have often been solved by teams that are cohesive, adaptable, and prepared. Although a fair share of the literature has explored the effect of personality on various other types of teams and tasks, little is known about how it contributes to teamwork when teams of strangers have to cooperate ad-hoc, fast, and efficiently. This study explores the dynamics between 120 crowd participants paired into 60 virtual dyads and their collaboration outcome during the execution of a high-pressure, time-bound task. Results show that the personality trait of Openness to experience may impact team performance with teams with higher minimum levels of Openness more likely to defuse the bomb on time. An analysis of communication patterns suggests that winners made more use of action and response statements. The team role was linked to the individual’s preference of certain communication patterns and related to their perception of the collaboration quality. Highly agreeable individuals seemed to cope better with losing, and individuals in teams heterogeneous in Conscientiousness seemed to feel better about collaboration quality. Our results also suggest there may be some impact of gender on performance. As this study was exploratory in nature, follow-on studies are needed to confirm these results. We discuss how these findings can help the development of AI systems to aid the formation and support of crowdsourced remote emergency teams.

High-Demanding Environment

Instances of crisis constitute a large part of what emergency teams have to deal with and radically define their functional and structural properties. Demanding environments have critical requirements with tangible consequences for poor performance (e.g., accidents, errors, stress). By portraying the element of urgency in the form of a virtual bomb and increased time pressure (Bell et al., 2018) we focus on a single objective—reaching the bomb on time—and deliver the results of a study task that is critically cooperative and built for productive communication. In our setting, virtual crowd teams must deliver innovative solutions and deliver them quickly. The typical environmental constraints of high-demanding tasks (time, urgency, risks) command for independent, stable, role-defined teams sharing mutual trust, values, and focus. As we reduce and inter-mediate communication through digital means, we impose an even further reliance on mutual objectives, well-defined roles and obligations, effective communication, and commitment.

Enforced Role Division

During cases of emergency, each team member has a distinct and specific role to play (Baldwin and Woods, 1994), which is typically a-priori and externally defined. Emergency and periods of crisis often create the need for established protocols of interaction respective to each part (Harrison and Connors, 1984). Although role division is typically fixed for these response units (e.g., medical, logistic, security, public relations, etc.), it must nonetheless be adaptable when facing unpredictable outcomes. By assigning strangers to pre-defined roles, we replicate a scenario where team roles are agreed upon yet flexible and interposed. Through well-defined roles and responsibilities, we evaluate the matching capabilities of crowd workers and investigate what are the constituents that fundamentally determine the execution of role-based virtual teamwork emergency response.

Performance Pressure and Stress

nik shuliahin BuNWp1bL0nc unsplash1 | How Personality and Communication Patterns Affect Online ad-hoc Teams Under Pressure | Coletividad

Prior work has shown that users involved in games such as the crowdsourcing task exhibit various forms of stress (Sabo and Rajčáni, 2017) and heightened emotional states (Hart et al., 2018). These teams are more susceptible to allostatic load, i.e., the process of “wear and tear” experienced by team players facing stressful conditions (Davaslioglu et al., 2019). Regarding the definition of stress, there are two kinds of stressful conditions and stressors (Ma et al., 2021). One definition follows the general assumption that a stressor (the triggering factor) negatively affects the person by degrading performance; the other sees stress as a challenge that improves performance and individual gains (Zhang and Lu, 2009). In this research, we stripped the task from several elements of the original video game with the intent to transverse from multiple sources of hindering stressors [that increase environmental demands and exceed the available resources (Salas et al., 1996; Gardner, 2012)] to a unique challenge to inspire and motivate collaborators. Finally, virtual teams experience stress differently than offline ones as they tend to experience lessened social support (Su et al., 2012) which exacerbates predispositions to stress and anxiety (Tarafdar and Stich, 2021). For this reason, even though we adjusted the task to limit encumbrance, we still regard the individual and team response to a stressful task as the determining factor for whether personal characteristics and/or team compositions help handle the challenge successfully.

By engaging the players in this high-pressure challenge, we examine whether personality characteristics (Conscientiousness, Extraversion, Neuroticism, Agreeableness, and Openness) may make individuals more prone to cooperation under time pressure. We further evaluate which, if any, combination of personalities results in better than average team performance. Similarly, we examine whether additional factors such as the participants’ socio-cultural background affect their actual ability to work together and their satisfaction with teamwork. Understanding the crowds perception of the collaboration (and not only performance) will help the development of AI agents to support their needs—and not only effectiveness—in times of crisis. Additionally, perceptions on the collaboration may provide insights into why certain teams are more effective than others, and what teams may be willing to work together again on the next task. Thanks to the heterogeneous data gathered during the experiment, we look at the dyadic communication to unravel indicators of a given team’s potential to cope with a high-demanding task under time pressure.

A focus of this research is the impact of participants’ personality on ad-hoc online teamwork, that is crowd-sourced, brief, and under pressure. We use the Big Five personality model (Goldberg, 1990), also known as the Five-Factor model, to model and comprehend the relationship between crowd workers’ personality traits and their disposition for online teamwork in emergency contingencies. We selected the Big Five model as it is most commonly used for personality analysis [e.g., Highhouse et al., 2022; Ikizer et al., 2022; Mammadov, 2022] and for artificial intelligence systems that automatically adapt to personality [see (Smith et al., 2019) for a review of personality models used for personalization in persuasive technology, intelligent tutoring systems and recommender systems]. Additionally, many validated instruments exist to measure the Big Five traits, including the brief version of the Big Five Personality Inventory (Rammstedt and John, 2007) which we use in this paper. The Big Five model distinguishes between 5 traits1, each of which has multiple facets (see Table 1)

frai 05 818491 t001 | How Personality and Communication Patterns Affect Online ad-hoc Teams Under Pressure | Coletividad

Positive and negative facets of the BIG-5 personality traits (Neuman et al., 1999).

Operational Setting and Problem Scope

Significant research effort has been placed over the years on teams that need to perform in situations that require spontaneous, ad-hoc decisions and short-term planning, to resolve ambiguous or uncertain events, and where the consequences of failure are significant (Reuter et al., 2014). The scope of the problems that such teams are called to deal with is broad. It can include responding to natural disasters, like floods, hurricanes, and fires, but also managing crises (King, 2002), such as terrorism events (Longstaff and Yang, 2008), events occurring in long-duration spaceflights (Salas et al., 2015), nuclear plant control rooms (Stachowski et al., 2009), or situations taking place in a military context (Driskell et al., 2014). It can also include more benign everyday workplace settings, such as on-call software teams dealing with organizational incidents, like security or service failure events (for example the recent Google outage (Bergen, 2020), journalist teams for the immediate coverage of unexpected events (Archibold, 2003), but also short-term project teams (Galbraith and Lawler, 1993) and task forces (Hackman, 1990). Their size can vary, from dyads and triads (Foushee, 1984), to dozens (Helmreich, 1967), to twenty or more (Stuster, 2011).

Differences From Normal Teams

What separates these teams from teams in “normal” settings, is the extreme, atypical environment within which they operate, which overall entrails time pressure, high levels of risk, increased consequences for poor performance (Driskell et al., 2018), no previous work experience with one another, and the need to perform their task almost immediately on team formation (Mckinney et al., 2005; Mendonça, 2007). Harrison and Connors (1984) use the term exotic environment to describe a work setting that is marked by hostile environmental demands, restricted working conditions, isolation from those outside the setting, and confinement and enforced interactions for those inside it. Using the related term extreme environment, Bell et al. (2018) add that these settings are also characterized by limited time to finish the task. Performance pressure and severe consequences for ineffective performance are also characteristic of these settings, and this pressure can act as a double-edged sword that can lead the team to outstanding performance, or cripple it Gardner (2012). The tasks that teams in these settings must solve are usually characterized by ambiguity and urgency (Yu et al., 2008; Stachowski et al., 2009).

Factors Affecting the Success of Emergency Teams

Which factors determine team success in this high-demand, high-stress environment? Skill and expertise are the primary factors. Teams traditionally trained as emergency response units rely on the specialized expertise of the stages of the incident response and carry insider knowledge of the organizational policies, their obligations, the communication channels, and the tools supplied by the hiring organization. Thereof, the effectiveness of traditionally formed emergency response teams relies to a great extent on the level of preparedness and competence of the hiring body (or authority) that trained and assembled them, with multiple historical incidents providing evidence for the need for precise training programs and hiring criteria (Alexander, 2003). Examining command and control teams, Ellis et al. (2005) find that team members with higher training demonstrated greater proficiency in planning and task coordination activities, as well as in collaborative problem-solving, and communication. The study also found that it is the knowledge competencies of the team member with the most critical position that benefited the team the most.

The second factor of interest is the allocation of roles and authority. A prominent characteristic of typical high-stake teams, such as STAts (swift-starting action teams), is that they comprise experts (Mckinney et al., 2005) with specific roles and responsibilities. Multiple studies confirm the value of stable role structure in the division of labor and in enhancing the predictability of team interactions, allowing each team member to know what to expect from their teammates in critical situations (Hackman and Morris, 1975; Stachowski et al., 2009). The reason is that misunderstandings or disagreements about authority and role accountability (especially non-desirable roles like clean-up) may lead to team conflict, especially in the presence of unprecedented emergency response tasks (Quarantelli, 1988; Weick, 1993). The meta-analysis of De Wit et al. (2012) further confirms the negative relationships between process and role conflict, and team results such as cohesion, commitment, and performance. On the other hand, flexibility, the ability to improvise, and entrusting functional requirements to determine roles, rather than relying on titles may also be of benefit (Briggs, 2005; Mendonça, 2007). A highly defined role structure with clear roles seems to benefit more tasks that are structured. On the contrary, a flatter structure may be better for ambiguous tasks for which no apparent solution can be easily found (Worchel and Shackelford, 1991) (such as the task of responding to the 2001 World Trade Center attack Mendonça, 2007).

Personality is another prominent factor affecting the success of high-stakes teams, in line with the broader personnel selection literature which indicates that if relevant personality factors are identified for a specific job, future performance can be predicted (Borman et al., 1980). Using the occupational personality questionnaire to study the emergency command ability of offshore installation managers, Flin and Slaven (1996) finds significant correlations between command abilities in critical situations and certain personality elements. From their results, it appears that the highest-rated performance came from those who (a) like to take charge and supervise others (high score on controlling), (b) consider themselves to be fun-loving, sociable, and humorous (high score on outgoing), (c) are less interested in analyzing human behavior (low score on behavioral), (d) are more interested in practical than abstract problem solving (low score on conceptual), and (e) prefer to make decisions quickly rather than take time to weigh up all the evidence (high score on decisive).

Flin and Slaven (1996) contribution, however modest in size, is only pertinent to emergency command responsibilities and applicable only within a specific type of organization (offshore installation managers). Other researchers have focused on the possible existence of a “rescue personality,” in multiple additional domains where emergency services and occupational stress are pivotal. Kennedy et al.’s (2014) research on how personality influences the workforce decisions of emergency nurses reveals that certain traits matter more than others. High Extraversion, Openness to experience, and Agreeableness were especially common amongst emergency nurses. Extraversion was also present among emergency department senior medical staff (Boyd and Brown, 2005) as part of the controversial ENTJ (Extrovert, Intuitive, Thinking, Judging) personality type2 (Myers, 1962).

Partially supporting these findings is the work of Wagner et al. (2009) on the personality traits of paid professional firefighters. Although high Conscientiousness was not a determinant factor in this vocational role, Extraversion had significance. Certain personality traits seem to cluster under particular types of emergency professions; the differentiation between correlation and causality between these two variables is not always easy to untangle. Feelings of anxiety and insecurity, as well as heightened levels of Neuroticism and Openness, were seen to be most likely the results, and not the cause, of the repetitive exposure to experiences of loss and distress (Pajonk et al., 2011). By broadening the sample to the general public (virtual crowd), we aim at decoupling the effects that a specialized profession could have on one’s propensity to emergency response.

Finally, certain interaction patterns are useful predictors of whether an ad-hoc team that has been brought together for immediate task performance will succeed or not, in classical emergency response teams. Although swift-start teams have little time to build their group processes before starting to work on the task, it is also known that team routines get established early in the team’s lifecycle. The same initial interactions have an effect on subsequent communication and norms (Gersick and Hackman, 1990). The study of Zijlstra et al. (2012) reveals that there are certain early patterns of communication that distinguish effective from less effective teams. Specifically, they find that effective teams engage in communication that is more stable in duration and complexity, more balanced, and less monopolized by a single participant compared to inefficient teams that exhibit frequent mono-actor patterns, consisting of a single team member posing and answering their questions and commenting on their observations. They also found that efficient teams exhibit more reciprocity and trust, with the team members engaged and in the same direction of action toward the task goal. The presence of trust as a crucial factor is also highlighted (Wildman et al., 2012). The study of Waller et al. (2004) reveals that efficient teams in non-routine situations focused their actions on information collection and task prioritization. Finally, Kanki et al. (1991, 1989) complement the above by showing that the communication of effective swift-start two-person crews focuses on immediate task execution, expressed as low-complexity, straightforward action statements, and is less focused on other non-standard communication.

Although classical rapid-action teams are widely studied, these literature findings do not necessarily translate to online crowd rapid-action teams. Traditional emergency teams comprise highly trained professionals with a shared understanding of the crisis domain, and often a shared loyalty to an organization. In contrast, crowd teams mainly consist of non-experts, and they are more volatile and heterogeneous regarding the motivators that draw their members to the particular task. Considering the multiplication and globalization of the events that require swift action, it is likely that in the future, we will need to turn more and more to crowd workers and volunteers to form ad-hoc online teams that can deal with high-stake situations under pressure. In this light, the extensive study of classical rapid-action teams can provide us with the first grounded indications of specific parameters to look at to identify predictors of successful team formation in online crowd action teams. Given that in a crowd setting, the allocation of roles is likely to take place based on arrival and availability, in this work, we focus on the parameters of personality and communication patterns as predictors of forming a successful crowd team to tackle unforeseen situations under time pressure.

This is our RSS Feed and this story was found here by our Project ADA. Make sure to visit the site and original post!

Compare items
  • Total (0)
Shopping cart