Architectural design creativity in Multi-User Virtual Environment: A comparative analysis between remote collaboration media

Abstract

Previous studies have argued that a Multi-User Virtual Environment has the potential to foster creative collaboration, but empirical studies that examine this claim are rare. With a focus on architectural design, this study has investigated the affordance of Multi-User Virtual Environment for the production of novel and appropriate solutions in remote collaborative environments. Forty-four participants produced design solutions using Multi-User Virtual Environment and online sketching in remote collaboration. Four expert-judges assessed the novelty and appropriateness of the collaborative results, following the Consensual Assessment Technique. The results analyzed by paired samples T-test indicated that in remote collaboration, Multi-User Virtual Environment’s scores for both novelty and appropriateness were partially higher than those of online sketching. In remote collaboration, the immersion by avatars and co-presence with others in Multi-User Virtual Environment were perhaps more effective due to explicit communication cues used to share spatial information and collaborative procedure, compared to the two-dimensional, static representation of online sketching.

Keywords

Creativity remote collaboration Multi-User Virtual Environment architectural design

Introduction

There is a growing need for creative collaboration in architectural design due to rapid urbanization, depletion of natural resources, and growing awareness of the effects of human actions on the environment. New Media, particularly Multi-User Virtual Environments (MUVEs), have been hypothesized in recent years as having the potential to enhance such collaboration. Nonetheless, this hypothesis rests mostly on theoretical investigations rather than empirical research. The present paper documents our empirical investigation of the affordance of MUVE in the production of creative solutions in collaboration. In a comparative experimental setting, we have assessed MUVE against the online sketching medium, the latter being the most prevailing synchronous collaborative medium in remote collaborative environments (e.g., Gu et al., 2011; Kristensson and Norlander, 2003). The comparative investigation enabled us to comprehend enablers and barriers of MUVE in the particular collaborative environments.

This paper comprises three main sections. The first section starts with a theoretical investigation of the notion of creative collaboration. We break it into its two components, creativity and collaboration, and analyze them independently. We then discuss the affordances of MUVE, as an architectural design medium, and the ways in which it can open up new possibilities for creative collaboration. The second section documents our experimental procedure, participants, and results. In the last section, we interpret the results put forward by the experiment in light of our earlier theoretical discussion.

Creative collaboration and communication media

The need for collaboration arises when the limits of their individual abilities prevent architectural designers from completing a given task on their own (due to the lack of knowledge, power, or resources) or when collaboration can help them complete the task more quickly and more effectively. In addition to helping individuals undertake larger and more complex tasks, a collaborative arrangement helps them gain new perspectives on the shared enterprise, which they would not have been able to perceive on their own. It enables them to learn from others, and to be motivated by them. Collaboration, as such, is an enabling force. At the same time, it can also be a restrictive force, in the sense that the action best suited to satisfy the goals and needs of one designer may not also be best-suited to satisfy the goals and needs of another collaborator, thereby raising potential conflicts and the need to compromise or even to yield to the will of others. The difficulty of collaboration can also increase as collaborators most often are spatially—and sometimes temporally—distributed and collaboration must be undertaken remotely. This difficulty is especially apparent when collaboration is multi-disciplinary or multinational in nature, or both. This paper, nevertheless, focuses on remote synchronous collaboration among architects.

Collaboration is a complex, interpersonal synergetic process (Kvan, 2000). The ease of communication can be seen to provide a fundamental component for a collaborative design process, without which no collaboration is possible (Sonnenwald, 1996). Communication is a means to share different opinions and world views among collaborators, to build up a shared mental model, and to manage conflicts (Sonnenburg, 2004). In architectural design collaboration, miscommunication arises from the abstraction and ambiguity of visual representation in particular, which conveys the information that describes three-dimensional buildings (Cuff, 1991). Participation of varying professionals who have their own specific worldviews and mental models causes arbitrary interpretation of visual representation (Kalay, 1997).

Creativity, one of many goals of collaboration in design, has been defined as an ability to produce products that are both novel (original, or possessing a new quality which people have not experienced before) and appropriate (useful and adaptive to the constraints of the task) (e.g., Amabile, 1996; Kaufman et al., 2008; Paletz and Peng, 2008; Plucker and Makel, 2010; Runco and Jaeger, 2012).

Extending Amabile’s (1996) definition of creativity, creative collaboration can be defined as a process of shared creation where the exchange of ideas among the participants helps to stimulate and enrich their own creativity to the extent that the solution they arrive at is novel and appropriate (Kalay, 2004; Schrage, 1995).

Previous studies stated that the unique mechanisms of collaboration can promote creativity by reflective feedbacks (Schön, 1983), generating ideas (Osborn, 1963), stimulating cognitive processes and divergent thinking (Badke-Schaub et al., 2010), and complementariness and interdependence (John-Steiner, 2006). In addition, collaboration also enables designers to resolve the given problems and synthesize novel solutions (Goldschmidt, 1995; Kalay, 2004).

Celebrated examples of creative collaboration include Picasso and Braque, who created Cubism through complementary and integrative collaboration, and Watson and Crick, who discovered the structure of the DNA molecule through collaborative exchanges between themselves (and indirectly with Rosalind Franklin), and some of the more recent architectural collaborations between Foster and Arup.

Yet, creative collaboration is challenging due to communication hurdles and conflicts. Collaborators have their own different knowledge, expertise, worldviews, and mental models developed through training, education, and social and cultural backgrounds (Jehn, 1997). While conflicts can stimulate divergent and cognitive perspectives to foster creativity (Badke-Schaub et al., 2010), they can also hamper task performance, collaborative process, and interpersonal relationship (Amason and Sapienza, 1997).

To benefit from the advantages of collaboration for creativity and conflict management, a process of communication based on shared understanding and equality among the collaborators is essential (Kratzer et al., 2004). The significance of the medium arises from the fact that it conveys a message from a sender to a receiver in a communication process (Shannon, 2001). The affordance of communication media, its perceived potential for action, enables or restricts collaborators to encapsulate and interpret information (Chastain et al., 2002).

In architectural design collaboration, the medium used should convey descriptions of three-dimensional forms and space, building performance, and usability to support users’ activities in buildings. The visual representation is a vehicle of communication in architectural design collaboration. Visual representation needs appropriate levels of abstraction to exchange essential information (Kalay, 2004). When a represented output is too abstract, and at the same time collaborators do not have appropriate knowledge to interpret the representation, miscommunication happens (Cuff, 1991), and it hampers collaboration performances. Varying representation media, for instance, sketching, study models, and rendered images, are used for communication in architectural design collaboration. However, using these media, participants have to interpret and re-organize the represented information in their minds. Miscommunication is often caused due to arbitrary interpretation.

Potential of MUVEs for creative collaboration

New Media, particularly MUVEs, have been assumed in recent years as having the potential to offer a possible solution to resolve the communication barriers and thus enhance the production of creative solutions in collaboration (Burkhardt and Lubart, 2010; Chastain et al., 2002; Uribe Larach and Cabra, 2010; Ward and Sonneborn, 2009). An online MUVE is a new form of communication, made possible through Web 2.0 technologies: an alternative to physical co-presence. Once the province of science fiction writers (like Neil Stephenson’s Snow Crash, 1992 and William Gibson’s Neuromancer, 1984), thanks to advances in computing and telecommunication, MUVEs have become an increasingly relevant, real, and economically viable reality. MUVE was applied to re-creations of physical places and culturally significant human experiences that have existed in the past, like 1950s West Oakland (Kalay and Grabowicz, 2006), or Medieval Cairo (AlSayyad, 1999); or future places, yet to be built—the sine qua non of architectural practice. The technology was also applied to create imaginary places, in the form of games like World of Warcraft (by Blizzard Entertainment) and Lineage (by NC soft), or alternative realities like Second Life (by Linden Lab).

As a multi-user 3D environment, MUVE provides an information-rich medium for creative collaboration. Previous Computer Supported Cooperative Work research shows that MUVE provides an immersion that produces the feeling of “being in the design environment” (Riva and Ijsselsteijn, 2003). The immersion and presence in MUVE is initiated by a first- or third-person view in a 3D surrounding environment (Slater and Wilbur, 1997), where people can socialize much like they do in physical space, albeit represented through proxies—anthropomorphic representations called “avatars.” In addition, in MUVE, users can be aware of other users’ presence and interact with them (Harrison and Dourish, 1996). The presence of others through their avatars and the ability to interact with them produce the feeling of co-presence (Riva and Ijsselsteijn, 2003). Co-presence may create unexpected, serendipitous events, which in turn may inspire new ideas.

Creativity research also shows that co-presence and shared objects in MUVE provide divergent stimuli and thus perhaps inspire creativity (Ward and Sonneborn, 2009). Burkhardt and Lubart (2010) note that “MUVE represents heterogeneous constructs, emphasizing specific properties of the stimuli/system considered at one extreme and characteristics of the psychological experience of users at the other.” At this point, the previous studies seem to benefit from David Canter’s (1977) construction of “place” as the interaction of physical attributes, actions, and conceptions. A successful MUVE is thus a digital place that replaces physical attributes in conventional environments with shared digital objects and actions with co-presence (Dourish, 2006). In an empirical study, Uribe Larach and Cabra (2010) investigate the potential of MUVE for groups’ brainstorming in teleconferencing. They argue that the characteristics of MUVE, immersion and presence of avatars, synchronous and shared objects, recreate environmental atmosphere and stimuli in a physical meeting place. Thus, these characteristics help the participants feel psychological affiliation and proximity regardless of the remote distance. The authors argue that MUVE can be a successful alternative place for creative collaboration.

In addition, previous design studies also examined the effects of MUVE on collaboration. Gu et al. (2011) compared habitual collaborative design protocols, using (1) face-to-face sketching; with (2) online sketching in remote collaboration, and in (3) MUVE. They found that there were no significant differences among those three tools in terms of communication content and operations. However communication in face-to-face sketching was higher than in MUVE and online sketching.

In another study, Koutsabasis et al. (2012) investigated three authentic design cases using MUVE. Using qualitative methods, they evaluated design cases in terms of the quality of communication, situation awareness, and problem-based collaborative learning. They concluded that while designers were immersed in MUVE, they produced more complex outputs by adding new content and by instantly manipulating and arranging 3D objects. In addition, they argued that MUVE effectively supported design review and customer-centered evaluation of conceptual design, active communication, and awareness of others within the collaborative process.

Dossick et al. (2012) argued empirically that some of MUVE characteristics, including the flexible, active, and informal shared visualization support mutual discovery, critical engagement, knowledge exchange, and synthesis in remote collaboration among architecture, engineering, and construction teams. In a later work, they argued that because MUVE affords verbally mediated shared visualization that are mutable in real time, it supports real time joint problem-solving in remote collaboration (Dossick et al., 2015).

However, while these precedents show that MUVE can potentially support creative collaboration, they did not investigate the relationship between the characteristics of MUVE and the psychometrics of creativity, novelty and appropriateness. These studies focused mainly on the brainstorming, information processing, observable behaviors, and interface operations in MUVE.

Indeed, the previous studies concentrated more on task performances and information processing in a collaborative process rather than the creativity of collaborative results. Such potential of MUVE to assist collaborative creativity rests mostly on theoretical assumptions, and thus, in the present paper, we have empirically investigated the affordance of MUVE in the production of novel and appropriate solutions. We assumed that the immersion of collaborators through their avatars and their co-presence with other collaborators in a shared MUVE have the potential to foster creative collaboration in architectural design.

To investigate the research topic, we have compared MUVE with online sketching media, the most prevailing synchronous collaborative medium in remote collaborative environments (e.g., Gu et al., 2011; Kristensson and Norlander, 2003). Sketching is not simply the act of visually representing architectural products. Sketching is an activity that involves a dialogue between the designer and himself, or the designer and a collaborator, in which design ideation keeps evolving as more sketches are produced (Goldschmidt, 1991; Schön, 1983). This dialogue and its accompanying verbalization—the communication aspect of sketching—are more pertinent to designers than the quality of their sketching skills (Jonson, 2005; Menezes and Lawson, 2006). Depending on the available hardware and software, online sketching may or may not afford skillful sketching. Yet, almost all online sketching services afford verbalization of design intents through text and voice communications. As such, according to Gu et al. (2011), there are insignificant differences between sketching and online sketching in design collaboration environments in terms of communication content and operations on external representations.

The comparative investigation between MUVE and online sketching enables us to comprehend enablers and barriers of MUVE in the particular collaborative environments. The potential of MUVE on the production of creative solutions has not been examined using empirical research settings (Burkhardt and Lubart, 2010). We take creativity as one the most prized qualities of architectural design.

Research questions

The present study aims to investigate the affordance of MUVE for the production of novel and appropriate design solutions in remote collaborative environments. Previous studies assumed that MUVE has the potential to support remote collaboration (Gu et al., 2011; Uribe Larach and Cabra, 2010). Deduced from the previous studies, this paper also presumes that the immersion by avatars, co-presence, and shared objects in MUVE perhaps can help interpret visual representations and thus enhance explicit communication regardless whether designer are co-located or apart. The importance of such explicit communication cues of MUVE rests on that they can be a potential means to support nature of creative collaboration between architects. The challenges of architects’ collaboration arise from that potential design solutions are inherently numerous and diverse, while every architect has her own unique way of problem-solving (Rittel and Webber, 1973). Therefore, in-depth interpretation of collaborators’ ways of problem-solving and mutual co-examination of potential solutions are essential to propose a fully matured, converged collaboration result that all participants’ creativity is melted into.

In theoretical deduction, the immersion by virtual body of avatars is a direct means of perceiving and examining physical and functional properties of design solutions proposed by collaborators. Such examination and evaluation possibly support in-depth interpretation of the aims and methods of collaborators’ problem-solving and inspire further unexpected design approaches. Co-presence, represented by explicit body actions of avatars, also enables the awareness of collaborators’ design processes and thus can also stimulate divergent and novel ways of problem-solving. In addition, co-examination, using the presence of self and co-presence, possibly promotes the opportunities to discover errors of design performances, and thus can enhance appropriateness of design solution.

This study hypothesizes that MUVE, featured by the immersion of avatars and co-presence in a shared 3D environment, promotes the novelty and appropriateness of final design solutions, produced in remote collaboration between architects. To examine the hypothesis, this paper compares MUVE and online sketching media. The comparison between these two synchronous collaborative media is intended to investigate ways in which the immersion and co-presence in MUVE can foster or hinder the production of creative solutions, compared to the static and non-immersive representation of sketching when computer-mediated, restricted communication cues are shared. Thus, in this study, independent variables are (1) using MUVE and (2) online sketching media. The presence of avatars and co-presence in a shared 3D environment are the main difference between the two independent variables (Table 1). In addition, dependent variables are (1) novelty and (2) appropriateness of collaboration results, measured by experts. The experiment and its controlled variables are listed in Table 2.

Table 1.

Communication cues, representation, presence, and time using MUVE and sketching media.

Comparison	Remote collaboration
Comparison	MUVE	Sketching
Communication cues	Avatar	Verbal and Written
Communication cues	Verbal and Written	Verbal and Written
Representation	3D objects (immersive)	2D lines and shapes (static, non-immersive)
Presence	Restricted, mediated	Restricted, mediated
Presence	Presence	Presence
Time	Synchronous	Synchronous

MUVE: Multi-User Virtual Environment.

Table 2.

Experiment and controlled variables for creativity assessment.

Design tools (independent variables)	Novelty and appropriateness of collaboration solution (dependent variables)	Controlled variables
Online sketching (group board) MUVE (second life)	Exterior Form 1.1 Exterior Form to support users’ activities 1.2 Exterior Form to support building performances 1.3 Exterior Form as a landmark Interior Space 2.1 Interior space to support users’ activities 2.2 Interior space to support building performances	Participants’ skills to use tools Participants’ design ability that generate creative design solutions Participants’ academic levels Complexity of design tasks Task swapping to prevent affection of prior task to later task Individuals’ solutions were required at the beginning of collaboration A converged solution was required at the end of collaboration

MUVE: Multi-User Virtual Environment.

The hypothesis assumes that the presence and co-presence of avatars in MUVE make explicit the role of communication cues, regardless of remote distance. Such cues reduce inherent communication hurdles such as (1) misrepresenting and misinterpreting early design intentions, preferences, and prototypes; (2) misinterpreting physical properties and performances of design solutions represented by partners; (3) misinterpreting shared goals in converging different solutions, and (4) clear tasks distribution in manipulating and examining a final, collaborative solution. As a result, MUVE supports fluent co-creation and mutual examination for novel and appropriate solutions. We also plan to observe such inherent communication hurdles and ways in which the representation and communication cues in MUVE enhance collaboration to resolve and manage such challenges. We intend to use this observation to help interpret the results of the statistical analysis.

Method

Experiment procedure

To examine the above questions, this research designed comparative experiments comparing Second Life (SL), a popular commercial MUVE platform, to online freehand sketching in remote collaboration. Typical of most MUVEs, SL is a synthetic online 3D MUVE where users don anthropomorphic avatars to access the environment. SL allows users to communicate via written messages and audio (verbal) communication. It also allows users to command their avatars to perform gestures, communicating a limited vocabulary of body language. On the other side, Group Board (GB) is a commercial online blackboard environment. It affords sketching using a stylus and audio communication.

Participants were separated in different rooms. They were required to work together using SL and GB with voice and text chat. The design tasks were two street exhibition booths located in different sites but have similar difficulty and requirements. Participants started by designing an exhibition booth via SL, then after a 1-hour break, they designed the other exhibition booth via GB (Figure 1). Each design task required a maximum of 2 hours.

Figure 1.

Remote collaboration environment: MUVE.

At the beginning of collaboration, all participants were asked to propose their own solutions and analyzed and evaluated collaborators’ solutions. Later, they attempted to converge the individuals’ solutions and proposed a collaborative result. This experiment setting intended to stimulate active communication between the participants, including co-examination and co-manipulation. Related to the hypothesis of the present study, the immersion of the virtual body of the avatar and co-presence of MUVE, participants were inherently engaged in such communication. In addition, the experiment setting also planned to measure collaborative creativity, embedded into converged collaboration solutions, not individuals’ creativity.

In addition, we swapped the order of experiments. Half of the 22 teams started SL first, the other half started GB first. This swap was intended to regulate participants’ fatigue that is sometimes experienced by inexperienced users when working with new tools.

We observed and recorded each participant’s entire collaboration process, from individual’s ideation to final collaborative integration. We setup a video recorder behind each participant and also installed screen-capture software in order to record all events in MUVE and GB. In addition, whenever the participants made individual or collaborative decisions, represented in verbal communication and visual products, we noted the key events and took photographs. In particular, the observation noted in each medium (1) ways to understand and examine the partner’s opinions, (2) ways to manage conflicts and different opinions, (3) authentic communication hurdles, (4) ways to resolve such communication hurdles, (5) steps and attempts to converge different opinions in order to achieve a final result (such as co-creation, co-manipulation, and co-evaluation), (6) explicit products and communication contents when participants discover unexpected, useful solutions. After completing the experiments, we matched the observation notes to video recordings, and then used the qualitative data to interpret the results of the statistical analysis.

Participants

Twenty-eight senior undergraduate students and 16 graduate students in the Department of Architecture at the University of California (UC), Berkeley, participated in the experiments. Sixteen of them were female, and 28 of them were male. The age range of the undergraduate students was 21–23 (average 22.03), and the age range of the graduate students was 29–34 (average 29.56). They were teamed up at the same academic level in groups of two, and thus, the collaboration type is a paired partnership. We assumed that the paired partnership helps to increase activity of participants since passivity is increased by the increase in numbers of participants (Sonnenburg, 2004).

The range of the participants’ previous design studio grades was from 4.0 (A) to 3.5 (B) in order to regulate the individual ability to generate creative design solutions. The paired participants also had appropriate levels of intimacy because they were classmates or teammates in previous design projects. We supposed that the equivalent academic levels, grades, and intimacy diminish the pressure and dominance inside teams, caused by different competences, and encourage active participation in collaboration. This method can also reduce the anxiety of female students, teamed with male students. In other words, since our research is focused on creativity in collaboration and not conflict resolution, we introduced the moderating effect of shared identity and shared context of participants to conflict in remote collaboration (Hinds and Mortensen, 2005).

While the paired participants knew each other, each two given design tasks were inherently new and unique, and the participants held heterogeneous approaches to problem-solving and figured out diverse design solutions. Therefore, the participants developed and modified their previous design approaches in order to suit the local site conditions of design tasks and collaborators’ proposals, and thus, they were independent from their previous experiences with collaborators. This setting was intended to prevent developing shared design language and problem-solving patterns. In addition, until completing all sets of the experiments, the participants were prohibited from meeting each other in a face-to-face condition. The participants were trained to use SL, either in a prior class or in two-week tutorial sessions with assignments for practice. Previous studies noted that, for unskilled participants, tool operation represented a cognitive load, which hampers creative performances (Kristensson and Norlander, 2003; Uribe Larach and Cabra, 2010).

Design tasks

We asked participants to design a bus stop and an exhibition booth. Both tasks had different site conditions, but similar complexity and requirements. Two design studio instructors at UC Berkeley controlled the difficulty and complexity of the two tasks. In interviews, all 20 pairs of participants reported that any particular task was neither difficult nor easy compared to the other. In the experiments, the participants used site photographs for both MUVE and sketching tasks. Additionally, they used a 3D site model with adjacent buildings and street elements in SL, for the MUVE task. The solution, proposed in the prior task, could not be applied for the later task because the given local site conditions and requirements were heterogeneous, and thus the participants used diverse and different design approaches. The participants had to experience and manage new ways of collaborative problem-solving accordingly.

Creativity assessments

In the field of psychology, creativity can be assessed by experts’ judgments of the creative product, using Consensual Assessment Technique (CAT) (Amabile, 1996). While creativity stems in the creative person and the process, ultimately the actual product (e.g., poems, buildings, paintings, any tangible responses, etc.) reveals an inter-subjectively objective quality regardless of any creative mechanisms and theoretical interpretations. In CAT, like any evaluation systems in the real-world (e.g., Nobel Prize, Pritzker Prize), expert judges, who are familiar with the domain in which the product is created, assess the novelty and appropriateness of actual products and respond independently. They tend to agree on which products are highly creative and which are not. Thus, CAT mimics the social system in which creativity is recognized and adopted by others. Since tangible architectural design products, from early sketches to actual buildings, represent architects’ creativity, previous architectural creativity research also adopts CAT (e.g., Styhre and Gluch, 2009). In this present study, we examined the influence of media on these two psychometric components of creativity, novelty and appropriateness, rather than the sum or multiplication of the components to produce creativity scores.

Four design studio teachers in the Department of Architecture at the UC Berkeley independently assessed each 10 results of the students’ designs, produced from collaboration. The judges had average eighteen years education experience and professional practice, and they also reported that they did not have biased preferences to a particular type of media. The evaluation materials contained sketches and SL images of the final products. The total numbers of cases n per each design tool was 44. The judges were asked to assess the novelty and appropriateness of the exterior form, and the interior space. The creativity evaluation criteria referred to the exterior form and interior space of buildings, which are classification criteria used to comprehend the major changes of spatial conception in the history of western architecture (Giedion, 1982). To provide measurable concepts to the judges, the evaluation criteria were specified to the exterior form and interior space to support the behavior of the inhabitants or users occupying a building, building performances (e.g., light, ventilation, and sound transmission), and social aspects (e.g., a landmark) for exterior form. These criteria referred to Horst Rittel’s evaluation factors in order to measure building performances (Rittel et al., 1966).

The evaluation scores ranged from −6 (very banal or very inappropriate) to +6 (very novel or very appropriate). We asked the judges to focus only on the evaluation for novelty and appropriateness of products, rather than the quality of the representation related to a particular design medium. The consistency of the judges’ assessments was calculated using Cronbach’s alpha, which yielded an acceptable reliability coefficient α of 0.956 for the novelty scores and 0.933 for the appropriateness scores in face to face collaboration, 0.920 for the novelty scores and 0.945 for the appropriateness scores in remote collaboration (Kaufman et al., 2008).

The four judges’ novelty and appropriateness evaluations were analyzed by paired samples T-test, which compared the mean scores achieved for the two media used. It confirmed the statistical significance of the mean difference between the two media via testing the p value, which indicates the statistical probability distribution of the sample.

Results

The statistical analysis of the judges’ novelty assessments indicates the following results:

MUVE’s novelty scores for exterior form to support users’ activities and to support building performances were not statistically different from those of online sketching. However, the scores of MUVEs for exterior form to be a landmark were meaningfully higher than those of online sketching (Table 3).

MUVE’s novelty scores for interior space to support users’ activities and to support building performances were not statistically different from those using online sketching (Table 3).

Table 3.

Assessments for novelty and appropriateness.

Factors	M (SD)
Factors	MUVE	Sketching	Mean difference	P
Exterior form to support users’ activities
Novelty	0.79 (2.46)	0.13 (2.80)	0.65	0.08
Appropriateness	0.63 (2.85)	−0.27 (2.94)	0.90	0.04^*
Exterior form to support building performances
Novelty	0.41 (2.56)	−0.14 (2.70)	0.55	0.23
Appropriateness	0.32 (2.90)	−0.25 (2.91)	0.58	0.17
Exterior form as a landmark
Novelty	−0.03 (2.79)	−0.94 (2.79)	0.90	0.04^*
Appropriateness	−0.27 (2.68)	−0.75 (3.01)	0.47	0.29
Interior space to support users’ activities
Novelty	−0.36 (2.38) 0.42	−0.85 (2.50)	0.49	0.24
Appropriateness	(2.67)	−0.45 (2.52)	0.87	0.00^**
Interior space to support building performances
Novelty	−0.45 (2.00)	−0.92 (2.43)	0.47	0.23
Appropriateness	0.18 (2.58)	−0.66 (2.52)	0.85	0.00^**

MUVE: Multi-User Virtual Environment. *p < 0.05 (95%), **p < 0.01 (99%).

In addition, the statistical analysis of the judges’ appropriateness assessments indicates the following results:

MUVE’s appropriateness scores for exterior form to support users’ activities were higher than those of online sketching, but MUVE’s appropriateness scores for exterior form to support building performances, and to be a landmark, were not statistically different from those of online sketching (Table 3).

MUVE’s appropriateness scores for interior space to support users’ activities and to support building performances were meaningfully higher than from those of online sketching (Table 3).

Discussion and conclusion

The affordance of MUVE for novelty in remote collaboration

First, as one result of statistical analysis, in remote collaboration MUVE helps participants promote the novelty of exterior form to be a landmark. We interpret that the reasons stem in the explicit and mutual communication cues, represented by the virtual body of the anthropomorphic avatars. Observations of communication in MUVE, at the early stage of collaboration, show that participants created simple prototypes independently, and they examined partners’ shape compositions and compared how they are different from their own (Figure 2). Such mutual but independent examinations happened just after—or even during—creation, using a first or third person view and animated body actions of the avatars. In addition, when participants attempted to converge divergent design solutions, participants used to refer to partners’ design styles using the perception of avatars and tried to mix them with their own design preferences (Figure 3).

Figure 2.

Participant Y walked inside the cone-shaped exhibition booth prototype, proposed by partner S. Vice versa S also observed Y’s stacked wall. They examined partner’s early prototypes of design mutually.

Figure 3.

Participant G referred to the curvature of exhibition structure, proposed by partner S, and then he attempted to propose the atypical form of benches that resemble with his partners’ design preference.

Therefore, such body perception of avatars reduced participants’ cognitive loads to interpret and refer to partners’ prototypes regardless of the remote distance, and thus perhaps facilitated more attempts in the search for unexpected forms. In particular, exterior form compositions (landmark or sculptural objects) were easy to perceive and examine in the avatars’ pedestrian mode.

Previous studies reported that the awareness of the other’s design processes, relevant to the synchronous co-presence with others, probably inspired unexpected solutions for proposing novel forms via partners’ activity awareness (Carroll et al., 2003), and visual stimuli created by other collaborators (Ward and Sonneborn, 2009). The co-presence of other collaborators also probably allowed them to manipulate the forms together and share the collaborative processes explicitly like in the real world (Uribe Larach and Cabra, 2010).

In online sketching, the lack of communication cues caused time constraints and cognitive loads. As a result, participants were seemingly reluctant in their search for unexpected form compositions. While previous communication media studies pointed out that the body expressions of collaborators (e.g., hand gestures and facial expressions) have an important role in providing communication cues in face-to-face collaboration mode (Clark and Brennan, 1991; Olson and Olson, 2000), in online sketching, body actions, like in MUVE, were not represented. Thus, in online sketching, it took the participants more time and effort to share information about three-dimensional forms and spaces than it did in MUVE. Otherwise, misinterpretation and miscommunication obstructed collaboration.

In online sketching, when participants proposed any new prototypes, they needed to explain intended meanings, qualities, and performances of the shapes, using verbal and written communication. In addition, to prevent misinterpretation, when one participant started drawing shapes, her partner used to stop individual explorations and focus on the ongoing sketching process. Therefore, mutual independent explorations were rarely observed. Indeed, to adapt such inherent time constraints and cognitive loads, participants composed simple shapes and focused mainly on design schemes (Figure 4). They were reluctant to propose and explore complex shape composition.

Figure 4.

Participants G and S planned the underground exhibition booth using grid modules. This proposal had a novel scheme for problem-solving, but its form and shape composition was rarely explored and developed.

Second, statistical analysis also indicates that MUVE and online sketching were not different in promoting novelty of interior form. We reason that each design medium has its own unique, and equivalent affordances that support particular types of tasks (Bonnardel and Zenasni, 2010; Chastain et al., 2002). In MUVE, body perception of avatars is a useful means to examine forms independently and mutually. When, however, forms did not fit in the range of avatars’ views and actions, participants failed to perceive them. Indeed, heterogeneous perceptions among collaborators’ avatars required additional efforts to set up objective, shared representations. On the other side, the static representation of online sketching, with detached view from representation, is an efficient means for objective, shared plans (e.g., a top-view of site or floor plan). Therefore, as observed empirically, participants added their opinions on their partners’ problem analysis and schemes, represented as diagrammatic representations such as the arrows representing wind, sunlight, and air circulation, and attempted to organize design elements (e.g., zones and rooms in top-view or elevation plans).

The affordance of MUVE for appropriateness in remote collaboration

First, the statistical analysis revealed that MUVE facilitated more searches for appropriate and useful solutions that support users’ activities and building performances than online sketching. We deduced that the reason stemmed from independent co-examination between collaborators. Such independent co-examination enabled participants to accumulate more evaluation data and complement each other’s design competence for useful and appropriate forms. In empirical observations, when participants focused on converging different solutions, they examined and used not-yet completed form compositions using their avatars’ body and perspective. Then, they discussed ways in which the solutions support each other’s activities. Initially, the immersive, experience-based evaluation by avatars was a means of assessing the usability of physical design properties. Such independent co-examination captured subtle performance data regarding participants exploring the behaviors of prospectus users. For instance, participants would use their avatars’ body to measure the optimal height of a seat and wheelchair access (Figure 5).

Figure 5.

Participant Y used his avatar to assess the height of the seat (left), and participants J and H assess the scale and distance of the passengers’ seats independently (right).

Second, in online sketching, participants could not experience the inside of their solutions, and thus they could not rely on pre-knowledge for co-reasoning the feasibility of their solutions. As discussed in the previous section, the lack of communication cues increases time and effort to interpret partners’ representations (Figure 6). Previous research shows that the static representation afforded by sketching allowed collaborators to easily evaluate their shared proposals back and forth compared to their previous steps of problem-solving, and helped the search for errors in the design processes (Oxman, 1990). However, such advantages of sketching are rooted in co-located situations that have enriched communication cues, such as gestures, facial expressions and body language. In online sketching, participants mainly focused on validity of their planning schemes, rather than the usefulness of physical design properties. When participants could not infer the performance of their partners’ proposals, they tended to accept their partners’ original schemes, and thus active co-examinations were rarely observed.

Figure 6.

Remote collaboration between R and Y in online sketching: Participant R struggled in interpreting the form of partner S’s walls, thus she repeatedly erased and redraw her roof part (the red lines on the right bottom corner in all images). This strenuous interpretation was possibly caused by online sketching’s static and non-immersive view, which describes partial and fragmented information about one completed three-dimensional form.

On the other hand, despite the remote distance, the immersion of the participants in MUVE through their avatars’ body actions and the explicit presence of other collaborators enabled the participants to evaluate the functionality of the proposed forms and space, to fit users’ activities and building performances. As, the previous studies argued, co-presence, represented by explicit actions and locations of avatars, promoted the feeling of being together, and thus facilitated proximity and intimacy among collaborators in remote collaboration in ways similar to the real world (Riva and Ijsselsteijn, 2003; Uribe Larach and Cabra, 2010). While not in the scope of our current study, in a real-world setting, the appropriateness of a design solution produced in MUVE can be challenged. As Bailey et al. (2012) highlighted, the verisimilitude of the design solutions that MUVE can afford can alter its users’ understanding of the empirical world. This can lead to erroneous design decision.

Limitations and future research

Our study only assessed the production of creative products, rather than other topics of creativity research such as the creative process and personality characteristics. To comprehend the mechanisms of MUVE to foster creativity, future study can investigate the relation between MUVE and the creative process to bear creativity. Sonnenburg (2004) introduced the eight phases of the creative process in collaboration: problem-finding, problem acceptance, preparation, incubation, illumination, verification, modification, solution. Oxman (1990) also stated that prior knowledge, based on dynamic and episodic memory, is a primary resource for a solution synthesis, relevant to the illumination phase, and evaluation, as the verification phase. MUVE’s experience-based evaluation by avatars’ actions can perhaps promote these phases of the creative process compared to the mere knowledge-based reasoning of non-immersive media.

In addition, the present study only focused on architectural design, and thus, it is unknown whether the results can be generalized for the creative collaboration in other disciplines, or not. Future studies can investigate other fields of design and planning in order to confirm the results of this study. In order to investigate the effects of the immersion and co-presence by avatars in depth, future studies also need to compare MUVE to other types of synchronous 3D modeling tools, but without the presence of avatars.

Implication

Despite the above limitations, this present study demonstrated that the affordance of MUVE to foster architectural design creativity in remote collaboration, using empirical comparative experiments, and proposed several new findings, not yet reported from other research. Based on the findings, we proposed several guidelines to use MUVE for collaborative design. First, MUVE is adequate to be used in ergonomic, body-scaled design projects aiming to support users’ activities and relevant building performances. Second, MUVE is also applicable for design collaboration in search for highly novel formed design products. Both ergonomics and highly novel design collaborations require inductive experimentations, mutual co-examinations, and explicit communications. As this present study showed, the immersion by avatars’ body actions can help evaluate the physical properties of ergonomic design products, and thus architects can test and share the unknown usability of novel products with other collaborators explicitly, without distance constraints. Last, MUVE is an appropriate means for collective, qualitative assessment in remote design collaboration.

Footnotes

Acknowledgment

This study is based on the data collected during the PhD study of the main author, at the Department of Architecture, University of California, Berkeley. The authors appreciate the sincere support and advices of professors Galen Cranz, M. Paz Gutierrez, and Kaiping Peng, and the enthusiastic participation of the 44 students at the University of California, Berkeley.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (no. 2016R1C1B2011274).

Seung Wan Hong is an assistant professor at the Department of Architecture, Inha University, South Korea. He previously received Lady Davis postdoctoral fellowship at the Technion Israel Institute of Technology. Hong holds a PhD degree from the University of California, Berkeley. His research specialty focuses on the development and application of avatar-based virtual environment and human behavior simulation to promote the architectural design creativity and collaboration. He holds a 5-year research grant from the National Research Foundation of Korea, and also published more than 20 relevant articles in highly reputed journals.

Ahmed El Antably is assistant professor at the Department of Architectural Engineering and Environmental Design, the Arab Academy for Science, Technology and Maritime Transport (AASTMT). He received his PhD from the University of California, Berkeley, with a Designated Emphasis in New Media from the Berkeley Center for New Media. El Antably's research focus is on digital design, design media, virtuality and digital heritage. He published many scholarly papers and one book.

Yehuda E Kalay is professor of Architecture at the Technion, Israel Institute of Technology, where he holds the Henry and Marilyn Taub Academic Chair. He served as Dean of the faculty of Architecture and Town Planning from 2010 through 2016. He is also Professor Emeritus at the University of California, Berkeley, where for 18 years he served as professor of Architecture. Prior to his tenure at Berkeley, for 10 years he taught in the Department of Architecture at the State University of New York at Buffalo.

Kalay's research explores the implications and applications of digital technologies for architectural design, in particular as they pertain to evaluating the building performance in terms of the interactions between spaces, the people who inhabit them, and their impact on the activities they engage in, in particular in the field of hospital design. He has published more than 120 scholarly papers and authored or edited nine books.

References

AlSayyad

(1999) Virtual cairo: An urban historian’s view of computer simulation. Leonardo 32: 93–100.

Amabile

(1996) Creativity in Context: Update to the Social Psychology of Creativity, Boulder, CO: Westview Press.

Amason

Sapienza

(1997) The effects of top management team size and interaction norms on cognitive and affective conflict. Journal of Management 23: 495–516.

Badke-Schaub

Goldschmidt

Meijer

(2010) How does cognitive conflict in design teams support the development of creative ideas? Creativity and Innovation Management 19: 119–133.

Bailey

Leonardi

Barley

(2012) The lure of the virtual. Organization Science 23: 1485–1504.

Bonnardel

Zenasni

(2010) The impact of technology on creativity in design: An enhancement? Creativity and Innovation Management 19: 180–191.

Burkhardt

Lubart

(2010) Creativity in the age of emerging technology: Some issues and perspectives in 2010. Creativity and Innovation Management 19: 160–166.

Canter

(1977) The Psychology of Place, New York, NY: St. Martin’s Press.

Carroll

Neale

Isenhour

et al. (2003) Notification and awareness: Synchronizing task-oriented collaborative activity. International Journal of Human-Computer Studies 58: 605–632.

10.

Chastain

Kalay

Peri

(2002) Square peg in a round hole or horseless carriage? Reflections on the use of computing in architecture. Automation in Construction 11: 237–248.

11.

Clark

Brennan

(1991) Grounding in communication. In: Resnick

Levine

Teasley

(eds) Perspectives on Socially Shared Cognition, Washington, DC: American Psychological Association.

12.

Cuff

(1991) Architecture: The Story of Practice, Cambridge, MA: MIT Press.

13.

Dossick C, Anderson A, Iorio J, et al. (2012) Messy talk and mutual discovery: Exploring the necessary conditions for synthesis in virtual teams. In: Proceedings of the Engineering Project Organizations Conference, Rheden, the Netherlands.

14.

Dossick

Anderson

Azari

et al. (2015) Messy talk in virtual teams: Achieving knowledge synthesis through shared visualizations. Journal of Management in Engineering 31: A4014003, Available at: http://ascelibrary.org/doi/pdf/10.1061/(ASCE)ME.1943-5479.0000301.

15.

Dourish P (2006) Re-space-ing place: “Place” and “space” ten years on. In: Proceedings of the 2006 20th Anniversary Conference on Computer Supported Cooperative Work. Alberta, Canada: ACM, pp. 299–308.

16.

Giedion

(1982) Space, time and architecture: The growth of a new tradition, Cambridge, MA: Harvard University Press.

17.

Goldschmidt

(1991) The dialectics of sketching. Creativity Research Journal 4: 123–143.

18.

Goldschmidt

(1995) The designer as a team of one. Design Studies 16: 189–209.

19.

Kim

Maher

(2011) Technological advancements in synchronous collaboration: The effect of 3d virtual worlds and tangible user interfaces on architectural design. Automation in Construction 20: 270–278.

20.

Harrison S and Dourish P (1996) Re-place-ing space: The roles of place and space in collaborative systems. In: Proceedings of the 1996 ACM conference on Computer supported cooperative work. Boston, MA: ACM, pp. 67–76.

21.

Hinds

Mortensen

(2005) Understanding conflict in geographically distributed teams: The moderating effects of shared identity, shared context, and spontaneous communication. Organization Science 16: 290–307.

22.

Jehn

(1997) A qualitative analysis of conflict types and dimensions in organizational groups. Administrative Science Quarterly 42: 530–557.

23.

John-Steiner

(2006) Creative collaboration, Oxford, UK: Oxford University Press.

24.

Jonson

(2005) Design ideation: The conceptual sketch in the digital age. Design Studies 26: 613–624.

25.

Kalay YE (1997) P3: An integrated environment to support design collaboration. In: Proceedings of ACADIA’97 Conference, Cincinnati, OH, pp. 191–205.

26.

Kalay

(2004) Architecture’s New Media: Principles, Theories, and Methods of Computer-Aided Design, Cambridge, MA: MIT Press.

27.

Kalay

Grabowicz

(2006) Oakland blues: Virtual preservation of seventh street’s 1950s’ jazz scene. 3DVisA Bulletin. Available at: http://3dvisa.cch.kcl.ac.uk/paper_kalay.htm.

28.

Kaufman

Plucker

Baer

(2008) Essentials of Creativity Assessment, Hoboken, NJ: Wiley.

29.

Koutsabasis

Vosinakis

Malisova

et al. (2012) On the value of virtual worlds for collaborative design. Design Studies 33: 357–390.

30.

Kratzer

Leenders

OTAJ

van Engelen

JML

(2004) Stimulating the potential: Creative performance and communication in innovation teams. Creativity and Innovation Management 13: 63–71.

31.

Kristensson

Norlander

(2003) The creative product and the creative processes in virtual environments. Creativity and Innovation Management 12: 32–40.

32.

Kvan

(2000) Collaborative design: What is it? Automation in Construction 9: 409–415.

33.

Menezes

Lawson

(2006) How designers perceive sketches. Design Studies 27: 571–585.

34.

Olson

(2000) Distance matters. Human-Computer Interaction 15: 139–178.

35.

Osborn

(1963) Applied Imagination: Principles and Procedures of Creative Problem-Solving, New York, NY: Scribner.

36.

Oxman

(1990) Prior knowledge in design: A dynamic knowledge-based model of design and creativity. Design Studies 11: 17–28.

37.

Paletz

Peng

(2008) Implicit theories of creativity across cultures: Novelty and appropriateness in two product domains. Journal of Cross-Cultural Psychology 39: 286–302.

38.

Plucker

Makel

(2010) Assessment of creativity. In: Kaufman

Sternberg

(eds) The Cambridge Handbook of Creativity, Cambridge, MA: Cambridge University Press, pp. 48–73.

39.

Rittel

Sanoff

Fessenden

(1966) Evaluative Methods to Measure the Performance of Buildings, Berkeley, CA: Department of Architecture, College of Envirnmental Design, University of California.

40.

Rittel

HWJ

Webber

(1973) Dilemmas in a general theory of planning. Policy Sciences V4: 155–169.

41.

Riva

Ijsselsteijn

(2003) Being there: The experience of presence in mediated environments. In: Riva

Davide

Ijsselsteijn

(eds) Being There: Concepts, Effects and Measurements of User Presence in Synthetic Environments, Washington, DC: IOS Press, pp. 3–16.

42.

Runco

Jaeger

(2012) The standard definition of creativity. Creativity Research Journal 24: 92–96.

43.

Schön

(1983) The Reflective Practitioner: How Professionals Think in Action, New York: Basic Books.

44.

Schrage

(1995) No More Teams! Mastering the Dynamics of Creative Collaboration, New York: Currency Doubleday.

45.

Shannon

(2001) A mathematical theory of communication. ACM SIGMOBILE Mobile Computing and Communications Review 5: 3–55.

46.

Slater

Wilbur

(1997) A framework for immersive virtual environments (five): Speculations on the role of presence in virtual environments. Presence: Teleoperators and Virtual Environments 6: 603–616.

47.

Sonnenburg

(2004) Creativity in communication: A theoretical framework for collaborative product creation. Creativity and Innovation Management 13: 254–262.

48.

Sonnenwald

(1996) Communication roles that support collaboration during the design process. Design Studies 17: 277–301.

49.

Styhre

Gluch

(2009) Creativity and its discontents: Professional ideology and creativity in architect work. Creativity and Innovation Management 18: 224–233.

50.

Uribe Larach

Cabra

(2010) Creative problem solving in second life: An action research study. Creativity and Innovation Management 19: 167–179.

51.

Ward

Sonneborn

(2009) Creative expression in virtual worlds: Imitation, imagination, and individualized collaboration. Psychology of Aesthetics, Creativity, and the Arts 3: 211.