Assessment of the Perception of Sustainability for Occupants of Residential Buildings: A Case Study in the UAE

Abstract

The residential sector is multi-faceted by nature. Although evidence shows that the UAE is among the countries in the world that take sustainability seriously, there is a lack of information about the perception of sustainability by occupants in the residential sector in the UAE. The aim of this paper is to assess the perception of sustainability of the residential sector in the UAE, which is achieved by following a methodological framework using the relevant literature review and experts’ knowledge. An online survey was distributed to the targeted population, followed by a statistical analysis to fulfill the aim of the paper. Results confirm the correlation between social, economic, and environmental aspects of sustainability. Additionally, structural equation modeling reveals that the perception of sustainability is significantly influenced by economic and environmental aspects in the residential sector in the UAE. Comparative analysis shows a statistical difference in the perception of sustainability among gender, educational level, employment status, and monthly income. Finally, a predictive classification model is built to classify the perception of occupants based on their attributes using decision tree algorithms. The outcomes of this study would be beneficial to policy and decision makers, developers, contractors, designers, and facility management entities to enhance overall sustainability in the residential sector.

1. Introduction

Sustainable Development (SD) is related to the effect of our current resource use on our planet’s natural balance and regenerative power [1]. It is concerned with prudent economic growth and economic development that maintains the resources we need for future sustenance; i.e., how do we live today so that our planet and future generations can thrive [2]? This denotes the need to improve efficiency in resource use and focus on renewables [3].

To achieve a sustainable system, a sustainable mindset needs to be adopted and used to address obstacles and measure progress towards a set of targets or goals. This involves the evolution of a way of thinking as well as a collective lifestyle. Therefore, sustainable systems encompass responsible care, sustainable consumption, and sustainable production [4]. SD has been defined by several studies and entities, however, the initial definition in the United Nation’s Brundtland Report, published in 1987, indicates that SD is the “development that meets the needs of the present without compromising the ability of future generations to meet their own needs” [5]. The definition stresses that all growth should take SD at its core, i.e., the environmental, economic, and social well-being of humanity. Accordingly, the “three-pillar” paradigm was introduced by Elkington [6], which is social, economic, and environmental as demonstrated in Figure 1. To attain a comprehensive and truly sustainable system, all the associated activities should be sustainable in all three dimensions [7,8,9].

Despite a focus on providing a sustainable built environment in general and residential areas in specific, the sustainability target is not usually set to achieve sustainability pillars in the residential sector. Moreover, amongst all the sectors in the UAE, the residential sector is known to have the highest energy consumption, which implies the highest contribution to the inverse effect on the environment [10,11,12,13]. Chartrand [14] noted that awareness must precede attempts at control. Thus, the creation of occupants’ awareness is a necessary prerequisite for a change towards responsible behavior and sustainable development [15]. It is suggested that awareness is not an “all or nothing” phenomenon [14]. Hence, occupants’ awareness is impacted by to which degree the occupants consciously reflect on their behaviors. Consequently, it is vital to enhance the awareness of this sector regarding sustainability, and this cannot be attained without an initial assessment of the perceptions of sustainability in the housing sector.

Due to the importance of this issue, the following research question will be answered in this study. How do people in residential neighborhoods in the UAE perceive sustainability principles? Thus, this paper aims to measure the perception of sustainability in the UAE residential sector. To achieve this aim, this paper intends to (i) develop an understanding of the progress in sustainability concepts, (ii) identify the main dimensions/indicators based on related literature studies, green rating assessment tools, and experts in the field, (iii) collect data from occupants of residential buildings in the UAE based on the adopted sustainability indicators, and (iv) test the proposed hypotheses and draw conclusions. The hypotheses are for testing the correlation among the three pillars of sustainability, as well as the positive impact of each pillar on the perception of sustainability by occupants in the residential sector in the UAE.

The review of the literature is drawn from several relevant sources. First, it investigates green building rating systems, then it focuses on awareness and perceptions of SD in the UAE. Subsequently, it highlights the link between awareness and behavior, and finally, it investigates the related sustainability indicators in the context of the UAE residential sector. The following subsections elaborate on each component of the literature review.

1.1. Green Building Rating Systems

The impact of buildings on the environment has become increasingly important in recent decades as the construction sector contributes significantly to carbon emissions, global warming, and environmental degradation due to its high consumption of natural resources and high energy use [16]. Worldwide, the residential sector is one of the major consumers of energy–mainly electricity and natural gas [17]. It consumes natural resources and impacts the environment. However, sustainability is not only concerned with environmental issues as it involves an interaction between a triple bottom line framework comprised of social, economic, and environmental factors. Therefore, the importance of residential buildings has been increased, which raises attention to the importance of enhancing their sustainability.

Government agencies and organizations have adopted green building rating systems to minimize the consumption of natural resources and control pollution. In order to evaluate the impact of the building sector on the environment, various sustainability assessment methods (assessment systems) have been implemented, such as Leadership in Energy and Environmental Design (LEED), Building Research Establishment Environmental Assessment Methods (BREEAM), Sustainable Building Tool (SBTool), DGNB (Deutsche Gesellschaft für Nachhaltiges Bauen-German Sustainable Building Council), SBTool (Sustainable Building Tool), and the Comprehensive Assessment System for Built Environment Efficiency (CASBEE) [18]. In recent decades, the need arose to develop specific tools for different countries and different types of buildings, such as assessment tools for residential buildings, hospitals, and schools.

1.2. Sustainable Development in the UAE

The UAE has been one of the pioneering countries in considering sustainability as a serious issue that concerns the future of the country. Several landmark projects are supporting the movement to a more sustainable and livable future. This is achieved by focusing on the Sustainable Development Goals (SDGs), which consist of 17 interconnected goals to attain a better and more sustainable future for all [19]. Consequently, the UAE has put in place several initiatives to drive the country to a more resilient and sustainable future by enabling ways to health care, sufficient and affordable food, economic growth, quality education, a healthy environment, and efficient use of resources [20]. To illustrate, the UAE energy strategy targets increasing the share of clean energy resources by 50%, increasing consumption efficiency by 40%, and reducing its carbon footprint by 70% by 2050 through the implementation of several initiatives across the country [21].

The main initiatives toward SD in the UAE include: (i) ESTIDAMA Pearl Rating System (PRS), which was established in 2009 by Abu Dhabi government to be one of the leading sustainability frameworks in the Middle East, where all the developments are required to meet at least the minimum number of pearls based on this sustainability rating system [22], (ii) Al Sa’fat initiative, which replaced Dubai Green Building Regulations and Specifications in 2020; it includes a comprehensive set of requirements for all new developments to secure a more sustainable environment in the future [23], (iii) Masdar City—Abu Dhabi, which was initiated in 2006 to help implement sustainability practices in an urban development [24], (iv) Sustainable City—Dubai, which was established in 2015 to focus on the three dimensions of sustainability by implementing sustainability practices that improve the occupants’ lifestyle [25], (v) Sharjah Sustainable City, which was established in 2021 as a continuation of the sustainability success of the Sustainable City—Dubai [26], (vi) EXPO2020, which is designed, built, and operated in a sustainable manner that makes it one of the most energy efficient development [27], (vii) Mohammad Bin Rashid Al Maktoum Solar Park, which is anticipated to reduce the carbon emission in the UAE by 6.5 million tons per year [28], and other governmental initiatives that support sustainable development goals (SDGs), such as Al Jalila foundation, Dubai cares, food bank, Dubai cares water, sanitation, and hygiene, international humanitarian city, Mohamed Bin Rashid Al Maktoum Global Initiative, Noor Dubai foundation, sustainable schools, UAE water aid initiatives, and the Zayed Bin Sultan Al Nahyan charitable and humanitarian foundation [29]. As per the SD report of 2021, the UAE has an overall score of 70.17% in achieving the 17 SDGs [30]. In addition, according to Asif [31], the UAE is leading the sustainability trends among the GCC countries, also having the highest share of green buildings in the whole Middle East and North Africa (MENA) region.

According to [32], residential buildings are among the essential community conditions that describe the healthy quality of life and well-being of people. In addition, residential buildings consume about 40% of the world’s electricity and account for up to one-third of global greenhouse gas emissions in developed and emerging economies [33]. In the UAE, the residential sector has seen a remarkable increase in electricity consumption–the consumption has increased from 6580 GWh to 12,795 GWh from 2006 to 2017 [34]. Therefore, it is apparent that special attention should be given to this sector to enhance the overall sustainability of the three sustainability pillars.

1.3. The Link between Awareness and Behavior

Awareness of environmental problems has grown in the second half of the twentieth century [35]. In addition to environmental problems, human society has also undergone dynamic changes due to advances in communication, technology, and globalization. The social structure and economic conditions of the current world have led to an increase in global inequality and polarization in society. Understanding the integration of environmental, social, and economic aspects of human and natural systems has become crucial [36]. Consequently, people need to understand how their activities impact the environment, including social and economic aspects, and what they can do to minimize the adverse impacts [37,38]. This position claims that increasing people’s awareness of sustainability within their communities can encourage them to implement sustainability principles and practices in their day-to-day activities. Several related terms are often used interchangeably with sustainability awareness; these include green awareness, environmental awareness, environmental knowledge, carbon literacy, energy literacy, environmental literacy, and ecological literacy [39].

A review of current literature reveals that considerable attention is devoted to sustainability in general and environmental sustainability in particular, which some authors describe as having an enviro-centric view [40,41,42]. Vasconcelos [43] argues that “for too many people the word environment still signals ‘green’ and fails to convey its social, economic, political, and cultural components”. Research into awareness, perceptions, and behavior in terms of sustainability in the residential sector is limited, but there is more evidence that behavior is built on perceptions and attitudes when addressing energy saving and recycling, energy efficiency, and climate change [44,45,46]. Environmental behavior remains the most widely and systematically studied subject within the existing socio-psychological literature [47,48,49].

Several studies have been conducted to develop strategies aimed at enhancing people’s awareness, perceptions, cognitive abilities, motivations, and behaviors to promote household sustainable practices. For instance, a study conducted by [50] in 2008 showed that there is a problem with people’s perceptions and behaviors when using sustainable energy and water technologies. Later, in 2010, a need for further research was identified by [51] in the area of users’ perceptions, behaviors, attitudes, and expectations for sustainable measures. Previous studies have shown that building occupants play a critical role in the built environment, as 50% of energy consumption in residential buildings depends on the behavior of buildings occupants [52].

Moreover, an investigation conducted by [53] suggested that energy-efficiency behaviors account for 51%, 37%, and 11% of the variance in heat, electricity, and water consumption, respectively, between dwellings. Abrahamse and Steg in [54] argued that in order to change household energy consumption behavior, it is important to understand the impact of sociodemographic and psychological factors on household attitudes and perceptions toward green building development, particularly to identify barriers and opportunities to achieving sustainable homes from the household perspective. Later in 2013, several studies were conducted in different countries to address the contribution of occupants of residential buildings to actual energy consumption from different perspectives [55,56]. In addition, a framework to improve the sustainability of residential buildings was proposed by [57], which investigated the motivation and behavior of residential building occupants in terms of energy savings and its impact on greenhouse gas emissions. Canale et al. [58] contend that the physical properties of the buildings alone are not enough to reduce their energy consumption and that occupants’ behavior is an influential factor in the real energy consumption of a building.

In addition, several studies have assessed the impact of environmental awareness on pro-environmental behavior [59]. For instance, Sekhokoane et al. [60] claimed that individuals with high environmental awareness are more likely to behave in an environmentally sustainable manner. Similarly, Yilmaz et al. [61] found that lack of awareness could lead to insignificant and minimal changes in individual behavior, and, consequently, dependence on government actions. Nevertheless, awareness does not necessarily result in environmental behavior [62]. Hadlock and Beckwith [63] indicated that individuals do not willingly behave sustainably and in favor of the environment unless they experience opposing impacts of environmental problems. Accordingly, there is a lack of consensus in the literature on the environmental awareness–behavior relationship [64]. Most of these studies have concentrated only on certain and selected aspects of environmental awareness. This study argues that this approach to identifying awareness and perception has limitations as it only captures a part of this whole construct. It attempts to create a comprehensive model, which provides a holistic means of assessing the perception of sustainability taking into consideration its three pillars (economic, social, and environmental).

Despite the complex and multi-dimensional nature of sustainability in terms of balancing the three pillars, environmental, economic, and social, prior research often focused on a single dimension–the environment. Although a good deal of research has been published in the area of sustainability, there are very limited studies that have addressed sustainability awareness and application as a whole [38]. Hence, investigating stakeholders’ awareness and behaviors in the residential sector is essential to understand the dynamic interactions and promote coordination among them [65]. The significance of such research can be a base for improving the decision-making process and promoting sustainability in the built environment [66].

Although studying occupants’ perceptions and behaviors toward more efficient resource use can have an impact on energy conservation and the reduction of emissions, a clear understanding of occupants’ perception of the big picture of SD including the three dimensions (economic, environmental, social) helps policymakers, designers, and building professionals not only identify gaps in current policies and develop potential new technologies, but also understand the attitudes, knowledge, preferences, and behaviors of building occupants toward sustainable measures and improve their behaviors through educational programs.

1.4. Sustainability Indicators

Sustainability perceptions by residential occupants are assessed based on the extracted indicators from the relevant literature, where the final selection of the indicators is verified by experts in the field to assure their compliance with the prevailing standards and regulations within the UAE. Several studies have discussed the sustainability of residential buildings; however, it is an evolutionary process that needs to be continually enhanced, considering any change that impacts the social, environmental, or economic pillars of sustainability.

According to [67], the first category is health and safety, and it involves the prevention of virus propagation using new smart, innovative, and touchless technologies [67,68,69,70], self-cleaning spaces [71], proper selection of indoor materials [70,72], allowing natural daylight (as sunlight enables a healthy environment and prevents the viability of viruses) [73,74], and the adjustability of indoor temperature and humidity; these are all influential factors in virus propagation, yet they are subjective to the perception of occupants [75,76,77]. The second category is mental health for occupants, which includes the availability of greenery and gardens such as indoor gardens, green views, and green balconies, which enhance relaxation and reduce the risk of stress-related diseases [78,79], availability of outdoor spaces in the building (balconies) as they improve the mental health of occupants [80,81], access to common building spaces with sufficient safety and social distance, and household-level activity/sports spaces to boost immunity and reduce stress.

The third category is air quality, and it includes the efficiency of air filtration systems to monitor and control indoor air pollution and develop a moderate fresh air circulation [67,82]. Ref. [83] highlighted that people’s perceptions of the indoor environment depend on an interaction between the physical properties of the room (temperature) and the psychological associations with the ‘green’ certification. The fourth category is water quality and availability, and it includes safety measures for drinking water and/or tap water from contamination [67,84]. Finally, the fifth category in this group is related to wastewater management and it includes specific measures to limit virus propagation at a household level to avoid transmission of viruses through wastewater [85,86].

Moreover, the environmental resources consumption category includes three aspects [67], where the first aspect is related to energy use, promotion of sustainable and alternative energy sources to improve the environmental effect, and the use of energy-efficient appliances to resolve the increase in energy consumption. The second aspect is for waste management [87], and the last is related to water consumption including the access to alternative water sources as well as the use of water-efficient appliances and fixtures.

In addition, the last category of sustainability as discussed by [67] is comfort, which has two aspects: personal comfort and local services. Personal comfort is related to a specific emphasis on household-level ICT infrastructure access to enable studying, working, food delivery, and medical consultation [88,89], levels of indoor space adjustability [78], and acoustic comfort [90]. On the other hand, local services include the availability of self-dependent services in residential complexes, especially for medicine and food [91], and urban/community farming to reduce the paths for food supply and enhance mental health [92]. Based on the aforementioned sustainability indicators, expert groups from academia, industry, and medicine were surveyed in 17 countries [93], and the results showed that health and safety have the highest importance, with more attention paid to the prevention of virus propagation, mental health, and air quality. Additionally, the experts agreed that touchless technologies, self-cleaning spaces, and smart technologies contribute to avoiding the spread of viruses.

According to global definitions of social sustainability, the needs of social groups should be addressed to improve overall satisfaction [94]. In terms of occupants’ needs in housing, the social aspect of a well-known sustainability assessment scheme only considered indoor environmental quality in most of the studies in the literature. However, social sustainability includes various issues from safety to other physiological and psychological needs of stakeholders [95]. Based on this, many researchers exploited social indicators for the sustainability evaluation of buildings based on the context of their countries. For example, Ahmad and Theheem [96] developed a social sustainability assessment framework for residential buildings in Pakistan to address the lack of social aspects in the assessment framework of this country. Likewise, Ullah et al. [97] developed a framework focusing relatively more on social issues for assessing the sustainability of residential buildings in Pakistan.

Similarly, a recent study conducted by Fatourehch et al. [94] addressed sustainability issues in the assessment framework of residential buildings in Iran. This study considered more detailed indicators in terms of social aspects of sustainability issues regarding Iranian residential buildings. It investigated the most common social aspects, which were identified by prior literature for the assessment of residential buildings worldwide. In addition, Maleki et al. [98] formulated a sustainability assessment model highlighting the social sustainability aspects of residential high-rise buildings (RHRB). The authors argued that emphasizing the social indicators is important, as they are effective for assessing the human relationship with the physical environment (housing).

Another study conducted by Karji et al. [99] sought to rectify the evaluative inequity between the social bottom lines and the economic and environmental bottom lines of sustainability within the existing sustainability rating systems for buildings (LEED, BREAM, CASBEE, etc.). The indicators were tested in one of Iran’s largest mass housing projects. In previous research, some researchers have reviewed existing literature regarding sustainability assessment for buildings. For example, Sierra et al. [100] conducted a comprehensive review of various Green Building Rating Systems (GBRS) adopted by the current sustainable construction community. The authors note that despite the vast research regarding GBRS, little has been done to examine the stakeholders’ perceptions of GBRS.

In addition, [101] has built a multi-criteria framework to assess the sustainability of buildings in Kazakhstan; the sustainability categories and indicators were extracted from four assessment tools (LEED, BREEAM, CASBEE, and SBTool) as well as the related literature, then experts were interviewed to specify the indicators that are applicable in Kazakhstan, where 35 indicators were validated for the analysis. Alyami et al. [102] designed a measurement framework to evaluate the principles of sustainable construction for residential buildings in Saudi Arabia by analyzing the leading international sustainable assessment schemes: BREEAM, LEED, SBTool, and CASBEE. The authors concluded that international schemes are inapplicable for the Saudi context, and there is a need to develop further categories and criteria for the assessment of the built environment in Saudi Arabia.

Interestingly, another study has been conducted in Kazakhstan to assess sustainability in residential buildings and how far is it from the green building indicators [103]; the study was based on a survey of occupants that covers different sustainability categories. The result of this study indicated that old buildings tend to have a poor sustainability level based on the responses of their occupants, while new buildings have an increasing interest in sustainability, especially when it is linked to economic aspects. Kamali and Hewage [104] conducted qualitative and systematic research for sustainable building rating systems and journal/conference articles to develop a list of sustainability criteria related to each sustainability pillar (TBL).

In addition, six case studies were conducted in Brno and Vienna [105] to study how open spaces between residential buildings affect sustainability. The results support the importance of green open spaces in enhancing the SD for the residential sector, however, the indicators were obtained from a comprehensive literature review. Winston [106] outlined an analytical framework for assessing sustainable housing and regeneration in Dublin, Ireland. This framework included characteristics that are associated with each of the key aspects of housing, which were derived from literature on sustainability. Another customized sustainability assessment system for multi-apartment buildings in Slovenia was proposed by [107]; although various aspects were included in the system, such as functionality, safety, and security aspects, the economic aspect was not considered.

Moreover, a case study was performed in Portugal to assess the sustainability of new, existing, and renovated residential buildings [108]. The indicators were obtained from the SBTool assessment system by customizing the set of indicators that are related to the context of the study, and they covered the three dimensions of sustainability; the proposed methodology supports the direction towards sustainability by identifying the objectives for the contractors working in residential sectors. Another study was conducted in Malaysia to identify and rank sustainability indicators for assessing green buildings [109]. This was accomplished through employing appropriate GBRSs and their associated sustainability indicators to evaluate and confirm the level of greenness and sustainability of buildings in the manufacturing sector.

Al-Jebouri et al. [110] developed a framework for sustainable building construction in Oman by proposing an assessment system for the construction industry. The indicators of the proposed rating system for Oman were derived by reviewing the literature on SD and buildings as well as analyzing international and regional sustainable building-rating systems, such as LEED, BREEAM, and UAE Estidama. Moreover, Karaca et al. [76] argue that existing methods in the literature mostly suggest indicator scores in construction sustainability assessments using highly technical data and require a high level of engineering expertise. Therefore, the authors developed a detailed method called A Rapid Sustainability Assessment Method (RSAM) that uses residents’ opinions-based sustainability in Kazakhstan.

A set of ratio-based indicators and rating functions for building rating systems in a global context was developed by [111,112], where the building sustainability rating system was proposed with two main features, namely, the use of quantifiable indicators instead of subjective ratings for aspects, and the use of continuous functions for indicator ratings because such functions allow better comparisons between buildings than step functions. In addition, a contextual assessment approach for residential buildings in Hongkong was proposed by [113]. The proposed approach aimed to assess and improve the integrated sustainability of public rental housing to improve the livelihoods of middle-and low-income families.

Given that the majority of the proposed assessment tools and methods were either complex or focused only on single issues, this encouraged researchers to develop simplified methods for assessing building sustainability. Markelj in [114], for example, developed a simple method for assessing the sustainability of Slovenian buildings in the early design phase and included economic and safety aspects in addition to the basic categories. In 2016, Raut [115] proposed a system for assessing the sustainability of residential buildings in Mumbai. The proposed methodology was used in the selection of sustainable residential buildings, using both relative benchmarking and absolute benchmarking, which provides a tangible means of evaluating residential buildings.

A study performed by Janjua et al. [116] proposed a holistic life cycle sustainability assessment (LCSA) framework based on the TBL indicators to address the sustainability performance of residential buildings, specifically for assessing Australian buildings. The TBL indicator selection was carried out systematically based on a literature review. After that, the stakeholders directly or indirectly involved in the building sector were given a platform to provide their opinions and become a part of the selection process of indicators. Another framework was developed by Kamili et al. [117] for the sustainability assessment of modular buildings. The authors conducted a comprehensive literature review including reviewing different rating systems such as LEED, Green Globes, and LBC, as well as journal/conference articles.

In addition, an assessment scale was proposed to evaluate sustainability from the perspectives of residents–as the main stakeholders–in Thailand. The authors extracted their indicators using theoretical bases from a literature review of related concepts, such as sustainable cities, urban sustainability, sustainable urban forms, and eco-cities [118]. Furthermore, it was noted that construction and buildings’ operation are, from the perspective of SD, unsustainable energy processes. Therefore, the authors developed an overall sustainability index based on the TBL indicators. They note that defining and determining the economic, social, and environmental indicators was developed based on the energy consumption of residential buildings in Belgrade city in Serbia [119]. In addition, Amasuomo et al. [120] established an assessment tool for residential buildings in Nigeria based on international rating tools to adapt it to the Nigerian requirements.

In the context of UAE, a few papers have looked at sustainability assessment and indicators in the residential sector, although there is a good deal of research in this area. For instance, Ref. [121] assessed the sustainability of three types of building (heritage, ordinary modern, and sustainable modern buildings) using the Analytical Hierarchy Process (AHP) based on conducting a survey that targeted experts only. The purpose was to integrate the sustainable principles and enhance the overall sustainability of these buildings. Moreover, the results showed that sustainable modern buildings are the most preferred alternative, followed by heritage buildings and ordinary modern buildings. Furthermore, amongst the sustainability dimensions, economic and environmental have the highest priority. In a study conducted by Baird [122], a worldwide set of commercial and institutional buildings, with well-recognized sustainability credentials, were evaluated to find out the users’ perceptions of their performance. One of the main conclusions of this study is that users perceived sustainable buildings to be significantly better than conventional ones, which indicates that building occupants are quite capable of assessing the performance of the buildings they occupy, and to a relatively fine degree. To summarize, the considered indicators in this study are summarized in

Table 1. Sustainability Indicators for UAE’s Residential Sector.

Dimensions
Social		Environmental		Economic
Indicator	Label	Indicator	Label	Indicator	Label
Visual comfort	SVC	Sustainable, alternative, and renewable energy sources	EES	Initial cost	EIC
Acoustic comfort	SAC	Energy efficiency	EEE	Operation and maintenance Cost	EOM
Thermal comfort	STC	Waste management	EWM	Affordability (rent and utilities)	ERU
Indoor air quality	SIQ	Water efficiency	EWE
Green and open spaces within the building	SGW
Green and open spaces between buildings (neighborhood)	SGB
Aesthetic and pleasant view	SAV
Parking capacity	SCP
Smart technologies	SST
Proper (non-toxic) indoor materials	SPM
Education and awareness	SEA
Cultural identity	SCI
Safety and security	SSS
Health of occupants	SHO
Social interaction	SSI
Accessibility to transportation, amenities	SAT
Accessibility for occupants with special needs	SAS

; they were drawn from numerous conceptual frameworks and models that are built and applied in different contexts in the literature.

Afterward, a validation process by experts was performed to fit the UAE’s residential sector, as illustrated in

Table 2. Extracted Indicators.

Dimensions	Indicators	References
		[70]	[71]	[72]	[73]	[76]	[77]	[78]	[79]	[80]	[81]	[82]	[83]	[84]	[85]	[87]	[88]	[89]
Social	Visual comfort	X				X	X		X
	Acoustic comfort	X							X
	Thermal comfort								X
	Indoor air quality	X							X						X
	Green and open spaces within the building	X									X	X	X	X
	Green and open spaces between buildings (neighborhood)
	Aesthetic and pleasant view
	Parking capacity
	Smart technologies	X	X	X	X			X	X	X							X	X
	Proper (non-toxic) indoor materials	X			X
	Education and awareness
	Cultural identity
	Safety and security
	Health of occupants								X							X
	Social interaction								X
	Accessibility to transportation, amenities								X
	Accessibility for occupants with special needs								X
Environmental	Sustainable, alternative, and renewable energy sources	X
	Energy efficiency	X							X
	Waste management	X							X
	Water efficiency	X							X
Economic	Initial cost								X
	Operation and maintenance Cost								X
	Affordability (rent and utilities)
Social	Visual comfort			X		X	X	X		X			X			X		X
	Acoustic comfort			X		X	X	X		X			X	X		X		X
	Thermal comfort			X		X	X	X		X			X			X		X
	Indoor air quality			X		X	X	X	X	X	X	X	X			X
	Green and open spaces within the building									X
	Green and open spaces between buildings (neighborhood)				X	X		X						X	X		X	X
	Aesthetic and pleasant view		X							X				X
	Parking capacity									X				X		X
	Smart technologies									X		X
	Proper (non-toxic) indoor materials									X		X
	Education and awareness									X				X			X
	Cultural identity			X	X	X	X						X		X			X
	Safety and security				X			X	X				X
	Health of occupants			X	X		X	X	X		X		X	X	X		X	X
	Social interaction						X	X					X
	Accessibility to transportation, amenities				X	X		X	X				X		X	X	X
	Accessibility for occupants with special needs			X
Environmental	Sustainable, alternative, and renewable energy sources									X		X				X
	Energy efficiency					X	X		X	X	X	X	X		X	X	X	X
	Waste management					X	X		X	X	X	X			X	X	X	X
	Water efficiency	X					X		X	X	X	X			X	X	X	X
Economic	Initial cost					X	X			X
	Operation and maintenance Cost					X	X			X			X
	Affordability (rent and utilities)									X	X		X		X
Social	Visual comfort	X	X	X	X					X	X			X		X		X
	Acoustic comfort	X	X	X	X		X			X	X					X		X
	Thermal comfort	X	X	X	X		X			X	X			X		X		X
	Indoor air quality	X	X	X	X		X			X	X					X	X	X
	Green and open spaces within the building													X
	Green and open spaces between buildings (neighborhood)	X						X	X		X			X				X
	Aesthetic and pleasant view													X
	Parking capacity					X
	Smart technologies													X
	Proper (non-toxic) indoor materials
	Education and awareness												X
	Cultural identity	X						X	X		X			X				X
	Safety and security				X	X		X						X
	Health of occupants	X			X	X		X						X				X
	Social interaction	X																X
	Accessibility to transportation, amenities			X	X	X		X	X		X			X		X
	Accessibility for occupants with special needs													X
Environmental	Sustainable, alternative, and renewable energy sources		X	X	X	X			X			X			X	X
	Energy efficiency	X	X	X	X	X	X	X		X	X	X	X	X	X	X		X
	Waste management	X	X	X	X	X	X	X			X	X		X	X	X		X
	Water efficiency	X	X	X	X	X	X	X	X		X			X		X		X
Economic	Initial cost													X
	Operation and maintenance Cost	X	X		X	X		X			X			X				X
	Affordability (rent and utilities)		X		X	X	X	X	X	X	X			X

. For practical purposes, four criteria were used to evaluate the indicators adopted in this research and to determine their eligibility for inclusion, where the indicators are chosen based on their relevance, recency, being the most cited in the literature, and where studies that have identified indicators must look at different contexts, not just one particular context. Table 2 indicates that this is the first study that comprehensively and simultaneously assesses the sustainability perceptions of residential occupants in the UAE in addressing the environmental, economic, and social sustainability aspects, which signifies the importance of this paper.

In addition, the initial building assessment systems were limited to basic aspects such as location, energy, water, materials, waste, pollution, and indoor environmental quality, while the new ones developed in the last decade added several categories to the field of rating systems. Most of the new systems also consider cost and economic aspects. Other added categories include safety, usability and functionality, local culture, outdoor environmental quality, impact on adjacent properties, time, low-carbon aspects, and project design aspects that have improved and enriched the existing systems. Although some building assessment systems have been proposed for some countries, no studies have been conducted to present an assessment system for buildings in other countries such as the UAE.

Moreover, Figure 2 shows the published research work in the sustainability and SD areas in the UAE. It can be observed that their importance has increased across the years, especially in the engineering, environmental, social sciences, and energy—as per Scopus research analysis, which is considered one of the premium databases and peer-reviewed journals [123].

Table 3. A Summary of Perception and Assessment of Sustainability Studies.

No.	Ref.	Context	Method/s	Selection Criteria		No.	Ref.	Context	Method/s	Selection Criteria
				Lit.	GBRTs					Lit.	GBRTs
1	[69]	Spain	ANOVA *	X		20	[106]	Ireland	Descriptive Analysis	X
2	[70]	-	Regression Analysis	X		21	[107]	Slovenia	Case Studies—MCDM	X
3	[73]	Copenhagen	Agent Based Models	X		22	[108]	Portugal	SBTool ****	X	X
4	[76]	Kazakhstan	RSAM **	X		23	[109]	Malaysia	MCDM		X
5	[78]	Spain	Descriptive Analysis	X		24	[110]	Oman	MCDM		X
6	[79]	The Netherlands	Correlation Analysis	X		25	[111]	Sri Lanka	Correlations Analysis		X
7	[81]	Italy	Regression Analysis	X	X	26	[113]	Hong Kong	Case Studies		X
8	[85]	Italy	Descriptive Analysis	X		27	[114]	-	MCDM	X	X
9	[90]	USA	Regression Analysis	X		28	[115]	-	MCDM		X
10	[94]	Iran	MCDM ***	X		29	[116]	Australia	Descriptive Analysis	X	X
11	[96]	Pakistan	ANOVA-MCDM	X		30	[117]	Canada	Case Studies—MCDM	X	X
12	[97]	Pakistan	Cluster Analysis	X		31	[118]	Thailand	Factor Analysis	X
13	[98]	-	MCDM	X		32	[119]	Serbia	Case Studies—MCDM	X
14	[99]	Iran	Content Analysis		X	33	[121]	UAE	MCDM	X
15	[101]	Kazakhstan	MCDM	X	X	34	[124]	-	MCDM		X
16	[102]	Saudi Arabia	MCDM		X	35	[125]	Iran	MCDM	X	X
17	[103]	Kazakhstan	Case Studies		X	36	[126]	-	Descriptive Analysis		X
18	[104]	-	Content Analysis		X	37	[127]	Jordan	MCDM	X
19	[105]	Czech Republic -Australia	Case Studies

*: Analysis of variance, **: Rapid Sustainability Assessment Method, ***: Multiple-Criteria Decision-Making, ****: Sustainable Building Tool.

illustrates a summary of studies found in the literature concerning sustainability perceptions and the methods adopted for each study in different contexts. It can be noticed that only one study took place in the UAE [124] to assess buildings as heritage, ordinary modern, or sustainable modern buildings concerning defined sustainability criteria. More importantly, none of the studies in all the different contexts, including the UAE, aimed at providing a comprehensive framework to assess the perception of the sustainability of the residential sector in the UAE, which embodies and addresses environmental, economic, and social sustainability aspects. In terms of the methods, it can be noted that none of the studies used Structural Equation Modeling (SEM) to measure sustainability perceptions. Therefore, these findings signify the importance of this paper.

In conclusion, the significance of this study lies in targeting the residential sector, which has a major influence on the environmental, social, and economic aspects due to it being the highest reported statistic in consuming energy resources in addition to contributing to greenhouse gas emissions and air pollution. Moreover, since the targeted population represents a large segment of the community, the outcomes of this study would be beneficial to policy/decision-makers, developers, contractors, designers, facility management entities, etc., to allow regulating and applying supportive sustainability practices to enhance overall sustainability in the residential sector. To the best of the authors’ knowledge, this is the first study of its kind that represents a comprehensive assessment of the perception and awareness of occupants in the residential buildings in the UAE based on the TBL paradigm, which is in line with the strategic direction of the UAE towards sustainable development.

2. Materials and Methods

This section presents the research methodology in this paper to fill the literature review gap and meet the aim and objectives of this study, as demonstrated in Figure 3. The first step is concerned with defining the problem of this study to fulfill the anticipated outcomes, and this is achieved by limiting the literature review to cover the aspect of interest, which includes the related sustainability indicators that describe the perception of sustainability by occupants in the residential sector.

After collecting the previously considered indicators from several relevant studies in different countries, subject matter experts (SMEs) were involved to come up with a verified list of indicators that are applied to the context of occupants in the residential sector in the UAE; the indicators were chosen based on their applicability, relevance, recency, and being the most cited in the literature. In addition, the SMEs were consulted regarding the design of the initial survey, and to give the final validation of the questions after piloting the survey by checking the clarity of the questions and revising them accordingly.

A cross-sectional online survey was adopted by using random sampling to avoid sampling bias [128,129]; it included two sections: (i) demographic information of respondents and (ii) a set of questions to measure each sustainability indicator amongst the TBL based on a five-point Likert scale (ranging from 5 = Strongly Agree to 1 = Strongly Disagree). The survey’s questions are outlined in Appendix A. In addition, after collecting the responses, several validation and reliability tests were conducted followed by applying processing techniques to prepare the data for the analysis and enhance the reliability of the results.

Consequently, several sequential analytical steps have been conducted to understand the responses, draw beneficial conclusions, and fulfill the aim of the study. These steps include: (i) descriptive data analysis, where general conclusions about the responses, their demographic information, and their overall assessment are understood, (ii) confirmatory factor analysis (CFA), used to measure–in the context of the residential sector in the UAE–how significant each indicator is to its relevant sustainability pillar (social, environmental, economic), and to confirm the existing theory regarding the correlation among the TBL of sustainability [130,131]; this can be confirmed by testing the hypothesis of having a statistical correlation among them [132], and (iii) structural equation modeling (SEM); to measure the occupants’ perception of sustainability based on the statistical significant indicators that resulted from the previous step. Using the extracted indicators along with the revealed correlation among the three dimensions of sustainability, the initial model is illustrated in Figure 4.

From the figure, it can be noticed that the overall perception of sustainability of occupants in residential buildings is measured using three main variables: social, economic, and environmental pillars. The perception of sustainability is a latent variable in the model, which means that it cannot be measured directly [133].

Therefore, SEM is used to measure it by employing the responses of the survey; it serves as a statistical tool to measure the overall perception of sustainability by measuring the social, economic, and environmental pillars’ contribution to overall sustainability for the targeted population (residential sector). In addition, each pillar constitutes a latent variable that is measured through the related depicted indicators from the literature (iv) To investigate the differences in occupants’ perceptions of sustainability according to their demographic characteristics, a comparative analysis (Kruskal-Wallis test) was performed in SAS. This test was used to test the differences in the overall perception of sustainability among the different class levels based on demographic information–namely age, gender, education level, employment status, monthly income for professionals, marital status, number of family members living together, accommodation type, and residence status.

Accordingly, several Kruskal-Wallis tests were conducted to test the null hypothesis (H₀: There are no differences between the different classes) from the alternative hypothesis (H₁: At least one class level is significantly different). Accordingly, multiple comparison tests were performed using PROC GLM in SAS to identify the sources of the differences. Least squares mean differences and confidence intervals were also used to gain further insight, and finally, (v) building a predictive model for the perception of sustainability based on the collected data of occupants, where the results from each of the aforementioned steps contribute to the benefits of the associated stakeholders including policy/decision-makers, developers, contractors, designers, facility management entities, etc., and allow regulating and applying supportive sustainability practices to enhance the overall sustainability in the residential sector.

3. Results

Since there are differences in the age and educational background of the target audiences, it is possible that some terms are not well understood by a variety of people in the targeted population. Therefore, the survey began by defining perception, sustainability, social sustainability, environmental sustainability, and economic sustainability to clarify these terminologies and thus minimize sample bias that might result from such differences in respondents’ backgrounds. In addition, administration bias was reduced by conducting online surveys, which increases respondent transparency and avoids any interviewer influence, as participants are more likely to display a positive attitude toward an interviewer [134,135]. Additionally, each indicator was represented by a question in the survey. The survey was distributed online for one month (March 2022), targeting people in the residential sector in the UAE regardless of their backgrounds, nationalities, educational levels, etc. In total, 406 responses were received, and they are fully completed. The following subsections represent the results for each part of the analysis.

3.1. Respondents’ Demographic Information

The demographic information for the respondents is summarized in

Table 4. Demographic Information.

No.	Demographics	%	No.	Demographics	%
1	Gender		6	Employment Status
	Male	30.0		Unemployed	30.8
	Female	70.0		Employed	49.7
				Business Owner	13.5
2	Age			Student	4.2
	18–24	6.0		Retired	1.8
	25–30	15.0
	31–40	41.0	7	Number of Family Members Living Together
	41 or older	38.0		1	7.4
				2–4	47.8
3	Marital Status			5–7	41.4
	Single	17.2		More than 7	3.4
	Married	80.3
	Divorced	2.5	8	Accommodation Type
				Rental House	28.8
4	Education Level			Rental Apartment	49.3
	Unschooled	0.5		Owned House	16.0
	High School	4.4		Owned Apartment	5.9
	Bachelor’s Degree	68.0
	Master’s Degree	22.6	9	Monthly salary Range “if employed”
	Ph.D.	4.5		Less than AED15,000	43.0
				AED15,000–AED25,000	19.6
5	Current Residency Status			AED25,001–AED35,000	14.0
	Local	25.0		AED35,001–AED45,000	10.0
	Expat	75.0		More than AED45,000	13.4

. It can be observed that the majority of respondents are females (70%), which implies that females tend to complete online surveys more than males; this was noticed by several researchers as females have a higher response rate for such surveys than males [136,137]. For the age groups of respondents, it was found that 79% of respondents are more than 30 years old, 15% are between 25–30 years old, and 6% are less than 25 years old. In addition, most respondents are married (80%), and most of the respondents’ family members who live in the same house range from 2–7 in number, with around 48% of them having 2–4 members and 41% having 5–7 members. Regarding the educational level, 68% of the respondents hold a bachelor’s degree, 22.6% hold a master’s degree, and the remaining are high-schooled (4.4%), Ph. Ds. (4.5%), and the unschooled yields the lowest percentage (0.5%).

Nevertheless, around 31% of the respondents are unemployed, which might be represented by housewives, since the majority are females and married. However, almost half of the respondents are employed, while the remaining are business owners, students, or retired. Interestingly, 78% of the respondents live in rental accommodations, which might be justified by the percentage of expat respondents (75%) as they tend to live in rental accommodations, unlike the local people. For monthly income, 43% of respondents earn less than AED15k, around 20% earn AED15k-25k, 14% earn AED25k-35k, 10% earn AED35k-45k, and the remaining (13%) earn more than AED45k.

3.2. Data Validation

Testing the validity and reliability of the data is a crucial step prior to starting the analysis. Since the responses are of a qualitative nature, the survey’s questions were reviewed and amended by SMEs to avoid bias in the questions’ formulation [138]. Moreover, random distribution (sampling) of the survey was conducted as another action toward bias avoidance [139], and several measures are taken to test the collected data and ensure their suitability for the required analysis. Firstly, the analysis in this paper requires an adequate sample size, hence, Kaiser-Meyer-Olkin (KMO) test is conducted, where any value between 0.7 and 1 implies that the data is adequate for the analysis [140]. The KMO value of the collected data in this paper is 0.95, which indicates that the sample size is adequate. Secondly, to test the reliability of the collected data, Cronbach’s α and composite reliability (CR) are measured for each construct (social, environmental, and economic) [141]. Cronbach’s α is used to test the internal consistency of responses, while CR is recommended when applying SEM to test how each indicator is related to the other within the same construct.

Table 5. Data Validation.

Dimensions	Correlation within Each Dimension	Cronbach’s α	CR
Social Sustainability		0.934899	0.933204
SVC	0.679425
SAC	0.530025
STC	0.674074
SIQ	0.660227
SGW	0.643177
SGB	0.670736
SAV	0.704612
SCP	0.643326
SST	0.660699
SPM	0.705919
SEA	0.890951
SCI	0.585643
SSS	0.589520
SHO	0.637328
SSI	0.578443
SAT	0.619643
SAS	0.634990
Environmental Sustainability		0.843111	0.842117
EES	0.677030
EEE	0.724276
EWM	0.694750
EWE	0.616997
Economic Sustainability		0.757581	0.767014
EIC	0.463926
ERU	0.657470
EOM	0.653138

shows the values for the correlation between each indicator and other indicators within the same dimension, which shows that there is a correlation among the indicators within each dimension. It is vital to check the correlation among indicators of each dimension as they are assumed to be correlated and coordinately contribute to the measurement of the related dimension [142,143]. Furthermore, Cronbach’s α and CR values range between 0.75 and 0.93, which indicates a high acceptance level for the data; hence, it can be used in the analysis with the set of indicators shown in Table 5.

3.3. Confirmatory Factor Analysis

The CFA is concerned with analyzing the covariance structure of the model. This step helps in testing the goodness of fit for the model, the correlation among the main constructs, and the significance of each indicator within the related construct. After analyzing the covariance structure of the model in Figure 4, the goodness of fit values indicates a statistically acceptable fit where the goodness of fit index (GFI) = 0.9097, the adjusted goodness of fit index (AGFI) = 0.8827, and the Bentler comparative fit index (CFI) = 0.9529, while the standardized root mean square residual (SRMR) error shows an acceptable error (<0.05) with a value of 0.0476.

Additionally,

Table 6. CFA Results.

Social Sustainability			Environmental Sustainability			Economic Sustainability
Indicators	Standardized Estimates	p-Value	Indicators	Standardized Estimates	p-Value	Indicators	Standardized Estimates	p-Value
SVC	0.68022	<0.0001	EES	0.75539	<0.0001	EIC	0.50280	<0.0001
SAC	0.52838	<0.0001	EEE	0.81617	<0.0001	ERU	0.79835	<0.0001
STC	0.66343	<0.0001	EWM	0.75498	<0.0001	EOM	0.84390	<0.0001
SIQ	0.66590	<0.0001	EWE	0.69443	<0.0001
SGW	0.65593	<0.0001
SGB	0.66042	<0.0001
SAV	0.71592	<0.0001
SCP	0.69320	<0.0001
SST	0.68110	<0.0001
SPM	0.72534	<0.0001
SEA	0.92299	<0.0001
SCI	0.61395	<0.0001
SSS	0.62090	<0.0001
SHO	0.64899	<0.0001
SSI	0.58250	<0.0001
SAT	0.63427	<0.0001
SAS	0.68888	<0.0001

shows the CFA results. It is noticed that all the indicators are significant within each construct, with a standardized estimate that is higher than 0.5 and a p-value that indicates statistical significance (<0.0001). Moreover, from the covariance matrix of the three main constructs, there is a significant correlation between social, economic, and environmental sustainability, as demonstrated in

Table 7. Standardized Covariance Matrix.

	Social	Environmental	Economic
Social		0.65126 <0.0001	0.66354 <0.0001
Environmental			0.59156 <0.0001
Economic

.

3.4. Structural Equation Modelling

The SEM was performed to measure how overall sustainability is perceived by the represented population by evaluating the impact of each sustainability pillar (economic, social, and environmental) on the overall sustainability. Each indicator was represented by a question in the survey, and SEM employs the qualitative responses to each indicator to measure the unmeasurable variables (social, economic, and environmental sustainability). Afterward, these variables are used to measure the overall perception of sustainability based on their standardized estimate values and their statistical significance [144]. This analysis is conducted by fitting the hypothetical model that is demonstrated in Figure 4 using PROC CALIS in SAS software; the results from the SEM analysis are illustrated in

Table 8. SEM Results.

Path			Estimate	p-Value
Social	→	SVC	0.67995	<0.0001
Social	→	SAC	0.52814	<0.0001
Social	→	STC	0.66252	<0.0001
Social	→	SIQ	0.66659	<0.0001
Social	→	SGW	0.65593	<0.0001
Social	→	SGB	0.66043	<0.0001
Social	→	SAV	0.71602	<0.0001
Social	→	SCP	0.69297	<0.0001
Social	→	SST	0.68080	<0.0001
Social	→	SPM	0.72538	<0.0001
Social	→	SEA	0.92350	<0.0001
Social	→	SCI	0.61375	<0.0001
Social	→	SSS	0.62161	<0.0001
Social	→	SHO	0.64897	<0.0001
Social	→	SSI	0.58221	<0.0001
Social	→	SAT	0.63420	<0.0001
Social	→	SAS	0.68928	<0.0001
Environmental	→	EES	0.75078	<0.0001
Environmental	→	EEE	0.80843	<0.0001
Environmental	→	EWM	0.75115	<0.0001
Environmental	→	EWE	0.70489	<0.0001
Economic	→	EIC	0.50969	<0.0001
Economic	→	ERU	0.80336	<0.0001
Economic	→	EOM	0.83967	<0.0001
(H1) Social	→	Sustainability	−0.10631	0.1047
(H2) Environmental	→	Sustainability	0.53691	<0.0001
(H3) Economic	→	Sustainability	0.27513	<0.0001

.

The covariance matrix of the indicators is used in this analysis [132,145] as the model is based on the existing theory that needs to be tested using the following hypotheses:

Hypothesis 1 (H1).

Social sustainability has a positive impact on the perception of overall sustainability for occupants in the residential sector in the UAE.

Hypothesis 2 (H2).

Environmental sustainability has a positive impact on the perception of overall sustainability for occupants in the residential sector in the UAE.

Hypothesis 3 (H3).

Economic sustainability has a positive impact on the perception of overall sustainability for occupants in the residential sector in the UAE.

The model fit indicates an acceptable fit with a GFI = 0.9078, AGFI = 0.8802, CFI = 0.9523, and SRMR = 0.0474. From the results in Table 8, it is revealed that each indicator is significant within each construct with a positive standardized estimate and significant p-value. Nevertheless, it is concluded that the overall sustainability of occupants in the residential sector in the UAE is significantly and positively impacted by environmental and economic sustainability aspects (H2 and H3), while social sustainability (H1) plays an insignificant role in defining and understanding the overall sustainability for this targeted population. Furthermore, the impact of environmental sustainability is more intense than the impact of economic sustainability, which would refer to the traditional perception of defining sustainability in terms of environmental aspects.

3.5. Comparative Analysis

Multiple comparative processes are conducted among the different categories of demographic information. The purpose of this comparison is to test if there is any difference in perceiving or understanding the concept of sustainability based on the demographic attributes; these comparisons are carried out using the nonparametric Kruskal test in SAS software. The results of this test are shown in

Table 9. Comparative Analysis.

Demographic Groups	p-Value
Gender	<0.0001
Age	0.1916
Educational Level	<0.0001
Employment Status	0.0020
Marital Status	0.5929
Family Members Living Together	0.1497
Accommodation Type	0.1512
Residency Status	0.4578
Monthly Income	0.0026

; from the p-values, it can be seen that gender, educational level, employment status, and monthly income categories differ in their perception of sustainability, while age, marital status, family members, accommodation type, and residency status categories have not shown any significant difference in their perception of sustainability.

Additionally, pairwise comparisons are done using PROC GLM in SAS for the attributes where a significant statistical difference is found among their categories. Pairs with significant p-values (<0.05) are shown in

Table 10. Pairwise Comparison.

Class	Levels Comparison		Difference between Least Squares Means LSMean(i)-LSMean(j)	Simultaneous 95% Confidence Limits for Difference		p-Value
	i	j		Lower Limit	Upper Limit
Gender	Males	Females	61.402	$39.113$	$83.690$	<0.0001
Educational Level	Bachelor’s degree	Master’s degree	−117.709	$-187.705$	$-47.713$	<0.0001
	Bachelor’s degree	Ph.D. degree	−190.960	$-260.174$	$-121.746$	<0.0001
	Master’s degree	Ph.D. degree	−73.251	$-99.853$	$-46.649$	<0.0001
Employment Status	Unemployed	Employed	−35.750	$-69.986$	$-1.515$	0.033
	Unemployed	Business Owners	−61.890	$-110.569$	$-13.212$	0.004
Monthly Income (AED)	Below 15,000	35,000–45,000	−62.366	−108.771	$-15.960$	0.0017

. Differences between least-square means and 95% simultaneous confidence limits for differences were reported and illustrated in Table 10. The difference in least-square means shows that a higher educational level contributes positively to the perception of sustainability; this finding confirms the vital role of education in perceiving sustainability and that more schooling years would lead to a higher probability of taking sustainable actions, as suggested by [146]. Therefore, including this factor as an indicator for sustainability assessment is extremely important and should not be ignored.

With regards to employment status, the analysis shows a significant difference between the unemployed and employed participants as well as the unemployed and business owners, with p-values of 0.0276 and 0.0051, respectively, where the difference in least-square means shows that the effect size of employed and business owners is higher than unemployed participants. These findings validate the fact that the UAE is one of the pioneering countries in adopting sustainability practices and regulating the behavior of companies and business owners by raising their awareness of sustainability. In addition, most workplaces in the UAE provide specialized training to improve employees’ perception of sustainability and sustainable development [147,148]. The previous findings also justify the significant difference in the level of awareness between males and females, where males’ data showed better results in terms of perceiving sustainability, as most of the unemployed respondents were females.

Similarly, pairwise comparisons among monthly income levels show that there is a significant difference in perceptions between low-income participants (below AED15k) and high-income participants (AED35k-AED45k), with high-income participants scoring higher least-square means in the perception of sustainability. This confirms the existence of a good balance between economic growth and environmental-social aspects of sustainability in the UAE. Such a balance was observed in several studies in Portugal, China, and Indonesia [149,150,151], as an essential factor to pave the way to a realistic path toward sustainable development.

3.6. Predicting the Perception of Sustainability for Occupants of Residential Sector in the UAE

In this part, the perception of the sustainability concept is predicted for occupants of the residential sector in the UAE. The proposed predictive models aim to predict whether the sustainability concept is perceived or not by building a function of all their demographic attributes. Since the output variable is categorical (perceived, not perceived), the prediction takes the form of classification. The collected dataset is used to train the model that represents the relationship between the attributes of occupants (demographic information) and the output (perceived, not perceived), and once the model is built, it can be used for prediction.

The dataset is divided into two sets–training and testing; the training dataset is 75% of the original dataset, while the testing is 25%. The training part is used for fitting the model, which will be applied to the testing dataset for assessing the performance of the prediction. In this paper, decision tree (DT) algorithms are applied for predictions; these models are easy to understand, construct, and interpret, and they are also inexpensive, extremely fast in classifying, robust to noise, and can handle redundant attributes [152,153,154,155]. Three decision tree algorithms are used–logistic regression, CART, and C4.5–where each one uses a different technique to choose what attribute to split and what inputs to split upon them; thus, each algorithm provides a different decision tree.

Moreover, this analysis is conducted using the CARET package in the R software, where nine predictors are used from the demographic attributes to predict an output of two classes; sustainability is perceived, or not perceived. Cross-validation with 10 folds is used for resampling;

Table 11. DT Algorithms Results.

Decision Tree Algorithm	Train Method	Accuracy
		Training	Testing
CART	rpart	78.48%	80.77%
Logistic Regression	glm	78.06%	78.21%
C4.5	J48	78.02%	78.06%

shows the accuracy for each algorithm. The results indicate good and similar values for the training and testing accuracies in each algorithm, which implies that the model performs well for both training and testing. This indicates that the models are neither over-fitted nor under-fitted; nevertheless, the highest accuracy is for CART algorithm (80.77%). It is important to highlight that the accuracy can be significantly improved if the dataset has more records; with more data points the training model and the testing model will perform better.

4. Discussion

Building on the sustainability literature, the present study investigated the perception of sustainability in the residential sector in the UAE. The residential sector is multi-faceted, and it influences the three main pillars of sustainability, namely: economy, environment, and social. In countries that focus on sustainability initiatives through their policies, programs, and projects, policymakers and associated entities must understand the perception of stakeholders in all sectors, especially the residential sector which is on the top of high energy-consuming sectors. The UAE is among the countries in the world that prioritize sustainability and puts intensive initiatives towards achieving it.

Of the various insightful findings of the study, one is a significant correlation among the social, economic, and environmental sustainability pillars. This result confirms the findings in previous studies in different parts of the world, such as Spain [156], Sweden [40], Poland [157], Korea [158], and Ukraine and other EU countries [159,160], that there is a clear and significant correlation between economic, social, and environmental dimensions of sustainable development. Another key finding of this research is that, overall, the survey respondents indicated that sustainability is significantly and positively impacted by environmental and economic aspects. However, their responses clearly showed that they do not perceive the impacts of the social aspect of sustainability. These findings are consistent with previous research suggesting that the social dimension is the least perceived aspect of sustainability, compared to the environmental and economic dimensions [160].

Moreover, the results showed that the impact of environmental sustainability is more intense than the impact of economic sustainability, which is in line with the literature [40]. In this regard, Zwickle et al. [161] noted that this is because the social dimension is the most elusive pillar of the TBL and that economic sustainability is similarly under-assessed. In addition, Anderson [162] justified this assessment by the strong influence of the media on the perception and policy agendas regarding global warming and climate change. To have effective and practical sustainable practices, this study suggests adopting a balanced approach in which environmental issues are harmonized with economic and the social issues. Interestingly, the results of the comparative analysis indicated that a higher educational level contributes positively to the perception of sustainability. This finding is in accordance with recent studies [146], indicating the fundamental role of education in perceiving sustainability. This would lead to increased awareness, and therefore, enhance attitudes and sustainable behaviors.

Additionally, the results also showed a significant difference between unemployed and employed respondents in perceiving sustainability. In agreement with this result, previous studies have demonstrated that with better employee training, employees are more aware of the need for environmental control, and increase their ability to adapt to change and develop a proactive attitude toward sustainability issues [163]. This is also evident in the UAE, where building awareness and capacity across stakeholders and regular workshops are organized in most workplaces [147,148].

The previous findings regarding employment justify the significant difference in the level of awareness between males and females, where males’ data showed better results in terms of perceiving sustainability; this would relate to the fact that most of the unemployed respondents were females. However, previous studies have found that awareness of sustainability does not vary by gender [164,165]. Nevertheless, other studies such as Glass et al. [166] and Kassinis et al. [167] noted that women tend to demonstrate a stronger commitment and tend to be more aware and concerned about sustainable behavior than men. In terms of monthly income, high-income participants have a higher level of perception of sustainability than low-income participants; this highlights the balance between economic growth and socio-environmental aspects of sustainability in the UAE. Prior studies in different parts of the world such as Portugal [149], China [150], and Indonesia [151] have noted the importance of this balance as being a crucial factor to pave the way towards sustainability. Nonetheless, Leiserowitz et al. [168], Bloodhart [169], and Harlan et al. [170] have found that wealthy societies are characterized by overconsumption, materialism, and unsustainable lifestyles.

5. Conclusions and Future Work

This study set out to determine the perception of occupants in residential buildings in the UAE based on the three-pillar paradigm (environmental–economic–social). Each of these pillars is measured by deducing indicators from relevant studies in different countries and validating them by the SMEs. Overall, it was found that the three pillars of sustainability are correlated; the perception of sustainability is more significantly influenced by economic and environmental aspects than by social aspect in the residential sector in the UAE. Moreover, the conducted comparative analysis showed that there is a statistical difference among the multiple demographic attributes including gender, educational level, employment status, and monthly income. Lastly, a predictive classification model was built to predict whether the sustainability concept is perceived or not by the targeted population. This study makes several theoretical and practical contributions.

In terms of theoretical contributions, the research formulated a novel comprehensive framework to assess the perception of sustainability of the residential sector in the UAE, which embodies and addresses environmental, economic, and social sustainability aspects. In addition, the framework can also be useful for defining performance indicators for sustainability assessment frameworks and can be integrated into multi-criteria decision methods to improve sustainability and performance. In terms of practical contributions, the theme of this study is closely integrated with the status quo and planning of sustainable development in the UAE. In this research, planners and policymakers can utilize this framework of sustainability perceptions to make decisions on designing and constructing residential building projects that would enhance the sustainability perception and awareness of the occupants. Furthermore, this framework would be beneficial to policy/decision-makers, developers, contractors, designers, and facility management entities to enhance the overall sustainability in the residential sector.

For future work, it is recommended to collect more data for the predictive model to enhance its accuracy and usefulness. In addition, the study can be conducted in other countries with similar demographic attributes such as GCC as well as dissimilar countries to either validate the findings or withdraw new insights. Furthermore, this study is based on data from a single industrial setting, i.e., the residential industry in UAE. Therefore, including other industries would result in different findings and insights regarding the perception of sustainability.