Implementing Affordance ‐ Based Design Review Method Using Virtual Reality in Architectural Design Studio

: In recent years, researchers in the architecture field have explored VR technology for var ‐ ious purposes in the architectural design process. Current efforts have supported the growing knowledge of VR utilization, specifically for the architectural design review process. In this study, we adapt the affordance concept from the ecological psychology study to add further evidence on VR utilization for the design review process. This study aims to develop an affordance ‐ based design review method in architectural design by utilizing immersive VR technology. It sits at the crossroads of architecture design, virtual reality, and affordance research. This project aimed to create an af ‐ fordance ‐ based review method framework, a VR system to support the method, test both the method and the VR system, and assess the VR system’s effectiveness as a companion system for the affordance ‐ based design method process. The study was conducted in the scope of architectural education settings only. It used a third ‐ year architectural design studio course as a case. This study confirms that the affordance ‐ based design review method using virtual reality helps students im ‐ prove their design work. The method reveals the presence of positive and negative affordances in their work. It also shows the differences between a student and supervisor in perceiving the af ‐ fordances for reviewing design works.


Introduction
Virtual reality (VR), as defined by Isdale [1], is an idea of how people "visualize, manipulate, and interact with computers and extremely complex data."It is formed as a three-dimensional spatial environment generated by a computer where a human participates inside it in real-time [2].VR is also a medium for simulating a new reality.People interact with three-dimensional objects inside it using their senses to feel immersed in the shaped reality [3][4][5].A virtual environment (VE) within VR is immersive when users feel fully involved inside it as if they were in an actual environment.This technological capability makes VR interesting to be explored further in the built environment, specifically in the scope of architectural design.Moreover, there is a new enthusiasm for the emergence of virtual reality technology, which is now easily accessible and more familiar among endusers.
space with rules and relationships that govern them; immersion or the sensation of being in the environment, including physical immersion, mental immersion, and a sense of presence; sensory feedback or feedback that can be felt directly by users with their senses; and interactivity that users can respond to through direct actions.As a technology, VR can produce an illusion of being in an environment acceptable to a user as a trusted place to exist with sufficient interactivity and perform various tasks efficiently and comfortably [4].
VR technology has the potential to assist architects in unexpected scenarios.It enables them to use the inhabitants' personalities as a starting point for design exploration [6] and conduct a usability assessment during conceptual design [7].Moreover, VR aids them in understanding how architects construct building façade geometry based on user reactions to lighting patterns [8], as well as providing a superior spatial perception to comprehend the architect's spatial arrangement [9].Designers can employ VR as a very successful aesthetic tool that allows them to manipulate space to elicit human emotional responses, according to Naz et al. [10].
From a technical perspective, VR has been proven to be a reliable and stable technology and helped them facilitate model exchange.They found that VR technology can extend students' learning processes and improve their motivation and awareness.VR provides "being there" with immersive interaction between students and their design works.It is vital because the architectural designer must consider behavior, cognitive outcomes, and users' subjective experiences when evaluating a building design using VR [11,12], primarily when focused on building usability [13].Tsou et al. [14] proposed an integrated system supporting architectural design education, which lets users explore, discuss, and manipulate 3D objects within the VE.Users operated the system using a VR controller, addressing many issues during the trial.
Wickens [15] highlighted that virtual reality is a concept "created by an impressive, exciting technology that readily engages the user's interest."With that concept, VR opens an opportunity for those involved in the education business to improve or expand their learning environment.Describing VR as an "instructional medium", Wickens saw some justification regarding the additional cost of the technology.VR could give a motivational value or show users a different or novel perspective.Additionally, there would be a "transfer of the learning environment" from physical to virtual, which a "natural" interface should support.VR technology also has a substantial capacity to be implemented for the design review process, especially in the architectural education setting.
Researchers are also looking into how VR technology is used in a professional context [16].VR, for example, aids persons who are unfamiliar with architectural representation in perceiving a correct design project [17].VR also helps architects visualize the relationship between design and thermal quality in different settings [18].Similarly, virtual reality (VR) can give delicate imagery for review sessions without requiring a trip to another location [19].Architects can also validate their designs by paying attention to how people react to specific sound settings in the space.VR provides architects with real-time feedback, helping them improve their designs and gain a more profound knowledge of the environments [20].Finally, VR may help designers make better project judgments [21] and better communication during meetings [22].

Theory of Affordances in Architecture
In his seminal work entitled "The Theory of Affordance", Gibson [23] described affordances as the relationship between the features offered by the environment and the ability of individual surroundings to take advantage of the environment.Therefore, affordances are neither in the individual nor in the environment.Humans, as an individual, recognize each affordance as a relational property deriving from the relationship between them.At the same time, stairs can also cause humans to fall or get injured.Thus, simultaneously, objects in the environment affordance both desirable (positive) and undesired (negative) affordance.Each desired and undesired affordance must be easily perceived from the individual's environment.
Inspired by Gibson's work, Norman [24] introduced affordances to product design and later inspired the field of human-interaction design as the set of action opportunities a product provides.With his seminal work entitled "The Psychology of Everyday Things", which was revised later as "The Design of Everyday Things", Norman [25] collected the anxieties that users face when interacting with objects they use every day.This situation was described [26] as an affordance-based error, representing the discontinuity between the designer's motive and the user's perception.
For experts and researchers in ecological psychology, the concept of affordance initiated by Norman is contrary to the idea presented by Gibson [27][28][29][30][31][32].Gibson claimed that affordance is independent of the experience and culture of the individual or user.In many cases, however, individual actions and interactions arguably presupposed the individual's previous experiences with the same environment.Unlike Gibson, Norman emphasized the individual's perceptual and mental abilities.In his book, Norman [29] conceptualizes affordances based on three behavioral constraints: physical, logical, and cultural.Physical constraints are closely related to real affordances, logical constraints use reasoning to determine the alternatives, and cultural conditions are based on conventions shared by a specific cultural group.In short, Gibson's affordance should be a direct one, while Norman's affordance needs a cognitive effort by an individual.
For architects, affordance provides an alternative way to view a design by emphasizing the relationship between the environment and its users.This relationship is similar between the building form and the end-users' behavior as part of the building function.Using an affordance-based approach, architectural designers put their design vision into account by identifying the affordances.
Pagano et al. [33] reviewed how designers should prefer the concept of affordances in the architecture dan design scene.The main goal of architectural design is to produce an artifact with which people can interact productively and safely.The scope of interaction between humans includes motor control and human perception.Relatively, the relationship between humans and their environment needs to be defined.Designers can do it quantitatively with the concept of affordances.It supports the definition of user interaction with objects and built space with minimal usage interpretation requirements.So, the design will become more inclusive since it allows the end-users to perceive the design without further cognitive elaboration and mental representation effort.
McGrenere and Ho [30] also stated that to achieve the purpose of a design, the object itself, an architect must define the required affordances that a design must supply.In addition, they should optimize information integrity by detailing affordances and the ease of interacting with them.Users can set the affordances for completing required tasks utilizing information such as the physical features of the design and artificial indicators.It corresponds to Gibson's idea of direct perception [34].It is in line with Gibson's concept of direct perception [23].To reduce users' cognitive efforts, the architect should have their designs' affordances perceived as quickly and directly as possible.They must eliminate the need to add signifiers or other information to direct users to the affordance's presence.
Koutamanis [27] went into greater detail on how the affordances of building components and places define how users engage with the design.Affordances in architecture, he claims, promise to handle functionality and usability in a solid, clear, and visible manner.Building component affordances are derived from defined functional and structural restrictions.The scale and user interaction with most items are comparable in developing component affordances.As a result, architects can intuitively cope with design issues.
In the meantime, spatial entities' affordances are obtained through the regular use of space and spatial conditions.In contrast to architectural components, spatial entities deviate from the generic example used in most affordances research.They provide a less physical form that allows anyone to map functions.As a result, structures should not need a handbook to explain how to use them, which is consistent with Gibson's definition of affordance as "direct perception."Despite cultural differences, building affordances should support the desired purposes in general.As a result, the affordances of their nature can be used to evaluate architectural design and refinement.Recognizing the spatial affordance of museum design [35] or maximizing domestic design [36] are two examples.

Problem Statement
We learned that the trend of VR adaption in architectural design has increased in recent years due to increased interest in VR technical capability.It lets architects and other design stakeholders do initial previews or perform initial reviews on the design itself.In the architecture education segment, adaptation efforts have been conducted by researchers and lecturers to use VR in the design studio, starting from the technical adaptation process [14,[37][38][39][40][41], pedagogy exploration [42], gamification [43,44], performing qualitative appraisal on the student work [45,46], collaborative learning [47], and visualizing construction process [48].The efforts supported the growing knowledge on VR utilization for improving architectural designs.This study could add additional evidence on how VR could help designers to perform the design review process, especially in educational settings [11,45].
We have also explored the study of affordances concept from ecological psychology study and found that researchers have explored the possibilities of affordance concept adoption in architectural design.It ranged from the analysis of affordances used in architecture [27] until the development of an affordance-based approach to architectural theory, design, and practice [49] that led to the affordance-based design (ABD) approach that offers a shift in design thinking from functional-driven to user-driven [50,51] and its operational method [52].After that, the adoption of ABD was performed by researchers to enhance the design process focused on the designers [26,32,53,54] and the design outputs [36,55,56].With that, the concept of affordance-especially the ABD-could be adopted as an approach for the design review method.Then, we looked through the study on using VR technology and the affordances approach in architectural design.The studies explored affordance perception in VR to help architects understand the effects of designs on users [57] and urban space assessment [58].Our review found insufficient research on the triangulation of these topics, specifically in the architectural education and design review process, which becomes our focus.

Research Objectives and Methodology
This research aimed to develop an affordance-based design review method in architectural design by utilizing IVR technology.This study focused on exploring the third prepositions of using affordances in architecture, as Maier et al. [49] proposed, which is the design review method using an affordance-based design approach.In this study, formulating a design review method framework using an affordance-based approach and developing an IVR application will be carried out.Adapting the concept of affordances as part of the architectural design process is not new [59][60][61].However, few studies have addressed affordance-based design for architecture, although this approach has been applied in various disciplines, such as engineering, industrial design, and human interface design [56].So, similar design mistakes can be avoided.
Furthermore, a trial implementation of the design review method framework and IVR applications was carried out in educational settings only.This was done to determine the effectiveness of the design review method in the architectural design review process.The trial was conducted in an ongoing third-year architectural design studio course, including the design iteration process that utilized the design review method.Thus, the effectiveness of the affordance-based design review method can be seen based on the results of architectural design iterations developed by the student.The student's supervisor used a similar design review method to obtain comparison results, and a further confirmation study was also carried out.We asked third-year Bachelor of Architecture students in Bandung, Indonesia, to participate in this research.As for the apparatus, we used an IVR head-mounted display (HMD) device due to its portability and ease of use.

Affordance-Based Design as Design Review Method
The framework of affordance-based design, established by Maier and Fadel [51], began with the basic notion of a structure's affordances, as we regarded a building as an artifact.The affordances were decided by the users' demands and the artifact's structure (spatial elements and building components).A design's success is determined by the existence and absence of positive and negative affordances.Simultaneously, the design process may identify and eradicate negative affordances from the building design.At the same time, it is also developing or maintaining the desired good affordances.The final design will be based on predicted affordances to enable desired user behavior while avoiding undesirable affordances.
The framework is represented by Maier and Fadel's Designers-Artifacts-User system (DAU system) [51].Artifact-user affordance (AUA) and artifact-artifact affordance (AAA) are two unique kinds of affordances suited for architectural and surrounding environmental aspects in the framework.The normal affordances of interest between objects and users [50] are what AUA stands for.The affordance represents a potential behavior, but it is not required to actualize that action.User behavior between the artifact and the user can occur that neither the user nor the artifact can exhibit on their own.The artifact, in this case, is the building design, and the users are the anticipated residents.AUA mapping aids in determining the possible use of structures as artifacts.Architects can describe AUA as a user-artifact interaction in which the artifact provides an affordance to the user.It is inextricably linked to Gibson's original notion of affordance, in which affordance always refers to a relationship between a person and his surroundings.
In the meantime, there are artifact-artifact affordances (AAA) between the two artifacts.These indicate that an artifact can conduct a behavior toward another artifact.Designers could figure out what kinds of behaviors the artifacts allowed.These behaviors must adhere to physical principles, of which architecture designers with a foundation of design knowledge should already be aware.Both structural beams and columns, for example, can hold the load.Furthermore, these affordance groups should consider direct perception for minimal cognitive processing.The groups should also consider the inherent units of artifacts (such as size and shape) when assessing the quality of affordances and examining the design's essential elements [55].
The architect defines the structure of the building components, their spatial entities, and the affordances they can offer, be it AUA or AAA.Architects should note that AUAs should disclose relationships that are directly beneficial to users, and AAAs should disclose relationships that are indirectly beneficial to building users.In turn, affordance determines how a system in an artifact behaves.Additionally, affordances should be intended to be perceived directly without any cognitive effort (direct perception) to increase their usability.However, as an architect, specific pieces of knowledge should be acquired first to define "intrinsic unit related to user characteristics and artifact's properties" to let anyone as a user have affordances with direct perceptions.
The affordances definition procedure in the affordance-based design framework necessitates architectural designers with experience or knowledge of the designer building's surroundings.It comprises information about what users may and cannot do in the building and what the building can and cannot do.The team utilized the affordance structure matrix [53] to determine whether a design contains the needed affordances to enable a specific behavior while avoiding individual behavior.Positive AUA (AUAP), negative AUA (AUAN), positive AAA (AAAP), and negative AAA (AAAN) are the four types of affordances in the ASM.In this study, we used a reduced version of the ASM, omitting the "roof" and "side" components because intradomain relationships were not considered, as seen in Figure 1.

Virtual Reality Design Reviewer (VRDR)
As part of the study, we created the virtual reality design reviewer, or VRDR, a gameengine-based VR system for analyzing architectural design studio outcomes, as shown in Figure 2. It was created to assist students in an architectural design studio course in understanding and evaluating their concepts.The system was created using the Unity game engine and is optimized for standalone VR head-mounted displays such as the Oculus Quest.We decided to use a game engine because it provides real-time 3D rendering, interactivity, and collision detection as its underlying strengths [62].VRDR was designed to run without requiring high-end personal computer (PC) specs.In a VE, VRDR allows the user to view design studio outcomes in a BIM model.We wanted to offer VRDR as a VR system for analyzing studio design solutions utilizing the affordance-based design approach, depending on the objectives assigned in the design brief.The VRDR system has been discussed as a whole separate discussion [63].

Case Study
The case study we utilized to create an affordance-based design review approach employing VRDR in an ongoing architectural design process was introduced in this part.The case study took place in the second semester of the third-year architectural design studio course at Institut Teknologi Bandung, Indonesia, as part of the Study Program for Bachelors in Architecture.The course required students to complete two design projects: lifestyle center (LC) and apartment (APT).The LC project was picked to facilitate students doing design exercises for building two to three stories in the sloping areas.Each project had its objectives and student performance criteria (SPC) that must be achieved by each student, which are described further in the following sections.
Meanwhile, the APT project facilitated students to design six to eight stories of buildings in the flat area in the urban context, focusing on building systems.The course focused on building functions, barrier-free design, building structure, and utility systems.The complexity is medium, and the design brief provided the spatial programs for students as a design guide.Both projects were situated in the city of Bandung (coordinates: 6°54′43″ S 107°36′35″ E), which is located 140 km southeast of Jakarta-the capital city of Indonesia-and 768 m above sea level.

First Project: Lifestyle Center (LC)
In the first project, students designed a lifestyle center (LC) building facility in Bandung.This project stressed the place-making issue by developing a safe and comfortable place for the people of Bandung with facilities such as retail spaces, convenience stores or minimarket, ATM center, cafe, restaurant, and supporting facilities.The overall site area was approximately ±6200 m 2 , including the site used for the second project, as seen in Figure 3. Furthermore, the course expected students to design a building through form and spatial composition and integrate them with site context, functional, and constructability aspects through this project.The course coordinator formulated studio objectivesderived from the goals of the curriculum, compiled in the second column of Table 1, and student performance criteria (SPC)-based on the accreditation quality standard that students must fulfill to help supervisors measure student achievements, further defined in Table 2.For final design submission, each student must submit a physical design study model on a scale of 1:200 using monochrome-colored materials.

OBJ1
Able to develop the architectural program, including site studies, user studies, and precedent studies, to formulate design goals and objectives, spatial programs, relationships between spaces, and initial ideas or concepts that can provide a picture of the success of the designed building.
Able to design an eight-floor midrise apartment with additional two floors of basement

OBJ2
Able to develop alternative spatial zoning on the site that guides the building design through the site analysis process, including building regulations, site context analysis, site access analysis, site topography analysis, and site potential analysis.
Able to design a site plan which considers the existing building mass composition and function, site circulation and accessibility, greening areas, and site context OBJ3 Able to develop schematic designs that show spatial organization and ideas in a 3D form.
Able to draft a schematic plan showing the layout of typical rooms, circulation area, shared spaces, and services area

OBJ4
Able to develop the form or spatial composition of a building according to aesthetic principles by considering the site context and integration with the spatial organization prepared previously.
Able to draft conceptual and schematic design in the form of building mass studies that follow the detailed urban design guideline OBJ5 Able to develop preliminary building designs by considering constructability and function ability.
Able to develop a preliminary design that integrates with the building systems, including circulation system, structural system, utilities, and facade system

SPC1
Able to formulate an architectural program that guides the functional design of the building.
Able to design a midrise building in an urban area with a building system approach SPC2 Able to understand the principles of visual aesthetics and apply them in the form of two-and three-dimensional architectural design.
Able to apply aesthetic principles in the building design, especially building form and facades

SPC3
Able to design architecture comprehensively based on environmental and sustainability aspects and utilize concepts generated from user analysis and environmental context.
Able to develop architectural programs by collecting relevant data and reducing them into concepts, able to design the site by considering the detailed urban guideline rules, building mass composition that responds to the urban design, and the open area in the outer building area, and able to consider the aspects of human behavior SPC4 Able to pick building materials, components, and structural systems integrated with the design.
Able to design a building structure system integrated with the spatial organization and building form.

SPC5
Able to design architecture with the principle of "barrierfree" design for the elderly and persons with disabilities.
Able to apply the principles of health and safety in the building facilities, including lighting and persons with disabilities

Second Project: Apartment (APT)
The students designed a mid-rise apartment or codenamed as APT eight stories tall for the second project.The building site was located on the northern side of the first project, as seen in Figure 3.The building was designed primarily to respond to the housing needs of the people of Bandung.Students should consider learning occupants' behavior when designing vertical dwellings.
The main competency that students must master is designing a building as a system.So, the design focused on combining the structural module with a typical spatial module and designing a structural system by paying attention to the standard spatial layout, circulation for building safety, and the aspect of mechanical/electrical utilities that support the building.This approach was suitable for developing a high-rise or midrise building with typical spaces, such as apartments.It also helped students design the building based on specific and relevant rules and integrate functional and constructability requirements as the main design focus.For final design submission, each student could submit a physical design study model on a custom scale (depending on the visualization needs) using monochrome-colored materials.Unlike the first project, the submission of a physical model was optional.The same as in the first project, the course coordinator formulated the studio objectives for this project (gathered in the third column of Table 1) and SPC that students must fulfill (defined in the third column of Table 2).

Research Mechanism
This section explains the research mechanism, which was performed in two parts.The first part was the implementation of VRDR in an ongoing architectural design studio.The second part was the confirmation study of the result obtained from the first part.

Part 1: VRDR Implementation in Architectural Design Studio
In an ongoing architectural design studio course, Part 1 implemented the VRDR system as an affordance-based design review tool.The main objective was to determine if VRDR effectively reviews architectural design, specifically in the educational setting.We hypothesized that a student could improve their design work using the affordance-based design review method using VRDR by discovering the perceived positive and negative affordances in the original design and making design decisions in the revised design work.At the same time, the supervisor could assist the student in improving their design by performing the same design review method using VRDR.For Part 1, we had a student and supervisor as participants.The supervisor we chose was a lecturer in the architecture program and a professional architectural design practitioner.His capability of understanding technical stuff, such as operating the Oculus Quest, installing the required software which is SideQuest version 0.10.21 made by SideQuest in Belfast, and simple troubleshooting, was also in our consideration in picking him as one of the participants.The supervisor chose the student who joined here and worked under his supervision during the design studio course.Both student and supervisor equipped Oculus Quest to access and review the design outputs in VR using VRDR.

Workflow
The workflow of Part 1 is briefly described in Figure 4.The grey and white boxes on the left side are procedures we performed.The flowchart on the right side is the workflow conducted by all parties-researcher, student, and supervisor-to complete the simulation.Each shape position represents which task each party performs during the process.Before the simulation started, we performed each project's "Affordance identification and ASM mapping process" procedure.So, we had one ASM for each project used for the design review process.Then, once the design was finished, the student sent the BIM model of the finished design to us as the researchers.The model was optimized and converted into the VRDR system through "Model optimization and conversion to VRDR" and "VRDR system improvements" procedures.Once it was ready, researchers deployed VRDR as an APK file and sent it to the student and supervisor-the participants-for installation.Then, participants performed the design review process-all the results were submitted online.To fill in the ASM form, participants checked whether each component's affordance was present or not inside the VRDR.They put a tick on the form for the affordance pair they perceived if it was present.
After they finished the design review process, we retrieved the ASM filled by the participants and performed the data analysis process.Once the analysis was concluded, the participants received the result as feedback.Then, the student and supervisor began revising the design based on the analysis result.When the revised design was completed, the workflow would restart from the beginning until they received the final analysis results.

Affordance Identification and ASM Mapping Process
To carry out the affordance-based design review process, we first had to identify the indicators (affordances sought in the design).The studio objectives and the SPC of the design projects were used to determine these affordances.We performed the affordance identification by conducting a content analysis on the objectives and SPC.Furthermore, we referred to building design objectives described in the Whole Building Design Guide [64] to help identify the affordances.These affordances were interpreted differently by each participant later in the simulation.
To finish the LC project, the student must meet seven goals and eight SPC, as outlined in the preceding section.To complete the APT project, the student must satisfy five studio goals and 10 SPC.Only SPC and numerous goals were used to derive affordances.The other objectives and criteria in this study on the students' presenting styles and working attitudes were discarded by ASM.We divided the affordances from goals and SPC into four groups using content analysis (AUAP, AUAN, AAAP, and AAAN).Since they were interconnected, we opted to map the projects' objectives and SPC into five objectives (Table 1) and five SPC (Table 2).Then, as shown in Table 3, we specified the affordances and combined them into an affordance structure matrix (ASM) that would be utilized in the design review process.The implementation began when each student and supervisor had successfully installed the VRDR application in the provided Oculus Quest HMD using the deployment procedure.We provided remote support on the installation and usage tutorial to ensure they could do the design review without hassle.
In Part 1, there were five design options reviewed by the participants.Two design options came from the lifestyle center project (LC), and three design options were from the apartment project (APT).For LC, the design options were taken from the submitted design for grading (LC1) and revised after the first design review process (LC2).For APT, they were taken from the snapshot model while the design process was ongoing (APT1), the proposed design for grading (APT2), and the revised design after the proposed design review process (APT3).However, since the student could not catch up with the tight schedule of the design studio course in the field, the APT1 design review was performed after the APT2 option was completed.So, we decided to use only APT2 and APT3 options for the data analysis.The model of all four design options can be seen in Figures 5 and 6.
To start the design review process, participants launched the VRDR app inside the Quest HMD.Each participant must explore each design model in a VE and review the design based on two design component groups: rooms and building components.Each room had its tag inside the VE that showed its parameters, such as room name, area, and volume.Once the participants found the room, they had to review it and fill in the ASM by marking which affordances they perceived from it.The design model had to apply the process to all available rooms and building components.After participants filled in the ASM, we started the data analysis process to determine the affordance-based design review result.The result was returned to the participants as the basis for the design revision process.

Part 2: Confirmation Study
In Part 2, we conducted a confirmation study to affirm the result in Part 1.We raised similar hypotheses to the previous study [65].First, the media used for the design review might affect how each participant perceived affordances in each design, leading to the review result.Second, with its immersive spatial capability compared to NVR, we expected that VR would be more effective in helping participants perceive affordances from spatial entities.Third, we also expected that each design component might or might not be perceived equally by both media.If an affordance was confirmed on both media, it was highly probable that the affordance was present.The architectural designer could create the design component paired with that affordance.Lastly, by having an affordance confirmed to be perceived on specific media across all design works, we expected that the affordance would be highly compatible with the media.Using the compatible media for review is suggested if a participant wants to review that affordance presence.
Revised design work in Part 2 study could help confirm results obtained from the previous study [65].Therefore, we used two mediums for the affordance-based design review process: non-VR (NVR) and VR.In the end, we compare the results both from Part 1 and 2 to determine whether VRDR is adequate to be utilized for the affordance-based design review process.

Participants
During the COVID-19 epidemic, we ran the simulation.As seen in Figure 7, the participants in this simulation were third-year architectural design students.We hypothesized that the students had appropriate spatial reading skills to analyze an architecture design by seeing affordances within a VE and are comfortable with BIM-based design authoring tools.Fifty-eight students, ranging in age from 20 to 24, (33.9% of whom were 22), took part in the simulation.In total, 72.9% had no prior experience with virtual reality equipment, 50.8% were myopic, and 47.5% had no vision problems.As a result, we needed all children who wished to participate to have received at least one vaccination (83.1% had received their second) and to follow the mandated health protocol.We also followed the Japanese Ethical Guidelines for Medical and Health Research Involving Human Subjects and the principles of the Declaration of Helsinki when conducting the simulation.Based on the examined design projects, we informed participants and placed them into two groups.Group A looked at the LC and APT projects that used NVR and VR material, respectively.Group B, on the other hand, looked at the APT project with NVR media and the LC project with VR media.As a result, each participant had a unique experience with various projects and media.

Data Analysis Process
The data analysis procedures employed in this investigation are discussed in this section.The association between affordances and design components and the relationship between affordances and the medium utilized for the design review process was investigated using a correlation and hierarchy analysis approach.For a medium effectivity comparison, we used a paired statistical t-test.This study introduced the PDS process, a novel data analysis method.It was developed using affordances theory and distribution analysis to determine which design alternative is superior based on the presence and disappearance of perceived positive and negative affordances.

Present-Disappear-Stagnant (PDS) Process
Gibson [23] created a potential for a person to recognize what action they could undertake by presenting affordance as a collection of action options supplied by an item.He considers affordance to be a form of "direct perception."Perceptual information offered by objects, which describes affordance, is critical in assisting users as living creatures in determining the tasks that must be completed [35].An architect and architectural designer should create a building with enough perceptual information to guarantee its users can adequately inhabit it.They have sufficient expertise to determine whether a building design has sufficient information for the expected affordances given by users.Ordinary users may be unaware that affordances exist since they are contingent on the presence of a living individual capable of seeing them [31].As a result, an architect and architectural designer should be aware of whether particular affordances are present in a building design.For an affordance-based design review process, we suggest the PDS process, which is a data analysis procedure.
PDS itself stands for present, disappeared, and stagnant.This process counts how many perceived affordances in the latest iteration of an object are currently present and disappeared compared to the object's previous iteration.It also calculates how many perceived affordances whose presence is still stagnant, present, and disappeared in two compared iterations.This process can be used to compare two design iterations between an original design and a revised design in architectural design.Compared to the original design, 1. an affordance that is now present in the revised design marked as "Present" or "P," 2. an affordance that is now disappeared in the revised design marked as "Disappeared" or "D," 3. an affordance that is still present in the revised design marked as "stagnantly present" or "S1", and 4. an affordance still disappeared in the revised design marked as "stagnantly disappeared" or "S0".
This process must be performed with perceived positive and negative affordances.So, the designer can use the PDS process to measure if the revised design is better or worse based on the amount of perceived positive and negative affordances.First, we summarize the perceived positive and negative affordances in percentages based on their PDS marks in Table 4 below.

Objective
Affordance Group P D S0 S1

OBJ1
A-P ∑ A-P (P) To determine whether a design's tendency is improved or not based on the number of perceived affordances, we create indexes named positive index (PI) and negative index (NI).Both are expressed with the following equations: 1. Positive Index (PI) A positive index (PI) is a sum of percentages of present positive affordances, disappeared negative affordances, stagnantly disappeared negative affordances, and stagnantly present positive affordances.This index shows a revised design or a design iteration tendency to be positively enhanced.

Negative Index (NI)
A negative index (NI) is a sum of percentages of present negative affordances, disappeared positive affordances, stagnantly disappeared positive affordances, and stagnantly present negative affordances.This index shows the tendency of a revised design or a design iteration to be negatively revised.
Then, to find out the improvement between two design iterations based on the amount of added perceived affordances, we propose other indexes named Imprv(+) and Imprv(−).These indexes might also help us to know if the tool or media we utilize for the design review process is helpful or not.Both are expressed with the following equations: Imprv(+) is a sum of percentages of present positive affordances and disappeared negative affordances.This index shows the improvement of a design iteration in a positive direction.

Imprv(−)
Imprv(−) is a sum of percentages of present negative affordances and disappeared positive affordances.This index shows the improvement of a design iteration in the negative direction.

Relationship between Affordances and Design Components
We used JMP software version 15 developed by JMP Statistical Discovery LLC based in North Carolina, USA, to perform correspondence and hierarchical clustering analysis methods between affordances and design component groups to investigate the association between observed affordances and evaluated design components.Based on the cubic clustering criterion, the analysis outcome is provided as dendrograms grouped in distinct colors.The clusters were then mapped based on the medium utilized.As a result, we can figure out if each affordance pair is experienced using a single medium (NVR or VR) or a dual medium (NVR and VR) (NVR and VR).When an affordance and a design component are connected inside a dendrogram of the NVR or VR, a single media detects an affordance pair-NVR or VR.Both media see the pair when the identical pair appears in both the NVR and VR dendrograms.

Relationship between the Affordances and Media
This data analysis aims to discover the relationship between affordances and medium for a design review in different design component groups.We perform correspondence and hierarchical clustering analysis between affordances and medium used during the study to find the relationship between perceived affordances and medium.Then, we determine the medium compatibility in assisting participants in perceiving affordance in each design project.The same as in the previous section, we used JMP software version 15 to process the data.The analysis output is presented in the dendrogram.Each dendrogram is clustered in different colors based on the cubic clustering criterion calculated by the JMP software version 15.We mapped the clusters based on the media used within each design component group and calculated the distribution of confirmed affordances on both design options in the design projects.

Media Effectiveness Comparison
We used paired t-tests to compare design review outcomes using NVR and VR mediums to see how beneficial the medium was for the design review process.Each affordance-design component pair was subjected to a paired t-test."There is no difference in the degree of perceived affordance between NVR and VR medium", we stated as our null hypothesis.If the null hypothesis is disproved, VR media is judged to have more affordances than NVR media.We summed the p-values from all tests in all design possibilities, which were grouped based on affordance groups.The distribution of couples with a significant p-value was then estimated.

Results and Discussion
This section discusses the result from Part 1, where VRDR implementation in an ongoing design studio process was carried on, and from Part 2, the confirmation study of the findings from Part 1.We also compare the data analysis results between Part 1 and Part 2.

PDS Process
This process aims to discover whether the latest design iteration is improved over the previous iteration based on the perceived positive and negative affordances and exercises if the tool or media we used in the study is effective for the design review process.This section discusses the result of Part 1 and Part 2 studies and determines whether LC and APT projects' revised design improves and how effective the VRDR system is for the design review process.
First, we see Table 5, showing the LC project PI and NI from Part 1 and Part 2 studies.In Part 1, we have a student and her supervisor as the participants, while in Part 2, we have a group of third-year students as the participants.Compared to the Part 1 study, the students in the Part 2 trend result align with the student in Part 1.All PI values from Part 2's and Part 1's students are higher than their NI counterparts.Both agree that the LC project's revised design has achieved the objectives and SPC.However, it is different from the result from the supervisor, where the NI values are slightly higher than PI values in the BC-OBJ and BC-SPC categories.This result implies that the supervisor considered that the room components in the revised design of the LC project have achieved the objectives and SPC but not for the building components.  2 show that PI values from all categories are slightly higher than the NI values.They argue that the revised design of APT has achieved the objectives and SPC with a slight margin.It contrasts with the result from the Part 1 study, where both student and supervisor have the same perception of the affordances they perceived when reviewing APT design options.It is shown that both of their NI values for BC-OBJ categories are slightly higher compared to PI values.This result indicates that they think that slightly building components in the revised design of APT has not yet achieved the objectives but has met the SPC.The results above suspect that the building typology might affect the trend pattern seen in the LC and APT projects.The student was asked to focus more on the LC project's spatial elements (rooms), while in the APT design, the student was focused more on the building systems (building components/BC).So, the student might be less aware of the BC in the LC than the APT.At the same time, the supervisor was supposed to review both spatial elements and building systems-consistently-regardless of the student-designed building typology.
Next, we check the improvement index to determine the effectiveness of the VRDR system for the design review based on affordance presence changes.Table 7 shows the improvement index for the LC project taken from Part 1 and Part 2 studies.As reviewed in Part 1, the results show that revised LC projects' Imprv(+) values are higher than their counterpart's Imprv(−) across all categories.The margin between Imprv(+) and Imprv(−) ranges between 0.01 and 0.11, which is narrow.This result tells us that the student and supervisor of Part 1 argue that the improvement in the revised LC tends in the positive direction.
The students argue that the improvement for room components in the revised LC has a significant tendency towards the positive direction, but not for the building components in trying to achieve the studio objectives.As seen in Part 2, only the Imprv(+) value in the BC-OBJ is lower than its Imprv(−) value, with a margin of 0.03.It also has the highest margin between Imprv(+) and Imprv(−) in the RM-OBJ with 0.60 and RM-SPC with 0.54.Lastly, Table 8 shows the improvement index for the APT project taken from Part 1 and Part 2 studies.We see that for the room component (RM-OBJ and RM-SPC) in Parts 1 and 2, the Imprv(+) values are higher than the Imprv(−) values.On the other hand, for the building components (BC-OBJ and BC-SPC), Imprv(+) values are lower than Imprv(−) values as reviewed by the supervisor and the Part 2's students.Results from the student from Part 1 show that the Imprv(+) values for building components grouped based on objectives are lower than Imprv(−) values.These results show that all participants agree that the improvement of room components in the revised APT design towards positive directions-but not for the building components.We found improvements in the Room components across all projects from the Improvement Index results above that tend toward the positive direction.We also found mixed results in the building components across the projects on the other side.The improvements in the building components of the revised LC project are mainly in the positive direction.In comparison, the APT project is in a negative direction.So, in terms of improvements-the condition where we have more positive affordance and less negative affordances perceived, the VRDR system is considered more effective in improving spatial elements in the design but less effective for improving building systems (building components).

Affordances vs. Design Components
This section shows the analysis process to determine the association between observed affordances and evaluated design components.In Part 1, we performed the distribution analysis from the ASM to determine how many affordances were present on each pair of design component groups and affordance category in every design option.Part 1 participants were the student and his supervisor.Then, in Part 2, we performed correspondence and hierarchical clustering analysis processes between affordances and design components in LC and APT projects using JMP software version 15 developed by JMP Statistical Discovery LLC based in North Carolina, USA.The analysis output is presented in dendrograms, as seen in Table 9, and we mapped the clusters based on the media used.
Then, we utilized the results of performed analyses above to examine the relationship between affordances and design components in NVR and VR media across all LC and APT project design alternatives.Analyzing their relationship lets you see which design elements have met their objectives and which need to be enhanced.The student discovered that positive affordances are more prevalent than negative affordances in Part 1 of the investigation.This result was seen in the LC and APT project design options.Meanwhile, the supervisor's findings varied, with positive affordances not always outnumbering negative ones.Then, by calculating the percentage margin of affordance presence between design options of LC and APT projects, the result implies that students found more positive perceived affordances and mostly less perceived negative affordances in the revised design on all design components.Table 9. Dendrograms from the hierarchical analysis method between affordances and design components in LC1.

LC1 Room (RM) NVR VR
Meanwhile, the supervisor produced an outstanding outcome.The supervisor analyzed the APT design and found that it is improved by reducing the presence of negative affordances, but it is worse in terms of AAAP affordances because the affordances in this category have vanished.As a result of the increased positive and negative affordances on all design components, the outcome demonstrates that LC design is both better and worse.
Next, we compare between Part 1 result and the Part 2 result.We found a similar trend in the Part 1 result in Table 10 and Part 2 in Table 11 Both results show that more positive affordances are perceived than negative affordances.It is different from the results that came from the supervisor (see Table 12).This finding is also in line with the theory of affordance, whereas Gibson [23] said that "affordances are animal-relative properties of the environment."The student and supervisor can be considered "different kinds of animals" since both have different levels of understanding and experience when reviewing the architectural design.

Affordances vs. Media
This section explains the correspondence and hierarchical clustering analysis between affordances and medium used during the Part 2 study.The output is presented as a dendrogram.Then, we mapped the dendrogram clusters based on the media used within each design component group and calculated the distribution of perceived affordances on both design options in LC and APT projects, as seen in Table 13.Table 14 presents the results.
Table 14 shows the number of perceived affordances in LC and APT project design options using specific media.These numbers show the many affordances that have high compatibility to be perceived in a design project using a specific media.Statistically, these affordances highly correlate with the media used for design review.For example, in the LC project, we found six AAAP affordances highly perceived using NVR media when users review building components.Another example is in the APT project.Thirteen AUAP affordances have high compatibility to be perceived using VR media when a user reviews room components in the APT projects.
Table 13.Dendrograms of hierarchical analysis method between affordances and media in LC1.

LC1 Room (RM)
Building Component (BC) We summarized the finding to narrow the result to see how many affordances are perceived in both NVR and VR medium and LC and APT projects.We identified each affordance and listed them in Table 15.There are five affordances associated more with the NVR media: affordance of blocking user access (AUAN03), the affordance of sufficient capacity for essential furniture (AAAP04), the affordance of internal layout flexibility (AAAP05), the affordance of weather protection (AAAP11) in building components, and affordance of repairability (AAAP14) in a room component.By logic, these affordances are more easily perceived by NVR media, where users can inspect the design from different views.Contrastingly, in VR, it is set up as a 1:1 scale first-person exploration.Meanwhile, in the VR media, four affordances are more associated with the affordance of noise cancellation (AUAP14), the affordance of a sense of boring (AUAN06), the affordance of expansion ability (AAAP01) in a room, and the affordance of providing a sense of tightness (AUAN05) in building components.Compared to the previous simulation [65], the VR models in this study have sufficient information for users reviewing the design with minimal cognitive effort or direct perception-especially in reviewing room components.VR has more associated affordances with room components compared to NVR media.This result is logically acceptable since VR technology brings immersive spatial information to the user and the ability to explore the space.Additionally, these affordances are more compatible with being perceived by certain media.The finding could help future users pick which media is used for the design review process to perceive specific affordances.

Comparison of Media Effectiveness
This section explains the results of paired t-tests comparing design review results using NVR and VR medium.The objective is to compare medium effectivity in helping user perceives affordances between NVR and VR medium in each design option.We used the following null hypothesis: "there is no difference in the amount of perceived affordance between VR and NVR media."Once the null hypothesis is rejected, the following alternative hypothesis is accepted: "there are more affordances perceived in the VR media than in the NVR media."A snapshot of the paired t-test results for the APT2 project is presented in Table 16.The results for the LC project are presented in Table 17.Meanwhile, for the APT project, the results are shown in Table 18.
Each percentage number in both tables represents the percentage of affordance pairs with a significant p-value based on our performed paired t-test results.To recall, an affordance pair means a pairing between affordance and a design component (a room or a building component).An affordance pair with a significant p-value indicates that the pair is more effective in being perceived using VR than NVR media.For example, as seen in Table 17, 25% of positive AUA-type affordance (AUAP) pairs from the room component of the LC1 project and only 12.5% of AUAP pairs from building components (BC) of the LC1 project.This result shows more significant AUAP pairs from the room component than from BC.The highlighted numbers in both tables below are categories with higher percentages of affordance pairs with a significant p-value compared to the room and BC groups on each affordance type.For the LC project, Table 17 found that four categories have more pairs with a significant p-value in the Room group than the BC group (LC1-AUAP, LC1-AUAN, LC2-AUAP, and LC2 AAAP).Two categories have more pairs with a significant p-value in the BC group than the Room group (LC2-AUAN and LC2-AAAN).The rest of the two categories have the same number of affordance pairs with significant p-value both in Room and BC groups (LC1-AAAP and LC1-AAAN).This result indicates that, regarding media effectiveness in project-wide, VR media is effectively used only in a handful of affordance types of LC projects.Meanwhile, in comparison between Room and BC groups, VR media is considered more effective for perceiving affordances in Room than BC groups.
Conversely, in Table 18 for the APT project, we found that all affordance categories have more pairs with a significant p-value in the Room group compared to the BC group.It implies that VR media is more effective than NVR media in perceiving all affordance types of APT projects.Additionally, in a comparison of Room and BC groups, VR media is more effective in perceiving affordance pairs in the Room group than in the BC group.More relevant pairings of rooms in the APT project match the prior study's conclusions [65].Apartment building typology has more defined and uniform physical qualities due to utility than lifecycle center type, which offers more diverse combinations of spatial PROGRAMMING.When an object's physical features are better defined, its affordances may be much more easily experienced by people who can afford to notice it.

Conclusions
By applying it in an ongoing architectural design studio course, this study investigated the experimental investigation of VRDR system usage for the affordance-based design review process.The research is divided into two parts.In Part 1, we implemented a VRDR system with a student and a supervisor in a third-year architectural design studio course.A lifestyle center facility (LC) and an apartment were assigned design tasks (APT).They utilized VRDR to assess the design outputs and then used the input to improve the design to meet studio goals and student performance criteria (SPC).They then went through another design review procedure to ensure improvement.Part 2 consisted of confirmation research with third-year architectural design bachelor students utilizing non-VR (NVR) and VR media to corroborate the design review results from Part 1 and the prior study.
This research also focuses on uncovering design improvement statistically through the suggested PDS method, the link between affordances and design components, the relationship between affordances and medium, and the comparison of media effectivity.According to the PDS method result, VRDR helps students improve the spatial elements to reach the objectives and SPC on both projects-indicated by higher value of positive index (PI) value than its negative index (NI) value.However, mixed results have been found across the simulation in improving building components.The second analysis identifies which design components are more improved and less improved based on affordances and illustrates how the student and supervisor see "various sorts of animals" when they observe affordances.The third analysis result highlights more easily recognized affordances while utilizing NVR or VR media.The fourth analysis emphasizes the significance of physical features in defining an object's affordances and capacity to be perceived by users.
Furthermore, several findings tied to the effect of using the affordance-based design review method were found in this study.When a building typology can have better defined physical characteristics, it is much easier for users to perceive their affordances.In the PDS process, we suspect that the building typology might affect how a user perceives positive and negative affordances.Later in the paired t-test results in media effectiveness comparison, the result shows an even more suspicious impact because of the building typology.Then, comparing affordance and design components, the results confirm that the design review method result is in-line with Gibson's Theory of Affordance in terms of the "animal-relative" relationship.Lastly, in comparing affordance and media, the results show that it is possible to pick a specific media for perceiving specific affordances to target.

Limitations and Suggestions for Future Studies
There are several limitations to this study.First, this research was still limited to the scope of education only.Using the affordance-based design review method in the scope of professional architects requires further adjustments and research.It is due to the need or achievement of real project targets that are more realistic.This research was also still limited in the scope of Indonesian culture.Thus, the statistical findings were local, but the pattern might be like that of participants from other regions.
Nevertheless, in the future, the implemented method should be tested in other regions to find similarities and differences.Adapting the theory of affordance in the realm of the built environment, especially architecture, is still an area that can be explored further.The availability of virtual reality technology that is increasingly sophisticated and easy to reach can help further develop this research.Thus, we can obtain a more holistic picture of the performance of the VRDR application and the effectiveness of the affordance-based design review method.Lastly, from a technical point of view, developing an improved VRDR system can be done by adding various features such as multiplayer, instant data analysis process, and exploring machine learning (ML) to estimate users' ASM results.In addition, the building model optimization process needs to be made more concise and streamlined to speed up converting the building model into a VR model and improve the VE performance in the HMD.Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Figure 2 .
Figure 2. VRDR was built using the Unity game engine.

Figure 3 .
Figure 3. Site aerial view (a) and site plan (b) were used to form the design studio projects.(Image source, (a): Google Earth).

Author
Contributions: Conceptualization, F.A.A. and M.S.; methodology, F.A.A. and M.S.; software, F.A.A.; validation, F.A.A., M.S., M.D.K., and A.I.; formal analysis, F.A.A.; investigation, F.A.A. and M.D.K.; resources, F.A.A. and M.S.; data curation, F.A.A.; writing-original draft preparation, F.A.A.; writing-review and editing, F.A.A., M.S., M.D.K., and A.I.; visualization, F.A.A.; supervision, F.A.A. and M.S.; project administration, F.A.A.; funding acquisition, M.S.All authors have read and agreed to the published version of the manuscript.Funding: This research received no external funding.Institutional Review Board Statement: All subjects gave informed consent for inclusion before participating in the study.The study was conducted following the Japanese Ethical Guidelines for Medical and Health Research Involving Human Subjects and the principles of the Declaration of Helsinki.

Table 1 .
Studio objectives of LC and APT projects.

Table 2 .
Student Performance Criteria (SPC) of LC and APT projects.

Table 3 .
List of affordances mapped to affordance structure matrix.

Table 4 .
PDS calculation for Positive Index (PI) and Negative Index (NI).

Table 5 .
Positive and Negative Index for LC Project (Part 1 and 2) Second, we look at Table 6, showing the PI and NI of the APT project from Part 1 and Part 2 studies.Students in Part

Table 6 .
Positive and Negative Index for APT Project (Part 1 and 2).

Table 7 .
Improvement Index for LC Project (Part 1 and 2).

Table 8 .
Improvement Index for APT Project (Part 1 and Part 2).

Table 10 .
Affordance presence for each design component in LC and APT projects reviewed by the student (Part 1).

Table 11 .
Affordance presence is perceived by double media for each design component in LC and APT projects reviewed by students (Part 2).

Table 12 .
Affordance presence for each design component in LC and APT projects reviewed by the supervisor (Part 1).

Table 14 .
Number of affordances perceived on both design options in each LC and APT projects.

Table 15 .
List of affordances perceived on both design options of LC and APT projects.

Table 16 .
Snapshot of p-values from paired t-test results in Room component at APT2 project.
Note: p-value with * is considered as significant.

Table 17 .
Percentages of affordance pairs that have a significant p-value in the LC project.

Table 18 .
Percentages of affordance pairs that have a significant p-value in the APT project.