Archives of Design Research

Portman et al. (2015) defined VR as a communication component in a computer-made synthetic space that attaches humans as an integral part of the system. The latest head-mounted VR devices, such as the Oculus Rift, provide users with an immersive experience. Immersion—through images, sounds, and other stimuli in a VR system—is the perception of being physically present in a non-physical world as if one is truly in that place (Freina and Ott, 2015). VR replaces the user’s perception of the surrounding environment with a computer-generated 3D environment so that the real environment is no longer needed for cognitive experiences (Kim et al., 2013; Lanier, 1992).

VR technology can provide a higher-quality simulation of the environment and interaction between humans and space. As such, there is an opportunity to develop simulation tools and interactions between humans and space in spatial design (Kuo and Cheng, 2008). The VR environment provides the sensation of a spatial or physical presence (Lee, 2004). These sensations affect the effectiveness of users in processing spatial relationships in their environment. Hence, the use of VR as an educational tool requires careful consideration of the relationship between spatial presence and spatial ability (Özgen et al., 2019). VR learning environments allow learning more immersive interactive features such as the creativity component of digital environments (Casakin, 2007; Dorst and Cross, 2001; Özgen et al., 2019). Users gain an emotional and rational perception of a realistic 3D environment after having an immersive experience in a VR environment. Therefore, there is potential here for developing new ideas to increase empathy concerning spatial needs (Ji et al., 2018). In cognitive learning domains such as thought processes, problem-solving, verbal information, concept formation, and information processing, digital media such as VR-based learning can strengthen the learning design (Dede, 2008).

The literature shows that the features, outcomes, and impacts offered by VR can be used as a means of consulting in the field of spatial design. VR can potentially be used in various forms of interaction in the field of spatial design, both professionally and educationally.

Based on previous research (Adharamadinka et. al., 2023), the Associative Words Generator^© (AWG^©) application has been developed for visually computing verbal expressions using a vocal database generation method integrated with interior visual assets. The application can help get the most powerful, significant, and unverbalized associative meanings based on the user’s preferences of the visual perception of space.

Based on the responses obtained previously, to determine the associative word that has the strongest relationship with the keyword, the selection is made by looking at the largest total distribution value of each associative word obtained. After getting the distribution value of associative words for each word order that comes up, weighting is conducted for each associative word order. The collected survey data are used as a database in which there are various associative meanings that are the same and different. The order of the frequencies in the database also affects the results to be generated. In the formation of a three-dimensional space, this is done by building a room that has four walls, floors, ceilings, and furniture items. Every wall, floor, ceiling, and furniture item is made according to the inherent associative properties of words, has weight and frequency value, and is combined with C# programming language algorithms. Associative words that have been collected from keyword stimuli in each perception dimension will correlate with other associative words and become a resource for the creation of three-dimensional visual assets.

The space generated based on the input in the sentence and program fields results in a mix of visual assets. The results generated change randomly each time the VR button is pressed. The resulting room atmosphere contains elements of the words associated with the perceptual dimension keywords that already exist in the AWG database. The final result can be viewed using the Oculus Rift (VR equipment) from a first-person perspective. Figure 1 below illustrates the interface of the 3D AWG^© spatial visualization generator application and an example of three-dimensional generation results. The interface design of AWG^© reveals three interface design bars in Figure 1 below. In the topmost column, there is a function to input sentences perceiving the desired space (with a database still in development). In Figure 1, it is written, “this space is very spacious and natural.” Consequently, in the second column below, upon pressing the “extraction” button, keywords listed in the AWG^© database, namely “spacious” and “natural,” will automatically appear. In the bottommost third column, upon pressing the “associative” button, associative words will emerge, accompanied by weight values indicating the proximity of profound impressions. Once all the data is input, the final button “VR” will provide a spatial visualization based on the input data in the database. Pressing the “VR” button more than once will present various spatial options in terms of spatial composition, objects, and others while maintaining the same associative weight score (Out-Degree Centrality Score).

Figure 1 Associative Words Generator interface

In previous studies on the dimensions of user resilience to interior spaces, researchers in the field of linguistic cognition and computer science believe that a word has a rich and broad essence of meaning when used to express an event (Junaidy et. al, 2020). The essence of a word can contain implicit and metaphorical associations (Taura et. al., 2010; Nagai et al., 2011; Georgiev and Nagai, 2011). Here, a common method for measuring human impressions of the interior environment is applied, together with the results of the SMB, which is intended to describe a systematic depiction of the environment. The investigated environment can be interior, exterior, or simulated spaces (Küller, 1991). According to Laike (1999), the SMB method is easy to manage and use, and has high validity. SMB measures the overall environment based on a questionnaire with a 1–7 (1 = little; 7 = very) scale to identify how well each adjective is grouped into the eight SMB factor dimensions. When used to assess real-world environments, SMB achieves a high degree of stability in group sizes of 15–20 individuals (Küller, 1991; 1975). SMB claims to be able to measure the average level of a person’s impression of the environment such as a room, building, or landscape in a simulation (Küller, 1991). In addition, SMB can also measure the level of a person’s impression of the perceived well-being. The scale in the SMB uses adjectives about the environment given an interval of 1 to 7; then, the adjectives are divided into eight dimensions of environmental quality: pleasantness, complexity, unity, closedness, social status, potency, affection, and originality. Table 1 below shows the division of these eight dimensions of environmental quality in the SMB theory.

According to Küller (1977), pleasantness is influenced by various things; therefore, it is difficult to predict the interpretation of pleasantness. However, pleasantness may be interpreted as being influenced by the balance between complexity and unity. For example, if the unity value is too high and the complexity value is too low, an environment may be considered boring and not very pleasant. Meanwhile, if the unity is too low and the complexity is too high, this will create a messy impression. Complexity has to do with the amount of detail. Unity relates to the assessment of the harmony of various elements in a single unit. Enclosedness is influenced by the actual size of the room, colour, window size, and the number of demarcations or boundaries in the environment. Social status is a social and economic evaluation of the environment. Affection is influenced by the element of familiarity according to the respondents. Originality is influenced by unusual objects according to the respondent’s experience.

In this study, we focus on measuring the spatial sensation and perception of each group of respondents by validating the generation of a three-dimensional virtual space formed through the Associative Words Generator (AWG) using the SMB dimension. We do not fully use the SMB dimension as a measuring tool for spatial sensation in AWG applications. Instead, we use six SMB dimensions, namely, pleasantness, complexity, unity, enclosedness, affection, and originality. This choice is based on the relevance of the dimensions to the context of our research objectives. The output generated through the AWG is projected using VR as a tool for sensing the manipulated scene. A virtual environment (VE) experience generated by VR can manipulate the experience of space (scene), and manipulating space will make people who use VR feel like they are in a simulated environment. When we move in a 3D environmental space, cognitive relevance will unite with our interactions with objects (Draschkow and Võ, 2016). Objects here play an important role when we can remember what objects there are and how our attention changes when manipulated objects are associated with design elements and design principles to form a different scene. The scene contains supporting elements that are related to associative word stimuli spoken during design consultations between clients and designers/architects as well as studio lecturers and students.

Analysis of the average distribution of the completed questionnaires was conducted to determine the respondents’ assessments based on the SMB dimensions. The questionnaire raised the following six SMB dimensions. Of the 36 research respondents, the average values of the dimension sorted from highest to lowest are originality (6.24), affection (5.38), pleasantness (5.10), unity (4.66), enclosedness (4.39), and complexity (4.21).

In Table 2, the originality dimension has the lowest deviation (0.75), meaning that its perception in each group has almost the same proportion (average of 6.24 for all groups). Meanwhile, the highest deviation is found in the pleasantness dimension, with the student group rating this dimension the highest (5.55).

Figure 2 shows that the pleasantness and affection dimensions in the designer/architect and studio lecturer groups have overlapping means, meaning that these groups have similar abilities in assessing aspects of beauty and familiarity in the interior design. The similarity of these assessments indicates that both groups have similar assessments of the suitability of the keywords with the respondents’ image and liking for the AWG^© application. In the enclosedness dimension, all groups have means that are close to each other, meaning that all groups rate the impression of enclosedness at almost the same level as the experience obtained when using the AWG^© application. This indicates that all groups similarly assess the completeness of interior elements and have a strong and rigid impression of the AWG^© application. This happens because of the uniform arrangement of space in the virtual simulation of the AWG^© application. On the originality dimension, the student, client, and designer/architect groups have overlapping means, while the mean of the studio lecturer group does not approach the mean of the other groups. This indicates that the students, clients, and designer/architect groups have similar assessment abilities of authentic elements. The similarity of these assessments suggests that these groups have similar assessments of the development motivation and curiosity towards AWG^© applications.

Figure 2 Comparison of the questionnaire average score based on the respondent groups and the six SMB dimensions

Meanwhile, the studio lecturer group has a higher critique of the exploration of spatial elements, meaning they think that the spatial visualization of the AWG^© application tends to be normative or ordinary. It is necessary to visualise an AWG^© application that is more relevant to the perception dimension keyword to make the studio lecturer group rate the originality dimension with a higher value. For example, the keyword perception dimension ‘shady’ should be visualised with associations such as ‘shadows’, ‘cool’, and the like by using Associative Concept Network Analysis (ACNA) method. The means of the student and client groups are higher than those of the studio lecturer and designer/architect groups on most dimensions, suggesting that the student and client groups are more easily impressed by the virtual space stimulus of the AWG^© application. The Associative Concept Network Analysis (ACNA) proves valuable for extracting profound insights from collected verbal inputs (verbalizations). This process involves generating word associations using The University of South Florida Word Association (USF Dictionary) Rhyme and Word Fragment Norms (Nelson, McEvoy, and Schreiber, 2004). This technique offers significant advantages as it enhances comprehension of the underlying meaning in verbal expressions (Nagai et al., 2011; Georgiev et al., 2012; Junaidy & Nagai, 2013; Junaidy et al., 2020).

Referring to the line graph of the comparison of word associations resulting from the Associative Concept Network Analysis (ACNA) data processing of student–studio lecturer on the use of the VR-AWG^© application (see Figure 3) that was examined in the previous study (Adharamadinka et al., 2023), the student group produced associative words with Out-Degree Centrality (ODC) score of: (0,18) house; (0,11) tree, door; (0,10) color, window, hard; (0,08) hot, green, blue, red, wall; (0,07) room, white, light, work, flower; (0,06) metal, square, open, soft, floor, top, up, art, bright, food, good, picture, see, clothing, smell, paper, cold, chair, sun; (0,05) water, straight, space, sex, beautiful, black, tile, high, desk, table, wood, thing, pretty, look, back, face, grass, play, nature, feel. Meanwhile, the studio-lecturer group produced associative words with ODC score of: (0,12) house; (0,10) paper; (0,08) room, tree, love, light; (0,07) open, hard, school, door; (0,06) square, up, flower, window, good, dog, nature, place, close, air, star, wood, money, green, sun, fly, life, sky, water, smell, space, black, wall, color, timeless, work, bright, nice, length, big, grass, play, yard, shine, fun, car. It indicates a visual gap or discrepancy in comparing the responses of the two groups. Students generated more associative words than studio lecturers when using the VR-AWG^© application.

Figure 3 Comparison line graph of word association results from ACNA data processing of student-studio lecturer groups on the use of VR technology through the AWG^© application (Adharamadinka et. al., 2023)

Adharamadinka et al. (2023) also examined a comparison line graph of the word association data processing of ACNA client–designer/architect on the use of VR-AWG^© application (see Figure 4). Similar to student–studio lecturer on the use of the VR-AWG^© application, the designer/architect group also produced fewer associative words than the client group when using VR-AWG^© application. The client group produced associative words with (ODC) score of: (0,22) house; (0,14) tree; (0,12) green; (0,11) color; (0,10) black, red, blue; (0,09) hard, white, wall, flower, room, money, water; (0,06) life, light, forest, paper, dog, art, glass, shade, window, home, nature, good, door, chair, picture. Meanwhile, the designer/architect group produced associative words with ODC score of: (0,14) house; (0,08) green, light, tree, room, good, paper, wall, door; (0,07) work, clothes, space, color, red, money; (0,06) hard; (0,05) friend, water, open, square, area, blue, sun, flower, food, car, people.

Figure 4 Comparison line graph of word association results from ACNA data processing of client-designer/architect groups on the use of VR technology through the AWG^© application (Adharamadinka et. al., 2023))

The distribution of the average assessment of all groups of respondents through the questionnaire shows that for most of the dimensions of the questionnaire, the studio lecturer and the designer/architect groups have means below those of the student and client groups. This indicates an alignment of data in the comparison of responses between the student–studio lecturer and client–designer/architect groups from both association word generation and assessment through the questionnaire. In other words, the data analysed through the ACNA method, where the results are latent data, are aligned with the data obtained through subjective assessments given by the respondents through the questionnaire. The alignment of the data shows a response pattern whereby the student and client groups feel more stimulated by the AWG^© application than the studio lecturer and designer/architect groups.

Correlation analysis was conducted to see the relationship in the assessment patterns between the groups, namely, S and L and C and D/A group pairs. Prior to this analysis, a correlation analysis was performed for each group on the six selected SMB dimensions and then, we connected pairs of groups by looking at the similarity of the correlation significance of each respondent. This way, the dimensions that bridge each pair of respondent groups can be seen.

Table 7 shows a comparison of the results of the SMB dimension correlation analysis from the questionnaires that were filled out after the respondents used the AWG^© application. By pairing the student and studio lecturer groups, and the client and designer/architect groups as parameters for assessing the relevance of the usability of the AWG^© application, we find that:

• a. The student and studio lecturer groups have a similar tendency towards the pleasantness–affection dimension, as indicated by a significant correlation. Both the student and studio lecturer groups tend to judge an environment or room to be more beautiful when the environment or room feels more familiar. Using the AWG^© application, the student and studio lecturer groups like the spatial visualization more when the spatial visualization is more in line with their imagination. This suggests that both student and studio lecturer groups prefer rooms with similar characteristics because they both have developed a similar understanding of the definition of beauty and familiarity. This can happen because the student group has been exposed to the knowledge of the studio lecturer group in their studio classes/sessions.
• b. The client and designer/architect groups have a similar tendency to rate the pleasantness–complexity dimension, indicated by a significant correlation. Both the client and designer/architect groups tend to judge an environment or room to be more beautiful when the contrast or variation of elements in the environment or room is higher. Using the AWG^© application, the client and designer/architect groups assess that the spatial visualization is more in line with their imagination when the spatial visualization is easier to understand despite the complexity of the variations in interior elements. This indicates that the client and designer/architect groups have similar assessments of the beauty of the room because both have built an understanding of the effect of contrasting variations of space elements on beauty. This can happen because of the influence of design trends and media that introduce contrasts to a wide audience such that the client and designer/architect groups have been influenced by similar views.