Keywords

1 Introduction

One (1) type of cognitive architecture in domain of team work and collaboration that has received substantial research attention in Human-computer Interaction is Shared Mental Model (SMM). SMM is derived from the root of mental model construct from the discipline of cognitive psychology (DeChurch and Mesmer-Magnus 2010). Shared mental model (SMM) refers to “knowledge structure held by members of a team that enables them to form accurate explanations and expectations for the task, and in turn, to coordinate their actions and adapt their behavior to demands of the task and other team members” (Cannon-Bowers et al. 1993, p. 228). According to Payne (2003), even though the research interest in SMM has become overwhelming in HCI, more emphasis on theoretical, empirical and conceptual work pertaining to the team cognition and computer interaction are still required.

Earlier in 1993, mental model has been discussed by Staggers and Norcio (1993), concerning the mental model formation, characteristics and models. Other reviews have been conducted by Isenberg et al. (2011), which it provides a detailed review on five (5) scenarios on how SMM being applied in collaborative visualisation tools. Other study conducted by Grimstead et al. (2005) provide a review of how SMM are being used in forty two (42) collaborative visualisation systems across four application areas: collaborative problem-solving environments, virtual reality environments, multi-player online games and multi-user enabling of single user applications. Most recently study is a review and analysis of shared visual representation for building a shared mental model (Nor’ain and Siti Salwah 2015). These studies only described how important is the SMM application in particular interactive systems, however, none of them provide a thorough understanding on the theoretical concepts behind the SMM processing such as properties, aspects and types of SMM, as well as techniques being used in evaluating SMM in visualization technologies.

The aim of this paper is to provide a theoretical review about this concept and guide to future research. The search for this review studies was performed through popular electronic databases i.e. Science Direct, ACM Digital Library, IEEE Explore Digital Library, ISI Web of Knowledge, Scopus Online, Taylor and Francis Online and Springer Link. The search strategy was limited to English articles only. The initial keywords search include: (‘shared mental model’ OR ‘shared cognition’ OR ‘visualisation technology’). The final step was to narrow down the search to the most relevant researched area in Human-Computer Interaction studies.

2 Shared Mental Model in Human-Computer Interaction

In HCI, team cognitive research is characterized as the study of a team as an information-processing unit (Salas et al. 2008). In order to understand the cognitive processes of the team that they want to study, HCI researchers learn from cognitive models that describe the fundamental concept.

Yusoff and Salim (2014) reported two (2) types of approaches in studying team cognition; socially shared cognitive approach (SSC) and shared situation awareness approach (SSA). SSC is a shared cognitive approach views “how dyads, groups and larger collectives create and utilize interpersonal understanding” (Thompson and Fine 1999, p. 3). On the other hand, SSA refers to “degree that team members possess the same awareness of shared situation awareness requirements, within a volume of time and space, as well as the comprehension of their meaning and projection of their status in the near future” (Endsley et al. 2003, p. 13).

SMM is a type of information-processing model which is developed underlying the SSC approach (Yusoff and Salim 2014). This model views that group members have a separate and independent memory structures. It suggests that group member who is able to access to other member’s memory stores can effectively expand their storage and retrieval, thus leading to development of group interaction. Conversely, SMM has also seen to support awareness situation. For example, works by Haig et al. (2006) shows a SMM development to support situation awareness among clinicians as well as work by Entin and Entin (2000) who found team situation awareness in SMM using simulated military missions. Hence, the SMM can be used to support both SSC as well as SSA.

Using an underlying input-process-output (IPO) framework, the greatest focus of concern on team cognitive research in HCI has been on enhancing communication and collaboration among team members as well as optimizing the performance of the team as a whole. The theoretical concept of SMM is discussed further in next section.

3 Theoretical Concept of Shared Mental Model

SMM refers to “knowledge structure held by members of a team that enables them to form accurate explanations and expectations for the task, and in turn, to coordinate their actions and adapt their behavior to demands of the task and other team members” (Cannon-Bowers et al. 1993, p. 228). SMM develops when team members interact and that converge the individual team member’s mental model, resulting in similar to, or sharing with, that of their team member’s mental model. The terminology SMM has also been introduced in many ways, for example, team mental models and compatible mental model (McComb 2008). Throughout this paper, the two (2) terms SMM and shared cognition are being used interchangeably to describe essentially the same concept.

3.1 Properties of Shared Mental Model

SMM consists of two (2) properties; similarity and accuracy (Mohammed et al. 2010). Similarity in SMM refers to “sharedness” or the “degree to which members’ mental models are consistent or converge with one another and does not signify identical mental models” (Mohammed et al. 2010, p. 880). Examples of SMM studies focusing on similarity are the study on similarity of knowledge structures between two (2) members (Mathieu et al. 2000), the effect of cross-training for the similarity of teammates’ team-interaction model (Marks et al. 2002) and the effect of similar mental models for high-performance team (Zou and Lee 2010). On the other hand, accuracy in SMM refers to the “true score” or “similarity of knowledge ratings about other members and one’s own corresponding self-ratings” (Espinosa 2001, p. 2103).

The studies in sharedness or similarity have given more emphasis than the accuracy in the literature even though some studies attempted to study both properties are also found (Mohammed et al. 2010). For examples; Burtscher et al. (2011) investigate how the similarity and accuracy and two (2) forms of monitoring behavior i.e. team versus systems interacted to predict team performance in anesthesia, and Resick et al. (2010) examined the relationships between team cognitive ability and personality composition in relation to the similarity and accuracy of team task-focused mental models.

3.2 Importance of Shared Mental Model

According to Cannon-Bowers and Salas (2001), constructing SMM is important due to following three (3) reasons:

  • Firstly, SMM provides an explanatory mechanism that helps to understand team performance. It explains the effectiveness of teams’ interaction with one another without the need to communicate.

  • Secondly, SMM construction can be valuable to predict variable in teams such as identifying potential performance problems and providing insight into how the problems can be fixed.

  • Thirdly, SMM can diagnose problems such as identifying poor communication that may derive from lack shared of knowledge.

Due to the importance of SMM, the application of SMM can lead to three (3) outcomes (Cannon-Bowers and Salas 2001) as below:

  • First, SMM could lead to better task performance, such as in terms of the accuracy, efficiency, quality of output, volume, timeliness. This outcome is defined as task-specific.

  • Second, SMM leads to better team processes, which in turn lead to better task performance such as more efficient communication, more accurate expectations and predictions, consensus, similar interpretations, and better coordination. This outcome is defined as task-related.

  • Another expected outcome from SMM is referred to motivational outcomes. This includes cohesion, trust, morale, collective efficacy and satisfaction with the team. However, (Cannon-Bowers and Salas 2001) stresses that the motivational outcomes have a looser association with task performance than the previous two.

Cannon-Bowers and Salas (2001) argued that there is a need to clarify which kind of outcomes that is expected from the SMM so that the types and aspects of shared cognition can be determined. The types and aspects of shared cognition are explained in the next sections.

3.3 Types of Shared Mental Model

Types of SMM refer to what cognitive processes are shared. There are four (4) types of cognitive categories on what is shared in team (Cannon-Bowers and Salas 2001):

  1. 1.

    Task-specific Knowledge - This type of shared cognition allows the team members to coordinate without the need to communicate overtly and act on knowledge without discussion. The nature of knowledge being shared is highly task-specific, which involves specific procedures, sequences, actions and strategies to perform a task.

  2. 2.

    Task-related Knowledge - This type of shared cognition allows team members to have common knowledge about task-related processes such as what it is, how it operates and its importance, which contribute to the team’s ability to accomplish the task. In contrast to task-specific knowledge, it is not task-specific, but it can hold variety of similar tasks.

  3. 3.

    Knowledge of Teammates - This type of shared cognition allows team members to understand each other in terms of their preferences, strengths, weaknesses, and tendencies in order to maximize performance. It views that team learns the distribution of expertise within the team over time. It is also a task-related knowledge but not necessarily task-specific.

  4. 4.

    Attitude or Beliefs - This type of shared cognition allows team members to have similar attitudes and beliefs that lead to effective decisions. It involves the notions of shared beliefs and cognitive consensus. This shared cognition type covers a broad category of knowledge, where it does not related to task-specific or task-related.

These four (4) types of SMM are categorized into two (2) major content domains; task-work, and team-work (Mathieu et al. 2000). Task work domain refers to the work goals and performance requirements, while the team work domain refers to the interpersonal interaction requirements and skills of other team members (Mohammed et al. 2010). The integration between the two (2) major domains and four (4) types of SMMs are presented in Table 1.

Table 1. Mathieu et al. (2000)’s major domains and Cannon-Bowers et al. (1993)’s types of SMMs
Full size table

3.4 Aspects of Shared Mental Model

Aspects of SMM refer to how cognitive processes are shared. There are four (4) different categories of ways of how cognition is shared in team (Cannon-Bowers and Salas 2001):

  1. 1.

    Shared vs. Overlapping - This refers to situations where two (2) or more team members need to have some common knowledge but should not be redundant.

  2. 2.

    Similar vs. Identical - This refers to the need to hold similar or identical knowledge. Team members need to hold similar attitudes and beliefs in order to draw common interpretations that can drive towards effective performance. For example, surgeon and nurse working together in an operation theater are not expected to have identical knowledge, but portions of their knowledge bases are needed to be shared (Undre et al. 2006). This category of shared cognition is associated with the task that must be common among members.

  3. 3.

    Compatible vs. Complimentary - This refers to team which possess specialized roles and knowledge that is crucial to task performance. A multidisciplinary team where each member possesses specialized expertise to solve a problem may have dissimilar knowledge, however still can lead them to complementary behavior.

  4. 4.

    Distributed - This refers to the knowledge that is distributed across members. This aspect of shared cognition is applied in many high performance teams, such as military combat teams, where the systems and tasks are complex and difficult. Therefore, if the team members’ knowledge is specialized and distributed, team members need to coordinate their knowledge effectively in order to achieve SMM.

4 Evaluation for Shared Mental Model

This section describes in detail approaches of evaluations in shared mental model.

4.1 Ways to Measure Shared Mental Model

Shared knowledge can be measured in two (2) ways (Cannon-Bowers and Salas 2001). Firstly is by assessing the structure of team member knowledge and secondly is to measure the content of team member knowledge. Mohammed et al. (2010, p. 884) refers the structure as “how concepts are organized in the minds of participants” whereas content is the “knowledge that comprises cognition”. It is stated that assessing the team knowledge structure is more straightforward; however in practical ways it is rather very difficult. On the other hand, measuring contents has been seen as more possible to conduct.

Steps to Measure Shared Mental Model

DeChurch and Mesmer-Magnus (2010) further have described steps to measure SMM, which involves three (3) aspects of characteristics: elicitation method, structure representation, and representation of emergence. These three (3) characteristics are needed as they can show the operationalization of SMM. They are as described as follows:

  1. a.

    Elicitation Method

    It refers to the technique used to determine the components or content of a mental model. Techniques include:

  • Similarity ratings - Participants are presented with a grid and they will be requested to consider each pair of task nodes and report their perceptions of the relation between the two (2) nodes.

  • Concept maps – Participants are asked to elicit contents and place the actions into a meaningful organizational scheme.

  • Rating scales – Participants are asked to elicit the content of the model and respond to questions about the task on fixed-response formats such as strongly agree to strongly disagree.

  • Card sorting tasks – Participants are asked to sort numbers of cards and categorize or list them based on their understanding of the structure and relationship.

  • Interactively elicited cause mapping - Participants are asked to provide data through questionnaires and/or interviews using interactive ways.

  • Text-based cause mapping – Participants are asked to provide post hoc analyses of data such as systematic coding of documents or transcripts.

  1. b.

    Structure Representation

    It refers to the organized knowledge structures corresponding between how the knowledge content is represented in the mind and how the knowledge representation can be modeled by the researcher (DeChurch and Mesmer-Magnus 2010). Techniques include:

  • Pathfinder – This technique is used to produce appropriate psychological scaling based on the underlying structure between concepts. It provides algorithm that can transform raw paired comparison ratings into a network structure where these concepts are represented as nodes, while the relatedness of the concepts are represented as links.

  • UCINET - This technique is developed by Borgatti et al. (2002) to support social network data analysis. It comes with a complete software package for data visualization.

  • Multidimensional scaling – This technique uses geometric models to represent proximity data spatially. It is used to identify unknown underlying dimensions in organizing cognitive stimuli.

  • Concept mapping/card sorting - as described.

  1. c.

    Representation of Emergence

    It refers to representation technique used “to reveal the structure of data or determine the relationships between elements in an individual’s mind” (Mohammed et al. 2000, p. 129). Techniques include; concept mapping, path-finder, UCINET, interactively elicited cause mapping, text-based cause mapping and Euclidean distance.

Of all the techniques described in each steps for measuring SMM, Mohammed et al. (2000) recommended only four (4) techniques because they encompassed both elicitation and representation. These techniques are referred to as pathfinder, multidimensional scaling, interactive elicited mapping and text-based cause mapping.

Next section describes Cognitive Task Analysis as another approach of evaluation in SMM.

4.2 Cognitive Task Analysis Approach

Cognitive task analysis (CTA) focuses on the difficulties in cognitive structures such as knowledge-based and representational skills as well as processes such as attention, problem solving and decision making (Stanton et al. 2005). The aspects of cognitive structures and processes in the CTA can provide a description of the knowledge and thought processes that are required at the expertise level (Schraagen et al. 2008; Seamster et al. 1997). They can also lead to a process for designing, developing and evaluating a better human–computer interface intended to amplify and extend the human ability to make good decisions (Crandall et al. 2006).

Most studies in CTA are concerned with expertise (Klein and Militello 2001, p. 180). Cognitive study is designed to elicit the knowledge and wisdom acquired (Crandall et al. 2006, p. 134). For example, during CTA interviews, interviewers will appreciate the nature of expertise when responses and feedback received are probed in detail. Some related expertise studies that have been conducted using CTA include experienced air warfare coordinators unpacking their expertise and coaching skills for the development of shipboard-based on-the job training for the Navy (Pliske et al. 2000), certified cytotechnologists detecting questionable cells and making sense of the clinical picture for the process documentation of tissue biopsies and cell samples for pathology (McDermott and Crandall 2000) and army ranger squad or platoon leaders describing the required skills for clearing buildings in urban combat settings for the development of training software (Phillips et al. 1998).

Cooke (1994) found more than hundred (100) types of CTA methods and techniques. Due to the growing number of CTA methods, extensive CTA reviews by Stanton et al. (2005), Schraagen et al. (2000) as well as Wei and Salvendy (2004) offer a broad exploration of the difference among these methods and techniques in a number of ways. Stanton et al. (2005) present five (5) selected CTA methods based upon their popularity and the application used, while Schraagen et al. (2000) described a comprehensive review of reviews and classifications to guide researchers interested in exploring and applying the CTA techniques. On the other hand, Wei and Salvendy (2004) classify the CTAs into four (4) broad families, namely: 1) observation and interview 2) process tracing 3) conceptual techniques and 4) formal models. This CTA family classification is meant to guide researchers who are aiming for particular outputs, to select appropriate techniques.

5 Role of Shared Visualisation to Support Shared Mental Model

One way to externalize the individual mind is through the representation of visualization. Many studies have also demonstrated that visual representation that is shared among the users can lead to the development of SMM. Visualization is referred to as “a method of computing…offers a method for seeing the unseen, enriches the process and unexpected insights” (National Science Foundation’s Visualization 1987). According to McGrath et al. (2012), visualization is a graphical representation of data to aid human cognition. These two (2) definitions explain that visualization that is shared can enrich the process as well as the unexpected insights performed by many users.

The application and effect of visualization to shared cognition have been studied by many researchers in the domain of cognition and design studies. For example, the effectiveness of visual representation for the purpose of externalizing and communicating the design process has been demonstrated by Goldschmidt (2007). In this study, two (2) experiments are conducted to clarify how the visual representations have created a SMM of a new bicycle accessory meant to carry a backpack. The result shows that in order for all team members to arrive at a shared task model, it is necessary for them to see the design entity eye to eye in order to progress.

Other studies which had demonstrated the important of visual representation includes; collaborative knowledge construction via visual graphical representation (Suthers 2005), reducing the effort of explicit communication via shared white boards in emergency department (Xiao et al. 2007), and understanding different kinds of video representation and analysis via the use of video story (McNeese 2004).

Arias et al. (2000) opined that SMM can be visualized through the use of external artifacts. External representation can be used to make the knowledge available to all members explicitly as well as able to transcend the cognitive limitation across individual minds. This externalization is important as it creates what is vaguely resides in one’s mental efforts. In other word, artifact represents an externalization which can be communicated visually to the users.

Next section describes the significant role of using cognitive artifact to externalize visualisation.

5.1 Cognitive Artifacts

Artifact that is used as a tool for cognitive activities is called as the “cognitive artifacts”. According to Visser (2006), cognitive artifact comes in two (2) forms:

  • Internal cognitive artifacts or mental cognitive artifacts (i.e. mental representation) – Examples; such as rules of thumb, mnemonics, shopping lists, and other kinds of procedures.

  • External cognitive artifacts – There are two(2) types; physical being such as buildings, cars or garments or any results from mental representations and symbolic such as software, route plan, drawings, mock-ups or any results that symbolize the mental representation.

Visser (2006) explains that due to the nature of the mental cognitive artifacts that are less vague in terms of ideas or images that designers have “in their heads”, the use of external cognitive artifact to visually externalize the emergence of mental cognitive artifacts are therefore required. A research example of using external cognitive artifact is conducted by Nemeth et al. (2004) to study communication and information sharing among healthcare providers. In this study, external cognitive artifacts which are related to operating room or scheduling are used such as the availabilities sheet, master schedule, graph and board. This study finds that better computer-supported cognitive artifacts should benefit patient safety by making teamwork processes, planning, communications and resource management more resilient. Figure 1 shows the emergence process of external artifact to represent the visual externalization of internal cognitive artifact.

Fig. 1.
figure 1

The emergence process of an external cognitive artifact

Full size image

5.2 The Technological Approach Using Collaborative Visualisation

Salas et al. (2008) suggested that SMM can be improved using technological development and implementation. One (1) way to support SMM is by using technological visualization approach called as collaborative visualization (CoVis). In CoVis, a shared use of any forms of visual representations is required to enable any cognitive activities collaboration.

CoVis is referred to as “the shared use of computer-supported, interactive, visual representations of data by more than one (1) person with the common goal of contribution to joint information processing activities” (Isenberg et al. 2011, p. 312). CoVis is an approach that emphasis the shared use of interactive visual representations, which could be in a form of joint viewing, interacting with, discussing, or interpreting the presentation.

According to Isenberg et al. (2011), one (1) of the important aspects in CoVis is the focus on cognition and results. It emphasizes that CoVis is not concern about the creation of a “product” i.e. the shared representation, but the focus should involve unique cognitive activities i.e. shared cognition. Besides, Isenberg et al. (2011) also emphasizes the need to support social interaction process around the data. It concerns on the ability for a team to build each other’s insights, which in turn could reach a common understanding of the datasets. Examples of social interaction include arriving at a common understanding of the data as well as enhance knowledge construction by making use of interaction of data.

CoVis also includes distributed design environment which can be directed to the division or spread of resources such as design artifacts, design knowledge or design team. Distributed design can be operated in either synchronous or asynchronous mode. Synchronous mode in distributed design enables real-time communication and collaboration in a “same time-different space” environment; whereas, asynchronous mode enables the communication and collaboration in distributed design to be operated over a period of time through a “different time-different space” mode.

In order to understand the analogy of cognitive process in SMM and CoVis, we present the major processes of input, processing and output as shown in Table 2. As seen in CoVis process, data within the mind of users is acquired, represented and emerged using a form of visual representation (Isenberg et al. 2011). Similarly, SMM process also show the data or knowledge from the mind of individual or group is elicitated, represented in a structure form and finally emerge to form SMM (DeChurch and Mesmer-Magnus 2010).

Table 2. Analogy of shared cognitive process in shared mental model and collaborative visualisation
Full size table

These processes however do not show how the role of artifact can be used for each process. As such, we come out with a cognitive data process of SMM using the role of artifact. Based on these two (2) processes, the role of artifact can be used to map with each of the major process in order to understand what cognitive data can be acquired as an input, how the cognitive data can be externalized as a process, and what form of visualization can display the emergence of that cognitive data. From the previous review, we have mentioned the focus on elements of knowledge and/or needs which can lead to the development of individual mental model derived from Badke-Schaub et al. (2007)’s framework. The cognitive data process of SMM using the role of artifact is as follows:

  • Cognitive Data Acquisition - is the input process to identify internal cognitive artifact that represents the data from knowledge and needs of users.

  • Cognitive Data Process - refers to how that internal cognitive artifacts being processed.

  • Cognitive Data Emergence - is the output process which displays the emergence of that cognitive data in a form of visualization. It can be referred as the Visser (2006)’s symbolic form of external cognitive artifact.

Looking the importance of shared visualization for SMM, a systematic review study of shared visualizations focusing on SMM has been conducted by Nor’ain and Siti Salwah (2015). This paper is aimed to understand what strategy or techniques being applied in shared visualization to achieve SMM, which include how these strategies and techniques are being used in shared visualization.

Next section describes user interaction in shared visualization technology.

5.3 User Interaction in Shared Visualisation Technology

Brodlie et al. (2004) provide four (4) aspects that relate to how users interact when participating in visualization systems. The four aspects are:

  • Joining/leaving – shared visualization systems should have a facility to allow users to join and leave at any time.

  • Floor control – Shared visualization systems should offer different levels of access to a session for individual users such as allowing editing, or sharing or both editing and sharing authority.

  • Privacy – Shared visualization systems should allow users to work privately and at the same time, still remain in the conference. This is to protect some information and at the sometime can share other information.

  • Global view - Shared visualization systems should be able to allow users to view the network editor of other users in order to reassure that they understand what other user is doing.

6 Future Research Directions

This review discusses the fundamental understanding of SMM processing in visualization technologies. There are number of areas that can benefit from the SMM approach, which is briefly discuss in this section.

6.1 Shared Visualisation in Big Data and Visual Analytics

For big data processing, SMM can be explored on how it should work effectively and efficiently in both distributed and collaborative visualization. According to Brodlie et al. (2004), distributed visualization has some resource allocation problems such as location of processing close to data for minimizing data traffic. One example of an enabling technology is cloud computing and the link to web services could provide better enhancement for distributed visualization in the any visualization technology system, such as close coupling of simulations and visualizations in a real-time, interactive steering environment. A model of visual management system has been devised to support Lean Production in construction sites (Valente et al. 2018). Big data visualisation is needed to support intuitive tools as well as practices involved in very large and changing construction environments, teams and equipment that are often spread in large areas. This work shows that shared visualisation is needed to manage wide construction sites where there is a gap of visual languages, design, infrastructure, and mechanics of visual perception among the construction workers (Valente et al. 2016).

According to Alharthi (2016), a selected team can be visually transformed to enhance collaborative ideas for a specific purpose to enhance mental and overall capabilities as well as to maintain high performance of task and reduces likelihood of failure on a work mission. Mixed-initiative visual analytics system (MIVAS) is another work conducted by Makonin et al. (2016). The conceptual architecture of MIVAS consists of five (5) key components that support data wrangling, alternative discovery and comparison, parametric interaction, history tracking and exploration, system agency and adaptation to support human-human collaboration, multi-user collaboration as well as autonomous revisit system solution.

Work by Seipp et al. (2019) shows that combining information visualization in data mining techniques is an interactive decision-making process that can offer a visual guide to communicate uncertainty. This GeoVisual Analytics should assist the analysis of increasing availability and complexity of geographical information. A review by Chapeton et al. (2018) shows that GeoVisual analytics support visualisation sharing in hybrid collaborative scenarios, cross-device collaboration, time-critical and long-term analysis. Similar work in geographical domain is studied by Ruda (2015) about cartography visualisation to obtain precise results for spatial decision making.

6.2 Shared Visualisation in Multimedia Interface

Another area that requires great attention is involving the user interfaces design construction and development. For example, SMM processing can be increased using visualization and multimedia output capabilities through sophisticated multimodal interaction. According to Oviatt and Cohen (2000), multimodal input facility in a system could give more powerful interfaces for the user to access and manipulate information. Example of future work may include designing multimodal inputs such as speech and handwritten recognition from the user. These recognition techniques should be able to read, interpret and translate integrated data inputs in a form of visualization can provide better multimodal interaction facility in the visualization technology system. Redlich et al. (2017) found that creative virtual tools enhance SMM but still lack in perceived efficiency compared to physically present teamwork. It is suggested that to increase the quality, efficiency and satisfaction of virtual creative processes (such as in virtual agent, chatbot, live chat), further investigation on the usage of information communication technology (ICT), multimedia and virtual tools need to be conducted.

6.3 Shared Visualisation for Mobility

On the other hand, the support for SMM processing in the visualization technology system can be extended through the use of mobile-based application with particular interest to the shared visualization data. For example, the capability of the framework can be enhanced by allowing both desktop and mobile clients to simultaneously visualize the same data visualization in sharing a common view. A research contribution in this area has been done by Craig and Huang (2015). Interactive animated mobile information visualization in “Mobile Tree Browser” is developed to browse labelled hierarchies on mobile devices. The display is able to optimize readability and navigation on devices with limited space. This browser can be used to support multi-device co-located synchronous collaboration using animation to smooth the transition between views (Craig 2015; Craig et al. 2015).

6.4 Shared Visualisation for Disability, Awareness and Learning Analytics

In terms of multi-users participations, SMM can be further explored by incorporating different types of users such as those with certain kind of disabilities. Krishna et al. (2010) stated that social situational awareness is one of the important aspects for new technologies to enrich the social presence among remote disability users. For visually impaired users, portable and wearable interfaces can provide better access to non-verbal social cues through non-visualization medium. Specific research can be done to identify techniques that provide accessible SMM processing among different types of disabilities.

A shared route navigation that combines mobile devices with large display interfaces is designed to allow metro system navigation more convenient and at the same time, allowing a passenger to plan their route in the station using a large-display interface and follow the route using notifications on a smart-watch or similar wearable device (Craig and Liu 2019). This alternative view coordination methods that link smart-phone and large-display multi-device environments is a pervasive approach to facilitate different degrees of autonomous and collaborative working (Craig and Liu 2018). A Multi-Dimensional Visualization is an awareness visualisation system that reveals the level of perception of students towards learning of innovative skills in university (Muraina and Ibrahim 2016). This awareness system focuses on bringing the perception of students towards the learning of innovative skills into reality prior to the commencement of teaching.

Learning analytics is another specialized work in SMM to support education. Middleton (2020) examines the constructed visualization formats of small groups in a library learning commons. This study generates a hybrid and ubiquitous approach to integrating multiple visualization modalities within spatial settings in order to understand how visual and social affordances of group generated visualization formats. Liu and Nesbit (2020) study that visual dashboard facilitate group awareness, shared mental models and group cognition, and in turn fosters effective teaching and learning strategies.

As a conclusion, SMM provides a powerful predictive and explanatory for understanding the interaction in a common way. In order to interact with the world, people form unique share representations or shared mental models with which they can interact to perform better task and work performance.