A Framework for Assessing the Performance of Adaptively Reused Buildings from Users’ Perspective: A Case Study of School Building
Abstract
The positive relationship between building performance and users’ satisfaction and behavior in new buildings has been widely studied. However, in the context of adaptively reused buildings converted from historic buildings, how each performance aspect affects users’ perception is missing. To explore the design concerns needed to be considered related to the performance of adaptively reused buildings from user’s perspective, this paper proposed a framework to examine the effect of performance of adaptively reused buildings on users’ satisfaction and behavior using structural equation modeling. A total of 608 valid face-to-face interview questionnaires were collected at a school building converted from an abandoned industrial building. The empirical conclusions are as follows: (1) Building performance (including indoor environment quality, interior design, building services, and authenticity) all exerted a significant positive correlation with the learning behavior of students, among which indoor environment quality is the most critical factor affecting learning behavior of students. (2) Students’ satisfaction is positively affected by the building performance (including indoor environment quality, interior design, building services, and authenticity), among which authenticity has the greatest effect on students’ satisfaction. (3) The correlation between satisfaction and learning behavior of students is significant and positive. (4) Considering the mediator impact, authenticity indirectly affects learning behavior through satisfaction, while students’ satisfaction has a partially mediating effect between the other three performance aspects and learning behavior. This paper provides a framework for assessing the performance of adaptively reused buildings and gives practical strategies for the improvement of users’ satisfaction and behavior in similar projects.
1. Introduction
Considerable changes have been introduced in industrial structures and functional layouts of cities with the development of society and economy in China [1]. Many abandoned buildings in the city are being considered for adaptive reuse or demolition [2]. The demolition of these buildings would imply the end of its life, which will lead to a massive environmental burden and cause large amounts of debris [3]. Conversely, adaptive reuse, a process of transforming the use of the original built environment to satisfy the new function while preserving the original structure as much as possible [4], is in accordance with the idea of sustainable development due to its significant social, economic, environmental, and cultural benefits [5, 6]. Accordingly, many abandoned buildings have been adaptively reused for new purposes such as creative spaces, schools, museums, housing, offices, shopping mall, venue buildings, and parks [7, 8]. However, the research group conducted an in-depth investigation on 148 adaptively reused buildings in China’s 30 cities during 2015–2018 and found the problems such as high energy consumption, the unreasonable layout of space, and poor environmental quality exists in most investigated projects and users’ feedback towards the performance of adaptively reused buildings was dissatisfied. Previous studies have proved that uncomfortable building performance can lead to a decrease in users’ satisfaction and thus affect their work productivity negatively [9, 10]. Accordingly, it should focus on the performance of adaptively reused buildings.
Previous studies have researched the design strategies of performance of adaptively reused buildings. Bullen and Love [2] proposed that the improvement of the performance of adaptively reused buildings must be directly consistent with the goals of sustainable design (reuse and recovery of materials, reduction of resource and energy consumption, waste discharge, and generation). Celadyn [11] presents design strategies that apply circular design methods and techniques to interior design for achieving resource and waste management efficiency according to environmentally sustainable design principles. For example, the interior components of the adaptively reused building are acquired from original demolished building materials and products. Also, green design methods should be applied to the design of the performance of adaptively reused buildings [12], especially the green and energy-saving design of enclosures. Interestingly, the historical building witnessed the city’s development. Its architectural style, building material, and building structure show the development of science and technology level at that time, even the value orientation of social culture. Profound cultural connotation of the historical buildings has a positive impact on the emotional and spiritual life of modern city residents. Therefore, in the context of adaptive reuse of historical buildings as other functions, many scholars believe that the design of the built environment should show historical information, humanistic feelings, and cultural value of original buildings creatively [13]. Mısırlısoya and Günçe [14] put forward a comprehensive approach that designers should take the social and cultural aspects of historical buildings into consideration and expect physical and functional aspects during the design of adaptively reused buildings. Samadzadehyazdi et al. [15] used the concept of authenticity to identify the degree of cultural expression in adaptively reused buildings and proposed specific guidelines to improve the authenticity of the built environment. In order to realize the physical and functional performance of the adaptively reused building and protect the historical and cultural value of the historical building at the same time, Martinez-Molina et al. [16] reviewed the feasibility of improving thermal comfort and energy efficiency of adaptively reused buildings on the basis of maintaining the cultural value of historic buildings. The design strategies of building performance are being increasingly researched in previous literature in the context of adaptively reused buildings. However, it was mainly from the perspective of government and developers, lacking understanding of the performance of adaptively reused buildings from the user’s point of view. Further, how each building's performance aspects affect users’ satisfaction and behavior in the context of adaptively reused buildings is missing though the positive relationship between building performance and user satisfaction and behavior in new buildings has been widely studied [17, 18].
This paper aims to make up for the lack of influencing mechanism between the building performance, satisfaction, and behavior of users in the context of adaptively reused buildings and to provide a framework to assess the performance of adaptively reused buildings from the perspective of users. Its scientific questions are how each building's performance aspects affect the satisfaction and behavior of users and what is the mediating effect of satisfaction between each building performance aspects and behavior of users, respectively. This study will provide a theoretical framework for assessing the performance of adaptively reused buildings from the perspective of users. The study attempts to explore the design concerns needed to be considered related to the performance of adaptively reused buildings from the user’ s perspective. Users’ feedback on the performance of adaptively reused buildings can provide guidance to designers and managers on how to meet user needs and expectations better. The SEM technique is a multivariate statistical technique that combines multiple regression analysis, factor analysis, and path analysis using both qualitative hypothesis and quantitative data [19, 20]. Also, the SEM approach allows the establishment of structural relationships among theoretical factors by analyzing the structural model and the measurement model separately [21]. There are two main SEM approaches: partial least squares structural equation modeling (PLS-SEM) and covariance-based structural equation modeling (CB-SEM) [22, 23]. Compared with CB-SEM, PLS-SEM provides researchers with more flexibility in terms of data requirements, model complexity, and relationship specification [24–26]. Nowadays, the PLS-SEM approach has been widely used in many fields by scholars, such as psychology management [27–29], behavior [30–32], and social science [33–36]. Therefore, PLS-SEM was used to explore the influencing mechanism between building performance, satisfaction, and behavior of users in this study.
Taking adaptive reuse of an abandoned industrial building as a school building as an example, first, we discussed the theoretical framework to construct our hypotheses based on the design characteristics of school buildings and the principles of the adaptive reuse of industrial buildings, via field investigation, experts’ consultation, questionnaires, and literature review. Second, we established a theoretical framework using PLS-SEM and applied it to explore quantitatively the effect of building performance on satisfaction and learning behavior of students with reference to Xi’an University of Architecture and Technology, Huaqing College. Finally, we proposed strategies for designers and school administrators to enhance the performance of the adaptively reused buildings in a similar project in the future. The objectives of this study are fourfold: testing the relationships between (a) each performance aspect and students’ satisfaction, (b) each performance aspect and learning behavior, and (c) students’ satisfaction and learning behavior and further, proposing improvement measures of performance of adaptively reused buildings to improve satisfaction and learning behavior of students.
The paper is structured as follows. In “Literature review and hypotheses,” we established a theoretical model and proposed hypotheses. Then, the method was proposed in “Methodology.” We carried out data confirmatory factor analysis and correlated analysis in “Result.” Next, “Discussion” analyzed different hypotheses and explained the reason for the result. The final conclusions, contributions, and limitations were presented in “Conclusions.”
2. Literature Review and Hypotheses
The relevant literature consists of two main aspects. The first aspect was related to the performance evaluation of school buildings since it is the new purpose of the selected case study in this paper. The second aspect focused on the evaluation of the performance of adaptively reused buildings since it includes unique features of adaptively reused buildings. By reviewing literature, consulting expert’s opinions, and interviewing with users, factors affecting the performance of adaptively reused buildings are identified. These factors are grouped into four dimensions according to “habitability pyramid” suggested by Vischer [37], namely, indoor environment quality, interior design, building services, and authenticity. Among them, authenticity shows the performance aspect of adaptively reused buildings which is different from new construction, which reveals the historical and cultural value of the original building. The literature review and hypotheses of the theoretical model were shown as follows.
2.1. Indoor Environment Quality
Indoor environment quality includes four main aspects: thermal environment, acoustic environment, lighting environment, and indoor air quality [38]. Indoor environment quality is closely related to the comfort and health of occupants [39]. Many studies have researched the positive relationship between indoor environment quality and satisfaction and the learning behavior of students in school buildings. Issa et al. [40] found that indoor environment quality has a positive effect on occupants’ satisfaction in green Toronto schools. Lee et al. [41] investigated the relationship between indoor environment quality and learning behavior in air-conditioned university teaching rooms via subjective assessment and objective measurement. The result shows improving indoor environment quality can enhance the learning behavior of students. Improving indoor environment quality can enhance students’ performance during the cooling season [42]. Zuhaib et al. [43] also show that indoor environment quality has a positive influence on satisfaction and performance of students in a partially retrofitted university building in Galway.
2.2. Interior Design
Many researchers have emphasized the importance of interior design in buildings. Kwon et al. [44] examined the relationship between interior design and user satisfaction in workplaces, and the result showed the office layout and desk location contribute more to occupant satisfaction with thermal and visual comfort. Tekce et al. [45] also showed that interior design has a significant effect on occupant satisfaction by assessing the performance of office buildings. The elements of interior design, such as space layout, furniture, facilities, finishing, have been found to have a positive influence on students’ perceptions. Barrett et al. [46] revealed that the design of building spaces has obvious effects on the academic performance of students in primary school. Yang et al. [17] indicated that students were more impacted by the functionality and comfort of furniture than by the amount of furniture in higher education classrooms. Further, ergonomic and adjustable furniture is considered optimal to fit students of different sizes, and it allows students to stay focused and attentive in classroom environments [47, 48].
2.3. Building Services
Building services are closely related to the satisfaction and behavior of user [45]. The elements of building services, such as personal control, maintenance management, safety management, facilities services, and cleanliness, have been found to exert a positive impact on the occupants’ perceptions. Personal control is defined as the ability for users to actively control the indoor environment, which directly affects the user’s satisfaction [44]. Safety management plays an important role in the satisfaction and behavior of users. Lai [49] conducted a post occupancy evaluation on a large hostel building and found fire safety has a positive effect on users’ satisfaction. The structure safety of the building is an important predictive indicator for occupants’ satisfaction [50, 51]. Kang et al. [47] proved that facility services influence the satisfaction of students in educational buildings in two different countries and environments. Lacking inadequate cleaning practices can alter the ecosystem of school buildings and encourage the growth and spread of microorganisms that can put the students’ health at risk [52].
2.4. Authenticity
The concept of the authenticity of adaptively reused industrial buildings is referred to as the features that most truthfully reflect and embody the historical and cultural values of the place. Authenticity can be reflected by the design, materials, setting, and workmanship of a building, and it is worth noting that authenticity does not mean completely copying past without any creativity [53]. During the process of adaptive reuse of historical buildings, authenticity has become a prominent design factor in evaluating the satisfaction and behavior of occupants, especially in heritage tourism projects. Nguyen and Cheung [54] demonstrated that perceived authenticity has a positive influence on tourist satisfaction in the context of heritage tourism from the Asian tourists’ point of view. Improving authenticity can generate greater tourist satisfaction towards cultural heritage sites [55, 56]. According to Kolar and Zabkar [57] and Yi et al. [58], they demonstrated that perceived authenticity exerted a significant effect on destination loyalty in the heritage site.
2.5. Satisfaction and Learning Behaviour
Many studies have demonstrated that satisfaction affects the behavior of users in different buildings. Ziegler et al. [59] indicated that job satisfaction can predict the job performance of employees, and higher satisfaction is related to higher performance in office buildings. Students’ learning behavior can be improved by enhancing their satisfaction towards the indoor environment in the library [18]. In the context of heritage tourism, Adiwibowo et al. [60] indicated that the positive correlation between the satisfaction with historical buildings and visitor behavior according to the case of heritage building was reused for commercial purpose in Malaysia.
- (H1)
Indoor environment quality has a positive impact on the satisfaction of students.
- (H2)
Indoor environment quality has a positive impact on the learning behavior of students.
- (H3)
Interior design has a positive impact on the satisfaction of students.
- (H4)
Interior design has a positive impact on the learning behavior of students.
- (H5)
Building services have a positive impact on the satisfaction of students.
- (H6)
Building services have a positive impact on the learning behavior of students.
- (H7)
Authenticity has a positive impact on the satisfaction of students.
- (H8)
Authenticity has a positive impact on the learning behavior of students.
- (H9)
Satisfaction has a positive impact on the learning behavior of students.
3. Methodology
3.1. Study Site
A case study was conducted at the Xi’an University of Architecture and Technology, Huaqing College, which is the earliest example of the adaptive reuse of an abandoned industrial building in Xi’an city, Shaanxi Province, China. It resulted from the adaptive reuse of the Shaanxi steel mill and was once one of the ten famous steel mills in China. The process of adaptive reuse for the Shaanxi steel mill is shown in Figure 2. The selected 1 # and 2 # buildings are adaptively reused from the original two steel rolling workshops of the Shaanxi steel mill in this study. The location and details of the two buildings (before and after adaptive reuse) are shown in Figure 3. The specifications of the buildings are shown in Table 1.
| Location | 1# | 2# |
|---|---|---|
| Original function | Rolling mill | Rolling mill |
| New function | Teaching building | Teaching building |
| Areas | 16000 m2 | 14450 m2 |
| Floors of original function | 1 | 1 |
| Height | 10.7 m | 10.7 m |
| Floors of new function | 2 | 2 |
| Type of structure | Bent structure | Bent structure |
| Number of toilets | 4 | 4 |
| Number of stairs | 4 | 3 |
| Number of air conditioners | 0 | 0 |
| Heating condition | Boiler heating | Boiler heating |
| Material of windows | Aluminum alloy frame glass | Aluminum alloy frame glass |
| Material of doors | Plastic steel door | Plastic steel door |
| Material of internal load-bearing elements | Vertical: Reinforced concrete slab | Vertical: Reinforced concrete slab |
| Horizontal: Reinforced concrete ceiling slab | Horizontal: Reinforced concrete ceiling slab | |
| Enclosing structure | Glass curtain wall | Glass curtain wall |
3.2. Questionnaire
A 150-pilot study was submitted to the professors and students who stayed at the school for over two years. Face-to-face interviews were conducted between October 10–30, 2020, to ensure that the questionnaires could be understood easily, and a total of 118 valid samples were obtained. Appropriate adjustments were made to the questionnaires with the help of the pilot study.
Formal investigations were handed out at the school between November 15–25, 2020 (see Table 2). The respondents were specifically requested not to interact with others, and they were guaranteed to be anonymous. A total of 671 questionnaires were distributed in the survey and 608 were returned. Of these, 63 were excluded because of a number of missing values. The number of questionnaires met the PLS-SEM requirements for a large sample size (N = 200), and the effective rate was 90.61%. The questionnaires were designed with reference to previous studies and divided into two parts. Part I included the basic information: gender, grade, age, school, birthplace, and seat location. Part II included four latent variables of building performance with 19 measurement items and two latent variables: satisfaction and learning behavior with 6 factors. Part II was answered on a five-point Likert scale ranging from strongly disagree (1) to strongly agree (5). The basic information from the samples is listed in Table 3.
| Part I Individual information please mark the suitable option with “✓” | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| 1 | Gender: | (1) Male | (2) Female | ||||||
| 2 | Grade | (1) Sophomore | (2) Junior | (3) Senior | |||||
| 3 | Age | (1) <18 | (2) 18∼25 | ||||||
| 4 | School | (1) School of Architecture | (2) School of Civil Engineering | (3) School of Economics and Commerce | |||||
| (4) School of Art | (5) School of Management Engineering | (6) School of Mechanical and Electrical Engineering | |||||||
| 5 | Birthplace | (1) Northa | (2) Southb | ||||||
| 6 | Seat position | (1) Windows Seat (sitting near the window) | (2) Doors Seat (sitting near the door) | (3) Others (sitting neither near a window or door) | |||||
| Part II | |||||||||
| Specification: according to your perception, please mark score with “✓”. | |||||||||
| 1 = Strongly disagree; 2 = Slightly disagree; 3 = Neutral; 4 = Slightly agree; 5 = Strongly agree. | |||||||||
| Q11 | Temperature is comfortable. | 1 | 2 | 3 | 4 | 5 | |||
| Q12 | Humidity is comfortable. | 1 | 2 | 3 | 4 | 5 | |||
| Q13 | Lighting is comfortable. | 1 | 2 | 3 | 4 | 5 | |||
| Q14 | Indoor air quality is good. | 1 | 2 | 3 | 4 | 5 | |||
| Q15 | Noise is acceptable. | 1 | 2 | 3 | 4 | 5 | |||
| Q21 | Space layout is reasonable. | ||||||||
| Q22 | Furniture is designed for convenience. | ||||||||
| Q23 | Facilities are designed for convenience. | 1 | 2 | 3 | 4 | 5 | |||
| Q24 | Finishing is reasonable. | 1 | 2 | 3 | 4 | 5 | |||
| Q31 | Personal control is convenient. | 1 | 2 | 3 | 4 | 5 | |||
| Q32 | Safety management is good. | 1 | 2 | 3 | 4 | 5 | |||
| Q33 | Maintenance management is good. | 1 | 2 | 3 | 4 | 5 | |||
| Q34 | Facility services are good. | 1 | 2 | 3 | 4 | 5 | |||
| Q35 | Cleanliness is good. | ||||||||
| Q41 | Appearance is full of sense of history and culture | 1 | 2 | 3 | 4 | 5 | |||
| Q42 | Material utilization of original building is adequate. | 1 | 2 | 3 | 4 | 5 | |||
| Q43 | Propaganda of architectural history and culture is adequate. | 1 | 2 | 3 | 4 | 5 | |||
| Q44 | Space utilization of original building is adequate. | 1 | 2 | 3 | 4 | 5 | |||
| Q45 | Newly designed structures have a good relationship with the past. | 1 | 2 | 3 | 4 | 5 | |||
| Q51 | I am satisfied with the overall environment. | 1 | 2 | 3 | 4 | 5 | |||
| Q52 | Comparing with the ideal design, I am more satisfied with the overall environment. | 1 | 2 | 3 | 4 | 5 | |||
| Q53 | Comparing with 4# teaching building, I am more satisfied with the overall environment. | 1 | 2 | 3 | 4 | 5 | |||
| Q61 | My learning efficiency is enhanced. | 1 | 2 | 3 | 4 | 5 | |||
| Q62 | My academic record is enhanced. | 1 | 2 | 3 | 4 | 5 | |||
| Q63 | My memory is enhanced. | 1 | 2 | 3 | 4 | 5 | |||
- Note: a Born in the north of the Huai River-Qin Mountains Line in China; b Born in the south of the Huai River-Qin Mountains Line in China.
| Type | Information | Number | Percentage (%) |
|---|---|---|---|
| Gender | Male | 317 | 52.1 |
| Female | 291 | 47.9 | |
| Grade | Sophomore | 204 | 33.6 |
| Junior | 212 | 34.9 | |
| Senior | 192 | 31.6 | |
| Age | <18 | 9 | 1.5 |
| 18∼25 | 599 | 98.5 | |
| School | School of Architecture | 101 | 16.6 |
| School of Civil Engineering | 117 | 19.2 | |
| School of Economics and Commerce | 105 | 17.3 | |
| School of Art | 82 | 13.5 | |
| School of Management Engineering | 109 | 17.9 | |
| School of Mechanical and Electrical Engineering | 94 | 15.5 | |
| Birthplace | Northa | 502 | 82.6 |
| Southb | 106 | 17.4 | |
| Seat position | Windows seat (sitting near the window) | 286 | 47 |
| Doors seat (sitting near the door) | 40 | 6.6 | |
| Others (sitting neither near a window or door) | 282 | 46.4 | |
3.3. Variable Measurement
There are five items used to determine indoor environment quality. All the items were obtained from the work of Zuhaib et al. [43]; Caro & Sendra [61]; Tahsildoost & Zomorodian [62]. As for interior design, four items were taken from Yang et al. [17]; Barrett et al. [46]. Building services dimensions were measured using five items adopted from Tekce et al. [45]; Kang et al. [47]. The five items for measuring authenticity were from the study of Samadzadehyazdi et al. [15]; Gao et al. [63]; Lu et al. [64]; Taketo [65]. Three items for measuring student satisfaction and learning behavior were from the work of Zhang [18]. The measurement items are listed in Table 4.
| Performance aspects | Measurements items |
|---|---|
| Indoor environment quality | Q11 Temperature is comfortable. |
| Q12 Humidity is comfortable. | |
| Q13 Lighting is comfortable. | |
| Q14 Indoor air quality is good. | |
| Q15 Noise is acceptable. | |
| Interior design | Q21 Space layout is reasonable. |
| Q22 Furniture is designed for convenience. | |
| Q23 Facilities are designed for convenience. | |
| Q24 Finishing is reasonable. | |
| Building services | Q31 Personal control is convenient. |
| Q32 Safety management is good. | |
| Q33 Maintenance management is good. | |
| Q34 Facility services are good. | |
| Q35 Cleanliness is good. | |
| Authenticity | Q41 Appearance is full of sense of history and culture. |
| Q42 Material utilization of original building is adequate. | |
| Q43 Propaganda of architectural history and culture is adequate. | |
| Q44 Space utilization of original building is adequate. | |
| Q45 Newly designed structures have a good relationship with the past. | |
| Satisfaction | Q51 I am satisfied with the overall environment. |
| Q52 Compared with the ideal design, I am more satisfied with the overall environment. | |
| Q53 Compared with 4# teaching building, I am more satisfied with the overall environment. | |
| Learning behaviour | Q61 My learning efficiency is enhanced. |
| Q62 My academic record is enhanced. | |
| Q63 My memory is enhanced. | |
3.4. Analysis Procedures
This study used the PLS-SEM technique to explore the relationship between building performance (including indoor environment quality, interior design, building services, authenticity), students’ satisfaction, and learning behavior. This method is used for establishing, estimating, and testing causality models; analyzing the effects of an individual factor on the whole system; investigating and quantifying the interactions between individual factors [66]. PLS-SEM is used because it involves a complex model setup, explains the variance of endogenous constructs, and enables exploratory and theory development, the use of nonnormal data and small sample sizes, and the formative measurement of latent variables [20]. Advanced Mortar System (AMOS) and Statistical Package for Social Sciences (SPSS) are popular software used in SEM. In this study, AMOS version 22.0 and SPSS 22.0 were applied to analyze the data. The detailed analysis procedures are shown in Figure 4.
4. Results
4.1. Common Method Bias
The common method bias (CMB) refers to an artificial covariation between criteria and predictive variables caused by the same data sources, measurement environment, and context. As a systematic error, this can lead to confusion and misdirection of the results. Harman’ s one-factor test was conducted to test the validity of data to avoid artificial covariation. The exploratory factor analysis showed four factors with eigenvalues greater than 1, and the cumulative variance contribution rate was 66.221%. The variance explained by the first factor is only 18.423%, which is less than the threshold value of 40%. Thus, the CMB is within the acceptable range.
4.2. Reliability and Validity
4.2.1. Reliability
For a more accurate evaluation of reliability, both Cronbach’s alpha and composite reliability were used in this study. Cronbach’s alpha coefficient was used to evaluate the reliability of internal consistency to prove that all questions measure the same phenomenon. The relative advantage of composite reliability over Cronbach’s alpha is that the reliability of structures is not evaluated absolutely; however, it is calculated to consider the correlation of the structures. Table 5 indicates that the Cronbach’s alpha coefficients for the six latent variables are 0.849, 0.890, 0.848, 0.854, 0.822, and 0.840, respectively. They are greater than the threshold value reported by Nunnally [67]. Through the confirmatory factor analysis of the six dimensions (Table 5), the factor loading of the observed variables corresponding to the six latent variables ranged between 0.655 and 0.850, and the square multiple correlations ranged between 0.429 and 0.723; the combination reliability (CR) was between 0.826 and 0.892, which met the requirements of factor loading >0.6, square multiple correlation >0.36, and CR > 0.7 [68, 69]. The Cronbach’s alpha and composite reliability were found to be relatively ideal, which proved that the reliability for this scale is good.
| Construct | Item | Significance of estimated parameters | Item reliability | Construct reliability | Convergence validity | Cronbach’s alpha | ||||
|---|---|---|---|---|---|---|---|---|---|---|
| Unstd. | S.E. | Unstd./S.E. | P-value | Std. | SMC | CR | AVE | |||
| IEQ | Q11 | 1.187 | 0.052 | 22.718 | ∗∗∗ | 0.850 | 0.723 | 0.892 | 0.623 | 0.890 |
| Q12 | 1.011 | 0.055 | 18.460 | ∗∗∗ | 0.716 | 0.513 | ||||
| Q13 | 1.027 | 0.051 | 20.073 | ∗∗∗ | 0.766 | 0.587 | ||||
| Q14 | 1.021 | 0.047 | 21.518 | ∗∗∗ | 0.811 | 0.658 | ||||
| Q15 | 1.000 | — | — | — | 0.797 | 0.635 | ||||
| Interior design | Q21 | 1.113 | 0.065 | 17.066 | ∗∗∗ | 0.775 | 0.601 | 0.849 | 0.585 | 0.848 |
| Q22 | 1.056 | 0.061 | 17.334 | ∗∗∗ | 0.791 | 0.626 | ||||
| Q23 | 1.120 | 0.066 | 16.965 | ∗∗∗ | 0.769 | 0.591 | ||||
| Q24 | 1.000 | — | — | — | 0.722 | 0.521 | ||||
| Building services | Q31 | 1.197 | 0.071 | 16.901 | ∗∗∗ | 0.780 | 0.608 | 0.855 | 0.541 | 0.854 |
| Q32 | 1.172 | 0.069 | 17.037 | ∗∗∗ | 0.788 | 0.621 | ||||
| Q33 | 0.943 | 0.060 | 15.598 | ∗∗∗ | 0.710 | 0.504 | ||||
| Q34 | 0.987 | 0.065 | 15.229 | ∗∗∗ | 0.691 | 0.477 | ||||
| Q35 | 1.000 | — | — | — | 0.703 | 0.494 | ||||
| Authenticity | Q41 | 1.098 | 0.068 | 16.253 | ∗∗∗ | 0.741 | 0.549 | 0.850 | 0.533 | 0.849 |
| Q42 | 1.114 | 0.065 | 17.021 | ∗∗∗ | 0.783 | 0.613 | ||||
| Q43 | 0.992 | 0.060 | 16.507 | ∗∗∗ | 0.754 | 0.569 | ||||
| Q44 | 0.902 | 0.062 | 14.548 | ∗∗∗ | 0.655 | 0.429 | ||||
| Q45 | 1.000 | — | — | — | 0.710 | 0.504 | ||||
| Satisfaction | Q51 | 1.000 | — | — | — | 0.849 | 0.721 | 0.826 | 0.613 | 0.822 |
| Q52 | 0.924 | 0.055 | 16.886 | ∗∗∗ | 0.738 | 0.545 | ||||
| Q53 | 0.807 | 0.047 | 17.126 | ∗∗∗ | 0.758 | 0.575 | ||||
| Learning behaviour | Q61 | 1.067 | 0.058 | 18.422 | ∗∗∗ | 0.787 | 0.619 | 0.841 | 0.638 | 0.840 |
| Q62 | 1.053 | 0.057 | 18.613 | ∗∗∗ | 0.807 | 0.651 | ||||
| Q63 | 1.000 | — | — | — | 0.803 | 0.645 | ||||
4.2.2. Validity
Validity is divided into convergent validity and discriminant validity. Convergent validity refers to the degree of similarity of the measurement results when different measurement methods are used to measure the same construct. Discriminant validity is used to evaluate the ability of a measurement model in differentiating latent variables and other latent variables in the model. When the average variance extraction (AVE) is greater than 0.5, the convergent validity is considered to be good [69]. As summarized in Table 5, the AVE for each dimension exceeds 0.5, which proves the research has good convergent validity. To test discriminant validity, the square root of the AVE was calculated. If the square root of AVE for each latent variable is greater than the internal consistency, the discriminant validity of the model is confirmed [69]. According to the values in Table 6, the correlation coefficients of all potential variables were less than the square root of AVE, which showed that the discriminant validity of all latent variables is very good.
| Construct | Learning behavior | Satisfaction | IEQ | Interior design | Building services | Authenticity |
|---|---|---|---|---|---|---|
| Learning behavior | 0.799 | — | — | — | — | — |
| Satisfaction | 0.684 | 0.783 | — | — | — | — |
| IEQ | 0.629 | 0.677 | 0.789 | — | — | — |
| Interior design | 0.611 | 0.699 | 0.597 | 0.765 | — | — |
| Building services | 0.652 | 0.720 | 0.687 | 0.685 | 0.736 | — |
| Authenticity | 0.540 | 0.717 | 0.392 | 0.519 | 0.527 | 0.730 |
4.3. Structural Model Testing
The maximum likelihood estimation method is used to estimate the degree of fit of the model, and the path coefficient is used to test the previous hypothesis. The acceptability of the theoretical model depends on the goodness of fit between the hypothetical model and the sample data. This paper adopted the nine most widely used fitting indices of the model to report the research results [70]: chi-square (χ2), the number of degrees of freedom for testing the model (df), χ2/df, root mean square error of approximation (RMSEA), comparative fit index (CFI), standardized root mean square residual (SRMR), goodness of fit index (GFI), adjusted goodness of fit index (AGFI), and Tucker–Lewis coefficient (TLI). As listed in Table 7, the results showed that the fitting indices of the model meet the threshold requirement proposed by relevant scholars [70], which proved that the model fits well.
| Statistics for goodness of fit | Standard value | Test value | Adaptability of the model |
|---|---|---|---|
| χ2 | The smaller, the better | 399.702 | — |
| df | The larger, the better | 260 | — |
| χ2/df | 1 < χ2/df < 3 | 1.537 | Qualified |
| RMSEA | <0.05 | 0.030 | Qualified |
| CFI | >0.90 | 0.983 | Qualified |
| SRMR | >0.90 | 0.028 | Qualified |
| GFI | >0.90 | 0.952 | Qualified |
| AGFI | >0.90 | 0.940 | Qualified |
| TLI | >0.90 | 0.980 | Qualified |
According to the results, indoor environment quality (ß = 0.255, P < 0.001), interior design (ß = 0.194, P < 0.001), building services (ß = 0.176, P < 0.01), authenticity (ß = 0.400, P < 0.001), all have a significant positive effect on the satisfaction of students. Authenticity has the greatest effect on the students’ satisfaction. Indoor environment quality has a greater influence on the satisfaction of students than interior design, followed by building services. Meanwhile, indoor environment quality (ß = 0.217, P < 0.001), interior design (ß = 0.131, P < 0.05), building services (ß = 0.177, P < 0.01), authenticity (ß = 0.134, P < 0.05), all positively affect the learning behavior of students. Indoor environment quality was given the greatest weight to influence the learning behavior of students, followed by building services, authenticity, and interior design. Besides, students’ satisfaction has a significant effect on the learning behaviour of students (ß = 0.204, P < 0.05). Therefore, H1, H2, H3, H4, H5, H6, H7, H8, and H9 are all confirmed, which indicates that each building performance aspects has a good ability to predict the satisfaction and learning behaviour of students. The standardized path coefficients are shown in Table 8.
| Hypothesises | Std. | Unstd. | S.E. | C.R. | P | Conclusion |
|---|---|---|---|---|---|---|
| H1: IEQ ⟶ Satisfaction | 0.274 | 0.255 | 0.045 | 5.631 | ∗∗∗ | Accept |
| H2: IEQ ⟶ Learning behaviour | 0.233 | 0.217 | 0.057 | 3.825 | ∗∗∗ | Accept |
| H3: Interior design ⟶ Satisfaction | 0.200 | 0.194 | 0.050 | 3.848 | ∗∗∗ | Accept |
| H4: Interior design ⟶ Learning behaviour | 0.135 | 0.131 | 0.059 | 2.207 | ∗ | Accept |
| H5: Building services⟶ Satisfaction | 0.179 | 0.176 | 0.058 | 3.051 | ∗∗ | Accept |
| H6: Building services⟶ Learning behaviour | 0.180 | 0.177 | 0.067 | 2.640 | ∗∗ | Accept |
| H7: Authenticity ⟶ Satisfaction | 0.412 | 0.400 | 0.043 | 9.200 | ∗∗∗ | Accept |
| H8: Authenticity ⟶ Learning behaviour | 0.138 | 0.134 | 0.060 | 2.222 | ∗ | Accept |
| H9: Satisfaction ⟶ Learning behaviour | 0.204 | 0.204 | 0.094 | 2.178 | ∗ | Accept |
4.4. The Mediating Effect of Students’ Satisfaction
All hypotheses are confirmed, which indicates that the satisfaction of students plays the role of a mediator between indoor environment quality, interior design, building services, authenticity, and learning behavior of students. To calculate the mediating effect more accurately, the assessment of the mediation effects of each building performance aspect on the learning behavior in the proposed structural model is based on PLS-SEM bootstrapping. Bootstrapping is a nonparametric resampling procedure that has been recognized as a more rigorous and powerful method for testing mediating effects.
Table 9 summarizes the analysis results of the mediator effect. The total effects of authenticity with bias-corrected 95% confidence interval (CI) (Lower level = 0.430, upper level = 0.655) and percentile 95% CI (Lower level = 0.430, upper level = 0.655) do not contain “0”, indicating that the total effect is statistically significant. The direct effects of authenticity with bias-corrected 95% CI (Lower level = -0.052, upper level = 0.239) and percentile 95% CI (Lower level = -0.046, upper level = 0.251) contain “0”, indicating that the direct effect is not statistically significant. The indirect effects of authenticity with bias-corrected 95% confidence interval (CI) (Lower level = 0.323, upper level = 0.571) and percentile 95% CI (Lower level = 0.320, upper level = 0.564) do not contain “0”, indicating that the indirect effect is statistically significant. Accordingly, the effect of authenticity on learning behavior is totally mediated by the satisfaction of students. Similarly, the effect of indoor environment quality, interior design, and building services on learning behavior is partially mediated by the satisfaction of students, respectively.
| Type | Point estimates | Product of coefficients | Bootstrapping | ||||
|---|---|---|---|---|---|---|---|
| SE | Z | Bias-corrected 95%CI | Percentile 95%CI | ||||
| Lower | Upper | Lower | Upper | ||||
| Direct effect | |||||||
| IEQ⟶ Learning behaviour | 0.288 | 0.059 | 4.881 | 0.178 | 0.405 | 0.174 | 0.401 |
| Interior design ⟶ Learning behaviour | 0.255 | 0.074 | 3.446 | 0.113 | 0.402 | 0.102 | 0.395 |
| Building services⟶ Learning behaviour | 0.331 | 0.073 | 4.534 | 0.189 | 0.471 | 0.187 | 0.470 |
| Authenticity ⟶ Learning behaviour | 0.104 | 0.073 | 1.425 | −0.052 | 0.239 | −0.046 | 0.251 |
| Indirect effect | |||||||
| IEQ ⟶ Learning behaviour | 0.300 | 0.045 | 6.667 | 0.220 | 0.392 | 0.218 | 0.389 |
| Interior design ⟶ Learning behaviour | 0.334 | 0.061 | 5.475 | 0.224 | 0.470 | 0.219 | 0.464 |
| Building services⟶ Learning behaviour | 0.316 | 0.059 | 5.356 | 0.199 | 0.436 | 0.207 | 0.441 |
| Authenticity ⟶ Learning behaviour | 0.424 | 0.061 | 6.951 | 0.323 | 0.571 | 0.320 | 0.564 |
| Total effect | |||||||
| IEQ ⟶ Learning behaviour | 0.588 | 0.049 | 12.000 | 0.493 | 0.686 | 0.491 | 0.681 |
| Interior design ⟶ Learning behaviour | 0.589 | 0.053 | 11.113 | 0.493 | 0.700 | 0.486 | 0.694 |
| Building services ⟶ Learning behaviour | 0.648 | 0.052 | 12.462 | 0.545 | 0.750 | 0.548 | 0.752 |
| Authenticity ⟶ Learning behaviour | 0.528 | 0.058 | 9.103 | 0.430 | 0.655 | 0.430 | 0.655 |
5. Discussion
5.1. Indoor Environment Quality
The results informed that indoor environment quality has a positive impact on student satisfaction. This means that the better the quality of the indoor environment, the more satisfied the students are with the performance of adaptively reused buildings. The results are in line with research conducted by Lee et al. [71] that indoor environment quality has a positive effect on satisfaction of students. Further, indoor environment quality has a positive and significant direct effect on the learning behavior of students and student satisfaction plays a partial intermediary role between them, which supports the findings that indoor environment quality shares a positive relationship with learning behavior of students in new teaching buildings [52], libraries [18], open-plan research offices [47]. This means a high level of indoor environment quality can increase the learning performance of students. Conversely, poor indoor environment quality can negatively affect the learning behavior of students in classrooms by decreasing attention, concentration, and students’ efficiency [52]. Besides, indoor environment quality is also proved to be the most influential factor for the learning behavior of students. The main reason is that indoor environment quality is directly related to the students’ physical perception, and it has a considerable impact on the students’ physical comfort and health [72]. In fact, the indoor environment quality of school buildings is poor at present. Historical buildings were designed by the old standards and their design standards do not meet the requirements for the new standards. And the energy consumption of original buildings is also large [73]. However, architectural designers neglect the indoor environment quality of the building with new function while focusing too much on the expression of exterior features of the building. Considering the Xi’an University of Architecture and Technology Huaqing College as an example, the case adopted a large number of glass curtain walls to show the aesthetic design of the contrast between the new and the old, while ignoring the thermal comfort to influence the satisfaction and learning behavior of students. Therefore, the performance of adaptively reused buildings should conform to the current building design codes and requirements. Also, the retrofit technology of green design and energy-saving design suitable for the adaptive reuse of historical buildings should be vigorously promoted and used.
5.2. Interior Design
The results agreed that interior design has a direct positive influence on the learning behavior of students. Students’ satisfaction has a partially mediating effect between them, which were similarly observed in the new buildings [44, 46]. From the perspective of environmental psychology, interior design is related to the physiological and psychological needs of occupants. Interior design elements influence occupants’ perception of amenity and efficiency [74]. Also, interior design has an interlaced impact on other building performance factors, such as visual comfort and acoustic comfort. Improper interior design can result in problems such as poor natural lighting and low ventilation rate, which may affect the students’ satisfaction and learning behavior directly. Kong et al. [75] confirmed interior designs have significant impacts on occupant satisfaction with the daylighting environment and also demonstrated the consequence of integrating interior layout and flexible furniture designs with interior designs to achieve better daylighting environments. Compared to building new constructions, the interior design of the adaptively reused buildings is an innovative secondary design within the original building space. A reasonable design scheme can not only save developers’ time and cost but also offer cultural and environmental benefits. Accordingly, interior design should maximize the use of the structure and materials of the existing buildings. Reusing and recycling materials can reduce resource consumption and waste generation, which is in line with the principle of sustainability. Furniture and equipment are ergonomically designed and upgraded regularly to enhance user comfort.
5.3. Building Services
The results indicate that building services directly affect students’ satisfaction and learning behavior, respectively, and satisfaction plays a partial mediating role between building services and the learning behavior of students. This is consistent with the finding of Tekce et al. [45]. The main goal of building services is to provide a comfortable and convenient environment to increase user satisfaction and productivity; building services with good quality contribute to the sustainability of the building. It is worth noting that the quality of building services has an interactive effect on the other performance aspects of buildings. For example, Users’ perceived ability to use and control building services affects their satisfaction with the indoor environment [76]. The level of cleanliness of a building has a positive influence on indoor air quality [39]. Safety is the basic need of users for building performance, ensuring the health and safety of users. Since original buildings were generally built early, the construction quality at that time may be far lower than the current standards or not meet the requirements of the new function. The developer needs to conduct structural inspection and reinforcement before formal use [6]. Besides, mechanical, electrical, and plumbing systems and fire systems are often redesigned due to the original systems being damaged or inefficient. However, the safety of adaptively reused buildings is uncertain due to the lack of original design data. Therefore, compared with new buildings, safety management and maintenance management of adaptively reused buildings play a very important role during the operating process.
5.4. Authenticity
The results revealed that authenticity has a significant and positive effect on the satisfaction of students, and this finding is in line with those of scholars in the field of heritage tourist buildings, who identified a close link between the two in terms of industrial heritage tourism [54], historic districts [64], historical town [77], which all confirmed that authenticity positively affected the visitor’s satisfaction. This study provides additional evidence on the critical role of authenticity, thereby extending previous findings to the adaptive reuse of historical buildings as nonprofitable buildings. At the same time, authenticity is found as the strongest factor affecting satisfaction of students. Authenticity is trying to show the historical value of historical buildings and provide a cultural identity for the adaptively reused buildings [14, 78]. Unique cultural experience creates a positive sentiment for students to contribute to the satisfaction of students. In the survey, we developed the interviews with students in school. They showed a strong requirement for a sense of history and culture of the buildings when they found that the teaching building was adaptively reused from an industrial building. As a result, students prefer buildings with a sense of history and culture rather than just meeting the requirements of new functions. Further, the result also shows authenticity has no significant direct influence on the learning behavior of students. It obviously affects the learning behavior through the students’ satisfaction. In this sense, students’ satisfaction is a complete mediator. This explains the influencing mechanism of authenticity affects the learning behavior of students in adaptively reused buildings. This is the main reason that academic achievement is always used to evaluate the learning behavior of students. The learning behavior of students depends on factors related to academic achievement and their learning attitudes. The design of authenticity may not be at the core of academic achievement. However, the results offered evidence that progress in the learning behavior of students can be achieved by increasing the students’ satisfaction towards authenticity. So, building design of adaptively reused buildings should preserve and exploit the history and cultural value of the original building.
6. Conclusions
- (1)
Building performance (including indoor environment quality, interior design, building services, and authenticity) all exerted a significant positive correlation with learning behavior of students, among which indoor environment quality is the most critical factor affecting learning behavior of students, followed by interior design, building services, and authenticity.
- (2)
Students’ satisfaction is positively affected by the building performance (including indoor environment quality, interior design, building services, and authenticity), among which authenticity has the greatest effect on students’ satisfaction.
- (3)
The correlation between satisfaction and the learning behavior of students is significant and positive.
- (4)
Considering the mediator impact, satisfaction is a complete mediator between authenticity and the learning behavior of students. Authenticity directly influences satisfaction and then impacts the learning behavior of students. While students’ satisfaction has a partially mediating effect between other three performance aspects and learning behavior.
This study has some theoretical contributions. First, this study established the framework for evaluating the performance of adaptively reused buildings from the user’s point of view, which enriches the literature on post occupancy evaluation of the adaptive reuse of abandoned industrial buildings as new functions. Second, taking the adaptive reuse of abandoned industrial buildings as school buildings, this paper verifies the influence mechanism among building performance, satisfaction, and the learning behavior of students. Further, the mediating role of satisfaction between each building's performance aspects and the learning behavior of students is also identified. This is conductive to explore this influencing mechanism in the context of adaptively reused buildings. Third, the authenticity dimension needs to be considered along with indoor environment quality, interior design, and building services when determining the building performance in the context of adaptively reused buildings. This study highlights the significance of preserving the historical and cultural value of original buildings in adaptively reused buildings.
This study also has practical contributions that it provides valuable information to the administrators for improving users’ experience of building performance timely by strengthening the management and maintenance of adaptively reused buildings during the operation phase. Further, it provides useful strategies for designers considering users’ needs and expectations towards adaptively reused buildings at the design stage to better enhance the satisfaction and work productivity of users in similar projects in the future. For example, indoor environmental quality has the greatest positive impact on students’ performance. Therefore, designers and administrators should give priority to the improvement of indoor environment quality and also balance the relationship between the enhancement of indoor environment quality and energy saving reconstruction to achieve the sustainability of adaptively reused buildings. Besides, in order to improve the authenticity of adaptively reused buildings, the structure, materials, and equipment of the original buildings should be preserved and utilized as much as possible so that the unique history and culture of the abandoned industrial buildings can enrich the campus culture. At the same time, energy-saving and carbon reduction measures of adaptively reused buildings should avoid damaging the historical and cultural sense of the original buildings as much as possible.
Although our study has some interesting findings and contributions, it is matter to emphasize the limitations of this research. The case in this study comes from a typical city in northwest China. There could be differences in the regional economy and cultural aspects in different countries. Therefore, in order to obtain a more general result, respondents expanded, and typical cities in different regions should be selected for further research. Besides, this paper took a school building converted from an abandoned industrial building as a case study. Other new functions should be investigated using the proposed framework to get more data in the future.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
The authors would like to thank Dr. Tian xiaoxu, Dr. Dong zhifang, and Dr. Liu bo for the warm help during the surveying and analysis. This work was supported by the National Natural Science Foundation of China (grant nos. 51808424 and 51677879); Ministry of Housing and Urban-Rural Development of the People’s Republic of China (grant no. 2018-R1-009); and Shanxi Provincial Department of Education (grant no. 18JK0458).
Open Research
Data Availability
All data used to support the findings of this study are available from the corresponding author upon request.
