Do overall findings correlate with granular findings? An analysis of factors influencing office workers’ thermal comfort and indoor air quality

Indoor Air Quality (IAQ) and Thermal Comfort (TC) are key components of Indoor Environmental Quality (IEQ) and strongly affect workplace comfort and productivity. Poor IAQ can cause respiratory issues, headaches, and fatigue, lowering morale and efficiency (Chen et al., 2023). Similarly, poor TC can lead to heat stress and reduced concentration (Foster et al., 2020; Saidi and Gauvin, 2023). In contrast, well-managed IEQ supports health and performance, with IAQ improvements linked to productivity gains (Felgueiras et al., 2023; Aniebietabasi et al., 2024).

Understanding the role of IAQ and TC in the workplace reveals the complex link between air quality, comfort, and productivity. Key variables include temperature, CO₂, and humidity. Extreme temperatures can significantly impact occupational health and productivity. Saidi and Gauvin (2023) argued for the need to monitor body temperature in real-time in the workplace to prevent high thermal stress. Elevated CO₂ levels impair cognition and cause drowsiness. Chen et al. (2023) demonstrated the difference in improved performance among sixty-nine healthy university students at varying CO₂ levels (600, 1,500, and 2,100 ppm). Poor humidity control can lead to poorer sleep quality and disrupt thermoregulation. McCabe et al. (2022) noted that humid heat is reported to particularly affect slow wave sleep in the earlier phase of the sleep period, and cold exposure is found to impact the quality of sleep in later segments of the sleep. Maintaining these within optimal ranges is crucial for a healthy and productive work environment.

Several factors influence IAQ and TC in workplaces. Office layout—open plan, cubicles, or private—affects air circulation, temperature, and pollutant dispersion, shaping occupant exposure (Torriani et al., 2025). Time spent at work is also key, as prolonged exposure to poor conditions can worsen health and reduce long-term productivity (Rasheed et al., 2021). Understanding IAQ and TC requires moving beyond broad averages to a more detailed analysis. While general assessments provide an overview, they often miss localised issues that affect specific areas. Granular analysis enables the precise identification of problems and targeted interventions (Gallego et al., 2025) to improve workplace health and comfort.

This study employs a granular analysis of IAQ and TC to investigate how workplace factors impact comfort and productivity. It contrasts overall and detailed measurements of temperature, CO₂, and humidity with occupants’ perceptions, and explores correlations with individual comfort responses.

Granular analysis of building performance data disaggregates complex information related to buildings into more specific components. Unlike overall IEQ measurements, which offer a general snapshot of environmental conditions, granular analysis provides aggregated data over a specified space or time frame, enabling a more nuanced understanding, uncovering local variations and patterns influenced by factors such as time of day, location within the building, and occupancy levels (Pollard et al., 2021; Melo et al., 2023; Imani et al., 2025).

One of the key advantages of granular analysis is its capacity to capture temporal and spatial differences in IEQ. For example, pollutants like CO₂ and VOCs can fluctuate significantly across various office areas or throughout the day, influenced by factors such as occupancy levels, ventilation rates, and human activities. By employing granular analysis that examines specific locations and timeframes, such as areas with poor environmental conditions or identifies times when conditions decline, targeted interventions can be enabled. Recent research has begun to evidence this.

For instance, Pollard et al. (2021) established the feasibility of synthesising continuous IEQ exposure based on high-resolution real-time location data. The authors measured IEQ using 12 autonomous desk-mounted devices spread across the office floorplate, a Real-Time Location System (RTLS) that tracked occupant location continuously over one month, with 47 location sensors across the research site (1,220 m²) and 45 tags attached to occupants’ staff ID access cards.

Likewise, Imani et al. (2025) employed a mixed-method approach, utilising handheld IEQ measurement devices and the Eco-Scout system, which autonomously navigated 10 workstations to retrieve IEQ levels, in determining occupants’ individual IEQ satisfaction levels. The authors demonstrated the superior performance of this mixed method in enhancing data quality and accuracy, producing more precise data collection capabilities than conventional approaches. Similarly, Melo et al. (2023) evaluated annual electricity data from 3,060 fully electric-powered buildings. The authors noted that Monthly and hourly factors establish energy benchmarks for finer time resolution.

These works argue the necessity of conducting hourly or daily measurements instead of relying solely on annual averages, as annual averages may obscure critical trends affecting indoor environments.

Subjective comfort assessments are vital in granular analysis, offering insights into occupants' real-time experiences. Studies show strong links between perceived comfort and environmental conditions. Kawakubo et al. (2023) found that the relationship between thermal sensation and reported productivity varied by gender. The authors noted that men reported being most productive when the temperature was cooler than thermally neutral, while women were most productive when the temperature was warmer than thermally neutral. Similarly, Lin et al. (2025) evidenced that the effect of raised temperatures on performance becomes significant after exposure durations exceeding one hour. Zhou et al. (2023) demonstrated that office workers experience a detrimental impact on cognitive performance when exposed to elevated indoor air quality levels of PM2.5. These findings suggest that when occupants perceive poor air quality or temperature, their satisfaction declines.

While objective IEQ measures are important, research highlights the role of contextual factors in shaping how occupants perceive comfort and satisfaction (Rotimi and Rasheed, 2024; Rasheed et al., 2021). These factors explain why people respond differently to similar conditions (Kawakubo et al., 2023). A key influence is perceived control—the belief that one can modify their environment. Even in suboptimal conditions, occupants who feel able to adjust settings report higher satisfaction (de Dear and Brager, 1998; Azzazy et al., 2024). In centrally managed, shared spaces, limited control can lead to discomfort, disengagement, and dissatisfaction.

Contextual situations of environmental comfort are shaped by temporal and spatial variability. For instance, occupants in naturally ventilated spaces typically tolerate wider temperature ranges than those in air-conditioned environments (Suerich-Gulick et al., 2022). Physical activity also influences comfort—the rate of walking and running affects thermal sensation (Liu et al., 2023). Comfort is thus dynamic and context-dependent.

This study explores how such variability influences perceived IAQ and TC, highlighting how office type and time spent indoors shape CO₂ levels, comfort, and health outcomes.

This study aimed to demonstrate how granular analysis enhances understanding of the relationship between IAQ, TC, and occupant comfort, satisfaction, and productivity. We compared overall and granular measurements of temperature, CO₂, and humidity with occupant perceptions, supported by qualitative comments. These variables were analysed across the following contexts.

1. Individual offices

2. Office type (private; shared offices)

3. Daily work hours (less than 8 h; 8 h or more)

4. Frequencies of IEQ Compliance and Exceedance

Three objectives were set to be achieved using both qualitative and quantitative methods (Rasheed et al., 2024; Zhao et al., 2024). Quantitative IEQ data provides objective, comparable evidence of conditions, while qualitative feedback explains how occupants experience those conditions.

1. Objective 1: To examine the influence of office type and individual offices on measured IEQ variables

2. Objective 2: To analyse the influence of office type, individual offices and daily work hours on perceived environmental conditions.

3. Objective 3: To ascertain whether the measured IEQ compliance and exceedance frequencies influenced perceived environmental conditions.

This exploratory study aimed to understand the experiences and influencing factors of office occupants (Rasheed and Rotimi, 2024; Kawakubo et al., 2023). It focused on temperature, humidity, and CO₂—key indicators of IAQ and thermal comfort that have direct effects on health and well-being (ASHRAE, 2017; Foster et al., 2020). Noise and lighting were excluded to maintain this focus, as they influence perception but do not directly impact IAQ or TC (Weerasinghe et al., 2024; ISO, 2012).

The case study involves a recently renovated university office, expanded from 2,484 m² to 4,373 m² and equipped with a new centralised environmental management system to enhance IEQ and energy efficiency. The inclusion of a modern mechanical ventilation system enabled assessment of comfort factors without the confounding effects of equipment age or maintenance. The space accommodates staff and PhD students, with both shared and private offices shown in Figure 1.

IEQ data were collected using Tether EnviroQ sensors integrated with Tether software (Rasheed et al., 2024). Sensors were calibrated prior to installation (Zuhaib et al., 2018) and placed 1.2 metres above the floor—within the EPA's recommended breathing zone (0.9–1.8 m) (EPA, 2024)—and positioned away from vents or pollution sources, near workstations to capture occupant-relevant conditions. Readings were recorded hourly from March 1, 2023, to February 28, 2024. Table 1 outlines the characteristics of each office, including occupant count and type.

We conducted a perception-based study using a self-evaluated questionnaire to assess workers' comfort, productivity, and satisfaction with their environmental conditions. Self-evaluation is widely used to gather workplace perceptions (Rotimi and Rasheed, 2024; Pastore and Andersen, 2022). Following a sequential development process, we identified key IAQ and TC variables from existing literature and designed primarily quantitative questions, complemented by qualitative ones for deeper insights (Rasheed and Rotimi, 2024). Ethics approval was obtained (Notification ID: 4000027990), and the final survey was deployed via SurveyMonkey. The questionnaire, covering the same period as IEQ monitoring, addressed comfort, control, perceived productivity, and behavioural responses (Weerasinghe et al., 2024) (see Table 2).

The questionnaire was distributed to the respondents, and out of the 56 occupants in the office space, a total of 30 responses (60%) were deemed viable for analysis. Table 3 shows the demographics of the respondents.

Quantitative data were analysed using IBM SPSS 28 (Khoshbakht et al., 2024; Rasheed et al., 2024), while qualitative comments were assessed manually. ANOVA with Tukey HSD was used to test for significant mean differences across contexts (Rasheed et al., 2021), and linear regression was employed to evaluate the influence of context on perceived environmental conditions (Onkangi et al., 2025). Analysis was limited to working hours (7 am–7 pm, Monday–Friday), and IEQ values were benchmarked against standards for temperature (18–24°C), humidity (30–60%), and CO₂ (≤800 ppm) (WorkSafe, 2019).

Table 4 below shows that the mean readings for all the IEQ variables were within the recommended thresholds (CO₂ = 522.18 ppm, temperature = 21.53 °C, humidity = 58.78%). These overall indices indicate that this office environment is operating within acceptable environmental conditions that should provide a conducive indoor climate quality for occupants.

Next, a granular examination is conducted based on individual offices and office types to provide a more representative evaluation of the office environment.

1. Measured IEQ levels per Individual Offices

We examined the measured IEQ levels in their offices, and Table 5 below shows that very high levels were recorded in all the workstations (i.e.> 800 ppm, >24 °C, and >60%). Specifically, Office A recorded the highest CO₂ level (2,730 ppm), while the highest temperature reading was in Office M (27.3 °C). Office D recorded the highest humidity (100%).

The standard deviations in the offices show a relatively low variation in CO₂ levels in offices A, B, C, G, I, J, and K–O compared to offices D, E, F, H, and P. In particular, higher variations in CO₂ were observed in offices F and H. Office M had the lowest SD, indicating better air quality. Most offices maintained consistent temperature control as the SDs were below 1°C. Offices J and L are the exceptions, with more varied temperature readings (SD above 1^oC).

Exceedance readings were recorded in 1914 cases for CO₂, 1,822 cases for temperature, and 33,652 cases for humidity. These readings were above 800 ppm, outside the range of 18°C–24 °C, and outside the range of 30–60%. Office D had the highest frequency of exceedance humidity levels (3,368 cases), Office M had the highest frequency of exceedance temperature levels (270 cases), and Office H had the highest frequency of exceedance CO₂ levels (246 cases).

Regarding the ANOVA test, the Tukey HSD Post Hoc Analysis showed that while there were significant differences in the measured IEQ levels (p < 0.05) among the offices, some offices recorded IEQ levels that were similar (p > 0.05) based on the mean comparisons. For example, while the CO₂ levels in Office P and the humidity in Office J were significantly different from those in all other offices, the temperatures in Offices B, E, H, L, N, and P were significantly different from those in all the other offices. However, no significant differences were observed in the measured CO₂ levels in Offices A, G, I, and L. Similarly, Offices A, I, and J had overlapping temperature levels, while Offices B, H, and I showed similar humidity levels.

1. Measured IEQ levels per office type

We also examined the mean difference between measured IEQ levels based on the types of offices occupied by occupants (private and shared office types).

Table 6 below shows that very high levels were recorded in both private and shared workstations (i.e.> 800 ppm, >24 °C, and >60%). The highest readings were recorded in a private office (CO2 = 2,730 ppm, temperature = 27.3 °C; humidity = 100%). Similarly, private offices recorded the lowest exceedance frequencies (temperature = 16.1oC; humidity = 0%).

The standard deviations in both office types indicate that shared offices exhibit higher variability in CO₂ levels, while both office types maintain relatively stable temperatures and humidity.

The ANOVA test Analysis showed that there were significant differences in all the measured IEQ variables between private and shared offices (p < 0.05).

For overall perceived IEQ, the mean values indicated that occupants felt more comfortable (m = 3.67), more satisfied with the TC (m = 3.37) and IAQ (m = 2.93), and believed their productivity was improved by the environmental conditions in their workstations (m = 2.6). Additionally, they did not believe their behaviour in the workplace was influenced by environmental conditions (m = 1.33). However, they were less satisfied with their control over the temperature and ventilation in their workstations (m = 1.20 each). Table 7 presents the mean values of occupants' perception of the environmental conditions.

Following suit, a granular examination is conducted based on individual offices, office types, and daily work hours to provide a more representative evaluation of occupants’ perceptions of the environmental conditions.

1. Perceived environmental conditions per individual office space

For perceived environmental conditions, Table 8 shows that all occupants were satisfied with their overall comfort (m > 2.50), except for the occupant in Office B (m = 1). The occupant in Office C noted their satisfaction with the TC in their workstation as very poor, whereas others were satisfied with it (m > 2.50). Occupants in offices A and H did not rate the TC in their workstations. Except for the occupant in Office N, all other occupants were satisfied with the IAQ in their workstations. Not all occupants were satisfied with the level of control they had over the temperature and ventilation conditions in their workstations.

Interestingly, their perception of whether the environmental conditions improved their productivity in their workstations was mixed. While seven occupants did not think so (m < 2.5), eight occupants felt so (m > 2.5). Regarding the impact on behaviour change, none of the occupants felt their behaviour in the workplace was affected by environmental conditions (m < 2.5).

As shown in Table 8, the ANOVA test revealed no significant differences in workers’ perception of the IEQ variables among the offices (p > 0.05).

1. Perceived environmental conditions per Office type and Time spent at work per day

We examined the mean difference in perceived environmental conditions based on the office types occupied by occupants and the amount of time they spent at work each day.

Table 9 shows that occupants in private offices were slightly more satisfied with their overall comfort and IAQ (m = 3.78; 3.11) than those in shared offices (m = 3.62; 2.86). However, they were less satisfied with the TC (m = 3; 3.50) and control over temperature and ventilation (m = 1.11; 1.24). Also, they felt their productivity was less improved by the environmental conditions in their offices. Neither group noted a significant behaviour change as a result of the environmental conditions.

As shown in Table 9, occupants who spent 8 h or more in their offices were more satisfied with their overall comfort (m = 4.57; 3.39) and TC (m = 3.50; 3.33) but were less satisfied with their IAQ (m = 2.43; 3.09) and their control over temperature (m = 1; 1.26) and ventilation (m = 1.14; 1.22). Additionally, occupants who spent 8 h or more in their offices reported that the environmental conditions improved their productivity (m = 3; 2.48) and experienced more behavioural changes (m = 1.86; 1.17) than those who spent less time in their offices each day.

The ANOVA test revealed significant differences in the perceived environmental conditions (p < 0.05). Workers' perceptions differed for behaviour change only between private and shared offices (p = 0.00). Meanwhile, between workers who spent less than 8 h per day at work and those who spent 8 h or more, their perceptions differed significantly regarding overall comfort (p = 0.032) and associated behavioural changes (p < 0.001).

1. Occupants' comments on the environmental conditions in their offices

In the perception study, more negative than positive comments were recorded about the office environment. Positive feedback—mainly generic—came from occupants in Offices A, H, K, and P, who appreciated the space's beauty and its support for thinking and writing.

Negative feedback was more detailed. Occupants in Offices C, F, I, B, and D disliked the inability to open windows, preferring natural ventilation over air conditioning, which they deemed unnecessary in a moderate climate. One occupant in Office I criticised its cost, disruptive noise, and strange vibrations. Office A also reported a persistent cold, a lack of thermostat control, and a reliance on artificial lighting, necessitating winter clothing during the summer. Office C cited frequent headaches from the fixed 22 °C setting and often left early seeking fresh air.

An occupant in shared Office E complained about overly cold summer conditions, while Office G noted fluctuating temperatures. In Office F, one occupant felt disconnected from the outdoors and preferred working from home, calling the building unproductive despite its appearance.

In Office I, another occupant described it as “working inside a box,” expressing the need for more fresh air, although they used a standing desk to stay active. Office L reported stuffiness and poor ventilation, even with the door open. The remaining offices gave no comments.

As described above, while the average measured IAQ and TC variables fall within acceptable limits, the occupant satisfaction scores reveal dissatisfaction with the environmental conditions in several locations. This suggests that occupants' perception of environmental conditions might have been influenced by periods of exceedance in the measured IAQ and TC variables, i.e. the periods when the measured IAQ and TC variables fell outside the recommended thresholds.

To ascertain this, we conducted a regression analysis to determine whether the frequencies of measured IEQ exceedances (readings outside the recommended comfort thresholds) and compliance (readings within the recommended comfort thresholds) influenced occupants' perceptions of the environmental conditions. For example, we wanted to ascertain if an increase in the frequencies of temperature exceedances predicted a decrease in occupants’ satisfaction. Table 10 presents the frequencies of exceedances and compliance temperatures, humidity, and CO₂ measurements.

Hence, we employed linear regression analysis to determine how satisfaction with environmental conditions (IAQ and TC) changed in relation to the frequency of measured IEQ complaints and exceedances, the significance of the relationship, and how well the regression model explained the outcome of the relationship. The results are provided in Tables 11 and 12. Figures 2 and 3 show the regression plot for the relationships described above. Each point represents a location, while the trend line shows the relationship (R² values). The wide 95% confidence interval (shaded areas) indicates high variability and low statistical significance, implying that exceedance frequency alone is not a strong predictor of the perceptions.

1. IEQ Complaint Frequencies and TC and IAQ Satisfaction

Table 11 shows a weak linear regression model of a positive Beta coefficient (β = 0.031) with a p-value of 0.916 (F = 0.012), indicating that the relationship between TC perception and temperature complaint frequencies is not statistically significant at the 95% confidence level. The R² value (0.001) means that the temperature complaint frequencies explain just 0.01% (less than 1%) of the variation in perception. Similarly, humidity complaint frequencies had a weak positive relationship (β = −0.103) with thermal comfort. The relationship was not statistically significant (p-value = 0.727; F = 0.128). The 0.011 R² value means that measured humidity exceedances explain just 1.1% of the variation in perception.

For IAQ perception, the regression Beta coefficient for measured CO₂ complaint frequencies and IAQ perception was −0.179, indicating a small negative effect. The p-value = 0.508 and F-statistic = 0.462 suggest a weak and statistically non-significant relationship. The R² value is only 0.032, meaning measured CO₂ complaint frequencies explain just 3.2% of the variation in IAQ perception. For humidity-compliance frequencies and IAQ perception, the regression analysis produced a Beta coefficient of −0.344, with a p-value of 0.192 and an F-statistic of 1.881, meaning the relationship is not statistically significant. The R² is 0.118, suggesting that only about 1.2% of the variance in IAQ satisfaction can be explained by humidity complaint frequencies.

1. IEQ Exceedance frequencies, TC and IAQ Satisfaction

For TC perception, the table indicates that the linear regression model yielded a weak negative Beta coefficient (B = −0.152) with a p-value of 0.605 (F = 0.282), indicating that the relationship is not statistically significant at the 95% confidence level. The R² value is only 0.023, meaning measured temperature exceedances explain just 2.3% of the variation in perception. Similarly, a weak negative effect (B = −0.094) was also observed between measured humidity exceedances and temperature perception. This means the relationship is not statistically significant at the 95% confidence level (p-value = 0.750; F = 0.106). The 0.009 R² value means that measured humidity exceedances explain just 0.9% (less than 1%) of the variation in perception.

The same outcome was the case for IAQ perception. The regression Beta coefficient for measured CO₂ exceedance and IAQ perception was −0.217, indicating a small negative effect. The p-value = 0.418 and F-statistic = 0.695 suggest a weak and statistically non-significant relationship at the 95% confidence level. The R² value is only 0.047, meaning measured CO₂ exceedances explain just 4.7% of the variation in IAQ perception. For humidity exceedance and IAQ perception, the regression analysis produced a Beta coefficient of +0.057, with a p-value of 0.835 and an F-statistic of 0.045, meaning the relationship is not statistically significant. The R² is 0.003 for the relationship between humidity and IAQ perception. This suggests that humidity exceedance frequency can explain only 0.3% (less than 1%) of the variance in IAQ satisfaction.

Granular data analysis reveals complex factors affecting building performance and enables more precise improvements in design and management (Pollard et al., 2021; Chen et al., 2023; Imani et al., 2025). Unlike traditional methods that rely on broad performance indices (Wei et al., 2020), our study demonstrates how detailed insights provide a more accurate explanation of variations in workplace comfort and productivity.

First, we assessed IEQ across office types, comparing overall indices with granular readings (Objective 1). While annual averages met acceptable limits (Table 4), some periods exceeded thresholds (Table 5). Office D, in particular, recorded 100% humidity and the highest number of exceedances (3,368), signalling a risk of mould due to prolonged high moisture.

ANOVA results showed significant differences in IEQ levels across offices (p < 0.05), especially for temperature in Offices B, E, H, N, and P, and CO₂ in Office P. Private offices had higher average IEQ levels but fewer exceedances. This finding supports previous work, such as Moazami and Sterud (2025), who studied 7,968 Norwegian office workers and found that shared offices had significantly poorer indoor climate conditions than private offices. Additionally, Parkinson et al. (2023) observed that occupants of open-plan offices with low or no partitions were nearly twice as likely to complain about their workspace as those in a private, enclosed office.

Next, we explored how office type, individual offices, and work hours influenced occupants' perceptions (Objective 2). While most were satisfied with thermal comfort (TC), indoor air quality (IAQ), and overall comfort, many were dissatisfied with their control over temperature and ventilation. At the office level, occupants in Office B reported low overall comfort, Office C noted TC issues, and Office N rated IAQ poorly. This finding is similar to that of Ha et al. (2021), who found carbon dioxide levels exceeded 800 parts per million in a small conference room with 8 occupants and an office with 3 occupants.

No significant differences were found in occupants’ perceptions based on office type, likely due to only slight variations in comfort and IAQ ratings (<0.5). However, work hours showed a significant effect (p = 0.32): those working ≥8 h daily reported higher overall comfort and TC satisfaction but lower satisfaction with IAQ and control over temperature and ventilation. This aligns with findings that prolonged office exposure increases vulnerability to poor IAQ and TC effects (Vellei et al., 2023; Rasheed et al., 2021).

Overall, occupants felt that the office's environmental conditions did not impact their productivity. This view was stronger among those in private offices and those working fewer than eight hours per day. Rasheed et al. (2021)’s study supports this, as they noted that no worker in their 5,149 dataset thought their productivity was substantially affected by the IEQ factors in their offices.

The granular analysis also revealed mixed occupant experiences. While temperature discomfort was the most common issue, other concerns included disconnection from nature, a preference for remote work, and lack of control over temperature and ventilation—particularly in shared offices. Some respondents expressed satisfaction with their workspace, while silence from others suggested possible indifference to environmental conditions.

Regarding IEQ exceedances and satisfaction (Objective 3), we observed weak, statistically insignificant relationships. As CO₂ and temperature exceedances increased, satisfaction with IAQ and TC decreased, while humidity exceedances showed an inverse trend. Humidity exceedances had the strongest link to IAQ satisfaction, followed by CO₂. Overall, exceedances had a greater influence on satisfaction than compliance, but other factors likely played a more significant role. Past works, such as Onyeizu and Byrd (n.d.) and Rotimi and Rasheed (2024), argue that other factors in the workplace contribute to or may be more important for workers’ comfort, well-being, and productivity.

Relying solely on averaged IEQ indices masks critical variations. Our study showed that while annual mean readings were within acceptable limits, exceedances occurred at specific times. Significant IEQ differences were observed across office types—private offices had higher readings but fewer exceedances. Granular analysis also revealed variations in comfort based on daily work hours. Although measured IAQ and TC were within limits, many occupants reported dissatisfaction, particularly with temperature and air quality. Regression analysis revealed no significant relationship between satisfaction and IEQ exceedances, suggesting that other influencing factors warrant further exploration. In addition to the limitations of a single-building case and a one-time survey, which constrain the generalisability of the findings, we also acknowledge that statistical significance thresholds do not quantify the magnitude of the effect or its practical relevance, and they do not imply causality. That said, our findings highlight the value of granular data in capturing nuanced building performance. High-resolution analysis, even on a small scale, supports more precise and effective building management.

Ultimately, a shift toward granular, high-resolution data analysis is expedient, even in small-scale research. Practically, our findings promote the integration of granular data for smarter building management by identifying and addressing inefficiencies that average measurements often overlook.

We are indebted to the Ph.D. students who participated in the data collection phase of this project – Ali Hashemi and Masoud Mahmoodi. We are also grateful for the support from the Massey University Fund – Strategic Research Endeavour Fund (SREF) for this project.