Abstract
Introduction:
Team situation awareness (SA) is essential for effective team performance. Yet, current methods rely on potentially biased self-reports and obtrusive tools such as task probes. Objective, unobtrusive, and continuous methods for monitoring team SA could support its improvement. In contrast to individual SA, unobtrusive measures of team SA have not been systematically reviewed. Because team SA is an umbrella concept, we adapted a framework proposed by Salmon et al., identifying two components that can potentially be measured unobtrusively: shared mental models (SMMs) and team coordination. This systematic review examines behavioural and physiological technologies that may enable unobtrusive measurement of these components.
Methods:
We conducted a systematic literature review of studies investigating behavioural technologies, such as verbal communication and eye-tracking, and physiological technologies, such as electroencephalography, for unobtrusive measurement of team SMMs and coordination. We searched Web of Science (May and July 2025) for relevant literature published between 2000 and 2025. We included peer-reviewed articles published in English. In total, 19 studies addressing team SMMs and 33 addressing team coordination were included. We extracted sample and team size, task environment, operationalisation of team SMMs and coordination, and their relationship with unobtrusive measures.
Results:
Cardiac measures (42%) and verbal communication (47%) were most popular for estimating team SMMs, while functional near-infrared spectroscopy (27%) and cardiac measures (30%) were most popular for estimating team coordination. Interpersonal synchrony—the temporal contiguity and interdependence of physiological and/or behavioural activity between teammates—was applied in 70% of team coordination studies. Studies covered diverse tasks and settings, from laboratory-based cognitive tasks and controlled physical tasks to real-world environments, including surgery. Various methods were used to operationalise SMMs and coordination, including participant experience, expert assessment, and self-reports.
Discussion:
Findings suggest that team SMMs and coordination cannot be measured as single, unified constructs using one unobtrusive measure. Instead, different measures may provide information about specific aspects of these processes. Future research should begin with task analysis to identify relevant aspects of team SA and derive appropriate measures for the specific task environment. Task-related behavioural actions also represent a potentially valuable, yet often overlooked, category of unobtrusive measures. A key limitation is the substantial variation in how SMMs and coordination were operationalised and measured, making direct comparisons difficult. Future work should investigate how identified measures relate directly to team SA using validated measures.
1 Introduction
Professionals increasingly work in teams. An essential component of good teamwork is situation awareness (SA), especially in time-critical situations involving dynamically changing information (; e.g., aircraft maintenance teams: ). SA is often defined as the perception of a stimulus, its comprehension, and the projection of the future environmental state (the 3-Level model of SA; ). In a team, not only should each teammate have adequate SA to perform their individual task, but also SA about the needs of their teammates—referred to as team SA (Salmon et al., 2008). A team is defined as a group of at least two individuals with a shared goal and specified roles working interdependently and adapting to a dynamic situation (Salas et al., 1992). Continuous measurement of team SA during teamwork may help reduce errors and improve safety. Yet, current measurement methods are not optimal for this purpose.
Currently, team SA is typically assessed using questionnaires or interviews gathered after task completion (e.g., ). A key limitation of these after-action reviews is that teammates may not accurately recall all details of their performance. An alternative method is to collect a probe during brief interruptions of the task (e.g., SAGAT, Situation Awareness Global Assessment Technique; ). However, this approach may alter team SA itself (Salmon et al., 2006). To overcome these flaws, team SA should ideally be assessed in a more unobtrusive, objective, and continuous way—hereafter referred to as unobtrusive measurement. “Unobtrusive” refers to not interfering with or altering teaming processes by teammates answering questions during the task (Salmon et al., 2006). “Objective” means that the method should produce quantifiable data, as subjective judgements of SA (e.g., self-ratings or external assessors) may be influenced by cognitive bias (Salmon et al., 2006). Finally, “continuous” indicates that the measurement allows for the evaluation of team SA development throughout a task. To advance the unobtrusive measurement of team SA, the most likely option is the use of behavioral and physiological tools (Zhang et al., 2023).
1.1 Unobtrusive measurement of team SA through behavior and physiology
Technological advancements in wearable technology enable the continuous recording of physiological signals and behavioral data in a relatively unobtrusive manner (reviews: ; ; Zhang et al., 2023). Such approaches are attractive for monitoring team SA as they do not interrupt the task at hand to do so. These methods have been investigated relating to individual SA. At the individual level, eye-tracking and electroencephalography (EEG) show promise—eye-gaze is a proxy of (visual) stimulus perception, and EEG may be indicative of stimulus comprehension (review: Zhang et al., 2023; application: Yang et al., 2024). However, team SA is more complex. It depends not only on each teammate's ability to develop their own SA, but also on teaming interactions. Consequently, unobtrusive measures of team SA need to capture these interactions (; ). Below, we overview how behavioral data and physiological signals may relate to team SA.
1.1.1 Behavioral data's role in measuring team SA
Human behavior is not uniform across situations—instead, it is influenced by the context in which the human is situated (). Thus, observable—and measurable—behavior can be indicative of changes in the way we interact with elements and with other participants in the environment. In this review, we will explore the use of verbal communication, eye-tracking, and bodily behavior as indicators of team SA. Verbal communication facilitates the sharing of information, potentially directly revealing the team's knowledge, and aids in coordinating team actions. For example, teammates that have good awareness of each other's SA-related needs may ask their teammates less questions (). Next, as shown for individual SA, eye-tracking is an objective measure of (visual) attention allocation. For example, this can inform whether teammates are attentive to each other's needs. Last, bodily behavior may be influenced by other teammates' behavior. For example, it has been shown that when observing others, individuals may begin to mimic their actions (i.e., motor mimicry) or feel and express the same emotions (i.e., emotional contagion). This happens as a result of empathy—the process by which humans (and other animals) understand others' cognitive or emotional states when observing them (). Humans are especially attentive to others that are socially significant to them, such as family, friends, or teammates. Thus, similarity in teammate bodily behavior may be indicative of a close social bond and active social processing of teammate needs.
1.1.2 Physiological signals' role in measuring team SA
Physiological signals stem from autonomous (ANS) and central nervous system (CNS) activity. While physiological processes are less directly related to higher-order constructs like team SA, they provide objective information about stress, mental workload, cognitive resource allocation, attentional engagement, and others. These factors are indirectly associated with (the capability to develop) team SA. In addition, similar to behavior, physiological activity occurs within a social environment. As such, it is affected by and reflective of changes in the environment (Waldman et al., 2019). The tools explored in this review are EEG, functional near-infrared spectroscopy (fNIRS), cardiac and breathing responses, and electrodermal activity (EDA). EEG and fNIRS reflect CNS activity, such as executive functioning and the engagement of socially relevant brain areas. These tools can inform about the extent to which teammates are processing information or actively interacting. Cardiac and breathing responses reflect sympathetic and parasympathetic ANS activity, while EDA primarily reflects sympathetic ANS activity. In general, lower sympathetic and higher parasympathetic ANS activity is associated with reduced stress and, consequently, greater availability of executive function resources and improved task performance. In contrast, higher sympathetic and lower parasympathetic ANS activity reflects elevated stress, which can hinder team processes. In some tasks, optimal performance has been shown to be associated with an intermediate level of sympathetic ANS activity (i.e., stress). Furthermore, the type of stress experienced (e.g., acute or chronic) and the nature of the task can also play an important role (Sandi, 2013). Alternatively, parallel stress responses can inform about similarity in teammates' mental processing, indicating a shared interpretation of the task environment. Overall, cardiac and fNIRS responses have lower temporal resolution, while EEG and EDA have higher temporal resolution, being able to follow rapid physiological changes (). These characteristics are especially relevant in highly dynamic tasks.
1.1.3 Interpersonal synchrony's role
As briefly mentioned, measuring team SA would require a method that may capture teaming interaction processes. While we do not specifically target interpersonal synchrony (IPS) in our review, we observed a strong presence of this construct in literature aimed at quantifying team processes. IPS refers to interdependent variation and temporal alignment in physiological or behavioral activity between at least two people (Palumbo et al., 2017; ; ). IPS can be applied to various measurement modes, such as synchrony in ANS activity (e.g., heart rate, breathing rate, or EDA; Palumbo et al., 2017) or in brain activity (Montague et al., 2002). This can inform about various processes, such as common coordination styles or joint teammate attentiveness behavior. IPS has already been explored extensively in relation to teaming processes such as team coordination ().
1.2 Goal of systematic review
The goal of this paper is to uncover behavioral and physiological measurement technologies of team SA. To achieve this, we systematically reviewed the literature for a range of behavioral and physiological indicators that may co-occur with strong team SA. The results of this review may advance unobtrusive measures of team SA and be used as a guide for future studies when selecting the most appropriate measures of the construct.
An initial search for literature on behavioral and physiological measurement of the construct “team SA” identified only one suitable study—linking team SA to electrodermal activity (EDA; ). This apparent lack of literature likely reflects the complexity of team SA, which cannot be easily captured by a single unobtrusive measure and across all task contexts. To overcome this challenge, the perspective of this review shifted toward identifying behavioral and physiological measurements of underlying team processes of team SA that may be more readily measurable. The theoretical framework by Salmon et al. (2008) was used to identify these relevant processes. In this way, we uncovered potential measures that in future may be compared with traditional measurements of the broader construct of team SA (e.g., SAGAT).
1.3 Theoretical framework of team SA
Various conceptual frameworks of team SA have been proposed (see reviews: Salmon et al., 2008; Stanton et al., 2017). From these, we selected the framework described by Salmon et al. (2008) because of its relative simplicity and because we deemed the team processes involved to be potentially unobtrusively measurable. The framework has overlap with the team SA framework by ), which has been described as most well-known in the field (Salmon et al., 2008), but seemed too elaborate for our current purpose. In Salmon et al. (2008) framework, data from the environment are first processed by each individual teammate (based on Endsley's 3-Level model) and then transferred between teammates through team processes. These team processes, as well as the final product (team SA), may in turn influence the individual SA of each teammate.
The framework distinguishes four team processes: (1) shared mental models (SMMs), (2) collaboration, (3) co-operation, and (4) communication. An SMM is a common knowledge structure about the goals, expectations, and execution of a team task (). Salmon et al. (2008) discuss collaboration, co-operation, and coordination as relevant processes. These concepts all have strong conceptual similarities (reviews: ; ; dictionary definition: Merriam-Webster (n.d.)). They are encompassed by the definition—to work jointly with another party to achieve a common, valued goal. Salmon et al. (2008) do not distinguish or elaborate on why all three processes are uniquely relevant. Thus, we summarize them by the term team coordination—as it is the most used term in this field (e.g., ; ). We note that SMMs and team coordination are theoretically intertwined: (1) strong shared knowledge enables the alignment in team expectations and information needs, resulting in better team coordination (), and (2) better team coordination allows for better planning and transmission of information, therewith strengthening SMMs (Stout et al., 1999).
Last, communication represents transactive behaviors (verbal or non-verbal) by which teammates continuously and concurrently exchange information to advance shared meaning (). We considered communication to be a behavioral measure of the other two processes. Therefore, our adapted framework focused on SMMs and team coordination (Figure 1).
Figure 1
2 Literature review
2.1 Methods
Based on our adapted model from Salmon et al. (2008; Figure 1), we searched the literature for unobtrusive measures reflecting the team processes SMMs and team coordination. We performed four separate scans of the scientific literature using different sets of keywords (see Table 1 for the keyword sets). Set 1 contained synonyms of SMMs and Set 2 contained synonyms of coordination (coordination synonyms derived from the reviews: ; ). Set 3 contained behavioral and physiological measures (keywords selected after Zhang et al., 2023). We include verbal indicators even if they are currently marked by raters (i.e., not objective), due to their potential to be automatically transcribed (e.g., the Whisper tool, Radford et al., 2023) and analyzed (e.g., the Linguistic Inquiry and Word Count—LIWC-22—tool; Pennebaker et al., 2001). Set 4 contained keywords related to task-related behavioral actions as a measure of the team processes. Keywords inside a set were combined using an OR operator. Using Set 0, we filtered out non-human agents teaming. We searched Web of Science for peer-reviewed, English-language journal articles and conference proceedings published between 2000 and 2025.
Table 1
| Search set | Keywords |
|---|---|
| Set 1 SMMs | “team cognition,” “team mental model*,” “shared mental model*” |
| Set 2 Team coordination | teamwork, collaboration, cooperation, coordination |
| Set 3 Measures | physiolog*, “eye?track*,” respiration, “blood pressure,” electrocardioac*, “heart rate,” “heart rate variability,” electroencephalography, “brain signal,” “brain activity,” “event?related brain potential*,” “functional near?infrared spectroscopy,” electrodermal, “tissue blood volume,” “skin conductance level,” “skin conductance response*,” “galvanic skin response*,” “facial expression*,” speech, verbal |
| Set 4 | “user interface,” UI, GUI, “system interface,” “control input*,” “user interaction*,” dashboard, “display interface,” “computer interface,” “digital interface” |
| Set 0 | robot, AI, synthetic, autonomous, human-machine |
Keywords used in the systematic literature search of unobtrusive measures of the team processes in team SA.
An asterisk (*) denotes a wildcard, meaning that all words that begin with the indicated root are to be included in the results (example: physiolog* indicates that physiology, physiological, physiologically will all be included). Quotes (“”) indicate that the exact written phase must be matched in the results (example: papers that contain mental model will not be included, but papers containing shared mental model will be). A question mark (?) indicates that any symbol may be included in its place (example: eye?track indicates that eye-track, eye track will all be included). UI, user interface; GUI, graphic user interface; AI, artificial intelligence.
The first two scans were performed on May 14th, 2025, titled Search 1 and Search 2. A simplified visualization of the two literature scans is shown in Figure 2. In Search 1 on measures of SMMs, we combined (with an AND operator) Set 1 (SMMs) with Set 3 (measures) and the keyword “team” (to limit records to studies on team dynamics) and excluded records with Set 0 keywords (NOT operator). In Search 2 on measures of team coordination, we combined Set 2 (team coordination) with Set 3 (measures) and with “team,” excluding records with Set 0 keywords. These first two scans revealed 47 records regarding team SMMs (Search 1) and 1,107 records regarding team coordination (Search 2). Upon observing a lack of task-related behavioral actions, such as key presses and display interactions, we ran two additional scans on July 17th, 2025, titled Search 1.2 and Search 2.2. Search 1.2 combined Set 1 (SMMs) and Set 4 (task-related actions) with “team,” excluding Set 0 keywords. Search 2.2 combined keywords from Set 2 (team coordination) with Set 4 (task-related behavioral actions) and “team,” excluding Set 0 keywords. In both additional scans, the keyword “human-machine” was removed from Set 0 to avoid interference with the added keywords in Set 4. We identified five records on team SMMs (Search 1.2) and 385 records on coordination (Search 2.2).
Figure 2
2.2 Study selection
We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (Moher et al., 2009), aimed at standardizing the reporting of systematic reviews. We utilized four selection phases: (1) identification of records, (2) initial screening (title and abstract), (3) eligibility assessment, and (4) included papers (flowchart shown in Figure 3). In the initial screening, we excluded all records that: (1) did not relate a measure from Sets 3 or 4 to a process from Sets 1 or 2; (2) were purely theoretical or conceptual; or (3) tested human and non-human agent teaming. All records were screened, and all information was extracted from the papers by one reviewer (EZ). To include a record in the full analysis, we adhered to the following inclusion criteria:
SMMs or team coordination were discussed as the construct of interest, and not as an adjacent process (e.g., team performance, engagement, trust, etc.); and
At least one unobtrusive measure was tested for a possible association with the targeted team process.
Figure 3
This approach led to the exclusion of records that: (1) only addressed correlates of general team performance or cohesion (e.g., ) and: (2) the exclusion of the single paper that targeted team SA measures directly (). Following full-text assessment, we reallocated six records that were originally found under the team coordination search to SMMs (; Vrzakova et al., 2019; ; Stewart et al., 2021; Pugh et al., 2022; ).
2.3 Extracted factors
For our review, we extracted the following information from each paper:
Sample and team size: the paper's total sample size and the number of individuals in a team;
Task: task environment;
Measure: the main type of measure (e.g., “communication content”), where necessary, accompanied by how it was determined (e.g., “verbal content types were manually coded”). A relatively large number of papers (five for team SMMs and 23 for coordination) included measures of IPS;
Operationalisation of SMMs/coordination: We recognized the following methods by which the authors estimated the strength of SMMs and team coordination: (1) subjective assessment by experts; (2) time on task, where it is assumed that SMMs and team coordination improve with task progression; (3) self-report questionnaire; (4) participant experience, where experienced teams (comprised of teammates who are experts on the task, or of teammates who have worked together before) are assumed to have stronger SMMs and coordination than inexperienced teams; (5) experimentally controlled methods, for example by manipulating the amount of information shared within a team; (6) methods associated with a task phase, where it is assumed that certain task phases require strong SMMs; and (7) performance-associated methods, where it is assumed that better task performance indicates a stronger SMMs. For team coordination, there was an additional category: (8) teaming conditions (task conditions requiring active teamwork, compared to individual work or other baseline conditions); and
Result: relationships between the unobtrusive measure and SMMs/team coordination.
Results are discussed and grouped according to measuring modes, starting with behavioral measures: (1) verbal communication; (2) eye-tracking; (3) bodily behavior (facial expressions, head movements). Then, the results of all physiological measures will be described: (4) cardiac measures; (5) EEG; (6) functional near-infrared spectroscopy (fNIRS); (7) breathing measures; (8) EDA. Finally, (9) we describe the papers that combined multimodal measurement into a unified measure (all other papers are discussed in each corresponding mode). All findings on individual physiological and behavioral activity, as well as IPS findings, are discussed within their respective measurement modes, while overarching IPS findings that link the different modes are presented separately in section (10) IPS findings across measuring modes. These global trends in IPS can reveal general patterns of convergence across team processes.
A brief overview of the papers, categorized by team process(es) (SMMs and/or team coordination), measurement mode(s) (behavior, physiology, and/or multiple modes of measurement), and type(s) of measurement(s) (e.g., verbal communication) can be found in Supplementary Table 1.
3 Results
3.1 Study characteristics
After applying the inclusion criteria and paper reallocation, 19 studies remained on team SMMs and 33 on team coordination. There was a large variation in the investigated tasks, including laboratory cognitive tasks (gaming, programming, creative design, moral dilemma); controlled physical tasks (piano playing, juggling); and research conducted in real-life environments with large freedom in behavior (surgery, basketball, firefighter training). We observed both virtual (e.g., CS:GO players in virtual teams; ) and non-virtual teaming (e.g., teams in one room solving a product design problem; ) in all measurement modes—meaning that the types of measures chosen were not specific to virtual or non-virtual teaming (Vorreuther et al., 2026).
The most common methods to estimate the strength of SMMs were: participant experience (used in 26% of the SMM studies); association with task phase (21%); and subjective assessment by experts (21%). For coordination, these were: teaming conditions (48% of the coordination studies), and subjective assessment by experts (30%). An overview of all methods used to estimate SMMs and team coordination is shown in Figure 4.
Figure 4
Most studies on SMMs investigated verbal communication measures (47% of the SMM studies) and cardiac measures (42%). For team coordination, the most popular modes were cardiac measures (30% of the coordination studies) and fNIRS (27%). We did not find studies that specifically mentioned task-related behavioral actions as indicators of team processes. The frequency of the measuring modes described is shown in Figure 5.
Figure 5
Ten studies explicitly tested whether team SMMs or team coordination were related to team performance. Most of these relations were positive (two studies for SMMs: ; ; three for coordination: Xue et al., 2018; ; Sun et al., 2021). Some studies reported negative associations (one for SMMs: , Study 3; and for coordination: ), or non-significant (SMMs, though the trend was positive: ; two studies on coordination: ; , Study 1). The rest of the included studies did not explicitly test the relation between team processes and team performance.
In the subsequent sections, we synthesize the promising findings concerning measures of SMMs and, respectively, team coordination per measuring mode (e.g., eye-tracking, fNIRS). In the text, we also summarize non-significant results. All statistically significant measures are outlined in Supplementary Tables 2 (SMMs) and 3 (team coordination), grouped by measuring mode, summarizing information about sample size and the size of a team, the task, the operationalisation of the team process, the unobtrusive measure, and the observed association of the measure with the team process. All resulting associations are related to strong SMMs or coordination, unless stated otherwise.
3.2 Behavioral measures of SMMs
We now consider the literature on how measurements of behavior have been used to elucidate SMMs between team members, which is the first main component of the adapted model of team SA (Salmon et al., 2008).
3.2.1 Verbal communication and SMMs
Verbal measures used to estimate team SMMs can be broadly divided into communication content (content of the communicated information), communication dynamics (e.g., transitions from one content category to another), and communication frequency (e.g., overall frequency of verbal utterances). We overview the findings in this order, starting with communication content and ending with communication frequency.
Although the included studies used slightly different coding schemes and terminology, many of the verbal communication categories overlapped conceptually. Therefore, to facilitate comparison across studies, we grouped similar categories under a common set of labels based on the terminology used in the literature:
task-relevant statements (clarifications, task questions, explanations);
knowledge exchange (e.g., factual task statements);
planning (procedures, strategising, predictions, performance monitoring, like announcing intended actions);
action-oriented (e.g., commands);
reflective (evaluations);
motivational/emotional;
uncertainty (expressing doubt);
acknowledgment (confirmatory—“Yeah,” or appreciative statements—“Good idea”); and
any other statements, referred to as “task-irrelevant” by the studies (e.g., social or other statements without a clear task function).
According to three studies involving cognitive tasks (e.g., gaming, naval simulation), teams with strong SMMs (according to the authors' operationalisation of SMMs) exhibited more knowledge exchange statements (; ; ), acknowledgment statements (), as well as more action-oriented, motivational/emotional, and uncertainty statements (). ) observed increased knowledge exchange only in physically distributed teams (i.e., where teammates worked in separate physical locations), but not in co-located teams. At the same time, teams with stronger SMM teams produced fewer task-irrelevant statements in two studies (; ); in ), this effect was only observed in co-located, but not in physically distributed teams. Teams with stronger SMMs also used fewer planning statements (). A study in real-life football found that teams with self-reported strong SMMs exchanged fewer task-relevant statements ().
On communication dynamics, we found:
communication patterns (how teams transitioned from one content category to another);
content overlap (congruency in the communication content between teammates); and
what we refer to as communication strategy (unsolicited, compared to requested information).
We found two studies addressing communication patterns: one in online gaming (), and another in tennis players (). These studies suggest that transitions between content categories can provide insight into how teammates coordinate and respond to each other's contributions. For instance, action-oriented statements were followed by other action-oriented statements in teams with strong SMMs (operationalised through more experienced teams in ; and through higher performing teams in ). Another study () investigated communication overlap in clinical teams, finding that during task phases requiring strong SMMs—such as patient handovers—teams exhibited a high degree of overlap in the communicated content.
Regarding communication strategy, ) found that teams with strong SMMs (i.e., more experienced teams) pushed more information (offered unsolicited information), compared to requested information (explicitly asked teammates for information). ) also found a decrease in requested information in strong (i.e., more experienced) SMM teams.
Finally, regarding verbal communication frequency, studies used:
overall verbal frequency (total speech units, regardless of verbal content);
pace (word count per second); and
distribution of responsibility between members for sharing and discussing information.
We found varying results on overall verbal frequency. Three studies reported a decrease in verbal frequency in teams with strong SMMs (; ; ). Another study on online gaming found that verbal frequency increased with strong SMMs (). In the naval simulation study by ), distributed teams with stronger SMMs demonstrated a higher frequency of task-relevant statements compared to co-located teams. This effect was even more pronounced under challenging conditions. According to the authors, teams with stronger SMMs were better able to adapt to the distributed teaming environment. Weaker SMMs teams maintained the same amount of verbal communication despite their teaming environment.
) also found that strong-SMM teams communicated at a faster pace. ) related (self-reported) strong-SMM teams to equal communication distribution, suggesting knowledge similarity, which was not apparent in the study of Eldadi and Tenenbaum (2025).
Pugh et al. (2022) investigated the accuracy of a language-based model in distinguishing SMM-related verbal content from other verbal content. The model successfully distinguished instances of shared knowledge communication from other types of communication (coordinating and maintaining team function).
Overall, teams with stronger SMMs appear to exhibit more structured communication, characterized by regular communication patterns. They also tend to be more action-oriented and affirm the contributions of their teammates. We found indications of high communication content overlap and a faster speaking pace in strong-SMM teams. Stronger SMMs were associated with a communication strategy (more pushing, compared to requesting information). Some of the verbal measures used in the studies, such as communication content, overall verbal frequency, and distribution, showed inconclusive associations with SMMs. We overview potential reasons for differences between the studies in the discussion section.
3.2.2 Eye-tracking and SMMs
Two studies employed eye-tracking in relation to SMMs. ) found that teams with stronger SMMs, compared to weaker, more frequently directed their gaze toward areas of the digital display associated with their teammate's actions or other points of interest. This is an indicator of performance monitoring. In a study in visual programming, Vrzakova et al. (2019) found that teams exhibited stronger SMMs (as rated by experts) when teammates consistently directed their gaze toward changes on the display (e.g., the appearance of a new stimulus). However, IPS-eye-gaze between teammates in response to display changes was not significantly associated with SMMs.
3.2.3 Bodily behavior and SMMs
We found only one paper relating bodily behavior to SMMs. In a basketball game, ) constructed a team social network by interviewing teammates and analyzing video recordings of the bodily behavior of teammates. In this way, they determined whether individuals directed their attention toward partner teammates while planning their actions. This analysis encompassed elements such as body direction, head movements, and gaze direction, as well as teammates' reported interpretation of their actions at the time. In teams with stronger SMMs (operationalised through participant experience), teammates did not explicitly direct their attention toward other teammates while planning their actions. This finding was reflected by low network density (i.e., how many connections out of the total possible existed in the network), low network outdegree (i.e., how many connections a node has), and low network reciprocity (i.e., the extent to which connections are mutual). The authors concluded that teammates relied on SMMs and anticipation of future developments, thereby reducing the need for continuous, overt monitoring of one another.
3.3 Physiological measures of SMMs
We now follow with physiological measures and their relationship to team SMMs.
3.3.1 Cardiac measures and SMMs
In several studies on SMMs, authors examined both individual and IPS trends in HR (beats per minute) and HRV (fluctuations in HR quantified using measures such as the root mean square of successive differences between inter-beat intervals, RMSSD). Higher stress or arousal (i.e., sympathetic ANS activity) is related to increased HR and decreased HRV, while calmness (i.e., parasympathetic ANS activity) is associated with decreased HR and increased HRV. Two studies found relatively stable HR in teams with strong SMM throughout a naval simulation task despite increasing task challenge (, ). HR also quickly recovered after challenging conditions (). In accordance with this, teams with stronger SMMs showed high HRV (SMMs were operationalised through self-reports in Sætrevik, 2015: Study 2; and through time on task in , Study 3). According to the authors, this reflected greater availability of cognitive resources.
Regarding IPS-HR, one paper found positive correlations between teammates' HR in a psychomotor task (juggling; ). Similarly, ) reported IPS-HR between teammates during real-time surgery. Although ) did not explicitly examine SMMs, they examined shared motivation, emotions, and stress, where shared motivation falls under SMMs (). IPS indicators were also found for HRV. ) found IPS-HRV, namely, synchronized negative RMSSD peaks between teammates during challenging real-time surgery phases. They argued that this reflected SMMs and coordinated action. Finally, one paper found that teammates had a statistically similar average HRV during a challenging surgical phase that required strong SMMs (). The authors interpreted this to support a similar psychological state between teammates during the formation of SMMs.
Overall, the studies associate cardiac measures with SMMs, suggesting that reduced stress, greater availability of executive function resources, and faster recovery following challenging task phases are linked to stronger SMMs. IPS-HR and IPS-HRV indicated shared mental states.
3.3.2 Electroencephalography and SMMs
Using EEG, Hoyle et al. (2020; Study 3) found increased absolute theta power in the frontal regions with prolonged engagement in an online football game. Absolute theta power represents the spectral power density in the theta 4–8 Hz power band per EEG channel or a cluster of channels—in this case, in the frontal channels. The authors reasoned that this reflected heightened attention, workload, and stronger SMMs.
3.3.3 Breathing measures and SMMs
No studies examined the relationship between individual breathing activity and SMMs; however, one study related IPS in breathing rate (IPS-BR) between teammates to SMMs. Breathing rate refers to the number of breaths taken per minute. IPS-BR positively correlated with SMMs among juggling teammates, but only in the easy condition (). The authors interpreted these findings as indicating that the challenging condition significantly constrained the team's ability to establish sufficiently strong SMMs, thereby leading to decreased IPS-BR.
3.4 Multimodal integration of SMMs measures
Six studies assessed SMMs across multiple modes (, ; ; ; : Study 3; Stewart et al., 2021), two of which further integrated these modes into a single composite measure (; Stewart et al., 2021). As this approach extends beyond multimodal measurement alone, this section evaluates the results of the unified measures.
) integrated data on body movements, speech rate, and display changes to create a single measure of behavioral regularity during a computer programming task. Behavioral irregularity was negatively related to SMMs, as rated by an observer. Stewart et al. (2021) employed multimodal classification models incorporating linguistic, acoustic-prosodic, facial expression, and task context features to identify different aspects of collaborative problem-solving. Based on these features, their model could distinguish ‘construction of shared knowledge' from ‘negotiation/coordination,' and ‘maintaining team functions.'
Overall, a few multimodal approaches successfully integrated complementary measures, which showed promise in measuring the complexity of SMMs.
3.5 Interpersonal synchrony findings across measurement modes for SMMs
Out of the 19 total studies, three examined how ANS activity co-varied between teammates (; ; ), while two investigated IPS in teammates' behavior (; Vrzakova et al., 2019). There were different approaches to calculating IPS (see Supplementary Table 2 for details per paper), each with their own advantages and disadvantages—which is beyond the scope of the current review. Across approaches, IPS was positively associated with team SMMs. In two studies, IPS was a particularly good indicator of SMMs during challenging conditions (; ), while others observed high IPS throughout tasks (; ; Vrzakova et al., 2019).
3.6 Behavioral measures of team coordination
Now, we consider the literature on how measurements of behavior have been used to elucidate team coordination, being the second important component of our adapted model of team SA.
3.6.1 Verbal communication and team coordination
We found the following verbal measures to estimate team coordination:
speaking (detected voice activity) before action;
prosodic accommodation (adjusting, e.g., pitch range and speech rate, such as to match their teammate's);
focused listening (one teammate speaking at a time);
simultaneous speech (teammates speaking at the same time),
and total speaking (spoken communication units) or silence time (silence units).
Teams with strong coordination (as assessed through each paper's operationalisation method) showed several communication behaviors. They spoke before acting in firefighter training () and showed prosodic accommodation in a crew resource management task (). Additionally, they showed more focused listening in an online football simulation game (; Study 1).
Simultaneous speech, total speaking time, and total silence time showed conflicting results. Strongly coordinated teams had both more simultaneous speech () and less simultaneous speech in an online football simulation game (; Study 1). Hoyle et al. (2020; Study 1) further found a steady decrease in silence time with stronger coordination, while ) found no significant link between speaking or silence time and coordination (assessed through experts). In Hoyle et al. (2020; Study 1), the relationship between total speaking time and team coordination was likewise inconclusive.
As with SMMs, Pugh et al. (2022) demonstrated high language-based model accuracy in distinguishing coordinating communication from other types of communication (e.g., sharing knowledge), indicating that verbal communication can be a useful measure of team coordination.
In summary, promising measures of team coordination include prosodic accommodation, focused listening, and speaking before action. In contrast, measures such as total speaking time, total silence time, and simultaneous speech appear to be more sensitive to contextual factors or to the specific operationalisation of team coordination.
3.6.2 Eye-tracking and team coordination
Vrzakova et al. (2019) used eye-tracking to estimate team coordination (assessed through self-report and by experts). The authors presented individual activity in eye-gaze and display changes, as well as patterns in IPS-eye-gaze (see Supplementary Table 3 for a full overview). These measures reflect whether teammates direct their gaze toward relevant display changes; and whether they do so in a synchronized way. Overall, strong IPS in eye-gaze to display-changes behavior, as well as high regularity in eye-gaze-to-display-changes behavior, were positively related to team coordination. However, high complexity in eye-gaze (high fixation entropy) and display change patterns were negatively related to team coordination. These findings suggest that team coordination benefits from more synchronized and regular eye-gaze-to-display-changes behavior, while more complex and variable behavioral patterns are associated with poorer team coordination.
3.6.3 Bodily behavior and team coordination
Two studies examined IPS in bodily behavior. In the first study, Strang et al. (2014) observed higher IPS in head-postural sway movements in real teams compared with an artificial team composed of individuals performing the puzzle task alone. No differences were found between real and false-matched teams (i.e., a member of one real team artificially matched with a member from another team). The authors interpreted this as reflecting a common coordination style across all teams, potentially driven by task constraints. In the second study, Mønster et al. (2016) examined IPS in facial muscle activity using electromyography, focusing on facial muscles related to smiling (zygomaticus activity) and frowning (corrugator supercilii activity). IPS-zygomaticus (smiling) was higher in real, compared to false-matched assembly-line task teams. No differences were observed for IPS-corrugator-supercilii (frowning). The authors stated that smiling may be more contagious than frowning. IPS-zygomaticus (smiling) correlated with positive, self-reported emotional coordination (i.e., a good relationship with teammates), but showed no relationship with performance-related coordination aspects.
3.7 Physiological measures of team coordination
We now elaborate on measures stemming from physiology and their connection to team coordination.
3.7.1 Cardiac measures and team coordination
As for SMMs, team coordination was related to individual and IPS trends in HR and HRV.
During real-time surgery phases, demanding strong team coordination, teammates showed statistically similar averaged HRV (). In addition, high-coordination phases aligned with synchronized negative peaks in HRV between teammates (). Increased IPS-HRV in teams with strong coordination was also reported in two additional studies, conducted in a programming task and in real-time surgery, respectively (Veerabhadrappa et al., 2021; Powezka et al., 2022). During medical training sessions, higher IPS-HRV (calculated via the standard deviation of normal-to-normal intervals, SDNN) was observed in teams with stronger, compared to weaker coordination (operationalised through teaming conditions and time on task; Wespi et al., 2025). Among the two, subsequent high-coordination conditions, IPS-HRV was higher in the second task than in the first, suggesting a potential timing effect of IPS-HRV. Wespi et al. (2025) also reported lower IPS-HRV (calculated via RMSSD) in high- compared to low-coordination tasks. They interpreted this as reflecting higher stress and lower cognitive resources in the high-coordination tasks. Further, the authors noted that the RMSSD metric tends to be more variable under stress, which can reduce HRV-based IPS.
Regarding IPS-HR, Wespi et al. (2025) found increased IPS-HR in teams with strong coordination, interpreted as reflecting shared goals, effort, and strong coordination. In the same study, teams in closer physical proximity showed consistently higher HR-IPS. Timing effects in IPS-HR were also found within the high-coordination tasks, with lower IPS in the second high-coordination task, compared to the first—the opposite pattern of what was observed in IPS-HRV (HRV calculated through SDNN). ) found high HR IPS during expert-team juggling.
Strang et al. (2014) reported higher IPS-HR in real teams compared to false-matched teams in a puzzle task. IPS-HR was elevated in real teams relative to false-matched teams only when teammates shared the same role, whereas no such increase was observed when teammates held different roles. The authors interpreted this result as indicating that IPS-HR reflected similar intra-role coordination processes, but not general teaming processes.
Fusaroli et al. (2016, Study 1) found that IPS-HR was higher in false-matched teams than in real teams, and higher during individual construction task phases than during collective (team) task phases. However, IPS-HR was more stable in real teams and increased over time only in real teams engaged in collective tasks. The authors suggested that both task constraints (e.g., individual engagement) and coordinative effects contributed to this. Furthermore, Study 2 showed that IPS-HR was positively associated with behavioral and speech coordination ().
Two studies found no significant link between IPS-HR and team coordination, and the authors of both suggested that their experimental designs influenced this (; Mønster et al., 2016). ) discussed their potential imprecise operationalisation of team coordination. For this reason, )'s results are mentioned in the remaining relevant sections but not considered in the qualitative summaries of this review. Mønster et al. (2016) reasoned that their assembly-line task lacked sufficient arousal to elicit high IPS-HR.
Overall, findings on cardiac measures appear to be context-dependent and to have timing effects. Two studies indicated that strongly coordinated teams maintain synchronized levels of calmness, reflected in parasympathetic activity (measured through IPS-HRV). While IPS-HR may provide meaningful insights into coordination during challenging tasks, it also appears to be influenced by task-driven factors and not solely by coordinative processes.
3.7.2 Electroencephalography and team coordination
All five EEG studies on team coordination examined IPS. IPS was associated with brain activity in regions related to social processing (through anterior cingulate cortex activity), attentional coordination and problem-solving (frontal areas), and sensory integration (parietal areas). Synchrony was observed in alpha (8–12 Hz, relaxation), beta (12–30 Hz, alertness and focus), and/or gamma (30–40 Hz; complex problem-solving) bands.
) and Reinero et al. (2021) operationalised team coordination through teaming conditions. ) found that high IPS in the prefrontal cortex (PFC) was positively linked to team coordination in simulated combat. Reinero et al. (2021) reported a smaller IPS decline over time in frontal, occipital, and parietal areas (in the alpha and beta bands) for problem-solving teams with strong coordination, compared to weak.
Some studies reported asymmetrical IPS when one teammate assumed a leadership role, and the other followed. ) found IPS in the leader's dorsolateral prefrontal activity with the follower's anterior cingulate cortex and parietal activity in a collaborative card game task. No IPS occurred between non-teammates. The authors suggested that this pattern reflected that followers inferred the leaders' strategies, while leaders were actively formulating those strategies. In a virtual room clearing task, ) reported IPS in low leader beta power with high follower beta power; and in low leader gamma power with low follower beta power during task phases with strong, compared to weak team coordination (team-performance-associated). Both effects were observed in medial-centro parietal locations. The authors state this reflected the unidirectional information flow from leader to follower, indicating high team coordination.
Finally, Zamm et al. (2021) linked temporal coordination in duet piano players to high IPS in oscillatory amplitude fluctuations at the performance frequency. This shows that IPS may be indicative of coordination dynamics.
Overall, despite varied study procedures, IPS in frontal (e.g., executive functioning), socially relevant regions, and in sensory integration regions (parietal) emerged as a key measure of team coordination. In decentralized teams—where no leadership hierarchy exists—coordination may be supported by symmetrical IPS, as members collaboratively propose and endorse plans of action. In contrast, in centralized teams, it is expected that complementary, rather than identical, brain regions synchronize between individuals. This shows the direction of information from leader to follower.
3.7.3 Functional near-infrared spectroscopy and team coordination
fNIRS formed the largest category of team coordination measures. In the brain, fNIRS measures levels of oxygenated (HbO) and deoxygenated (HbR) hemoglobin, with higher HbO indicating greater neural activity in a specific region. Across studies, activity involved frontal areas linked to executive control, goal maintenance, coordination and the mirror neuron system (specifically—in the frontopolar and PFC; dorsolateral PFC; inferior frontal gyrus), motor areas (primary motor cortex), and social areas (specifically—temporo-parietal junction, TPJ, for perspective taking and joint attention; superior temporal gyrus, related to the mirror neuron system). The mirror neuron system (MNS) involves neurons that track and interpret the actions of others.
) and ) operationalised team coordination through teaming conditions. ) found that in a risk-taking task, strong-coordination teams exhibited reduced activity in primary motor cortex (related to motor functions) and bilateral dorsolateral PFC. These regions are associated with goal maintenance, coordination, executive control, and the interpretation of others' actions. Left TPJ activity did not strictly relate to stronger team coordination. Yet, it was active when individuals were inferring others' goals and actions in the context of the task. Machine learning identified the left TPJ and primary motor cortex as the most discriminative regions for coordinative processes. ) used network and graph theory analyses in a group drumming task. Strong-coordination teams showed higher medial PFC Global and Local Network Efficiency, and higher left TPJ Global (but not Local) Network Efficiency. This indicated that strongly coordinated teams shared information broadly rather than only with neighboring teammates.
Two studies examined individual brain activity and synchronized activity between teammates (Xu et al., 2019; Zhang et al., 2021). Xu et al. (2019) tracked frontal activity (anterior PFC), related to the functions of executive control, goal maintenance, and interpreting others' actions. Strong coordination (operationalised through teaming conditions) was associated with increased activity related to these functions. This was expressed through slightly increased HbO in the regions toward the end of the joint drawing task requiring higher coordination. Three IPS-fNIRS findings also emerged in teams with strong coordination: increased IPS-fNIRS in the anterior PFC (expressed through high IPS-HbO); high IPS-fNIRS in decreased anterior PFC activity (IPS-HbR, although this effect was evident not only in stronger coordination, but in general in all task periods of shared information processing, even without active team coordination); and finally, high IPS-HbO/HbR in anterior PFC, increasing with challenging conditions, interpreted as signaling greater coordination demands. In a moral dilemma task, Zhang et al. (2021) found that individuals showed high dorsolateral PFC and middle frontopolar activity, suggesting increased inhibition of selfish behavior and increased coordination effort. Additionally, teams exhibited high IPS in the right inferior frontal gyrus, which the authors interpreted as reflecting efforts to understand teammates' intentions and align on shared goals.
Most studies relied on IPS-only measures. In strong-coordination teams using a moral dilemma task, Zhang et al. (2020) observed elevated IPS-fNIRS in the inferior frontal gyrus. This region is implicated in executive functioning, goal maintenance, and the interpretation of others' actions. The authors also found IPS-fNIRS in the dorsolateral PFC, related to similar functions, but this finding was not associated with strong team coordination. In Xue et al. (2018), expert-rated strong-coordination teams showed elevated IPS-fNIRS in the right dorsolateral PFC (also found by ) and right TPJ—attributed to inferring teammate goals and actions during realistic creativity problem-solving. (Sun et al. 2021) studied joint drawing and found IPS-fNIRS in the left frontopolar regions, which implied synchronized executive functioning, goal maintenance, and interpretations of others' actions in performance-associated strong-coordination teams (a finding not significant in 's joint drawing task). ) observed high IPS-fNIRS in the medial PFC in teams with stronger coordination (operationalised through teaming conditions), compared to weaker. They also observed the same (but non-significant) trend in IPS-fNIRS in the left TPJ during real-time drumming.
) found that strong coordination in creative design was associated with asymmetrical IPS-fNIRS, meaning that linked activity occurred in different, rather than identical, brain regions between teammates. In this study, the asymmetrical IPS-fNIRS showed that teams synchronized between regions of executive control and the MNS/mentalizing regions. Strong-coordination teams showed high IPS-fNIRS between the dorsolateral PFC in one member and the posterior superior temporal gyrus in another member; IPS-fNIRS between the inferior frontal gyrus and the posterior superior temporal gyrus; and IPS-fNIRS between the anterior PFC and TPJ with marginal significance (p < 0.1). Anterior IPS-fNIRS in PFC to posterior superior temporal gyrus increased with stronger self-reported coordination, interpreted by the authors as reflecting verbal communication and coordination processes. Over time, IPS-fNIRS between the anterior PFC and TPJ decreased as externally rated coordination improved, suggesting less effort was needed to coordinate and that teammates increasingly anticipated each other's actions as they adapted.
While the dissimilar methods make summarizing results challenging, some findings can be generalized. Overall, IPS-fNIRS is strongest between frontal and social regions, reflecting synchronized attention, joint coordination efforts, and inferring others' actions. Asymmetrical IPS-fNIRS tends to occur between functionally similar or complementary areas, e.g., between dorsolateral PFC and MNS areas, such as the posterior superior temporal gyrus. These regions together aid the maintenance of the task goal and interpretation of others' actions and how these may suit the goal. Frontal and TPJ activity may increase during initial or novel task phases, where the team coordination style is not yet established, and may decrease as teams strengthen their SMMs and coordination.
3.7.4 Breathing measures and team coordination
Two studies examined IPS-BR. As for SMMs, ) found higher IPS-BR in strongly coordinated teams in the easy, not in the challenging juggling task condition. In the challenging condition, task difficulty may have negatively impacted coordination. Thus, IPS-BR may reflect stronger team coordination. Nothing of significance was found in Henning et al. (2001).
3.7.5 Electrodermal activity and team coordination
Three studies examined IPS-EDA. IPS-EDA entails that teammates' sympathetic ANS activity was temporally associated. First, ) found that high IPS-EDA was aligned with high-coordination phases of a computer problem-solving task, indicating shared arousal and affective states among teammates. Mønster et al. (2016) found IPS-EDA to be related to perceived team tension, but not to coordinative performance. ) found no significant results. In sum, IPS-EDA reflects shared tension, arousal, or stress, which may—but do not necessarily—contribute to team coordination.
3.8 Multimodal integration of team coordination measures
All three multi-modal studies combined their modes into a single, unified measure. van Eijndhoven et al. (2023) incorporated HR and EDA data into a machine learning model, which successfully recalled 96% of externally rated coordination breakdowns (i.e., instances of ineffective team coordination) but exhibited low precision ( ≤ 0.29) due to a high number of false positives. This meant that the model falsely identified unrelated team dynamics as coordination breakdowns. Two studies, already discussed in Section 3.5. Multimodal Integration of SMM Measures showed similar results for coordination. First, Stewart et al. (2021) achieved fairly good accuracy in distinguishing coordination from other aspects of problem-solving by integrating linguistic, acoustic-prosodic, facial expression, and task context features. Second, ) found that behavioral irregularity—measured using body movement, speech rate, and display changes—was negatively related to coordination in a visual programming task.
3.9 Interpersonal synchrony findings across measurement modes for team coordination
Of the 33 studies on team coordination, 23 used IPS, with the most common modes being cardiac measures (nine studies) and fNIRS (six studies). Again, diverse approaches to calculating IPS were used (see Supplementary Table 3 for details).
Studies using EEG and fNIRS reported that higher IPS was associated with stronger team coordination. In fNIRS and EEG, IPS was found in areas tied to mental state prediction, joint attention (social, temporal, frontal regions), and executive functions such as working memory and goal maintenance, consistent with prior reviews (; Réveillé et al., 2024; Wang et al., 2025).
For measures of the ANS and some bodily behaviors, such as facial expressions, the association between IPS and coordination appears to be mediated by teammates' co-varying stress and relaxation levels, or by the team's overall emotional climate (e.g., tension, trust, friendliness). Behavioral IPS, conceptualized as interpersonal regularity (e.g., in eye movements), is related to stronger coordination, whereas increased behavioral irregularity is related to weaker coordination ().
IPS as an indicator of coordination seems to vary by task phase. It appears to rise during teamwork or challenging conditions, but not consistently during individual work. This pattern emerged in a study on IPS-EDA by ), as well as in studies identifying high IPS as an indicator of synchronized effort during challenging conditions (HRV—; fNIRS—Xu et al., 2019).
In sum, the relationship of IPS to coordination depends on task phase, challenge level, and constraints, and on IPS calculation methods.
4 Discussion
In this review study, we decomposed team SA into two unobtrusively measurable components: SMMs and team coordination. We performed a systematic literature review into potential unobtrusive measures of SMMs and team coordination. In the discussion, we will overview the main findings, limitations, and practical implications.
4.1 Main findings
Across 52 studies, we identified and extracted measures associated with SMMs and team coordination.
The included studies strongly differed in multiple characteristics—ranging from the task environments utilized to the operationalisations of the team processes. Furthermore, the unobtrusive measures were often not the primary focus of the studies, particularly in research on team SMMs. These aspects made it difficult to compare findings. However, by bringing together findings from the various measurement modes, we could extract general patterns of behavioral and physiological activity that relate to the targeted team processes and suggest why these measures relate to the team processes. The identified patterns overlap between both team processes, which is not surprising given the theoretical co-occurrence of strong SMMs and coordination in well-performing teams (; ).
Teams with strongly developed SMMs were characterized by teammates predicting each other's needs and knowledge; exchanging knowledge; remaining calm; and attention to the task. This was reflected across measurement modes. Predicting needs and knowledge concern both the ability to predict, as well as teammates' behaviors that ensure team predictability. Both were primarily reflected in cardiac and breathing rate IPS, indicative of shared mental states and similar experience of the task events, signaling a similar understanding of the task. Other evidence of this was a decrease in information requests and increased offering of unsolicited information; regular communication patterns (e.g., transitions from action- to other action-oriented statements) and regularity in eye-gaze behavior. Next, while teammates may be able to predict partner teammates' actions and movements in more structured tasks, like seen in ), they may also need to monitor teammate performance (through eye-gaze direction) in high-risk tasks like naval threat scenario simulations, especially in highly challenging conditions (). Last, low-SMMs teams showed more irregular behavior (in body movement, speech rate, and display changes).
Exchanging knowledge was evident in verbal measures, including communication that focused on discussing and assigning actions, exchanging knowledge, and acknowledging teammates' statements, as well as potentially faster-paced communication to facilitate information transfer.
Regarding remaining calm, teams with strongly developed SMMs showed overall lowered stress levels (high HRV or low HR), or the ability to recover quickly from stress (evident through decreased HR) after highly challenging conditions; and greater executive function resources. Heightened attention or vigilance was reflected in EEG through increased global theta power.
Teams with strongly developed coordination can be characterized by similar and regular teammate behavior; active social processing; remaining calm; attention, and team planning. Similar and regular teammate behavior was evident in prosodic accommodation between teammates in verbal communication; high regularity in eye-gaze between teammates, and eye-gaze-to-display-changes patterns; coordinated and similar experiences of the task difficulty, especially under challenging conditions (through IPS-HR), like for SMMs; as well as high IPS-EDA, relating to shared tension, arousal, and stress; and synchronized head movements. Last, behavioral irregularity (in body movement, speech rate, and display changes) was related to low-coordination teams. Active social processing was primarily reflected in heightened synchronized activity in social brain regions, supporting the interpretation of teammate behavior and goals, as well as in emotional contagion expressed through shared smiling. Remaining calm was evidenced by high IPS-HRV. Attention and team planning were evident through (synchronized) brain activity in frontal regions, where cognitive effort to coordinate the planning of actions may decrease over time as teams develop SMMs; and through verbal measures, including speaking before acting and greater periods in which only one teammate was speaking.
Finally, it should be noted that some findings—e.g., ANS measures like IPS-HR—were observed only under task conditions characterized by specific (i.e. high(er)) levels of workload, stress, or difficulty. This was summarized using the word challenging. We interpret these findings to indicate that the task's challenge level needed to be high enough to demand strong SMMs and/or coordination, while not too high to cause a coordination breakdown.
4.1.1 The interpretation of team process measures depends on task context
Some measures were inconclusive. We believe this to be a result not only of the varied operationalisations of the team processes, but also of the variation in task demands. The greatest variations were observed in verbal communication and cardiac activity.
An example in verbal communication is that while stronger SMMs were most often associated with decreased verbal frequency, some studies, examining distributed teams, showed that stronger SMMs were associated with increased verbal frequency. This seems to indicate that while teams with strong SMMs do not need to verbally communicate as much, under certain circumstances (e.g., team members being spread across locations) more verbal communication is indispensable to develop and maintain a shared understanding. A similar pattern emerged for team coordination. For team coordination, its association with total speaking time also differed across task environments. We reason that teams in more standardized or non-changing tasks may require less verbal communication as they develop stronger SMMs. Alternatively, teams in non-standard tasks may go through phases of less or more verbal communication as the task evolves.
Similarly, cardiac measures showed inconsistencies. Both HR and IPS-HR demonstrated positive, negative, and inconclusive relationships with team coordination across studies. This variability may partly reflect the sensitivity of HR to multiple factors such as mental stress, physical activity, and many others (Valentini and Parati, 2009). For example, in low arousing tasks, no connection between IPS-HR and team coordination was found (e.g., Mønster et al., 2016); yet in high arousing tasks, IPS-HR was found in strongly coordinated teams (e.g., Wespi et al., 2025). A possible explanation is that in low-arousal tasks, HR never increases enough to be informative of coordination. Furthermore, as discussed in the introduction, certain tasks require an optimal level of sympathetic ANS activity (Sandi, 2013). This optimal stress level may vary according to the different tasks. Consequently, the interpretation of cardiac signals may depend on both task context and the specific team process being approximated.
Taken together, these findings indicate that behavioral and physiological measures cannot be interpreted uniformly across task contexts. Their relationship with SMMs and team coordination and, ultimately, team SA, likely depends on when and how a particular process is required by the task. To use behavioral and physiological measures to estimate team SA, we need a method or approach that takes this into account and therewith produces more consistent findings (cf. methodological research gap; Miles, 2017). Such an approach would need to incorporate aspects of the task at hand such as whether a task is currently in a phase of team members performing independent tasks. This would satisfy findings that reported in this review that IPS-EDA was associated with team coordination specifically during task phases requiring strong coordination; and ), who found that high IPS-EDA was positively associated with performance during phases of interdependent teamwork but negatively associated with performance during independent work. By developing more targeted, a priori hypotheses regarding when and how each measure should relate to a particular team process, we may advance toward a solution of the methodological gap (also argued by Waldman et al., 2019). To solve this gap, we argue that the selection and interpretation of team SA indicators should begin with an analysis of the task requirements.
4.1.2 . Task analysis as the foundation for team SA measurement
Establishing a set of informative markers of team SA should begin with a task analysis to identify SA-related requirements, task constraints, and information flow (Salmon et al., 2010). Task analysis is defined as the study of what is required of an individual or of a team toward the execution of a given goal (). Task constraints shape which measuring modes and indicators can be meaningfully applied to estimate team SA. In some tasks, verbal communication is not possible, making behavior (e.g., key presses or bodily behavior) more informative. Alternatively, distributed teams may have access to less information about their teammates and the shared task environment than co-located teammates. This could be described as low team SA, or it could be described as the very nature of the task. A team with high team SA may accurately recognize the limitations that the task environment (i.e., distributed teams) poses and compensate through additional verbal clarifications. The question is not simply which measures can be correlated to strong team SA, but rather which measures correlate to or show that a team is engaging in the relevant and necessary team processes demanded by any particular task.
An initial task analysis can be used to define the behaviors and mental processes that are expected at different stages of a task. Then, the appropriate indicators may be chosen to detect changes in those processes and behaviors. For example, distributed teams may begin a task and employ more verbal clarifications. These may decrease after a team establishes strong SMMs, despite the physical distance between the teammates. Thus, a task analysis method may map out the expected progress of SMMs throughout the task and the elements of the task requiring active team coordination (as opposed to elements where each teammate tackles an individual task). Next, the task analysis may map out the exact sub-elements of SMMs/coordination necessary per task element. For example, a challenging phase within a certain task would require shared effort and understanding to quickly process the environment and maintain high team SA. This would be relevant to the elements of predicting needs and knowledge of teammates or similar and regular behavior between teammates, and approximated through, e.g., IPS-HR. Further, a stressful task event may elicit IPD-EDA—if all teammates involved did notice and process that such an event had occurred.
Because different aspects of team SA may manifest through different behavioral and physiological processes, these considerations also point toward the potential value of a multimodal approach to its measurement.
4.1.3 Unobtrusive measurement of team SA may benefit from a multi-modal approach
Due to the complexity of team SA and the relevance of measures per task phase, we argue that a single mode cannot capture all relevant elements. Instead, team SA requires a multi-modal measurement approach. In this review, we decomposed team SA into processes expected to be more readily measurable. The results suggest that SMMs and coordination can be decomposed into yet more measurable subcomponents (section of Main Findings, in italics). This interpretation is consistent with theory indicating that processes such as SMMs are complex and encompass multiple behavioral and cognitive components (Mohammed et al., 2000; ). Therefore, measuring these subcomponents requires multiple modes, each providing complementary insight into team processes.
4.1.4 Interpersonal synchrony may reflect both strong team SMMs and coordination
We observed that IPS—especially in brain activity—was more frequently examined in relation to team coordination than to SMMs. Because coordination involves active and deliberate effort to understand teammates' mental state and align goals and actions, IPS may be elevated during coordinative phases. Still, IPS likely reflects a multitude of processes (Palumbo et al., 2017), and in our review, ten studies interpreted their IPS results as evidence of both team SMMs (shared understanding, goals, mental state awareness) and coordination (; Xue et al., 2018; Xu et al., 2019; Zhang et al., 2020; ; ; Sun et al., 2021; Zhang et al., 2021; ; Wespi et al., 2025). The idea that these two processes are more intertwined than distinct has also been suggested previously (; ). Thus, a team that has established high team SA may be characterized by high IPS, signaling both strong team SMMs and coordination (an argument also suggested by ) —while a team in the process of strengthening their SMMs may show weaker IPS (as seen in distributed teams, which are expected to have weaker SMMs; Wespi et al., 2025).
4.1.5 Task-related behavioral actions are understudied as indicators of team processes
Our initial literature scans revealed no studies that used task-related behavioral actions as indicators of team processes, prompting additional scans with targeted keywords. Still, no relevant studies were found. Two studies examined display changes alongside other factors, like eye-gaze (; Vrzakova et al., 2019), but did not use actions like key presses as primary drivers of the display changes. Such interactions with the display may nonetheless reflect coordinative actions and shared knowledge and can be extracted through an initial task analysis to fit each task environment. Therefore, we recommend further study of how task-related behavioral actions can be measured and used for the analysis of team processes and team SA.
4.2 Limitations
The search strategy we used has its limitations. First, while we included synonyms of the targeted processes (SMM and team coordination) as keywords for our search, we may have missed studies which had a different scope. For example, we could have missed studies that may have indirectly addressed our targeted team processes as intrinsic aspects of team performance. Yet, our goal was to remain focused on our target constructs. We included synonyms of the team processes to cover more literature. Second, because of practical limitations, only a single reviewer (EZ) performed the screening and selection of literature, as well as the information extraction. Although this is a deviation from the PRISMA protocol, we expect it to have little effect because edge cases were discussed with the full research team. Third, we only examined studies that were published in the year 2000 and later. Reasons to exclude earlier work were 1) wanting to focus on novel technologies, and 2) the oldest reference included in the outstanding review paper on SA by Zhang et al. (2023) was from 2001 ().
Another limitation, referring to the outcome of the review, is that the included studies used highly diverse approaches to estimating the strength of SMMs and team coordination, making direct comparisons between studies difficult. As discussed in the following section on future work, the reviewed measures should be investigated more systematically in relation to team SA.
4.3 Future work
Based on the above discussion, we recommend future work in the following direction.
First, the task analysis approach should be developed and tested as the basis of selecting unobtrusive measures of team SA. Rather than selecting measures ad hoc, future work may investigate the characteristics of any given task and determine which measures would theoretically be most informative. In this way, guidelines may be established for the selection and interpretation of these measures across tasks and contexts. These guidelines can be validated by examining whether the theoretically justified predictions were accurate about the resulting unobtrusive measurements.
Next, these predictions of team SA should be validated against an established and independent measurement or manipulation of team SA. For example, the SAGAT () can be used as a measure, or in a study where teammates have visual information about each other's current task, this visual channel may be experimentally interrupted. This would be a priori expected to weaken team SA—participants can no longer see what their teammate is doing. A task analysis may be performed to develop exact expectations about how the verbal behavior of participants will change. Researchers could then examine whether these hypothesized changes in verbal behavior explain variance in independently measured team SA. Such an approach would provide a stronger empirical basis for the proposed relation between unobtrusive measures and team SA. Additionally, it would further corroborate the framework by Salmon et al. (2008). More broadly, research in the direction of utilizing a task analysis as the first step to team SA measurement could advance the field beyond the ad hoc selection of unobtrusive measures.
4.4 Practical implications
In this paper, we propose an approach for unobtrusively measuring team SA by identifying indicators that can estimate the team processes involved (Salmon et al., 2008). We propose that these efforts begin with a task analysis; and that the interpretation of the measures will depend on the exact task environment and team processes progression. We identified further sub-parts of these team processes (e.g., remaining calm, heightened attention, etc.). These indicators could eventually be automated for real-time application, for example, in team training, which is a crucial component of team SA (). A continuous, unobtrusive and automatic measure of team SA could support the teaching staff, reduce training time, and improve error prevention. For example, for certain contexts, teams may be instructed to utilize closed-loop communication by affirming they have heard a teammate's instruction whereas in other contexts biofeedback may be used to inform the team on IPS, which may improve teammate awareness. Although the optimal combination and application of these measures remain open questions for future projects, this paper proposes an initial selection of measures and a task analysis approach.
5 Conclusion
In this systematic review, we decomposed team SA into team processes that are likely to be unobtrusively measurable. To this end, we adapted Salmon et al. (2008)'s framework of team SA and focused specifically on the team processes of SMMs and coordination. In the 19 selected studies on unobtrusively measuring SMMs and 33 on team coordination, measures of various modes were used, with cardiac and verbal measures being most used to estimate team SMMs, and fNIRS and cardiac measures used for team coordination. IPS (in various modes) was a popular measure for team coordination. We found that task-related behavioral actions, like key presses, are understudied as unobtrusive measures. Furthermore, we found that SMMs and coordination are broken down into further subcomponents, all measured through different but complementary modes. A multi-modal measurement approach may result in the most meaningful estimate, but it will still depend on the task and context. Therefore, we recommend performing a task analysis to determine how the measurable processes underlying team SA develop over the course of a task and, given task constraints, select appropriate modes and measures from those identified in this review.
Statements
Data availability statement
The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding authors.
Author contributions
EZ: Software, Conceptualization, Visualization, Investigation, Formal analysis, Validation, Writing – original draft, Methodology, Data curation, Writing – review & editing. EG: Methodology, Resources, Funding acquisition, Writing – review & editing, Supervision, Conceptualization, Project administration, Writing – original draft. PH: Project administration, Resources, Writing – review & editing, Supervision. A-MB: Resources, Project administration, Conceptualization, Writing – review & editing, Methodology, Funding acquisition, Supervision, Writing – original draft.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This work is part of EZ's PhD work, which is funded by the Dutch Ministry of Defence. The contributions of AMB and EG were supported by the Dutch defence research programmes V2415 “Team Metrics” and V2306 “NextGen Aircrew Performance,” respectively.
Acknowledgments
We thank Jan-Maarten Schraagen for his valuable initial consultation on the topic of team situation awareness, and Paul Salmon for clarifying aspects of his work. We acknowledge the use of grammar and spelling check software (Grammarly) and the use of generative language models (GPT-4) in the revision stages of the manuscript.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The author A-MB declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.
Generative AI statement
The author(s) declared that Generative AI was used in the creation of this manuscript. We acknowledge the use of grammar and spelling check software (Grammarly) and the use of generative language models (GPT-4) in the revision stages of the manuscript. Generative language models (GPT-4) were used to aid rewording, structure, and clarity of the original draft, and were not used to generate novel text or other content (e.g., figures, tables, or any others).
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnrgo.2026.1864518/full#supplementary-material
Abbreviations
ANS, autonomic nervous system; BB, bodily behavior; BR, breathing measures; CNS, central nervous system; EDA, electrodermal activity; EEG, electroencephalography; ET, eye-tracking; fNIRS, functional near-infrared spectroscopy; SAGAT, Situation Awareness Global Assessment Technique; IPS, interpersonal synchrony; MMM, multiple modes of measurement; MNS, mirror neuron system; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; SA, situation awareness; SMMs, shared mental models; VC, verbal communication.
References
1
AmonM. J.VrzakovaH.D'MelloS. K. (2019). Beyond dyadic coordination: multimodal behavioral irregularity in triads predicts facets of collaborative problem solving. Cogn. Sci.43:e12787. doi: 10.1111/cogs.12787
2
AstolfiL.ToppiJ.FallaniF. D.VecchiatoG.SalinariS.MattiaD.et al. (2010). Neuroelectrical hyperscanning measures simultaneous brain activity in humans. Brain Topogr.23, 243–256. doi: 10.1007/s10548-010-0147-9
3
Badke-SchaubP.NeumannA.LaucheK. (2011). “An observation-based method for measuring the sharedness of mental models in teams,” in Coordination in Human and Primate Groups, ed. M. Boos, M. Kolbe, P. Kappeler, and T. Ellwart (Berlin, Heidelberg: Springer), 177–197. doi: 10.1007/978-3-642-15355-6_10
4
BedwellW. L.WildmanJ. L.DiazgranadosD.SalazarM.KramerW. S.SalasE. (2012). Collaboration at work: an integrative multilevel conceptualization. Hum. Resour. Manag. Rev.22, 128–145. doi: 10.1016/j.hrmr.2011.11.007
5
BlaserM. A.SeilerR. (2019). Shared knowledge and verbal communication in football: changes in team cognition through collective training. Front. Psychol.10:77. doi: 10.3389/fpsyg.2019.00077
6
BönselR.MüllerR. (2021). Shared mental models for cooperative fault diagnosis: effects on communication and diagnostic performance. Int. J. Hum. Factors Ergon.8, 215–244. doi: 10.1504/IJHFE.2021.118214
7
BoucseinW. (2012). Electrodermal Activity.New York, NY: Springer. doi: 10.1007/978-1-4614-1126-0
8
BourboussonJ.R'KiouakM.EcclesD. W. (2015). The dynamics of team coordination: a social network analysis as a window to shared awareness. Eur. J. Work Organ. Psychol.24, 742–760. doi: 10.1080/1359432X.2014.1001977
9
Cannon-BowersJ. A.SalasE.ConverseS. (1993). Shared mental models in expert team decision making. Individual Group Decis. Making: Curr. Issues221, 221–246.
10
CastañerX.OliveiraN. (2020). Collaboration, coordination, and cooperation among organizations: establishing the distinctive meanings of these terms through a systematic literature review. J. Manage.46, 965–1001. doi: 10.1177/0149206320901565
11
CookeN. J.CohenM. C.FazioW. C.InderbergL. H.JohnsonC. J.LemattaG. J.et al. (2024). From teams to teamness: future directions in the science of team cognition. Hum. Factors66, 1669–1680. doi: 10.1177/00187208231162449
12
CzeszumskiA.LiangS. H. Y.DikkerS.KönigP.LeeC. P.KooleS. L.et al. (2022). Cooperative behavior evokes interbrain synchrony in the prefrontal and temporoparietal cortex: a systematic review and meta-analysis of fNIRS hyperscanning studies. eNeuro9:0268-21.2022. doi: 10.1523/ENEURO.0268-21.2022
13
daSilvaE. B.WoodA. (2025). How and why people synchronize: an integrated perspective. Pers. Soc. Psychol. Rev.29, 159–187. doi: 10.1177/10888683241252036
14
De LoozeC.VaughanB.KellyF.KayA. (2015). “Providing objective metrics of team communication skills via interpersonal coordination mechanisms,” in 16th Annual conference of the international speech communication Association (Grenoble: International Speech Communication Association (ISCA)), 1226–1230. doi: 10.21437/Interspeech.2015-309
15
de WaalF. B. M.PrestonS. D. (2017). Mammalian empathy: behavioural manifestations and neural basis. Nat. Rev. Neurosci.18, 498–509. doi: 10.1038/nrn.2017.72
16
DechurchL. A.Mesmer-MagnusJ. R. (2010). Measuring shared team mental models: a meta-analysis. Group Dyn.: Theory Res. Pract.14, 1–14. doi: 10.1037/a0017455
17
del RosarioK.WestT. V.MendesW. B.KunduziB.MamodeN.GogalniceanuP. (2024). How does surgeons' autonomic physiology vary intraoperatively? A real-time study of cardiac reactivity. Ann. Surg.279, 258–266. doi: 10.1097/SLA.0000000000006007
18
DemirM.McNeeseN. J.CookeN. J. (2017). Team situation awareness within the context of human-autonomy teaming. Cogn. Syst. Res.46, 3–12. doi: 10.1016/j.cogsys.2016.11.003
19
DindarM.JärveläS.NguyenA.HaatajaE.ÇiniA. (2022). Detecting shared physiological arousal events in collaborative problem solving. Contemp. Educ. Psychol.69:102050. doi: 10.1016/j.cedpsych.2022.102050
20
DodelS.TognoliE.KelsoJ. A. S. (2020). Degeneracy and complexity in neuro-behavioral correlates of team coordination. Front. Hum. Neurosci.14:328. doi: 10.3389/fnhum.2020.00328
21
DodelS.CohnJ.MersmannJ.LuuP.ForsytheC.JirsaV. (2011). “Brain signatures of team performance,” in Foundations of Augmented Cognition. Directing the Future of Adaptive Systems, eds. D. D. Schmorrow and C. M. Fidopiastis (London: Springer Nature), 288–297. doi: 10.1007/978-3-642-21852-1_35
22
EilerB. A.KallenR. W.HarrisonS. J.RichardsonM. J. (2013). Origins of order in joint activity and social behavior. Ecol. Psychol.25, 316–326. doi: 10.1080/10407413.2013.810107
23
EldadiO.TenenbaumG. (2025). Between-rounds verbal communication in expert and nonexpert esports team players. Sport Exerc. Perform. Psychol.14, 471–488. doi: 10.1037/spy0000384
24
EndsleyM.JonesW. M. (2001). “A model of inter and intra team situation awareness: Implications for design, training and measurement,” in New trends in cooperative activities: Understanding system dynamics in complex environments (Santa Monica, CA), 46–67.
25
EndsleyM. R. (1995). Toward a theory of situation awareness in dynamic systems. Hum. Factors37, 32–64. doi: 10.1518/001872095779049543
26
EndsleyM. R. (1999). “Situation awareness in aviation systems,” in Handbook of Aviation Human Factors, Vol. 11, 257–276.
27
EndsleyM. R.RobertsonM. (2000). Situation awareness in aircraft maintenance teams. Int. J. Ind. Ergon.26, 301–325. doi: 10.1016/S0169-8141(99)00073-6
28
EndsleyM. R. (2020). “Situation awareness in teams: models and measures,” in Contemporary Research, 1 Edn. (Boca Raton, FL: CRC Press), 1–29. doi: 10.1201/9780429459733-1
29
EspevikR.JohnsenB. H.EidJ. (2011). Communication and performance in co-located and distributed teams: an issue of shared mental models of team members?Milit. Psychol.23, 616–638. doi: 10.1080/08995605.2011.616792
30
EspevikR.JohnsenB. H.EidJ.ThayerJ. F. (2006). Shared mental models and operational effectiveness: effects on performance and team processes in submarine attack teams. Milit. Psychol.18, S23-S36. doi: 10.1207/s15327876mp1803s_3
31
FeeseS.ArnrichB.RossiM.TrösterG.BurtscherM.MeyerB.et al. (2013). “Towards monitoring firefighting teams with the smartphone,” in: 2013 IEEE International conference on pervasive computing and communications workshops (PERCOM Workshops) (New York City, NY: IEEE), 381–384. doi: 10.1109/PerComW.2013.6529521
32
FilhoE.PieriniD.RobazzaC.TenenbaumG.BertolloM. (2017). Shared mental models and intra-team psychophysiological patterns: a test of the juggling paradigm. J. Sports Sci.35, 112–123. doi: 10.1080/02640414.2016.1158413
33
FusaroliR.BjorndahlJ. S.RoepstorffA.TylénK. (2016). A heart for interaction: shared physiological dynamics and behavioral coordination in a collective, creative construction task. J. Exp. Psychol.-Hum. Percept. Perform.42, 1297–1310. doi: 10.1037/xhp0000207
34
GormanJ. C.CookeN. J.WinnerJ. L. (2006). Measuring team situation awareness in decentralized command and control environments. Ergonomics49, 1312–1325. doi: 10.1080/00140130600612788
35
GuastelloS. J.MirabitoL.PeressiniA. F. (2020). Autonomic synchronization under three task conditions and its impact on team performance. Nonlinear Dyn. Psychol. Life Sci.24, 79–104.
36
GuastelloS. J.BednarczykC.HaganR.JohnsonC.MarscisekL.McGuiganL. M.et al. (2022). Team situation awareness, cohesion, and autonomic synchrony. Hum. Factors66, 1186–1200. doi: 10.1177/00187208221118301
37
HałgasE. A.van EijndhovenK. H. J.GeversJ. M. P.WiltshireT. J.M. WesterinkJ. H. D.RispensS. (2023). A review of using wearable technology to assess team functioning and performance. Small Group Res.54, 41–76. doi: 10.1177/10464964221125717
38
HarrisT. E.SherblomJ. C. (2018). Small Group and Team Communication.Long Grove, IL: Waveland Press, Inc. doi: 10.4324/9781351019583-10
39
HayneL.GrantT.HirshfieldL.CarterR. M. (2023). Friend or foe: classifying collaborative interactions using fNIRS. Front. Neuroergon.4:1265105. doi: 10.3389/fnrgo.2023.1265105
40
HenningR. A.BoucseinW.GilM. C. (2001). Social-physiological compliance as a determinant of team performance. Int. J. Psychophysiol.40, 221–232. doi: 10.1016/S0167-8760(00)00190-2
41
HollnagelE. (2012). “Task analysis: why, what, and how,” in Handbook of Human Factors and Ergonomics, 4th Edn. ed. G. Salvendy (Hoboken, NJ: John Wiley and Sons), 55–79. doi: 10.1002/9781118131350.ch13
42
HoyleB.TaylorJ.ZugicL.FilhoE. (2020). Coordination cost and super-efficiency in teamwork: the role of communication, psychological states, cardiovascular responses, and brain rhythms. Appl. Psychophysiol. Biofeedback45, 323–341. doi: 10.1007/s10484-020-09479-8
43
KaziS.KhaleghzadeganS.DinhJ. V.ShelhamerM. J.SapirsteinA.GoeddelL. A.et al. (2021). Team physiological dynamics: a critical review. Hum. Factors63, 32–65. doi: 10.1177/0018720819874160
44
Kennedy-MetzL. R.DiasR. D.StevensR. H.YuleS. J.ZenatiM. A. (2021a). Analysis of mirrored psychophysiological change of cardiac surgery team members during open surgery. J. Surg. Educ.78, 622–629. doi: 10.1016/j.jsurg.2020.08.012
45
Kennedy-MetzL. R.DiasR. D.ZenatiM. A. (2021b). “The cognitive relevance of a formal pre-incision time-out in surgery,” in Proceedings of the 32nd European conference on cognitive ergonomics (Siena: Association for Computing Machinery). doi: 10.1145/3452853.3452867
46
KleinmanD. L.SerfatyD. (1989). “Team performance assessment in distributed decision making,” in Proceedings of the symposium on interactive networked simulation for training (Orlando, FL: University of Central Florida), 22–27.
47
KolbeM.BurtscherM. J.ManserT. (2013). Co-ACT—a framework for observing coordination behaviour in acute care teams. BMJ Qual. Saf.22, 596–605. doi: 10.1136/bmjqs-2012-001319
48
LausicD.TennebaumG.EcclesD.JeongA.JohnsonT. (2009). Intrateam communication and performance in doubles tennis. Res. Q. Exerc. Sport80, 281–290. doi: 10.1080/02701367.2009.10599563
49
LiL.WangH.LuoH.ZhangX.ZhangR.LiX. (2020). Interpersonal neural synchronization during cooperative behavior of basketball players: a fNIRS-based hyperscanning study. Front. Hum. Neurosci.14:2020. doi: 10.3389/fnhum.2020.00169
50
LiuT.DuanL.DaiR.PelowskiM.ZhuC. (2021). Team-work, team-brain: exploring synchrony and team interdependence in a nine-person drumming task via multiparticipant hyperscanning and inter-brain network topology with fNIRS. NeuroImage237:118147. doi: 10.1016/j.neuroimage.2021.118147
51
MamykinaL.JiangS. L.CollinsS. A.TwohigB.HirshJ.HripcsakG.et al. (2016). Revealing structures in narratives: a mixed-methods approach to studying interdisciplinary handoff in critical care. J. Biomed. Inform.62, 117–124. doi: 10.1016/j.jbi.2016.03.025
52
MansikkaH.VirtanenK.HarrisD. (2023). What we got here, is a failure to coordinate: implicit and explicit coordination in air combat. J. Cogn. Eng. Decis. Mak.17, 279–293. doi: 10.1177/15553434231179566
53
MansikkaH.VirtanenK.UggeldahlV.HarrisD. (2021). Team situation awareness accuracy measurement technique for simulated air combat – curvilinear relationship between awareness and performance. Appl. Ergon.96:103473. doi: 10.1016/j.apergo.2021.103473
54
MayoO.GordonI. (2020). In and out of synchrony-behavioral and physiological dynamics of dyadic interpersonal coordination. Psychophysiology57:e13574. doi: 10.1111/psyp.13574
55
MayselessN.HawthorneG.ReissA. L. (2019). Real-life creative problem solving in teams: fNIRS based hyperscanning study. Neuroimage203:116161. doi: 10.1016/j.neuroimage.2019.116161
56
MengB.LuN.LinC.ZhangY. J.SiQ. M.ZhangJ. S. (2022). Study on the influencing factors of the flight crew's TSA based on DEMATEL-ISM method. Cogn. Technol. Work24, 275–289. doi: 10.1007/s10111-021-00688-7
57
MerchantS.YongjinK.SchnellT.EtheringtonT.VoglT. (2001). “Evaluation of synthetic vision information system (SVIS) displays based on pilot performance,” in 20th Digital avionics systems conference (New York City, NY: IEEE; AIAA), 2C1/1–2C1/12. doi: 10.1109/DASC.2001.963328
58
Merriam-Webster (n.d.). “Collaborate,” in Merriam-Webster.com dictionary.
59
MilesD. A. (2017). “A taxonomy of research gaps: identifying and defining the seven research gaps,” in Doctoral student workshop: finding research gaps-research methods and strategies (Dallas, Texas), 1–10.
60
MohammedS.KlimoskiR.RentschJ. R. (2000). The measurement of team mental models: we have no shared schema. Organ. Res. Methods3, 123–165. doi: 10.1177/109442810032001
61
MoherD.LiberatiA.TetzlaffJ.AltmanD. G. (2009). Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med.6:e1000097. doi: 10.1371/journal.pmed.1000097
62
MønsterD.HåkonssonD. D.EskildsenJ. K.WallotS. (2016). Physiological evidence of interpersonal dynamics in a cooperative production task. Physiol. Behav.156, 24–34. doi: 10.1016/j.physbeh.2016.01.004
63
MontagueP. R.BernsG. S.CohenJ. D.McClureS. M.PagnoniG.DhamalaM.et al. (2002). Hyperscanning: simultaneous fMRI during linked social interactions. Neuroimage16, 1159–1164. doi: 10.1006/nimg.2002.1150
64
PalumboR. V.MarracciniM. E.WeyandtL. L.Wilder-SmithO.McGeeH. A.LiuS. W.et al. (2017). Interpersonal autonomic physiology: a systematic review of the literature. Pers. Soc. Psychol. Rev.21, 99–141. doi: 10.1177/1088868316628405
65
PennebakerJ.FrancisM.BoothR. (2001). Linguistic Inquiry and Word Count (LIWC): LIWC 2001 (Mahwah, NJ: Erlbaum Publishers), 71.
66
PowezkaK.PettipherA.HemakomA.AdjeiT.NormahaniP.MandicD. P.et al. (2022). A pilot study of heart rate variability synchrony as a marker of intraoperative surgical teamwork and its correlation to the length of procedure. Sensors22:8998. doi: 10.3390/s22228998
67
PughS. L.RaoA.StewartA. E. B.D'MelloS. K.Assoc CompM. (2022). “Do speech-based collaboration analytics generalize across task contexts?,” in 12th Annual international conference on learning analytics and knowledge (LAK) (New York City, NY: Association for Computing Machinery (ACM)), 208–218. doi: 10.1145/3506860.3506894
68
RadfordA.KimJ. W.XuT.BrockmanG.McleaveyC.SutskeverI. (2023). “Robust speech recognition via large-scale weak supervision,” in Proceedings of the 40th international conference on machine learning, eds. K. Andreas, B. Emma, C. Kyunghyun, E. Barbara, S. Sivan, and S. Jonathan (Cambridge, MA: PMLR).
69
ReineroD. A.DikkerS.Van BavelJ. J. (2021). Inter-brain synchrony in teams predicts collective performance. Soc. Cogn. Affect. Neurosci.16, 43–57. doi: 10.1093/scan/nsaa135
70
RéveilléC.VergotteG.PerreyS.BosselutG. (2024). Using interbrain synchrony to study teamwork: a systematic review and meta-analysis. Neurosci. Biobehav. Rev.159:105593. doi: 10.1016/j.neubiorev.2024.105593
71
SætrevikB. (2015). Psychophysiology, task complexity, and team factors determine emergency response teams' shared beliefs. Saf. Sci.78, 117–123. doi: 10.1016/j.ssci.2015.04.017
72
SalasE.DickinsonT. L.ConverseS. A.TannenbaumS. I. (1992). “Toward an understanding of team performance and training,” in Teams: Their Training and performance (New York, NY: Ablex Publishing), 3–29. doi: 10.1037/e561492011-003
73
SalmonP.JenkinsD.StantonN.WalkerG. (2010). Hierarchical task analysis vs. cognitive work analysis: comparison of theory, methodology and contribution to system design. Theor. Issues Ergon. Sci.11, 504–531. doi: 10.1080/14639220903165169
74
SalmonP.StantonN.WalkerG.GreenD. (2006). Situation awareness measurement: a review of applicability for C4i environments. Appl. Ergon.37, 225–238. doi: 10.1016/j.apergo.2005.02.001
75
SalmonP. M.StantonN. A.WalkerG. H.BaberC.JenkinsD. P.McMasterR.et al. (2008). What really is going on? Review of situation awareness models for individuals and teams. Theor. Issues Ergon. Sci.9, 297–323. doi: 10.1080/14639220701561775
76
SandiC. (2013). Stress and cognition. WIREs Cogn. Sci.4, 245–261. doi: 10.1002/wcs.1222
77
StantonN. A.SalmonP. M.WalkerG. H.SalasE.HancockP. A. (2017). State-of-science: situation awareness in individuals, teams and systems. Ergonomics60, 449–466. doi: 10.1080/00140139.2017.1278796
78
StewartA. E. B.KeirnZ.D'MelloS. K. (2021). Multimodal modeling of collaborative problem-solving facets in triads. User Model. User-adapt. Interact.31, 713–751. doi: 10.1007/s11257-021-09290-y
79
StoutR. J.Cannon-BowersJ. A.SalasE.MilanovichD. M. (1999). Planning, shared mental models, and coordinated performance: an empirical link is established. Hum. Factors41, 61–71. doi: 10.1518/001872099779577273
80
StrangA. J.FunkeG. J.RussellS. M.DukesA. W.MiddendorfM. S. (2014). Physio-behavioral coupling in a cooperative team task: contributors and relations. J. Exp. Psychol.-Hum. Percept. Performance40, 145–158. doi: 10.1037/a0033125
81
SunB.XiaoW.LinS.ShaoY.LiW.ZhangW. (2021). Cooperation with partners of differing social experience: an fNIRS-based hyperscanning study. Brain Cogn.154:105803. doi: 10.1016/j.bandc.2021.105803
82
ValentiniM.ParatiG. (2009). Variables influencing heart rate. Prog. Cardiovasc. Dis.52, 11–19. doi: 10.1016/j.pcad.2009.05.004
83
van EijndhovenK. H. J.WiltshireT. J.HalgasE. A.GeversJ. M. P. (2023). A computational approach to examining team coordination breakdowns during crisis situations. J. Cogn. Eng. Decis. Mak.17, 256–278. doi: 10.1177/15553434231156417
84
VeerabhadrappaR.HettiarachchiI. T.BhattiA. (2021). “Using recurrence quantification analysis to quantify the physiological synchrony in dyadic ECG data,” in 15th Annual IEEE international systems conference (SysCon), 1–8. doi: 10.1109/SysCon48628.2021.9447059
85
VorreutherA.BrouwerA.-M.VukelićM. (2026). Reviewing digital collaborative interactions with multimodal hyperscanning through an ever-growing database. Front. Neuroergon.7:1756956. doi: 10.3389/fnrgo.2026.1756956
86
VrzakovaH.AmonM. J.StewartA. E. B.D'MelloS. K. (2019). “Dynamics of visual attention in multiparty collaborative problem solving using multidimensional recurrence quantification analysis,” in Proceedings of the 2019 CHI conference on human factors in computing systems (Glasgow: Association for Computing Machinery). doi: 10.1145/3290605.3300572
87
WaldmanD. A.WangD.FentersV. (2019). The added value of neuroscience methods in organizational research. Organ. Res. Methods22, 223–249. doi: 10.1177/1094428116642013
88
WangS. Y.HashemiG.ZhangY.ZhengB. (2025). Exploring a novel approach to assessing surgical team collaboration: evidence of brain activity synchronization. Laparosc. Endosc. Robot. Surg.8, 6–12. doi: 10.1016/j.lers.2024.10.001
89
WespiR.NeherA. N.BirrenbachT.SchauberS. K.FrenkelM. O.Schrom-FeiertagH.et al. (2025). Physiological team dynamics explored: physiological synchrony in medical simulation training. Adv. Simul.10:5. doi: 10.1186/s41077-025-00335-5
90
XuJ.SlagleJ. M.BanerjeeA.BrackenB.WeingerM. B. (2019). Use of a portable functional near-infrared spectroscopy (fNIRS) system to examine team experience during crisis event management in clinical simulations. Front. Hum. Neurosci.13:85. doi: 10.3389/fnhum.2019.00085
91
XueH.LuK. L.HaoN. (2018). Cooperation makes two less-creative individuals turn into a highly-creative pair. Neuroimage172, 527–537. doi: 10.1016/j.neuroimage.2018.02.007
92
YangJ.LiangN. D.PittsB. J.Prakah-AsanteK.CurryR.YuD. (2024). An eye-fixation related electroencephalography technique for predicting situation awareness: implications for driver state monitoring systems. Hum. Factors66, 2138–2153. doi: 10.1177/00187208231204570
93
ZammA.PalmerC.BauerA. K. R.BleichnerM. G.DemosA. P.DebenerS. (2021). Behavioral and neural dynamics of interpersonal synchrony between performing musicians: a wireless EEG hyperscanning study. Front. Hum. Neurosci.15:717810. doi: 10.3389/fnhum.2021.717810
94
ZhangM. M.JiaH. B.WangG. H. (2021). Interbrain synchrony of team collaborative decision-making: an fNIRS hyperscanning study. Front. Hum. Neurosci.15:702959. doi: 10.3389/fnhum.2021.702959
95
ZhangM. M.JiaH. B.ZhengM. X. (2020). Interbrain synchrony in the expectation of cooperation behavior: a hyperscanning study using functional near-infrared spectroscopy. Front. Psychol.11:542093. doi: 10.3389/fpsyg.2020.542093
96
ZhangT.YangJ.LiangN.PittsB. J.Prakah-AsanteK.CurryR.et al. (2023). Physiological measurements of situation awareness: a systematic review. Hum. Factors65, 737–758. doi: 10.1177/0018720820969071
Summary
Keywords
behavior, measure, physiology, review, shared mental models, team, team coordination, team situation awareness
Citation
Zhecheva E, Groen E, Haselager P and Brouwer A-M (2026) Review on unobtrusive estimation of team processes underlying team situation awareness: shared mental models and team coordination. Front. Neuroergonomics 7:1864518. doi: 10.3389/fnrgo.2026.1864518
Received
24 April 2026
Revised
11 September 2026
Accepted
24 September 2026
Published
14 October 2026
Volume
7 - 2026
Edited by
Riccardo Poli, University of Essex, United Kingdom
Reviewed by
Behzad Esmaeili, Purdue University, United States
Kyana Van Eijndhoven, Tilburg University, Netherlands
Updates
Copyright
© 2026 Zhecheva, Groen, Haselager and Brouwer.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Emanuela Zhecheva, emanuela.zhecheva@tno.nl
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.