Perceptual difference of urban public spaces between design professionals and ‘laypersons’: Evidence,health implications and ready-made urban design templates

Abstract

Urban Public Spaces (UPS) are important arenas for human interaction and social activities, and ensuring their quality and functionality is crucial for a successful urban design with public health benefits. However, mostly for insufficient public participation, user experiences of UPS are usually not what the designers were expecting. Therefore, the urgent need to investigate the difference in UPS perception between design professionals and ‘laypersons’, that is, non-professional users, has been increasingly highlighted. In this paper, we utilize Immersive Virtual Environment (IVE) and physiological measurement tools to obtain empirical observations on the psychological and physiological responses, as well as environmental preferences on UPS of the two groups, compare their perceptual similarities and differences, and consequently analyze the influencing factors and potential mechanisms. We find that the environmental perception of the two groups do differ, with design professionals showing a higher degree of ‘intolerance’ in the quality rating of UPS, and being more sensitive to scene features related to necessary than spontaneous and social activities. The findings reveal structural differences for the two groups in the mechanisms by which environmental features trigger perceptual differences, thus providing new support for designers to prepare ready-made UPS design templates that are evidence-based.

Keywords

Environmental perception urban public spaces psychological and physiological responses environmental preference immersive virtual environment physiological monitoring public health design guideline

Introduction

In the face of many environmental problems brought about by rapid urbanization,¹ the quality of the built environment shows increasingly strong effects on the health and well-being of residents. The potential to improve people’s quality of life through the provision of high-quality built environments has become one of the most prominent concerns of designers.² Urban Public Space (UPS), as an arena for various activities in cities, carries an essential function of promoting social inclusion and interaction. It also contributes to citizens’ cultural, ecological and health benefits.^3–5 Therefore, the creation of UPS plays a crucial role in the sustainability of cities, the quality of urban life and livability, and has become one of the central issues in sociological, geographic, architectural and urban design research.^5–8

Urban residents are the users of UPS, and meeting their needs is where the value of a successful UPS design lies.^9,10 However, traditional UPS design was usually determined by the designer’s perspective, which tends to follow a self-referential, structuralist paradigm, while ignoring the desires of ordinary users, resulting in a design practice that deviates from real needs.^10,11 The failure of designers to anticipate users’ expectations has prompted researchers to consider whether there is a fundamental difference in the perception of the built environment between design professionals and ‘laypersons’, that is, non-professional users.¹²

Previous studies had shown that design professionals and non-professional users do have some differences in experiencing and evaluating the urban landscape and architectural appearance,^12–18 resulting in a noticeable mismatch in environmental perception and preferences between the two groups.¹⁹ Extensive research has further discussed the issue that most architects are having a hard time predicting how non-professional users would react towards built environment accurately.^12–18 Nevertheless, the problem of inter-professional perceptual difference remains, especially in terms of urban landscapes, which to our knowledge, has not been thoroughly addressed in the existing literature. In particular, few studies have focused on inter-group perceptual differences in UPS regarding their forms, general patterns and causes, mostly due to limitations in research tools for directly observing human’s perceptual responses to the environment, leaving prominent knowledge gaps.

Recent developments in new research tools such as virtual reality (VR) and portable physiological sensors have provided new opportunities for research in this field, which opens up more possibilities for empirical investigations and theoretical development.^20–22 In this paper, we designed an Immersive Virtual Environment (IVE)-based experiment, integrating with physiological monitoring and conventional psychological scales. We try to answer the following scientific questions: (1) Do design professionals and non-professional users perceive UPS differently? (2) What are the specific perceptual differences between design professionals and non-professional users regarding the design of UPS? (3) How do UPS scenes with different physical features affect the perceptual differences between the two groups? To investigate the questions, we first present a critical review of the literature related to the differences in built-environment perceptions between the two groups, thus sorting out the specific research progress and knowledge gaps, and build the analytical framework of this paper accordingly.

Given the extant research related to the perception of UPS and perceptual differences in urban landscapes among users with different academic backgrounds, we propose to test the following hypotheses: H1. Perceptual differences regarding the design of UPS exist among different users, especially in this case, between design professionals and non-professional users. H2. The mechanisms of how environmental features contribute to the perception between the two groups differ, in which differences exist in their environmental preferences, physiological and also psychological responses. H3. Non-professional users might hold more positive attitudes when perceiving the same UPS than design professionals.

Literature review and analytical framework

Empirical Studies and Explanations on the Design Professionals’ Failure to Meet the Needs of Non-Professional Users in UPS Design

Overall, researchers had found that the design of many UPS does not meet user needs,^10,11 and they had been investigating the specific forms of these phenomena and their causes.

First, traditional urban design practice is usually guided by policies and guidelines mostly based on designers’ expertise and experience rather than an evidence-based approach.²³ Thus, this top-down approach suffers from insufficient public participation, resulting in design schemes that do not meet actual usage expectations.²⁴ Meanwhile, a large number of methods used to evaluate UPS are also based on expert perspectives,⁴ and the results often do not reflect broader public interaction and perception.²⁵ The biases and limitations here make them insufficient as a reliable measure of the urban quality of life.²⁶ Otherwise, they would ultimately lead to false and one-sided assumptions, predictions and practices. Therefore, in this study, we opt to analyze both subjective (stated) and objective (revealed) perceptual outcomes in an integrated manner, with a special focus on the quality of UPS from the perspective of non-professional users, thus, to compensate for the bias and limitations of the top-down approach.

Secondly, regarding the difference in the perception and evaluation criteria of the built environment between the two groups, empirical studies have been conducted on different aspects of this phenomenon, which have revealed the prevalence of perceptual differences. For example, in terms of understanding urban cultural ecosystem services, design professionals have a more ‘practical’ and ‘management-centred’ view of nature. In contrast, non-professional users have a focus on more ‘enjoyment of nature’.¹³ In researching perceptions of green stormwater infrastructure, design professionals prefer the newer forms of stormwater solutions.¹⁴ In terms of air quality evaluation and residential environment quality evaluation, design professionals cannot accurately predict the satisfaction and use frequency of residents.²⁷ In terms of preference and assessment of urban green space, such as derelict urban land or informal green space, studies have revealed that design professionals prefer ‘natural’ green spaces, while residents prefer ‘artificial’ green spaces.¹⁵ Meanwhile, the professionals typically better recognize the value of spontaneous vegetation compared to non-professionals,²⁸ although a subtlety different report exists in which local non-professional users were found to express preferences for both ‘wilder’ and more formal urban natural spaces towards managed natural spaces.²⁹ Evidence of assessment differences towards landscape designing between local non-professional users and experts were also found in research which indicates that the former have a more positive perception towards certain land-use associations compared to the latter.³⁰ Moreover, regarding architectural design, the difference of opinion between architects and users has also been confirmed by a growing body of empirical research: while there are some similarities between the two groups in terms of the aesthetic evaluation of architectural appearance,¹⁶ there are significant differences in the evaluation in dimensions such as ‘meaningfulness’ and ‘originality’.¹⁷ Other studies have shown that architects are more concerned with the comforts of the buildings,³¹ while they cannot accurately foresee the preference of non-professional users.^12,32 In addition, design professionals are found more critical in their evaluation of architectural appearance.¹⁸ These perceptual differences between groups with and without design-professional backgrounds have been suggested as potential causes of why user satisfaction and real-world use do not meet designers’ expectations,¹⁹ which in turn means that understanding the nature of such differences would help urban designers and decision-makers to improve UPS design.

The perceptual dimension of UPS and its quality evaluation

Generally, the quality of UPS can be assessed based on the objectivist paradigm, which focuses on the quality of intrinsic landscape attributes or the subjectivist paradigm, and includes the psychophysical, cognitive and experiential evaluation of users.³³ On the one hand, intrinsic landscape features include measurable landscape qualities such as scale, colour, texture, vegetation, etc.,²⁵ and many researchers have proposed quantitative systems for measuring UPS feature attributes after Lynch’s seminal work³⁴ which designated the five urban image metrics, including path, edge, node, district and landmarks. For a few examples, Ode and others³⁵ outlined the landscape characteristics from the perspective of visual perception using nine indicators: complexity, coherence, disturbance, stewardship, imageability, visual scale, naturalness, historicity and ephemera, and argued that they explain human’s experience of the landscape and visual preference. Ewing and Handy³⁶ focused on the street environment in the UPS and summarized the physical characteristics of the built environment as imageability, enclosure, human scale, transparency and complexity, which are measured through physical attribute auditing and expert rating. The measurement systems above provide an authoritative basis for the objective assessment of UPS and are of great reference value for environmental perception studies.

On the other hand, from the subjectivist paradigm, the complex and subtle relationship between physical features of the built environment and users’ subjective aesthetical perception and emotional responses can be assessed through environmental perception and affective appraisal.²¹ Previous research has taken the topics such as visual characteristics of public artworks,³⁷ spatial enclosure,³⁸ vegetation and infrastructure,³⁹ vegetation-created enclosure,⁴⁰ among others, and explore aspects of subjective perception, including user preference, behaviour and restorative potential of the physical environment.^21,41,42 Based on these specialized studies focussing on specific dimensions of UPS attributes, comprehensive subjective evaluation systems for UPS are also established, typically in the form of scale. For example, Mehta⁴³ established the Public Space Index (PSI), a five-dimensional UPS perception evaluation scale considering the inclusiveness, meaningfulness, safety, comfort and pleasurability of UPS. Currie⁴⁴ established the design principles of small urban parks considering their accessibility, specificity, authenticity, functionality and adaptability. Zamanifard proposed the Public Space Experiential Quality Index (PSEQI), which considers four dimensions of UPS: comfort, diversity and vitality, inclusiveness, image and likability, which are further divided into 15 variables and 83 specific measurements.⁴ The scale and index above are highly generic in-built environment research to capture the potential influencing mechanisms between the physical features of the environment and public perceptions. They are also important references for the design of the analytical framework of this paper.

Environmental perception research supported by virtual reality and physiological monitoring technologies

Since the introduction of consumer-grade, high-resolution Head-Mounted Displays (HMDs) in 2012, the IVE technology has been widely used in environmental psychology research in fields such as environmental restorativeness,^25,45 safety perception,⁴⁶ biophilic design,⁴⁷ environmental preferences and psychological responses to UPS,^21,22,48,49 form perception,^46,50 spatial navigation,⁵¹ etc. The IVE is capable of creating an immersive environment.^52,53 Thus, it not only compensates for the difficulty of precisely controlling the environmental variables and eliminating external distractions in field experiments,^2,54 but also ensures ecological validity in presenting subjects with scenes more vivid and complex,^46,55 stimulates a stronger sense of presence,^48,49,56 and in turn, effectively evokes physiological responses in subjects.^42,57

Meanwhile, for outcome measurement of environmental perception, the emerging physiological monitoring technologies can compensate for the ambiguity and low accuracy of results in traditional subjective evaluations caused by exogenous factors,⁵⁸ and have thus been increasingly used in environmental perception studies in recent years.⁵⁹ Specifically, some studies have used electrodermal activity (EDA) and electrocardiogram (ECG) metrics, which characterize the activities of the automatic nervous system (ANS) and thus capture psychological responses.^20,42,47,59 Some studies use electroencephalography (EEG), which characterizes central nervous system (CNS) activities in relevant environmental experiments. These new tools, combined with self-report psychological scale tables, allow comprehensive observation of the subject’s response to environmental exposures.^58,60,61 Latest research utilizing these new tools has explored topics including the effects of landscape images on adults’ psychophysiological restorative,⁵⁹ spatial perception and navigation performance in environments of different characteristics,⁶² the effects of buildings’ spatial interiors on occupants’ satisfaction and work efficiency,⁵⁸ and the effects of campus street trees in autumn on stress recovery.⁶³

In summary, the emerging measurement technologies provide researchers with a highly controllable experimental environment and a wealth of perceptual data acquisition options, enabling them to complete spatial evaluations and perceptual experiments in a more convenient, efficient, user-friendly manner.⁵⁸

Methodology

Scene selection

The experimental scenes were selected from the Old Bund in Ningbo, China. The site is an important local historic conservation area, containing several typical UPSs of different sizes and functions. At the same time, the site is rich in commercial activities, with a variety of urban functions such as recreation, entertainment, transportation and commerce, and is thus an important living and social area for the residents.

Based on the function of the site and the morphological characteristics of the built environment, we selected five representative scenes with distinctive appearances as environmental stimulus samples, namely A: the Cathedral Square; B: Square of Imperial Bank (an open waterfront square); C: Square in front of the Yu Residence (a historic building); D: Intersection of Bar Streets and E: Waterfront Art Promenade (Figure 1).

Figure 1.

Experimental scene.

We used a RICHO THETA V (Ricoh Company Ltd., Tokyo, Japan) dual fisheye panoramic camera to obtain panoramic video and audio in each scene. The video and audio were taken in clear weather during the same period on the morning of November 2020. 20 min of video and audio clips were captured from each of the five scenes, and 10 min of video and audio clips were retained after editing to remove clips that might have potential unintended effects on the subjects. In addition, we used a 5-min audio/video clip of the indoor laboratory environment as the exposure for the baseline. The immersive video and audio were displayed on an HTC VIVE VR headset (HTC Corporation, Taiwan, China) via Stream Media Player.

Indicators and questionnaire design

Indicator system for the physical environment features of the UPS scenes

Referring to the rationales in research on the physical features of urban streets and their edges,³⁶ elements of the micro-scale built environment,⁶⁴ landscape visual characters³⁵ and environmental attributes from the subjective environmental perception scale of public spaces,⁶⁵ and considering the built environment features of the study site, we designed a five-dimensional indicator system for auditing the physical features of UPS. The five dimensions include enclosure, transparency, complexity, imageability and human scale. They were then further divided into a total of 17 secondary indicators to quantify the physical features of UPS. After on-site audits, field surveys, manual identification and statistics computation, we assigned values to the environmental features of the study scenes to derive the corresponding indicator values for each scene. The structure of the indicator system and the descriptive statistics of the scenes are shown in Table 1.

Table 1.

Table for comprehensively quantifying the physical features of UPS scenes.

Perception	Physical feature	Indicator evaluation methodology	A The Cathedral Square	B Square of Imperial Bank	C Square in front of the Yu Residence	D Intersection of Bar Streets	E Waterfront Art Promenade
Enclosure	1. Scale	Event space area (m², obtained from geographic map data)	813	1012	267	314	426
	2. Proportion	Aspect ratio (field length/field width)	5.00	2.63	1.72	1.33	3.00
	2. Proportion	Height to width ratio (average height of surrounding buildings/width of site)	0.72	0.00	0.64	3.50	0.00
	3. Permeability relationship	Four-sided enclosure (4), three-sided enclosure (3), two-sided enclosure (2), one-sided enclosure (1), and four-sided open (0)	1	0	3	4	0
	4. Covert exposure relationship	Top shade share (shade area/total area)	0.00	0.05	0.20	0.40	0.30
Transparency	1. One-storey window opening ratio	Area of windows or openings on the first-floor façade/Total area of the first-floor façade	0.05	0.00	0.20	0.60	0.00
	2. Proportion of active use	Accessible or useable width of a floor/total width	0.20	0.00	0.33	1.00	0.00
	3. Functional openness	Semi-public (1), public (2)	2	2	1	1	2
Complexity	1. Main colours	The major use of colour grading	3	3	4	5	2
	2. Number of stores	Number of accessible stores in the surrounding area	0	0	2	8	0
	3. Pavement type	Single material and method (0), multiple materials or multiple methods (1), multiple materials and multiple methods (2)	1	1	0	2	1
	4. Type of greenery	The extent of use in planters, shrubs, trees, façade greening	3	2	3	1	2
Imaginability	1. Public artwork	Sculpture (1/0)	0	0	0	0	1
		Fountain pool (1/0)	1	0	1	0	0
		Art installation (1/0)	1	0	1	1	1
	2. Logo	Text logo (1/0)	0	1	1	1	1
	2. Logo	Image marks (1/0)	1	0	1	1	0
	3. Main landscape features	Open water (1/0)	0	1	0	0	1
Human scale	1. Rest facilities	Seats (1/0)	0	1	1	1	1
	1. Rest facilities	Table (1/0)	0	0	1	1	0
	2. Transportation facilities	Stairs (1/0)	1	1	0	1	1
	2. Transportation facilities	Isolation pile (1/0)	1	0	0	0	0
	3. Greening facilities	Flower pond (1/0)	1	1	1	1	1
		Shrubs (1/0)	1	1	0	0	1
		Arbour (1/0)	1	1	1	0	1
		Façade greening (1/0)	0	0	1	1	0
	4. Other facilities	Rubbish bins	0	1	1	1	0
	4. Other facilities	Streetlights (number visible within the field of view)	15	5	4	2	3

Stated Psychological Responses

A Chinese version of the Spielberger State-Trait Anxiety Inventory (STAI) was used to assess the short-term psychological status of the subjects.⁶⁶ Due to the large size of the questionnaire (full version), the six-item short form was used to reduce the impact of a lengthy experiment on the quality of the results.^67,68 The self-assessment questionnaire is efficient and concise, has proven to have good reliability and validity and has been widely used in different disciplines over the past decades.⁶⁹ The scale table consists of six words that measure intuitive expressions of current anxiety, including the negative words tense, upset and worried, and the positive terms relaxed, content and calm, in which the positive words were reversely scored. Subjects were asked to use four levels, including not at all, moderately, somewhat and very much to rate their feelings. The final psychological responses were the scores (total of six items) variation between exposure scenes and baseline phase, with higher scores indicating higher levels of anxiety gained (see Table 2).⁶⁸

Table 2.

Revised six-item short form of the State-Trait Anxiety Inventory.

Title	Rating
	Not at all	Moderately	Somewhat	Very much
I am calm	1	2	3	4
I am tense	1	2	3	4
I am upset	1	2	3	4
I am relaxed	1	2	3	4
I am content	1	2	3	4
I am worried	1	2	3	4
Minimum score 6	Highest score 24

Stated environmental preferences

In terms of the evaluation measures of stated environmental preferences from the user’s perspective, based on Mehta’s PSI and Zamanifard’s PSEQI, and considering the specific physical features and experimental conditions of the study site, we constructed a UPS quality evaluation system for Urban Public Space Evaluation (UPSE), which covers four dimensions (comfort, inclusiveness, sociability and attractiveness) of the user’s perspective. The table includes nine variables and 13 evaluation items (see Table 3 for details). The sum of the scores of each measurement item in each dimension yields the total score.

Table 3.

UPS quality evaluation scale table for environmental preference.

Dimension	Variable	Measurement item	Scoring criteria
Comfort	Walking comfort	I would like to take a walk here	1 = ‘Strongly disagree’ 2 = ‘Disagree’ 3 = ‘Not necessarily’ 4 = ‘Agree’ 5 = ‘Strongly agree’
		I think here 1. There is enough walking space 2. There are obstacles to obstruct the walking space	(The three items are summed)
			1 = ‘Yes’ 0 = ‘No’
		3. The road surface is suitable for walking	1 = ‘Yes’ 0 = ‘No’
	Rest comfort	I would love to rest here for a while	1 = ‘Strongly disagree’ 2 = ‘Disagree’ 3 = ‘Not necessarily’ 4 = ‘Agree’ 5 = "Strongly agree"
		I think here	(The three items are summed)
		1. There is enough rest space	1 = ‘Yes’ 0 = ‘No’
		2. Adequate shade
		3. Seating facilities suitable for rest
	Visual comfort	I think it has a great view	1 = ‘Strongly disagree’ 2 = ‘Disagree’ 3 = ‘Not necessarily’ 4 = ‘Agree’ 5 = ‘Strongly agree’
	Visual comfort	I think the scale here is comfortable
Inclusiveness	Security	I feel safe in this place	1 = ‘Strongly disagree’ 2 = ‘Disagree’ 3 = ‘Not necessarily’ 4 = ‘Agree’ 5 = ‘Strongly agree’
		I think this place is safe for the following people	(The three items are summed)
		1. Children	1 = ‘Yes’ 0 = ‘No’
		2. Women
		3. Elderly people
		4. People with disabilities
	Management level	I think the place is clean and tidy enough	1 = ‘Strongly disagree’ 2 = ‘Disagree’ 3 = ‘Not necessarily’ 4 = ‘Agree’ 5 = ‘Strongly agree’
	Sense of exclusion	I feel depressed when I stay here
Sociability	Event diversity	I would like to be here if I had the chance	(The four terms are summed)
		1. Rest	1 = ‘Yes’ 0 = ‘No’
		2. Walking
		3. Sports
		4. Photo taking
	Social potential	If given the opportunity, I would like to	(The three items are summed)
		1. Stay alone	1 = ‘Yes’ 0 = ‘No’
		2. Come here with family and friends
		3. Chat with other strangers here
Attractiveness	Visual appeal	I am satisfied with the colour environment here	1 = ‘Strongly disagree’ 2 = ‘Disagree’ 3 = ‘Not necessarily’ 4 = ‘Agree’ 5 = ‘Strongly agree’
		I am satisfied with the green environment here
		Here are some of the art facilities that appeal to me

Revealed physiological responses

The subjects’ physiological responses to UPS were obtained using the Emotiv EPOC+ device (EMOTIV Inc., San Francisco, CA, USA), which records the EEG data in real-time. The device has 14 electrode receptors and can calculate six measures of physiological states based on raw EEG data using specific algorithms. These states are Engagement, Excitement, Stress, Relaxation, Interest and Focus, measuring the degree of immersion in the present moment, the degree of having a positive arousal awareness or sensation, the comfort level of facing a current challenge, the ability to stop and recover from a highly focused state, the degree of attraction or aversion to a present stimulus, environment or activity and the degree of attentional focus on a specific task, respectively.⁷⁰

The physiological indicators were exported at a frequency of 2HZ, and each set of data consisted of raw values (RAW), which were compared to baseline data to calculate standardized values (SCALE). In this study, the mean of the SCALE values within each exposure interval was used to measure the corresponding physiological response status of the subjects. The standardized rate of change between experiment scenes and baseline phase was used as physiological responses outcome measurement.

Participants

The researchers recruited participants in the experiment through online social networks and on-site recruitment in December 2020, and 44 subjects were finally recruited. Subjects were divided into two groups according to whether they had professional backgrounds in disciplines such as landscape architecture, urban planning and architectural design. Due to the quality and completeness of the experiment data, we finally obtained 87 valid questionnaires and 66 valid EEG data (Table 4).

Table 4.

Descriptive statistics of subjects’ socio-demographic attributes.

Questionnaire participants			EEG experiment participants
Features	Quantity	Percentage (%)	Features	Quantity	Percentage (%)
Gender			Gender
Male	22	50	Male	17	52
Female	22	50	Female	16	48
Occupation			Occupation
Students	42	95	Students	31	94
Other	2	5	Other	2	6
Professional Background			Professional Background
Design professionals	16	36	Design professionals	13	39
Non-professional users	28	64	Non-professional users	20	61

The study strictly adhered to the Declaration of Helsinki and the ethical protocols of Peking University. All participants signed an informed consent form before the experiment and received compensation of 50 RMB after the end of the experiment. In addition, to ensure the objectivity of the experiment, enrollees with a history of epilepsy, colour blindness or other visual impairment conditions were not included as subjects. Also, subjects were asked to avoid smoking, drinking caffeinated or alcoholic beverages and strenuous exercise for 3 h before participating in the experiment to ensure data accuracy.

Experimental procedure

Reception: Subjects were led to a waiting area outside the laboratory to rest. After the subjects felt calm, they were guided to sign an informed consent form.

Preparation: After necessary inspections and the set-up of all the equipment, we guided the subjects into the laboratory, explained the procedure and helped them put on the Emotiv EPOC+ (detailed later) and HMD devices. After confirming their comfort and tolerance, we performed EEG baseline measurements of the subjects.

Scene perception phase: Each subject experienced a total of 2 stages of exposure to the UPS scene (i.e. the testing scene), with a baseline phase before the first UPS exposure and a resting phase between the two main testing exposure stages. In the baseline phase, we displayed the images of the laboratory to help subjects adapt to the VR environment, to improve the realism of the spatial feedback. After 2 min of exposure, subjects were asked to verbally fill in the psychological response questionnaire (using the STAI-6 questionnaire, detailed below). After the baseline stage was completed, we informed the subjects that they were about to enter the first testing scene and then started to play the testing video and audio. The allocation of the testing scene was based on the simple random principle. After 2 min of exposure at the first testing scene, the subjects were asked to verbally fill in the psychological response and environmental preference questionnaire (including the STAI-6 questionnaire and the UPSE questionnaire, detailed below), with questions of the same category asked randomly. At the end of the exposure, there was a 1-min break (i.e. the resting phase), during which the laboratory scene was replayed to calm the subjects. The second testing scene followed the same procedure as the first one.

End-of-experiment phase: After completing the exposure and questioning of both scenes, we informed the subjects that the experiment was over. We assisted them in removing the equipment, and guided them to fill in the basic demographic information questionnaire. Lastly, we organized the materials and guided the subjects to leave (Figure 2).

Figure 2.

Timeline of experimental procedure.

Statistical analysis

After basic descriptive statistical analysis of the experimental data, we first used the Mann–Whitney U non-parametric test to compare the scores of psychological/physiological responses and environmental preferences in each dimension with the overall scores, so as to examine the inter-scene and inter-group differences. Next, we used stepwise multiple linear regression to examine the effects of different physical features on environmental perception results, using the five dimensions of physical attributes of UPS as independent variables, psychological/physiological responses and environmental preference scores as dependent variables. The analysis was also conducted for both the whole sample and each group. In addition, we used the Spearman correlation index to analyze the consistency of the multi-source perceptual results, including psychological/physiological responses and environmental preferences.

Analytical framework

Based on the scope and methodological foundation synthesized in the literature review above, we designed the analytical framework of this study as follows (Figure 3). First, we recruited two groups of subjects with and without educational background on the built environment design. We then record their physiological and psychological responses, as well as environmental preferences in different UPS scenes with different built environment characteristics. Specifically, in synchronized monitoring and interviewing, we used a questionnaire to obtain the stated psychological responses and environmental preferences of the subjects, and simultaneously observed their revealed physiological responses to UPS scenes through physiological measurement. Meanwhile, we established a five-dimensional indicator system to quantify the physical features of UPS scenes. Next, with the empirical data and metrics obtained in the previous steps, we built statistical models to compare the differences in psychological/physiological responses and environmental preferences between the two groups, and then analyzed the correlation between the different environmental features and the subjects’ perception, so as to explain the possible reasons for the differences, and also the design implications.

Figure 3.

Analytical framework.

Results

Inter-scene perceptual difference

The results, including the psychological/physiological responses and the environmental preferences from the valid samples are shown in Figure 4 (a, c and d for the three types of indicators, respectively; b for pre-post psychological variation; see Table 5 for detailed statistics). In terms of psychological responses as reflected in the subjects’ state anxiety status, after exposure to the experimental scenes, the subjects’ state anxiety status demonstrated changes compared to the baseline scene, and the level of state anxiety in some scenes was increased compared to the baseline stage (Figure 4(a)), reflecting the arousal effect of the respective scenes on the subjects’ psychological states.⁵⁷ However, one-on-one Non-parametric Mann–Whitney U tests on the subjects’ changes in the state anxiety levels generally yielded insignificant results (Figure 4(b); see Table 6 for detailed statistics).

Figure 4.

Inter-scene perceptual difference. (a) Total STAI scores of different stages; (b) psychological responses; (c) physiological responses of all states; (d) environmental preference. Notes: ** p $<$ 0.01, * p $<$ 0.05, N = 87 in (a), (b) and (d), N = 66 in (c).

Table 5.

Descriptive statistics of psychological response (state anxiety levels change compared to baseline).

		N	Minimum value	Maximum value	Average value	Standard deviation
STAI	All scenes	87	−6	6	0.02	3.114
	A	19	−5	5	−0.37	2.692
	B	17	−5	6	1.18	3.087
	C	20	−6	4	−0.95	3.170
	D	14	−6	6	0.29	4.084
	E	17	−3	6	0.24	2.463

Table 6.

Results of the corresponding Mann–Whitney U test for inter-scene psychological response.

	Scene comparison grouping
	A–B	A–C	A–D	A–E	B–C	B–E	B–E	C–D	C–E	D–E
Status anxiety level	0.199	0.745	0.674	0.533	0.091	0.604	0.443	0.309	0.425	0.968

In terms of physiological responses as reflected in the EEG data, results from descriptive statistics (Table 7) and tests of variance (Table 8) on the 66 valid cases showed that there were statistically significant differences between subjects in Scene A and C, and Scene C and D on the two dimensions of stress and interest, respectively. People were more ‘stressed’ in Scene A than in Scene C, and less ‘interested’ in Scene C than in Scene D (Figure 4(c)). There were no significant differences in physiological responses in any of the other scene comparisons.

Table 7.

Descriptive statistics of physiological response (standardized rate of change).

		N	Minimum value	Maximum value	Average value	Standard deviation
EEG.En	All scenes	66	−0.384	0.877	0.042	0.233
	A	13	−0.038	0.248	0.078	0.027
	B	13	−0.034	0.877	0.048	0.086
	C	15	−0.302	0.507	0.008	0.053
	D	12	−0.269	0.825	0.032	0.081
	E	13	−0.225	0.671	0.050	0.070
EEG.Ex	All scenes	66	−0.72	13.943	0.506	2.190
	A	13	−0.235	10.777	1.109	0.821
	B	13	−0.476	0.608	0.074	0.080
	C	15	−0.720	13.943	1.072	0.938
	D	12	−0.580	1.817	0.071	0.178
	E	13	−0.276	1.409	0.172	0.115
EEG.St	All scenes	66	−0.691	0.890	−0.033	0.300
	A	13	−0.136	0.890	0.138	0.082
	B	13	−0.514	0.563	−0.010	0.078
	C	15	−0.625	0.499	−0.132	0.077
	D	12	−0.549	0.417	−0.028	0.089
	E	13	−0.691	0.199	−0.116	0.077
EEG.Re	All scenes	66	−0.439	4.103	0.278	0.771
	A	13	−0.129	1.548	0.275	0.129
	B	13	−0.439	0.876	0.083	0.115
	C	15	−0.432	0.921	0.102	0.104
	D	12	−0.436	4.103	0.467	0.343
	E	13	−0.273	3.955	0.507	0.301
EEG.In	All scenes	66	−0.29	0.479	0.058	0.126
	A	13	−0.034	0.479	0.082	0.039
	B	13	−0.068	0.392	0.084	0.038
	C	15	−0.290	0.127	−0.011	0.030
	D	12	−0.106	0.404	0.058	0.039
	E	13	−0.003	0.238	0.089	0.023
EEG.Fo	All scenes	66	−0.564	1.065	0.021	0.324
	A	13	−0.214	0.723	0.104	0.077
	B	13	−0.320	1.035	0.122	0.097
	C	15	−0.546	1.065	0.060	0.118
	D	12	−0.564	0.059	−0.097	0.053
	E	13	−0.527	0.189	−0.099	0.060

Table 8.

Results of Mann–Whitney U test for inter-scene physiological response.

	Scene comparison grouping
	A–B	A–C	A–D	A–E	B–C	B–D	B–E	C–D	C–E	D–E
EEG.En	0.270	0.279	0.218	0.103	0.945	0.720	0.828	1.00	0.903	0.849
EEG.Ex	0.397	0.333	0.701	0.157	0.629	0.608	0.586	0.475	0.770	0.328
EEG.St	0.209	0.007**	0.144	0.744	0.134	0.522	0.807	0.747	0.262	0.384
EEG.Re	0.209	0.259	0.939	0.913	0.800	0.228	0.277	0.222	0.495	0.786
EEG.In	0.939	0.205	0.397	0.663	0.134	0.663	0.849	0.016*	0.305	0.355
EEG.Fo	0.758	0.369	0.158	0.149	0.420	0.077	0.057	0.662	0.733	0.870

Notes: ** p $<$ 0.01, * p $<$ 0.05, N $= 66$ .

In terms of environmental preferences, results from descriptive statistics showed that Scene E, the Waterfront Art Promenade, received the highest preference score, while Scene A, the Cathedral Square, received the lowest (Figure 4(d); Table 9). Further, inter-scene Non-parametric Mann–Whitney U tests on the subjects’ total environmental preference scores yielded insignificant results (Figure 4(d); Table 10). However, there were between-group differences in the four perceptual evaluation dimensions.

Table 9.

Descriptive statistics of environmental preference score.

		N	Minimum value	Maximum value	Average value	Standard deviation
Comfort	All scenes	87	9	24	18.57	3.620
	A	19	11	22	17.16	3.500
	B	17	10	24	19.71	3.255
	C	20	9	22	17.45	3.762
	D	14	12	24	19.14	4.055
	E	17	14	23	19.88	2.913
Inclusiveness	All scenes	87	7	19	13.38	2.557
	A	19	8	16	12.26	2.207
	B	17	7	17	12.26	2.55
	C	20	8	17	14.3	2.536
	D	14	9	19	12.64	2.735
	E	17	10	18	14.12	2.446
Sociability	All scenes	87	0	7	4.64	1.705
	A	19	1	6	4.26	1.661
	B	17	1	7	5.06	1.713
	C	20	1	7	5	1.806
	D	14	0	7	3.79	1.847
	E	17	1	6	4.94	1.298
Attractiveness	All scenes	87	5	14	9.84	2.118
	A	19	5	14	10.89	2.355
	B	17	5	11	8.65	1.869
	C	20	7	13	10.2	1.795
	D	14	6	13	9.21	2.424
	E	17	8	12	9.94	1.56
Total score	All scenes	87	28	60	46.44	8.162
	A	19	32	54	44.58	7.705
	B	17	28	58	46.82	8.323
	C	20	28	57	46.95	8.532
	D	14	30	60	44.70	9.415
	E	17	34	59	48.88	7.088

Table 10.

Results of Mann–Whitney U test for inter-scene environmental preference.

	Asymptotic significance (two-tailed
Preference evaluation		A–B	A–C	A–D	A–E	B–C	B–D	B–E	C–D	C–E	D–E
	Comfort	0.037*	0.734	0.133	0.016*	0.068	0.889	1.000	0.125	0.042*	0.779
	Inclusiveness	0.145	0.004**	0.883	0.038*	0.146	0.238	0.532	0.031*	0.700	0.101
	Sociability	0.120	0.131	0.353	0.199	0.950	0.055	0.534	0.043*	0.625	0.042*
	Attractiveness	0.004**	0.212	0.060	0.107	0.032*	0.440	0.066	0.228	0.608	0.408
Total score	0.287	0.323	0.907	0.081	0.939	0.474	0.490	0.410	0.521	0.177

Notes: ** p $<$ 0.01, * p $<$ 0.05, N $= 87$ .

Overall, there were arousal effects but no significant differences in the subjects’ psychological responses between scenes. In contrast, some inter-scene differences were found in the subjects’ physiological responses and environmental preferences. The high revealed ratings of the Waterfront Art Promenade on the ‘comfort’ dimension and the low ratings of road intersections on the ‘sociability’ dimension are to some extent consistent with previous studies.⁷¹ These results indicate that the experimental scenes are diverse enough to arouse the participants’ responses and perceptions.

In addition, we used Spearman bi-variate correlation analysis to analyze the consistency of the psychological/physiological responses and environmental preference results (Table 11 for sample K-S test results, and Table 12 for the correlation test results). The tests generally yielded complicated results: psychological responses were negatively correlated with all items of environmental preference; only the ‘engagement’ and ‘interest’ dimensions in physiological response were weakly and negatively correlated with certain dimensions of environmental preference, but all physiological response indicators were not correlated with psychological responses. The results are consistent with the findings in previous studies,²² and imply the complex relationships among different measurements of human emotions, which we do not elaborate on here as they are not within the scope of this research.

Table 11.

Results of the one-sample K-S test of psychological response, physiological response and environmental preference.

	STAI	Comfort	Inclusiveness	Sociability	Attractiveness	Total score	EEG.En	EEG.Ex	EEG.St	EEG.Re	EEG.In	EEG.Fo
Number of cases	66	66	66	66	66	66	66	66	66	66	66	66
Average value	0.4545	18.1515	13.1061	4.4697	9.5455	45.2727	0.04244	0.5058	−0.03282	0.27841	0.05845	0.02095
Standard deviation	3.21622	3.58305	2.51842	1.81633	2.24761	8.27144	0.232786	2.189509	0.299799	0.77149	0.126388	0.324202
Asymptotic significance (two-tailed)	0.200^c,d	0.000 c	0.200^c,d	0.000^c	0.012^c	0.073^c	0.001^c	0.000^c	0.200^c,d	0.000^c	0.008^c	0.000^c

^aTest distribution is normal.

^bCalculated from data.

^cRiley’s significance correction.

^dThis is the lower limit of true significance.

Table 12.

Results of Pearson correlation analysis between phycho-physiological response and environmental preference.

	EEG.En	EEG.Ex	EEG.St	EEG.Re	EEG.In	EEG.Fo	STAI
STAI	0.044	0.008	0.165	0.14	0.128	0.105	1
Comfort	−0.046	−0.149	−0.047	−0.13	−0.132	0.056	−0.251*
Inclusiveness	−0.143	−0.171	0.01	−0.175	−0.254*	0.181	−0.373**
Sociability	−0.136	−0.077	−0.069	−0.234	−0.282*	0.134	−0.287*
Attractiveness	−0.156	0.044	−0.057	−0.068	−0.135	0.069	−0.359**
Total score	−0.112	−0.117	−0.037	−0.144	−0.195	0.13	−0.382**

Notes: ** p $<$ 0.01, * p $<$ 0.05, N $= 87$ .

Inter-group perceptual difference

The inter-group psychological responses in the five scenes were tested for differences (see Table 13 for detailed results). The results show that there is no significant difference in the changes in the state anxiety level between the two groups in most scenes. One exception is Scene C, which is enclosed by buildings on three sides, where there is a difference in psychological responses between the two groups (p = 0.001), and non-professional users have higher changes in anxiety levels in this scene (Figure 5).

Table 13.

Results of the Mann–Whitney U test for inter-group psychological response.

	Scenes
	Scene A	Scene B	Scene C	Scene D	Scene E
Asymptotic significance (two-tailed)	0.326	0.843	0.001**	0.887	0.286

Notes: ** p $<$ 0.01, * p $<$ 0.05, N $= 87$ .

Figure 5.

Inter-group psychological responses. Notes: ** p $<$ 0.01, * p $<$ 0.05, N $= 87$ .

Next, the results of the test for inter-group differences were obtained by comparing the changes in state anxiety levels in different scenes (see Table 14 for detailed results). The results showed interesting patterns. In the comparison between Scenes A and C, both the non-professional users and the design professionals showed differences, with non-professional users showing higher changes in anxiety levels in Scene C, while the design professional subjects showed the opposite (Figure 5). In addition, design professionals showed lower levels of anxiety changes in Scene C compared to Scenes B and E.

Table 14.

Results of Mann–Whitney U test for psychological response between groups at different scenes.

	Scene comparison grouping
Asymptotic significance (two-tailed)		A–B	A–C	A–D	A–E	B–C	B–D	B–E	C–D	C–E	D–E
	Non-professional users	0.184	0.038*	0.487	0.481	0.868	0.621	0.44	0.842	0.057	0.765
	Design professionals	0.439	0.009**	0.85	0.567	0.009**	0.698	0.736	0.122	0.008**	0.707

Notes: ** p $<$ 0.01, * p $<$ 0.05, N $= 87$ .

In terms of physiological responses between the two groups, a similar analysis (see Table 15 for detailed results) showed that only in Scene B, the open waterfront square, the two groups showed a difference in the ‘focus’ dimension, with non-professional users focussing on a lower level than design professionals (Figure 6(d)).

Table 15.

Results of Mann–Whitney U test for inter-group physiological response.

	Asymptotic significance (two-tailed)
	Scene A	Scene B	Scene C	Scene D	Scene E
EEG.En	0.884	0.306	0.908	0.643	0.465
EEG.Ex	0.380	0.380	0.563	0.643	0.935
EEG.St	0.770	0.057	1.000	0.643	0.808
EEG.Re	1.000	0.884	0.817	0.44	0.685
EEG.In	0.770	0.242	0.908	0.247	0.871
EEG.Fo	0.661	0.013*	0.908	0.123	0.935

Notes: ** p $<$ 0.01, * p $<$ 0.05, N $= 66$ .

Figure 6.

Inter-group physiological responses (a) EEG.Ex; (b) EEG.St; (c) EEG.In; (d) EEG.Fo. Notes: ** p $<$ 0.01, * p $<$ 0.05, N $=$ 66.

Likewise, inter-group difference tests on the outcome of the physiological response in different scenes (see Table 16 for detailed results) showed that non-professional users had stronger ‘excitement’ levels in Scene A versus B and E (Figure 6(a)), higher ‘stress’ levels in Scene A versus B (Figure 6(b)), and higher ‘interest’ levels in Scene D versus C and E (Figure 6(c)). In contrast, design professionals had higher ‘stress’ levels in Scene E versus D and E (Figure 6(b)), higher ‘interest’ levels in Scene E versus C (Figure 6(c)), and higher ‘focus’ levels in Scene B versus D and E (Figure 6(d)). Overall, the two groups showed completely inconsistent patterns in physiological responses to UPS scenes with different physical features.

Table 16.

Results of Mann–Whitney U test for physiological response between groups at different scenes.

	Scene comparison grouping
Asymptotic significance (two-tailed)			A–B	A–C	A–D	A–E	B–C	B–D	B–E	C–D	C–E	D–E
	EEG.En	Non-professional users	0.248	0.462	0.336	0.064	0.674	0.962	0.728	0.810	0.418	0.491
	EEG.En	Design professionals	0.754	0.465	1.000	0.917	0.685	0.462	0.917	0.705	0.808	0.389
	EEG.Ex	Non-professional users	0.036*	0.189	0.501	0.021*	0.462	0.149	0.817	0.248	0.728	0.153
	EEG.Ex	Design professionals	0.602	0.935	0.624	0.754	0.935	0.624	0.602	0.705	0.935	0.806
	EEG.St	Non-professional users	0.009**	0.059	0.700	0.418	0.462	0.290	0.602	0.773	0.908	0.711
	EEG.St	Design professionals	0.754	0.088	0.050	0.917	0.123	0.142	0.602	0.925	0.062	0.027*
	EEG.Re	Non-professional users	0.400	0.344	0.630	0.247	1.000	0.211	1.000	0.178	0.643	0.101
	EEG.Re	Design professionals	0.602	0.223	0.462	0.251	0.570	0.624	0.175	0.705	0.062	0.142
	EEG.In	Non-professional users	0.529	0.431	0.386	0.165	0.753	0.124	0.487	0.043*	0.643	0.017*
	EEG.In	Design professionals	0.465	0.372	1.000	0.251	0.088	0.327	1.000	0.257	0.042*	0.142
	EEG.Fo	Non-professional users	0.248	0.753	0.336	0.083	0.529	0.923	0.298	0.630	0.355	0.186
	EEG.Fo		Design professionals	0.076	0.167	0.142	0.754	0.167	0.014*	0.009**	0.571	0.570	0.086

Notes: ** p $<$ 0.01, * p $<$ 0.05, N $= 66$ .

In terms of environmental preferences, the overall score and the scores from each dimension between the two groups in the five scenes were tested for differences (see Table 17 for detailed results). In all dimensions where inter-group perceptual differences existed, including comfort, inclusiveness, attractiveness and the overall score, the mean scores from design professionals were uniformly lower than those from non-professional users (Figure 7).

Table 17.

Results of Mann–Whitney U test for inter-group environmental preference score.

Asymptotic significance (two-tailed)
	Scene A	Scene B	Scene C	Scene D	Scene E
Comfort	0.022*	0.657	0.293	0.721	0.261
Inclusiveness	0.008**	0.844	0.693	0.775	0.025*
Sociability	0.062	0.482	0.477	0.658	0.373
Attractiveness	0.051	0.501	0.144	0.392	0.057
Total Score	0.007**	0.922	0.333	0.777	0.064

Notes: ** p $<$ 0.01, * p $<$ 0.05, N $= 87$ .

Figure 7.

Inter-group environmental preference: (a) comfort score, (b) inclusiveness score, (c) sociability score, (d) attractiveness score, (e) total score. Notes: ** p $<$ 0.01, * p $<$ 0.05, N = 87.

Again, inter-group difference tests were performed on the environmental preferences in different scenes (see Table 18 for detailed results). In the comfort dimension, design professionals’ preferences in Scenes A and B differed significantly, with higher comfort scores in B (Figure 7(a)). In the inclusiveness dimension, design professionals’ evaluations of Scenes A, B and C differed, with higher scores in Scenes B and C than in A (Figure 6(b)). In contrast, non-professional users’ preference differences were reflected in the comparison of Scenes D and E, with higher comfort scores in scene D (Figure 7(a)). In the sociability dimension, non-professional users rated the former higher in the comparison between Scenes C and D (Figure 7(c)). In the attractiveness dimension, Scenes A, B and D, and Scenes B, C and E all had preference differences, with Scene A receiving higher attractiveness scores than both B and D, and Scene B scoring lower than C and E (Figure 7(d)). The implications of each finding in this section will be further elaborated in the discussion section.

Table 18.

Results of Mann–Whitney U test for environmental preference score between groups at different scenes.

	Scene comparison grouping
		A–B	A–C	A–D	A–E	B–C	B–D	B–E	C–D	C–E	D–E
Comfort	Non-professional users	0.287	0.907	0.666	0.067	0.151	0.674	0.894	0.442	0.081	0.405
Comfort	Design professionals	0.025*	0.382	0.184	0.074	0.144	0.242	0.682	0.233	0.339	0.539
Inclusiveness	Non-professional users	0.790	0.079	0.271	0.050	0.424	0.322	0.320	0.084	0.681	0.036*
Inclusiveness	Design professionals	0.012*	0.009**	0.151	0.157	0.188	0.630	0.213	0.228	0.103	0.457
Sociability	Non-professional users	0.319	0.289	0.136	0.580	0.918	0.089	0.464	0.036*	0.521	0.065
Sociability	Design professionals	0.090	0.197	0.846	0.136	0.768	0.210	0.734	0.388	0.940	0.215
Attractiveness	Non-professional users	0.001**	0.143	0.031*	0.051	0.013*	0.187	0.016*	0.270	0.633	0.457
Attractiveness	Design professionals	1.896	0.906	0.390	0.613	0.768	0.499	0.734	0.226	0.366	0.371
Total score	Non-professional users	0.947	0.580	0.355	0.156	0.791	0.520	0.355	0.234	0.434	0.113
Total score	Design professionals	0.034*	0.147	0.340	0.103	0.772	0.705	0.463	1.000	0.883	0.806

Notes: ** p $<$ 0.01, * p $<$ 0.05, N $= 87$ .

Influence of physical features on environmental perception

To further explain the inter-scene variability in psychological/physiological responses and environmental preferences found, we used physical features of the UPS scenes as explanatory variables and performed stepwise multiple linear regression for the perceptual indicators. The results are shown in Table 19. In general, physical environmental features induced changes in environmental preferences and physiological responses but had no significant effect on psychological responses.

Table 19.

Selected regression results for all samples.

Dependent variable	Independent variable	B	Beta	t
STAI	-	-	-	-
Comfort (Adjusted $R^{2}$ = 0.090)	(Constant)	19.604** (0.498)		39.328
Comfort (Adjusted $R^{2}$ = 0.090)	Fountain pool	−2.296** (0.745)	−0.317	−3.085
Inclusiveness (Adjusted $R^{2}$ = 0.043)	(Constant)	12.263** (0.574)		21.362
Inclusiveness (Adjusted $R^{2}$ = 0.043)	Text logo	1.428* (0.649)	0.232	2.199
Sociability (Adjusted $R^{2}$ = 0.048)	(Constant)	4.942** (0.220)		22.416
Sociability (Adjusted $R^{2}$ = 0.048)	Height to width ratio	−0.344*(0.149)	−0.242	−2.304
Attractiveness (Adjusted $R^{2}$ = 0.110)	(Constant)	9.837** (0.404)		24.329
	Fountain pool	1.151* (0.435)	0.272	2.648
	Rubbish bins	−0.877* (0.439)	−0.205	−1.997
EEG.En	-	-	-	-
EEG.Ex	-	-	-	-
EEG.St (Adjusted $R^{2}$ = 0.088)	(Constant)	−0.232** (0.082)		−2.829
EEG.St (Adjusted $R^{2}$ = 0.088)	Aspect ratio	0.074** (0.027)	0.319	2.690
EEG.Re	-	-	-	-
EEG.In (Adjusted $R^{2}$ = 0.075)	(Constant)	−0.011 (0.031)		−0.342
EEG.In (Adjusted $R^{2}$ = 0.075)	Stairs	0.090* (0.036)	0.299	2.508
EEG.Fo (Adjusted $R^{2}$ = 0.062)	(Constant)	0.134* (0.063)		2.142
EEG.Fo (Adjusted $R^{2}$ = 0.062)	Top shade share	−0.600* (0.261)	−0.276	−2.297

Notes: ** p $\leq$ 0.01, * p $\leq$ 0.05; N = 87 in measuring psychological responses and environmental preferences, N = 66 in measuring physiological responses; Standard errors in parathesis.

Specifically, in terms of the stated environmental preferences, fountain pools and the comfort score were negatively correlated (R² = 0.090, p = 0.003); text logos and the inclusiveness score showed a positive correlation (R² = 0.043, p = 0.0031); aspect ratio and the sociability score were negatively correlated (R² = 0.042, p = 0.032); fountain pools and the attractiveness score were positively correlated (R² = 0.110, p = 0.010); while rubbish bins were negatively correlated with the attractiveness score. In terms of the revealed psychological responses, stress (EEG.St) was positively correlated with a site aspect ratio (R² = 0.088, p = 0.009); interest (EEG.In) was positively correlated with stairs (R² = 0.075, p = 0.015); and attention (EEG.Fo) was negatively correlated with top shade share (R² = 0.062, p = 0.025).

Further, we construct similar models for the two groups to reveal the specific mechanisms through which the physical features of UPS influence human’s environmental perceptions, as well as the inter-group differences as reported. The results are shown in Table 20. Overall, differences in psychological responses, physiological responses and environmental preferences were found between the subjects of the two groups. First, the state anxiety levels change of design professionals was positively correlated with stairs (R² = 0.339, p < 0.001), while that of non-professional users did not show a significant correlation with environmental features. Second, the fountain pool (R² = 0.208, p = 0.010) and aspect ratio (R² = 0.241, p = 0.003) were negatively correlated with the comfort and inclusiveness scores of the design professionals’ environmental preference ratings, respectively, while those of the non-professional users did not correlate with any of the physical features in these two dimensions. In contrast, in the sociability and attractiveness dimensions, where the environmental preferences of the design professionals were not correlated with any of the physical features, non-professional users’ ratings were positively correlated with arbours (R² = 0066, p = 0.031) and art installation (R² = 0.260, p <0.001), and were negatively correlated with the top shade share (R² = 0.260, p = 0.006). Finally, in terms of physiological responses, non-professional users and design professionals’ stress levels (EEG.St) were positively correlated with streetlights (R² = 0.074, p = 0.050) and functional openness (R² = 0.218, p = 0.009), respectively. Also, design professionals’ interest (EEG.In) and focus (EEG.Fo) levels were positively correlated with image marks (R² = 0.200, p = 0.013), and negatively correlated with top shade share (R² = 0.118, p = 0.048), respectively.

Table 20.

Regression results for samples from the two groups.

	Non-professional users					Design professionals
Dependent variable	Adjusted $R^{2}$	Independent variable	B	Beta	t	Adjusted $R^{2}$	Independent variable	B	Beta	t
STAI		-	-	-	-	0.339	(Constant)	−4.000** (0.990)		−4.042
STAI							Stairs	4.652** (1.149)	0.601	4.049
Comfort		-	-	-	-	0.208	(Constant)	19.188** (0.905)		21.201
Comfort							Fountain pool	−3.587** (1.301)	−0.456	−2.757
Inclusiveness		-	-	-	-	0.241	(Constant)	15.294** (0.876)		17.451
Inclusiveness							Aspect ratio	−0.916** (0.282)	−0.516	−3.246
Sociability	0.066	(Constant)	3.800** (0.508)		7.476
Sociability		Arbour	1.243* (0.561)	0.289	2.217
Attractiveness	0.260	(Constant)	8.655** (0.569)		15.214
		Art installation	3.105** (0.688)	0.275812	4.510
		Top shade share	−5.093** (1.789)	−0.367	−2.847
EEG.En
EEG.Ex
EEG.St	0.074	(Constant)	−0.180*		−2.642	0.218	(Constant)	−0.478* (0.188)		−2.547
EEG.St		Streetlights	0.018*	0.312	2.024		Functional openness	0.321** (0.114)	0.500	2.826
EEG.Re
EEG.In						0.200	(Constant)	0.151** (0.040)		3.732
EEG.In							Image marks	−0.138* (0.051)	−0.481	−2.690
EEG.Fo						0.118	(Constant)	0.275* (0.121)		2.277
EEG.Fo							Top shade share	−1.092* (0.524)	−0.392	−2.086

Notes: ** p $\leq$ 0.01, * p $\leq$ 0.05; N = 87 in measuring psychological responses and environmental preference, N = 66 in measuring physiological responses; standard errors in parathesis.

Conclusion and discussion

Utilizing IVE and physiological measurement tools, this study revealed the inter-group difference in the perception of UPS with different physical features for people with or without design-related professional backgrounds. With three types of measurements for human’s environmental perception, including psychological/physiological responses and environmental preferences, we revealed the specific patterns of such differences and also their possible causes. Besides, the improved analytical framework for integrating subjective (stated) and objective (revealed) psychological/physiological responses and environmental preferences data developed in this study has methodological value. Meanwhile, the empirical findings of this study regarding the inter-group and inter-scene differences in environmental perceptions may provide a new basis for evidence-based UPS design practices. The study also suggests insights into residents’ perceptions and potential wishes, and sheds light on the plausible enlightenment that may help guarantee an active utilization of spontaneous and social activities as a basis for future UPS designing, in particular a more attractive UPS for non-professional users.

Aspects of Professional Background-driven Perceptual Differences: Psychological/Physiological Responses, Environmental Preferences and Health Implications

The primary finding of this study is that even within the exact same UPS scene, there are significant differences in psychological/physiological responses and in environmental preferences between the two groups who are with or without a professional background in built-environment design. Overall, design professionals’ scores are lower on average than non-professional users on the dimensions of comfort, inclusiveness and attractiveness of the UPS, which is consistent with findings from previous studies that the landscape and architectural designers tend to be more critical and even ‘demanding’ in judging the quality of the built environment, and are more ‘intolerant’ with low-quality design schemes.^14,18 Likewise, non-professional users of UPS tend to be more ‘indifferent’ than design professionals judging from their psychological and physiological responses. Meanwhile, the two groups differ in their aspects of concern on different physical environment features of UPS. Differences in the evaluation of UPS scenes by design professionals exist mainly on the comfort and inclusiveness dimensions, while those by non-professional users concentrate on the sociability and attractiveness dimensions. Based on these findings, it can be concluded that there are structural differences between the two groups regarding their sensitivity, priority of concerns and judging criteria of the built environment in psychological, physiological and preferential terms. Such differences may induce health hazards such as anxiety to the general users of the UPS as the designers’ anticipation of the places’ quality and the functionality may deviate from the former’s actual experiences.

Moreover, and for UPS particularly, under the theoretical framework of Gehl⁷² and Thompson,⁷³ activities in the UPS can be classified as necessary activities which are related to walking, spontaneous activities that occur under suitable outdoor conditions, and social activities that depend on the participation of others. In terms of subjective environmental preference, design professionals are more likely to notice site features associated with necessary activities, while non-professional users focus more on on-site features related to spontaneous and social activities. The finding is in line with previous studies on the two groups’ perceptions of natural landscapes, which state that design professionals have a more practical view of the landscape. At the same time, non-professional users are more concerned with the ‘enjoyment’ process.¹³

Physical features influencing factors for environment perception: Design takeaways

Beyond inter-group environmental perceptual differences per se, quantified analysis of the relationship between such differences and physical features of UPS reveals additional mechanistic details with distinct operational implications. Overall, for both groups, there are significant differences in the mechanisms of physical environmental features triggering perceptual differences. For design professionals, their state anxiety levels change, comfort scores and inclusiveness scores on the UPS improve with the presence of stairs and deteriorate with the presence of a fountain pool and increased aspect ratio of the site. However, non-professional users are insensitive to these physical features. Instead, for non-professional users, sociability scores are increased by the presence of arbours, probably for the reason as previous studies have concluded that the presence of trees creates small open or semi-enclosed spaces that provide good spaces for interaction.^39,74 Also, in terms of attractiveness scores, art installations increase the attractiveness scores of non-professional users, which is also consistent with previous research finding that public artwork enhances the excitement of urban environments and makes landscapes more interesting³⁷; while the increase of the top shade share reduces the group’s attractiveness scores. Further, the factors that elicit physiological response differences between groups are also quite different. For non-professional users, only the presence of streetlights increases their stress levels; whereas for design professionals, things are much more complicated: functional openness is positively correlated with their stress levels; image marks are negatively correlated with their interest levels; top shade share was negatively correlated with their focus levels. In conclusion, there is little similarity in the structure of physical environmental features eliciting perceptual responses and preferences between the two groups.

Practically, the above findings suggest that design professionals’ tendency of being ‘intolerant’ to building environmental quality and ‘sensitive’ to specific physical features may mislead their designs from ordinary users’ perspectives. On this, the revealed patterns and mechanisms of physical environmental features affecting user perception in this paper can, to a certain extent, provide new ideas and specific guidelines for UPS design practice. Designers should pay more attention to the social needs of users, as well as to improve the attractiveness and inclusiveness of UPS by enhancing the presence of arbours and quality art facilities, optimizing the top shade share, and providing necessary textual signage to encourage more people to enter and use UPS. On the contrary, the fountain pool, although being a popular design feature, has a contradictory effect on user perception. Further, the classic design parameter of width-to-height ratio affects the pressure level of users while also affecting the sociability of the site; the presence of stairs also has the dual role of triggering pressure and interest. The appropriate trade-off between these contradictory design elements to create appropriate place experiences in different built environment scenes is an important issue for designers, and the quantitative findings and analytical framework of this paper may provide empirical and methodological references for this task.

Towards a new paradigm of public participation in UPS design

The findings, however, should not only be interpreted within the scope of a particular case study, nor even be limited to specific design takeaways. Rather, they should indicate possible opportunities for building a new paradigm of public participation in UPS design.

The findings imply the potential utility of public participation in the creation of successful UPS. Admittedly, the experts’ knowledge is important for their expertise and awareness of the fundamental UPS quality control. Their focus on more necessary activity-related aspects guarantees the safety, security and the functionality of the UPS design. Users’ knowledge, however, provides useful information on the social and communicational aspects of the UPS design. It can also be a valuable resource to understand the context of the aimed area that awaits to be planned and designed. Therefore, designers should try to overcome the lack of understanding of user needs and their limitations of perception due to professional inertia and try to think from ordinary users’ perspectives as much as possible.

Public participation facilitates the flow of information between designers and users and thus has long been involved in the design process. Indeed, the arrogant early Modernist statements of ‘the design of a city is too important to be left to its citizens’ (Le Corbusier)⁷⁵ or ‘lay people have no capacity to choose’ (van der Rohe)⁷⁶ have been proven too much paternalism per the famous critics from Kevin Lynch, Christopher Alexander and Jane Jacobs,^34,77,78 and have been largely disregarded by today’s designers. Nevertheless, public participation to date has generally been limited within negotiations on the overall goal of the design, or communications of general design ideas between designers and users through simple diagrams, because ordinary users tend to lack the professional skills which designers possess to specifically express their design ideas. The professional thresholds here handicap detailed exchange of ideas between the two parties, and thus the effectiveness of public participation in this style is still questionable. Further, once a UPS has been built, there usually lacks room for remedying any pities caused as a result of the above design process due to potentially high costs, and Post-Occupational Evaluations (POE), whether the public participated or not, are on most occasions only meaningful to future projects, still leaving a prominent gap to fill between the designers and users.

Our approach, in contrast, hints at an opportunity of low-cost, a prior evaluations of detailed design schemes with the aid of the IVE and psychological-physiological monitoring technologies. Designers can now invite anticipated users to enter the IVE, and observe their revealed results, in addition to stated preferences on the alternative design schemes, such that the most genuine response of the users can be obtained which would be of great value in the evolution of subsequent designs. The process can be iterated for several rounds until the evaluations from the two parties converge to a mutually optimal state. In some sense, the procedure described above effectively turns the conventional POE to Pre-Occupational Evaluation (PreOE), which might constitute a somewhat revolutionary change in the paradigm of design.

We are inclined to be optimistic to this prospect as the necessary technological infrastructure involved appears to be ready. Of course, social, communicational and administrative infrastructures and institutions are equally important. For example, the participatory tools can be taken online for planners to ‘broadcast’ their projects to a larger population, and specific feedbacks can be provided that help inform citizens of how their ideas have been addressed.⁷⁹ On the policy level, as top-down bureaucratic approach is still dominant in public policy-making in China,²⁴ strategic approach to incorporate participatory decision-making on the policy level should be given more stress. Former research has found that the use of residents’ input continues to be challenged by categorizations such as seeing their input in a separate box within planning, meaning that the current legislation and policies encourage the reproduction of such categorizations rather than the integration of different types of knowledge.⁸⁰ Therefore, it is pertinent that governments must reinforce the role of users’ input and strengthen communication. The role of public participation should also transform from an ostensible stipulation to a real procedural and operative basis. Planners and urban designers should treat the participation input with seriousness and caution. Overall, it is of vital importance to ensure an appropriate balance between the overall planning in the perspective of designers and planners, and the coordination in residents’ input and participation, so as to appropriately strengthen the connection between user’s viewpoints and the policy-making on the normative level.

Limitations and future works

Methodologically, the study demonstrates to some extent the potential and effectiveness of the VR and physiological indicators monitoring technologies in empirical environmental perception research. With the support of these technologies, environmental perception researchers will no longer be limited by time and space constraints but will be able to obtain more accurate data in a more controlled, quantitative, efficient and cost-effective manner. However, this study still has some limitations, such as biosensor accuracy problems and missing data problems due to poor instrument contact. Also, the issue of consistency of multi-source measurement results and its possible neurobiological roots is also a matter to be explored in-depth in future studies. We hope that with future technological advances, the accuracy and stability of related research techniques will be improved, and their practical potential will thus be better explored to support a deeper exploration of the neurobiological mechanisms of human’s environmental perception. Finally, the inevitable lack of subject diversity associated with the convenient sampling method used in this study limits the generalizability of the findings and is also a point worthy of improvement in future studies.

Footnotes

Acknowledgements

The authors would like to thank Ms Ying Wu, Mr Chaoyue Tian and Ms Jielan Zhong for their contributions in early stages of experiment design and execution.

Authors contributions

All authors contributed equally in the preparation of this manuscript.

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Liyan Xu

References

Moore

Gould

Keary

. Global urbanization and impact on health. Int J Hyg Environ Health 2003; 206: 269–278.

Calogiuri

Litleskare

Fagerheim

Ry dgren

Brambilla

Thurston

. Experiencing Nature through Immersive Virtual Environments: Environmental Perceptions, Physical Engagement, and Affective Responses during a Simulated Nature Walk. Front Psychol 2018; 8: 1–14.

Carmona

Contemporary Public Space, Part Two: Classification. J Urban Des 2010; 15: 157–173.

Zamanifard

Alizadeh

Bosman

Coiacetto

. Measuring experiential qualities of urban public spaces: users’ perspective. J Urban Des 2019; 24: 340–364.

Askari

Soltani

. Determinants of a successful public open space: the case of Dataran Merdeka in the city centre of Kuala Lumpur, Malaysia. Landsc Res 2019; 44: 162–173.

Bendjedidi

Bada

Meziani

. Urban plaza design process using space syntax analysis. Int Rev Spat Plan Sustain Dev 2019; 7: 125–142.

Bodnar

. Reclaiming public space. Urban Stud 2015; 52: 2090–2104.

Giddings

Charlton

Horne

. Public squares in European city centres. URBAN Des Int 2011; 16: 202–212.

Alwah

AAQ

Alwah

MAQ

Shahrah

. Developing a quantitative tool to measure the extent to which public spaces meet user needs. Urban Urban Green 2021; 62: 127152.

10.

Francis

. Urban Open Space : Designing for User Needs. Washington: Island Press, 2003. Corpus ID: 106703520.

11.

Marcus

Francis

. People Places : Design Guidelines for Urban Open Space. 2nd ed. New York: John Wiley & Sons, 1994.

12.

Brown

Gifford

. Architects predict lay evaluations of large contemporary buildings: whose conceptual properties? J Environ Psychol 2001; 21: 93–99.

13.

Riechers

Noack

Tscharntke

. Experts’ versus laypersons’ perception of urban cultural ecosystem services. Urban Ecosyst 2017; 20: 715–727.

14.

Suppakittpaisarn

Larsen

Sullivan

, Preferences for green infrastructure and green stormwater infrastructure in urban landscapes: Differences between designers and laypeople. Urban Urban Green 2019; 43: 126378.

15.

Hofmann

Westermann

Kowarik

van der Meer

. Perceptions of parks and urban derelict land by landscape planners and residents. Urban Urban Green 2012; 11: 303–312.

16.

Ghomeshi

Jusan

. Investigating Different Aesthetic Preferences Between Architects and Non-architects in Residential Façade Designs. Indoor Built Environ 2013; 22: 952–964.

17.

Gifford

Hine

Muller-Clemm

Shaw

. Why architects and laypersons judge buildings differently: Cognitive properties and physical bases. J Archit Plann Res 2002; 19: 131–148.

18.

Akalin

Yildirim

Wilson

Kilicoglu

. Architecture and engineering students’ evaluations of house façades: Preference, complexity and impressiveness. J Environ Psychol 2009; 29: 124–132.

19.

Gifford

Hine

Muller-Clemm

Reynolds

Shaw

. Decoding Modern Architecture. Environ Behav 2000; 32: 163–187.

20.

Yin

Yuan

Arfaei

Catalano

Allen

Spengler

. Effects of biophilic indoor environment on stress and anxiety recovery: A between-subjects experiment in virtual reality. Environ Int 2020; 136(9 Suppl): 105427.

21.

Mouratidis

Hassan

. Contemporary versus traditional styles in architecture and public space: A virtual reality study with 360-degree videos. Cities 2020; 97: 102499.

22.

Zhang

Zhuo

Wei

Yin

. Emotional Responses to the Visual Patterns of Urban Streets: Evidence from Physiological and Subjective Indicators. Int J Environ Res Public Health 2021; 18: 9677.

23.

Koohsari

Badland

Mavoa

Villanueva

Francis

Hooper

Owen

Giles-Corti

. Are public open space attributes associated with walking and depression? Cities 2018; 74: 119–125.

24.

Shan

X-Z

. Attitude and willingness toward participation in decision-making of urban green spaces in China. Urban Urban Green 2012; 11: 211–217.

25.

Tabrizian

Baran

Van Berkel

Mitasova

Meentemeyer

. Modeling restorative potential of urban environments by coupling viewscape analysis of lidar data with experiments in immersive virtual environments. Landsc Urban Plan 2020; 195: 103704.

26.

Lotfi

Koohsari

. Analyzing Accessibility Dimension of Urban Quality of Life: Where Urban Designers Face Duality Between Subjective and Objective Reading of Place. Soc Indic Res 2009; 94: 417–435.

27.

Bonnes

Uzzell

Carrus

Kelay

. Inhabitants’ and Experts’ Assessments of Environmental Quality for Urban Sustainability. J Soc Issues 2007; 63: 59–78.

28.

X-P

Fan

S-X

Kühn

Dong

Hao

P-Y

. Residents’ ecological and aesthetical perceptions toward spontaneous vegetation in urban parks in China. Urban Urban Green 2019; 44: 126397.

29.

de Bell

Graham

White

PCL

. The role of managed natural spaces in connecting people with urban nature: a comparison of local user, researcher, and provider views. Urban Ecosyst 2018; 21: 875–886.

30.

Tudor

Iojă

Rozylowicz

Pǎtru-Stupariu

Hersperger

. Similarities and differences in the assessment of land-use associations by local people and experts. Land Use Policy 2015; 49: 341–351.

31.

Ilbeigi

KohneRoudPosht

Ghomeishi

Behrouzifard

. Cognitive differences in residential facades from the aesthetic perspectives of architects and non-architects: A case study of Iran. Sustain Cities Soc 2019; 51: 101760.

32.

Devlin

Nasar

. The beauty and the beast: Some preliminary comparisons of ‘high’ versus ‘popular’ residential architecture and public versus architect judgments of same. J Environ Psychol 1989; 9(4): 333–344.

33.

Lothian

. Landscape and the philosophy of aesthetics: is landscape quality inherent in the landscape or in the eye of the beholder? Landsc Urban Plan 1999; 44(4): 177–198.

34.

Lynch

The Image of the City. Cambridge: MIT press, 1960.

35.

Ode

Tveit

Fry

. Capturing Landscape Visual Character Using Indicators: Touching Base with Landscape Aesthetic Theory. Landsc Res 2008; 33: 89–117.

36.

Ewing

Handy

. Measuring the Unmeasurable: Urban Design Qualities Related to Walkability. J Urban Des 2009; 14: 65–84.

37.

Motoyama

Hanyu

. Does public art enrich landscapes? The effect of public art on visual properties and affective appraisals of landscapes. J Environ Psychol 2014; 40: 14–25.

38.

Shi

Gou

Chen

LHC

. How does enclosure influence environmental preferences? A cognitive study on urban public open spaces in Hong Kong. Sustain Cities Soc 2014; 13: 148–156.

39.

van Vliet

Dane

Weijs-Perrée

van Leeuwen

van Dinter

van den Berg

Borgers

Chamilothori

. The Influence of Urban Park Attributes on User Preferences: Evaluation of Virtual Parks in an Online Stated-Choice Experiment. Int J Environ Res Public Health 2020; 18: 212.

40.

Liu

Schroth

. Assessment of Aesthetic Preferences in Relation to Vegetation-Created Enclosure in Chinese Urban Parks: A Case Study of Shenzhen Litchi Park. Sustainability 2019; 11: 1809.

41.

Zhang

Cao

Han

. Residents’ Preferences and Perceptions toward Green Open Spaces in an Urban Area. Sustainability 2021; 13: 1558.

42.

Subiza-Pérez

Vozmediano

San Juan

. Welcome to your plaza: Assessing the restorative potential of urban squares through survey and objective evaluation methods. Cities 2020; 100: 102461.

43.

Mehta

. Evaluating Public Space. J Urban Des 2014; 19: 53–88.

44.

Currie

. A design framework for small parks in ultra-urban, metropolitan, suburban and small town settings. J Urban Des 2017; 22: 76–95.

45.

Browning

MHEM

Mimnaugh

van Riper

Laurent

LaValle

. Can Simulated Nature Support Mental Health? Comparing Short, Single-Doses of 360-Degree Nature Videos in Virtual Reality With the Outdoors. Front Psychol 2020; 10: 1–14.

46.

Baran

Tabrizian

Zhai

Smith

Floyd

. An exploratory study of perceived safety in a neighborhood park using immersive virtual environments. Urban Urban Green 2018; 35: 72–81.

47.

Yin

Zhu

MacNaughton

Allen

Spengler

. Physiological and cognitive performance of exposure to biophilic indoor environment. Build Environ 2018; 132: 255–262.

48.

Kim

. Immersive Virtual Reality-Aided Conjoint Analysis of Urban Square Preference by Living Environment. Sustainability 2020; 12: 6440.

49.

Birenboim

Dijst

Ettema

de Kruijf

de Leeuw

Dogterom

. The utilization of immersive virtual environments for the investigation of environmental preferences. Landsc Urban Plan 2019; 189: 129–138.

50.

Banaei

Ahmadi

Gramann

Hatami

. Emotional evaluation of architectural interior forms based on personality differences using virtual reality. Front Archit Res 2020; 9: 138–147.

51.

Bruns

Chamberlain

. The influence of landmarks and urban form on cognitive maps using virtual reality. Landsc Urban Plan 2019; 189: 296–306.

52.

Loomis

Blascovich

Beall

. Immersive virtual environment technology as a basic research tool in psychology. Behav Res Methods Instr Comput 1999; 31: 557–564.

53.

Fox

Arena

Bailenson

. Virtual Reality. J Media Psychol 2009; 21: 95–113.

54.

Smith

. Immersive Virtual Environment Technology to Supplement Environmental Perception, Preference and Behavior Research: A Review with Applications. Int J Environ Res Public Health 2015; 12: 11486–11505.

55.

Arellana

Garzón

Estrada

Cantillo

. On the use of virtual immersive reality for discrete choice experiments to modelling pedestrian behaviour. J Choice Model 2020; 37: 100251.

56.

Birenboim

Ben-Nun Bloom

Levit

Omer

. The Study of Walking, Walkability and Wellbeing in Immersive Virtual Environments. Int J Environ Res Public Health 2021; 18(2): 364.

57.

Niu

Wang

Sun

Yue

Zheng

. User Experience Evaluation in Virtual Reality based on Subjective Feelings and Physiological Signals. J Imaging Sci Technol 2019; 63: 60413–60511.

58.

Jin

Wang

. Building environment information and human perceptual feedback collected through a combined virtual reality (VR) and electroencephalogram (EEG) method. Energy Build 2020; 224: 110259.

59.

Jiang

Hassan

Chen

Liu

. Effects of different landscape visual stimuli on psychophysiological responses in Chinese students. Indoor Built Environ 2020; 29: 1006–1016.

60.

Gao

Zhang

Zhu

Gao

Qiu

. Exploring Psychophysiological Restoration and Individual Preference in the Different Environments Based on Virtual Reality. Int J Environ Res Public Health 2019; 16(17): 3102.

61.

Mauss

Robinson

. Measures of emotion: A review. Cogn Emot 2009; 23: 209–237.

62.

Hakak

Bhattacharya

Biloria

de Kleijn

Shah-Mohammadi

. Navigating abstract virtual environment: an eeg study. Cogn Neurodyn 2016; 10: 471–480.

63.

Guo

L-N

Zhao

R-L

Ren

A-H

Niu

L-X

Zhang

Y-L

. Stress Recovery of Campus Street Trees as Visual Stimuli on Graduate Students in Autumn. Int J Environ Res Public Health 2019; 17: 148.

64.

Zhang

Long

Zhejing

. The Impact of the Micro-Scale Built Environment of Historic Street on Visitor’s Walking Behaviors: A Case Study on Wudaoying Hutong in Beijing. Archit J 2019; 60: 96–102.

65.

. Scale Development for Environmental Perception of Public Space. Front Psychol 11: 596790. 2020. DOI: 10.3389/fpsyg.2020.596790.

66.

Dai

. Handbook of the Common Psychological Rating Scales. Beijing: People’s Military Medical Press, 2010.

67.

Auerbach

Spielberger

. The Assessment of State and Trait Anxiety with the Rorschach Test. J Pers Assess 1972; 36: 314–335.

68.

Marteau

Bekker

. The development of a six‐item short‐form of the state scale of the Spielberger State — Trait Anxiety Inventory (STAI). Br J Clin Psychol 1992; 31: 301–306.

69.

Barnes

LLB

Harp

Jung

. Reliability Generalization of Scores on the Spielberger State-Trait Anxiety Inventory. Educ Psychol Meas 2002; 62: 603–618.

70.

Emy . Performance Metrics. EMOTIV. https://www.emotiv.com/knowledge-base/performance-metrics/. Epub ahead of print 2019Accessed on August 10, 2021.

71.

Deng

Luo

E-K

Sun

L-X

Huang

Cai

S-Z

Jia

. Empirical study of landscape types, landscape elements and landscape components of the urban park promoting physiological and psychological restoration. Urban Urban Green 2020; 48: 126488.

72.

Gehl

Life between Buildings: Using Public Space. San Francisco: Goodreads, 1987.

73.

Ward Thompson

Activity, exercise and the planning and design of outdoor spaces. J Environ Psychol 2013; 34: 79–96.

74.

Wang

Zhao

Meitner

. Characteristics of urban green spaces in relation to aesthetic preference and stress recovery. Urban Urban Green 2019; 41: 6–13.

75.

Hall

Cities of Tomorrow: An Intellectual History of Urban Planning and Design since 1880. Chichester, West Sussex: John Wiley & Sons, 2014.

76.

Knox

McCarthy

. Urbanization: An Introduction to Urban Geography. 3rd ed. Upper Saddle River, NJ: Pearson Prentice Hall, 2005.

77.

Alexander

. A City is Not a Tree. Archit Forum 1965; 122: 58–62.

78.

Jacobs

. The Death and Life of Great American Cities. New York: Random House LLC, 1961.

79.

Afzalan

Sanchez

Evans-Cowley

. Creating smarter cities: Considerations for selecting online participatory tools. Cities 2017; 67: 21–30.

80.

Faehnle

Bäcklund

Tyrväinen

Niemelä

Yli-Pelkonen

. How can residents’ experiences inform planning of urban green infrastructure? Case Finland. Landsc Urban Plan 2014; 130: 171–183.