The global energy sector faces significant challenges, with human and social dimensions playing a critical role. Buildings account for approximately 40% of global GHG emissions, making the building sector a key area for intervention. Current efforts primarily focus on techno-economic assessments and legislation, neglecting the crucial aspect of human-building interactions (HBIs). Occupants interact directly and indirectly with the building environment, and failures to incorporate human dimensions lead to performance gaps. Existing literature shows that human factors significantly influence energy use; for instance, ‘wasteful workstyles’ can double energy consumption compared to ‘austere’ coworkers. Occupant comfort preferences also vary based on age, gender, and cultural factors, creating challenges for commercial building operation. Technology investments alone are insufficient to guarantee low-energy consumption or high comfort perception. The ‘energy efficiency gap’—the difference between cost-optimal and actual energy efficiency—highlights the need to address end-user behavior. Barriers to energy efficiency include risk, imperfect information, hidden costs, access to capital, split incentives, and bounded rationality. End-user behavior is a fundamental barrier at both individual and organizational levels, requiring policies aimed at enhancing energy efficiency. Empirical data from case studies, surveys, and real behavioral data are crucial for understanding this stochasticity. The discrepancy between predicted and actual energy consumption can be significant (up to three times higher), emphasizing the need for a multidisciplinary approach incorporating socio-personal parameters. This study aims to determine all aspects of building sustainability, including energy efficiency and optimal comfort conditions, recognizing the economic implications of productivity loss due to suboptimal comfort. The study will address the challenges of balancing energy consumption and human well-being, particularly in diverse populations such as those found in GCC countries, where rapid economic growth has led to high energy consumption due to Western-influenced architectural styles that are not suited to the local climate.

Existing literature highlights the importance of incorporating human factors into energy efficiency strategies. Studies have shown significant correlations between demographic factors (gender, age, ethnicity) and energy consumption patterns, as well as the influence of socioeconomic factors (income, homeownership) on energy efficiency behaviors. Research has also demonstrated the impact of occupant behavior on energy use, emphasizing the need for a multidisciplinary approach that integrates psychological and cognitive-behavioral methods to understand human decision-making processes related to energy consumption. The literature review also identifies various methods used in previous research for studying HBI, including surveys, monitoring data, and simulations. It highlights the challenges in capturing this data, particularly with regards to securing permissions for surveys and the difficulties of linking subjective data with data from building sensors. The review includes studies on the impact of energy subsidies, demand response (DR) programs, and energy pricing on occupant behavior. It identifies the limitations of relying solely on financial incentives and technical considerations for designing effective DR programs, underscoring the importance of considering factors such as comfort and user experience. Several studies exploring human attitudes and drivers towards energy efficiency are reviewed, revealing the significant influence of financial motivations, environmental concerns, and social norms on energy consumption behavior. The review concludes by stressing the need for case-specific assessments and research to address the stochasticity inherent in occupant behavior and to achieve accurate prediction and management of energy use in buildings.

This study used a survey administered to 2200 respondents in Doha, Qatar, to investigate the interdependencies in HBIs. Qatar, a GCC founder member, provides a relevant context with its diverse population (90% expatriate and migrant workers). The survey collected data on demographic (age, gender, ethnicity), socioeconomic (income, expenses, marital status, employment), and behavioral factors (consequence awareness, responsibility, habits, preferences). Building properties (construction year, floor area, type, AC type) were also collected to estimate Energy Use Intensity (EUI). Quantitative and qualitative measures assessed indoor thermal and lighting comfort and its influence on well-being and productivity. Respondents were asked about their interactions with building systems (window use, thermostat adjustments) and their willingness to participate in demand response (DR) programs. The k-means clustering method categorized respondents into high and low consumers based on EUI values. A feature importance analysis using random forest classifiers (Gini impurity, permutation importance, SHAP) identified key contributors to consumption behavior. Three specific techniques—Gini impurity, permutation importance, and SHAP—were used to determine the feature importance of demographic, socioeconomic, and behavioral factors in predicting energy consumption patterns. The random forest classifiers were trained using a bootstrapping algorithm and a random patching method. The accuracy of the model was evaluated using 10-fold cross-validation.

The study revealed profound insights into how human factors influence energy consumption. Key findings include: 1. **Consumer Clustering:** The k-means clustering method successfully categorized respondents into high and low energy consumers based on their EUI. 2. **Feature Importance Analysis:** Feature importance analysis using random forest classifiers showed that household expenses, age, and ethnicity were the most significant predictors of energy consumption patterns. Monthly household expenses were found to be a stronger predictor than household income. Consequence awareness showed less impact on energy consumption category. 3. **Thermal Comfort in Residential Buildings:** Women reported higher comfort levels and considered comfort more important than men. Higher household income correlated positively with comfort perception and its importance. Ethnic groups showed variations in comfort preferences, suggesting that targeted energy policies should consider these differences. Newer buildings, villas, and modern buildings showed higher indoor comfort levels than older buildings, flats, and traditional buildings. 4. **Thermal Comfort in Workplaces:** Women preferred colder temperatures than men. Arab nationalities and North Americans preferred cold environments, while Asians preferred warmer ones. Women were less inclined to override thermostat settings or open windows compared to men, while Asians showed a greater tendency to adjust the indoor environment. In general, a comfortable indoor environment positively impacted work performance. Open office spaces led to greater discomfort than personal offices. 5. **Awareness, Motives, and Responsibility:** Females showed greater concern about climate change than males. Financial motives were more prevalent across different ethnicities, and coercive actions were seen as more effective in improving responsibility. Females exhibited more responsible energy consumption at work than males. Lower-income groups reported higher instances of adverse interactions in work environments. 6. **Financial Drivers:** Higher electricity prices were more effective in reducing energy consumption among males, Arab, Asian, and Indian ethnic groups, and low-income households. DR program participation was primarily driven by financial incentives in many groups. Waived electricity bills were more likely to lead to increased energy consumption among lower-income groups. Apartment renters exhibited greater willingness to participate in DR programs than villa owners or homeowners. Old style buildings showed a higher participation rate for DR programs.

The findings address the research question by demonstrating the significant impact of social and human dimensions on energy consumption in the building sector. The study highlights the need for a human-centered approach to energy policy, emphasizing that technological interventions alone are insufficient to achieve substantial energy savings. The results underscore the importance of understanding the interplay between demographic, socioeconomic, and behavioral factors in shaping energy consumption patterns. The significant role of household expenses, age, and ethnicity in predicting energy use suggests that targeted interventions should be designed to consider these factors. The variations in thermal comfort preferences across gender and ethnic groups call for customized strategies in building design, operation, and energy pricing. The insights into human attitudes and behaviors can inform the development of effective awareness campaigns and incentive programs to promote energy efficiency and encourage participation in DR programs. The study's findings provide a strong foundation for designing more effective energy policies that address the human dimension in diverse societies, ultimately leading to increased sustainability and improved well-being.

This study offers valuable insights into the human dimensions of energy consumption in buildings, particularly in diverse societies like those in the GCC. The key contribution is the demonstration that incorporating human factors is crucial for effective energy policy and achieving sustainability goals. The use of machine learning methods allowed for the identification of key predictors of energy consumption, providing a robust basis for targeted interventions. Future research should focus on longitudinal studies to track behavioral changes over time and explore the impact of specific policy interventions. Analyzing the effects of cultural norms and social influence on energy consumption behavior would further enhance the understanding of the human dimension in energy systems. Further investigations should also delve into the specific barriers to DR program participation and explore innovative incentive mechanisms.

The study is limited to a specific geographic location (Doha, Qatar) and may not be generalizable to all GCC countries or other regions. The cross-sectional nature of the survey limits the ability to track changes in behavior over time. The reliance on self-reported data may introduce some biases. While the study utilized advanced statistical methods, it does not account for all the complexities of human behavior. The model accuracy, though high, can be improved by including further factors. The study did not examine the effectiveness of specific policy interventions and further research is needed to investigate the impact of various policy instruments on energy consumption behaviors.