© 2021 Children, Youth and Environments Children, Youth and Environments 31(1), 2021 Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes and Pro- Environmental Behavior Wanlu Liu Jin Chen Chinese Academy of Sciences Key Laboratory of Tropical Forest Ecology Xishuangbanna Tropical Botanical Garden Citation: Liu, W. & Chen, J. (2021). Green spaces in Chinese schools enhance children’s environmental attitudes and pro-environmental behavior. Children, Youth and Environments, 31(1), 55-87. Retrieved from http://www.jstor.org/action/showPublication?journalCode=chilyoutenvi Abstract Previous research has asserted that promoting connection to nature can enhance children's environmental attitudes and cultivate their pro-environmental behavior. School green spaces and educational programs have been recognized as important interventions toward this goal, although empirical study examining their effectiveness, especially in China, is rare. This study collected data from 1,597 students (9−12 years old) in southwest China. The study used the Two Major Environmental Values scale to measure two dimensions of children’s environmental attitude—preservation and utilization—and used a project-specific pro- environmental behavior scale to measure children’s self-reported pro-environmental behavior. The results found that students' perceived school environment is consistent with objective measures of “green space quality.” Meanwhile, students’ perception of their school environment and their interactions with natural elements have a positive correlation with students’ preservation attitude and pro- environmental behavior. School green activities have a negative correlation with a utilization environmental attitude. This study thus highlights the value of green campuses, and when supplemented by environmental pedagogical activities, they could significantly contribute to equipping children with environmental literacy. Keywords: school green space, perception, environmental attitudes, pro- environmental behavior http://www.jstor.org/action/showPublication?journalCode=chilyoutenvi Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 56 Introduction Interacting with nature benefits children in several ways (Chawla, 2015; Gill, 2014), including improved physical flexibility and balance (Fjørtoft, 2004; Sando & Sandseter, 2020), cognitive function (Dadvand et al., 2015; Ulrich et al., 2008), creativity (Samborski, 2010), social interactions (Baines & Blatchford, 2011; Titman, 1994), and mental health (Chawla, Keena, Pevec, & Stanley, 2014). Childhood interactions with nature also cultivate pro-environmental behavior (PEB) during both childhood and adulthood (Li & Chen, 2015; Whitburn, Linklater, & Abrahamse, 2020). Because of this, an increasing number of initiatives aim to enhance children’s PEB by promoting natural observation, camping, exploration (Zhang, Zhao, & Chen, 2019), and the like. However, although children benefit from experiencing nature, today they have fewer opportunities to experience it (Louv, 2008). Rapid urbanization (Miller, 2005), changing lifestyles that prioritize computers and mobile phones over outdoor activity (Pergams & Zaradic, 2008), and parents’ worries about outdoor safety (Dyment, 2005), lead to a deficit of nature experiences in children’s lives. The deficit may lead to a lack of basic familiarity about the natural environment (Atran, Medin, & Ross, 2004; Nabhan & Trimble, 1994) as well as physical and cognitive problems (Bell, Wilson, & Liu, 2008; Dadvand et al., 2015; Huh & Gordon, 2008). School campuses are vital to ameliorating this problem; school-age children spend most of their active time (eight hours per day on average) at school (Guo & Yang, 2008). Studies have shown that children who attend schools with green grounds have more opportunities to play (Boldemann et al., 2006; Jansson, Gunnarsson, Mårtensson, & Andersson, 2014; Malone & Tranter, 2003; Moore, 1996; Zamani & Moore, 2013), enjoy safer and less hostile outdoor environments (Dyment, 2005; Evans, 2001), develop enhanced social relations (Baines & Blatchford, 2011; Maller & Townsend, 2006; Titman, 1994), and have better cognitive development (Dadvand et al., 2015) and academic performance (Li, Chiang, Sang, & Sullivan, 2019; Kweon, Ellis, Lee, & Jacobs, 2017). Studies have also shown that teachers working at greener schools are more creative while designing their curriculum (Dyment, 2005; Moore, 1996). Other research has shown that green space provides important physical and psychological benefits to campus residents (Scholl & Gulwadi, 2018; Wee, 2017). Global Green School Efforts In recent decades, school green space development initiatives have been promoted around the world. In 1994, the Foundation of European Environmental Education developed the “Ecological School Plan,” which encourages energy-efficient and sustainable schools and emphasizes student-led management of the school environment to transform student behavior (Lysgaard, Larsen, & Læssøe, 2015). Other initiatives, such as the U.S. Green Building Council’s Leadership in Energy and Environmental Design (LEED) for Schools in the United States, Green Star Education in Australia, and the Deutsche Gesellschaft für Nachhaltiges Bauen in Germany, emphasize schools’ physical environments (Meiboudi, Lahijanian, Shobeiri, Jozi, & Azizinezhad, 2017). These efforts include increasing green space, improvement of indoor air quality and classroom acoustics, removal of toxic materials, sustainable waste management, and the like (Zhao, He, & Meng, 2015). Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 57 China began to promote green schools in 1996, when the government mandated their establishment to foster environmental education (Tan, Chen, Shi, & Wang, 2014; Wu, 2002). China’s Green Schools project emphasized both the physical environment, including energy saving and sustainable development, and environmental education in schools (Wu, 2002). By October 2000, there were 3,207 schools at various levels in 16 provinces named “green schools” (Zhiyan, Hongying, & Xuhong, 2004). However, few studies have evaluated whether the program’s outcomes have been successful. Potential Benefits of School Green Space on Students’ Environmental Orientation Studies investigating the relationship between general school green space with students’ environmental orientation are limited. A study that investigated the routines of students from 134 preschools in Stockholm found that among schools with similar demographics and teaching philosophies, those closer to urban green spaces had children who were more empathetic and concerned for other life forms and were more aware of human–nature interdependence (Giusti, Barthel, & Marcus, 2014). A quantitative study showed that in a school ground greening process, children’s physical use of the school ground was important for establishing a positive and caring relationship with the landscaped area (Jansson et al., 2014). Another study found a positive relationship between sustainable facilities at school campuses and students’ environmental attitudes and behavior (Tucker & Izadpanahi, 2017). Other research has demonstrated that students at an eco- school often showed higher environmental attitudes than students at a control school (Boeve-de Pauw & Van Petegem, 2013b). Studies that have included the school physical environment as a variable have discussed the area of green coverage at the school (Dadvand et al., 2015), the existence of a school garden (Blair, 2009), or the green landscape visible out of the school windows (Kweon et al., 2017). To our knowledge, there is a deficiency of investigations on the effects of other potentially important variables such as degree of plant diversity (Lindemann-Matthies, Junge, & Matthies, 2010; Sivarajah, Smith, & Thomas, 2018), diversity in vegetation structure, or landscape richness (Määttä et al., 2019). Likewise, few studies have examined whether variations in a school’s physical environment affect students’ perception of green space (but see Akoumianaki-Ioannidou, Paraskevopoulou, & Tacho, 2016; Li, Ni, & Dewancker, 2019; Hart, 1979). More empirical studies are needed to understand the specific features that make a green school most effective and the ways a green environment affects students’ environmental attitudes and pro-environmental behaviors. In this study, we aimed to add to the understanding of these processes by expanding the school green space variables into a more comprehensive perspective, considering variables such as plant species diversity, vegetation structure diversity, and landscape possibility (richness). In addition, we hypothesized that other features such as potted plants in classrooms and corridors (van den Berg, Wesselius, Maas, & Tanja-Dijkstra, 2017), and the existence of Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 58 recreational facilities (Akoumianaki-Ioannidou et al., 2016) are also important variables for the school physical environment. Considering these variables in combination, our aims for this study are to provide a clear vision on what makes a “good” school green space, and thereby to deliver suggestions on school green space practices. Potential Variables Connecting School Physical Environment and Pro- Environmental Behavior Environmental Perception People make sense of the objective environment through their own perceptions. Environmental perception is an integral component of human-environment interaction, providing a link between the physical context and individuals’ subsequent cognitive, affective, and behavioral responses (Ittelson, 1976; Stokols, 1978). Environmental perception has commonly been defined as awareness of, or feelings about, the environment, and as the act of apprehending the environment by the senses (Zube, 1999). It is also the first step in the process of developing environmental knowledge and awareness (Albuquerque & Alves, 2016). Studies have shown that a well-designed school environment can foster a dynamic relationship between students and the environment (Orr, 1997; Taylor, 1993). The school physical environment has the potential to train and sharpen students’ perception, enhance their sensitivity, and allow them to develop an aesthetic sense of the environment (Demirbaş & Demirkan, 2003). An empirical study found that students’ aesthetic perception of green space is positively correlated to the students’ feelings of security (Karimian, Samiei, & Kazemi, 2017). Students’ perception of their school environment is a potential variable between the physical school environment and students’ environmental attitudes and PEB. Interaction with Natural Elements at School Research suggests that adults’ pro-environmental attitudes and behaviors are largely shaped by childhood experiences in nature (Chawla & Cushing, 2007; Hinds & Sparks, 2008; Li & Chen, 2015). Experiences in nature may consist of any number of interactions with natural elements, such as catching fish in a creek, observing a caterpillar on a leaf, smelling flowers, etc. A green environment at school may provide children such opportunities to interact with elements in nature, observing insects or plants in the parterre, reading books under the shadow of a tree, or growing plants in the school garden (Akoumianaki-Ioannidou et al., 2016). These kinds of activities may allow students to develop an emotional affinity to nature (Collado, Staats, & Corraliza, 2013), which in turn can enhance their positive environmental attitudes and pro-environmental behavior. Environmental Attitudes Environmental attitudes refer to a person’s psychological tendencies when evaluating an environmental object (e.g., environmental protection, appreciation of nature, or the human-environment relationship) (Kaiser, Hartig, Brügger, & Duvier, 2013; Milfont & Duckitt, 2010). Many studies have reported strong correlations between environmental attitudes and PEB (Bamberg & Möser, 2007; Johnson & Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 59 Činčera, 2015; Pooley & o’Connor, 2000); however, a few other studies have found weak correlations (Gifford & Sussman, 2012; Müderrisoglu & Altanlar, 2011). There are two dimensions of environmental attitudes: preservation and utilization (Milfont & Duckitt, 2004; Wiseman & Bogner, 2003). The former refers to “a biocentric dimension that reflects conservation and protection of the environment,” while the latter refers to “an anthropocentric dimension that reflects the utilization of natural resources” (Wiseman & Bogner, 2003, p.787). Studies have shown that the preservation dimension of environmental attitude has a positive relationship with PEB (i.e., students who scored higher on the preservation dimension performed more environmentally friendly behaviors in their daily life), while the utilization dimension has a negative relationship with PEB (i.e., students who scored higher on the utilization dimension performed fewer environmentally friendly behaviors in their daily life) (Johnson & Činčera, 2015; Pooley & o’Connor, 2000). Environmental Education in Chinese Schools Students’ environmental attitudes and PEB depend not only on students’ perception of their school environment and their interactions with natural elements, but on other variables as well. Among these, environmental education activities may have a positive influence on students’ environmental attitudes and PEB (Collado et al., 2013; Soga & Gaston, 2016). Environmental education began to be integrated into China’s basic education curriculum in 1987, and the country is experiencing a rapid growth in the field of environmental education (Tian & Wang, 2016). The “Green School Program” has been promoted throughout the country since 1992. The Ministry of Education requires schools to integrate environmental knowledge, attitudes, and values into the compulsory school curriculum, for example, by offering environmental education courses as an independent subject, fully integrating environmental education into school schedules, offering teacher training, and conducting research on environmental education pedagogies (Tian & Wang, 2016). In addition to the efforts of the government and formal education systems, some Chinese schools also have opened their doors to environmental non-governmental organizations. According to an incomplete investigation, as of 2018, there were nearly 400 environmental education organizations in China, of which 54% were commercial companies and 21% were non-government organizations. Fifty-six of the organizations were newly established (Feng, Xiao, Zhou, & Wu, 2018). These environmental education organizations provide schools with more professional environmental-focused resources, such as natural observation tour guides in schools, themed course implementation, and undertaking school outings. Thus, a growing number of children have chances to participate in environmental education activities both within and outside of school, and we assume that Chinese children have benefited from this process. Research Questions Previous studies have outlined the relationship between school environment attributes and environmental attitudes and PEB (Tucker & Izadpanahi, 2017). However, few studies have explained the mechanisms by which this relationship can be built, and which factors among the campus green space variables are Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 60 significant predictors of educational outcomes. This study aimed to fill this gap and examine whether and how a school green environment impacts students’ environmental attitudes and PEB. Here, we propose the following hypotheses: • Hypothesis 1 (H1): A green school environment has a direct and positive influence on students’ perception of their school environment (PSE). • Hypothesis 2 (H2): Students’ PSE and their interaction with natural elements (INE) have a direct positive effect on the preservation dimension of students’ environmental attitude. • Hypothesis 3 (H3): Students’ PSE and INE have a direct negative effect on the utilization dimension of students’ environmental attitude. • Hypothesis 4 (H4): Students’ PSE and INE have a direct positive effect on their PEB. Materials and Methods Geographic Context and School Selection This study was conducted in the city of Chongqing, which is a central municipality city in southwest China in the Yangtze River upstream region. Chongqing has a humid subtropical monsoon climate, with an annual average temperature ranging from 16–18℃ and an average annual rainfall of 1000–1350 mm. Chongqing has more than 30.5 million permanent residents, and the urbanization rate is about 62.6%. The main body of the city includes nine districts with a total area of 5,472.68 km2, and there is a 39.8% green coverage ratio as of 2016 (Chongqing Statistics Bureau, 2018). In 2017, there were 425 primary schools in Chongqing, including 146 in rural areas and 279 in urban areas (Chongqing Statistics Bureau, 2018). Based on the school list from the local education department website, we randomly selected 20 schools for participation in this study. We then contacted the principal of each school, explained the purpose and contents of our study, and asked permission for surveying their schools. Three schools declined participation because of conflicts in schedule or distrust of us. We then selected another three to replace them. Figure 1 shows the geographic location of each school (red solid circles). (See Table 3 in Appendix A for basic information about the 20 participant schools.) Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 61 Figure 1. Map of participating schools Data Collection Physical Environment Based on the findings of previous studies (Burns, 2010; China Green Building Council, 2013; Voigt, Kabisch, Wurster, Haase, & Breuste, 2014), we investigated eight variables regarding the school physical environment. 1) Green space area (GSA) According to the Green Campus Evaluation Criteria (CSUS/GBC 04-2013) (China Green Building Council, 2013), school green space refers to “all kinds of green space within the scope of school land, such as public green space, dormitory green space, etc. The green space of primary and secondary schools should include concentrated green land, scattered green land, water surface, plantings for teaching practice, and small animal feeding gardens” (China Green Building Council, 2013, p. 33). We used Google Earth to calculate the area of the large green spaces at each school and field measurements to calculate the area of smaller green spaces. The total green space area (GSA) equals the area of the large green spaces plus the area of smaller green spaces. 2) Green space area per student (GSA/stu) Green space area per student (GSA/stu) measures the ratio of total green space area to the total number of students at each school. This indicator can show how crowded green space is at a given school. Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 62 3) Plant species diversity (PD) Studies have found that plant diversity has a positive influence on the perceived restorative properties and self-reported benefits of urban green spaces (Carrus et al., 2015; Zhao, Wu, & Wang, 2019). A more plant-diverse school ground has also been demonstrated to be more aesthetically appealing for students (Lindemann- Matthies & Köhler, 2019). In this study, we identified every plant species surrounding each school, including trees, shrubs, vines, grasses, aquatic plants, and herbs. We then took photos of each species to facilitate identification, which was accomplished with the help of a plant taxonomist. 4) Vegetation structure diversity (VSD) Vegetation structure diversity (VSD) refers to the different types of plant communities on a school ground: a single large tree, a single small tree, groups of trees, rows of trees, lawn, flower beds, low vegetation, natural shrubs, aquatic plants, vine plants, and potted plants (Voigt et al., 2014). We counted the number of vegetation structure types for each school. 5) Landscape possibility (LP) Nature realizes children’s ten “central capabilities” in several ways (Chawla, 2015, p. 3; Nussbaum, 2011). The richer the interaction possibilities with the environment, the more likely children can benefit from it. Landscape possibility (LP) refers to the landscape’s affordances of a wide variety of children’s activity (Burns, 2010). According to Kyttä’s definition, affordances are “the functionally significant properties of the environment that are perceived through the active detection of information” (Kyttä, 2002). Affordances include properties from both the environment and the acting individual, which means affordance is unique and relative for each child (Gibson, 2014). Researchers have found that each child adopts a different sensory learning modality (Smith, 1998). For example, according to Mahdjoubi and Akplotsyi (2012), children with a visual learning style prefer places with flowers, trees, landscaped fields, and multi-colored scenery. Meanwhile, children with an auditory learning style prefer environments such as bushy areas, where they can interact with their peers while remaining out of sight, and children with a kinesthetic learning style prefer spacious environments that encourage active exploration and movement. Children’s perception of the environment is also affected by their social and cultural context (Kyttä, 2002; Albuquerque & Alves, 2016), so the same environment may have different affordances for different children. We assumed that a school with more landscape possibilities will meet more children’s physical and psychological needs. We recorded the following landscape possibilities at each school: small gardens, woods, vegetation fields, roof gardens, pools, meadows, theme parks, vegetation gardens, backyard gardens, and single big trees. These landscape types are “non-prescriptive”; they stimulate the imagination and can be used in multiple ways. 6) Recreation facilities (RF) For the purposes of this study, we included the following as recreation facilities, and counted the types at each school: flower beds with seats, stools, chairs, open-air theaters, reading galleries, calligraphy galleries, pergolas, terraces, and pavilions. Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 63 7) Green area of corridors (GAC) We measured the green areas of the school’s interior corridors (GAC) using tape measures. This includes the area of planted plants as well as potted plants in the corridors of the school buildings. 8) Indoor plants (IPlant) We recorded the number of indoor potted plants in classrooms at each school and used the average to represent the indoor plants in the entire school. School Green Activities School green activities (GA) refer to educational activities related to the environment or nature, organized by schools with the goal of improving students’ environmental literacy. In this study, the following types of activities are included: natural notes activity, cleaning hosted by the school, art performances related to the environment or environmental protection, lectures, science competitions, hand- written activities (such as themed compositions or freely designed hand-written papers), and recycling projects. We conducted interviews with teachers to record the number of green activities held in the last two years at each school. Participants We conducted a survey of two randomly selected Grade 5 classes at each school from September to November 2017. The informed consent protocols were approved by the participating schools, and school principals provided approval to approach classes. Students had the freedom to choose whether to complete the questionnaire or not, and only students who agreed to participate were handed the questionnaire. A total of 1,860 students finished the questionnaire; we excluded 170 invalid questionnaires, resulting in a validity rate of 90.8% and a sample size of 1,690. We also omitted the results of students who had been in the school for less than a year (n = 93); we felt that it was too short of a time for them to be affected by the school. This resulted in a final sample of 1,597 valid responses: 829 boys and 861 girls ages 8–12 (97.8% were ages 10–11). Of these, 89.9% had been at their current school for at least four years. Self-Reported Variables The survey asked students to report on their perception of the school environment, their interaction with natural elements at school, other environmental activity in which they participated outside of school, their environmental attitudes, and their pro-environmental behavior. Each of these variables is described below. 1) Perception of the school environment (PSE) Environmental perception is the organization, identification, and interpretation of sensory information in order to represent and understand the environment (Gärling & Golledge, 1989). We relied on existing literature (Collado & Corraliza, 2012) to develop four items measuring students’ perception of their school’s physical environment (Table 1). Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 64 2) Interaction with natural elements (INE) at school Interaction with natural elements at school refers to activities related to the natural environment in which students engage during school free time. We used four items to measure INE (Table 1). 3) Other environmental activity (OEE) Apart from school, students’ environmental attitudes are influenced by their environmental education activities or experiences outside of school. To measure this, we included five items to measure the environmental education activity students joined outside of school (Table 1). 4) Environmental attitudes We used the Two Major Environmental Values model (2-MEV) scale developed by Bogner and Wiseman (2006) to measure students’ environmental attitudes. This scale consists of two subscales: preservation and utilization. It has been determined reliable and valid in multiple countries (Braun, Cottrell, & Dierkes, 2018; Castéra, Clément, Munoz, & Bogner, 2018; Johnson & Manoli, 2010; Boeve- de Pauw & Van Petegem, 2013a). Liu and Chen (2020) developed a modification of this scale for us in China, and in this study, we use their 16-item Chinese version. 5) Pro-environmental behaviors (PEB) We used existing scales to develop a project-specific pro-environmental behavior scale to measure children’s self-reported PEB (Liu & Chen, 2020). This scale consisted of ten items and measured children’s pro-environmental behaviors such as energy saving, resource saving and recycling, persuasion, and asking a teacher about environmental issues. Previous studies have reported on the reliability and validity of this scale (Liu & Chen, 2020). For all five scales mentioned above, we used a 5-point Likert-type scoring system with responses ranging from 1 (strongly disagree) to 5 (strongly agree). Data Analyses We used principal component factor analysis to extract major factors of the eight school physical environment variables. We calculated the score of each school’s principal factors based on an SPSS regression method and used these scores in linear regression models. To explore the influence of school environment and green activity on students’ perceptions, we calculated the mean scores of students’ perceptions of each school. We used linear regression models to test the standard estimate and significance of the influence. Furthermore, to understand how students’ perceptions, interactions with natural elements at school, and other environmental activities affected their PEB, we developed a pathway model using AMOS software. We used a bootstrap method with 5,000 resamples and bias-corrected 95% confidence intervals to estimate whether environmental attitudes significantly mediate the relationship between perception of the school environment, interaction with natural elements, and other EE activities and PEB (Preacher & Hayes, 2008). We assessed the model’s Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 65 goodness-of-fit using chi-square, root mean square error of approximation (RMSEA), and comparative fit index (CFI) (Hooper, Coughlan, & Mullen, 2008). Table 1. Constructs and measurement items Construct Item code Measurement items Perception of school environment: Do you agree or disagree with the following statements? (Strongly disagree to Strongly agree) PSE1 I feel there are diverse species of plants in our school PSE2 I think the plant species diversity in school is very high PSE3 I think the vegetation structure in school looks rich and beautiful PSE4 I think the environment of our school is very natural Interaction with nature environment in school: Please indicate your frequency of conducing the following activities in school in your spare time (Never, Seldom, Sometimes, Often, Always) INE1 Observing small animals INE2 Closely observing plants INE3 Playing under tree shadows INE4 Staying quiet and enjoying the tree shadows Other environmental education activity: Please indicate your frequency of participating in the following activities in the past year (Never, Seldom, Sometimes, Often, Always) OEE1 Participating in environmental education activities not organized by school OEE2 Going to the countryside or natural places OEE3 Going to city gardens OEE4 The time spent in housing estate gardens on average every day (zero, <0.5h, 0.5~1h, 1~1.5h, >1.5h) OEE5 Times visiting a zoo last year (zero, once, twice, three times, more than three times) Environmental Attitudes—Preservation: Do you agree or disagree with the following statements? (Strongly disagree to Strongly agree) P1 It upsets me to see the countryside taken over by building sites P2 I enjoy trips to the countryside (woods, meadow) P3 Humankind will die out if we do not live in tune with nature P4 Sitting at the edge of a pond watching dragonflies in flight is enjoyable P5 I save water by taking a shower instead of a bath P6 We must set aside areas to protect endangered species P7 It is interesting to know what kinds of creatures live in ponds or rivers P8 Dirty industrial smoke from chimneys makes me angry Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 66 Environmental Attitudes—Utilization: Do you agree or disagree with the following statements? (Strongly disagree to Strongly agree) U1 We need to clear forests in order to grow crops U2 Our planet has unlimited resources (e.g., potable water, wood, coal, or oil) U3 Nature is always able to restore itself U4 Only plants and animals of economic importance need to be protected U5 Humans have the right to change nature as they see fit U6 People worry too much about pollution U7 Human beings are more important than other creatures U8 We should remove garden weeds to help beautiful flowers grow Pro-environmental behavior: Please indicate your frequency of doing the following activities in the past year (never, seldom, sometimes, often, always) (Liu and Chen, 2020) PEB1 When I played outside, I bought bottled water PEB2 I bought more food than I really needed at a restaurant PEB3 I consciously picked up the garbage on the playground or corridor PEB4 I looked at books about the environment (nature, trees, and animals) PEB5 I collected and recycled used paper PEB6 I consulted my teacher about the environment PEB7 I helped clean the neighborhood in my free time PEB8 I talked with friends about problems related to the environment PEB9 I pointed out unecological behavior to someone PEB10 I paid close attention to environmental issues in the media (newspapers, magazines, and TV) Reliability of Scales We applied descriptive analysis and reliability analysis to the five scales used to measure students’ self-reported variables. The Cronbach’s alpha of PSE is 0.710 (M = 3.89), of the INE is 0.773 (M = 2.75), of the OEE is 0.670 (M = 2.98), and of the PEB is 0.770 (M = 3.02). For the 2-MEV scale, which measured students’ environmental attitudes, the Cronbach's alpha of the preservation subscale was 0.636 (M = 4.20) and of the utilization subscale was 0.682 (M = 2.28). School Environment Among the 20 schools sampled, there was great variation in the schools’ physical green environment: green space area ranged from 1,155 to 19,080 m2, the number Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 67 of plant species ranged between 21 and 135, and the vegetation structure diversity ranged between 4 and 11. Other variables are shown in Table 4 in Appendix B. We found high correlations between GSA, GSA/stu, PD, VSD, and LP; the correlation coefficients ranged between 0.48 and 0.85 (see Table 5 in Appendix C). Our principle component analysis showed two principal factors (eigenvalue > 1). We used the Kaiser−Meyer−Olkin (KMO) measure of sampling adequacy and Bartlett’s test of sphericity (BTS) to verify that the data was amenable to being factor analyzed, and confirmed the suitability of the ensuing factor model with a KMO measure of 0.643 and a significant BTS statistic (approximately χ2(28) = 96.40, p < 0.001). The Principal Component 1 (PC1; “Green Environment”) represented GSA, GSA/stu, PD, VSD, LP, and RF and explained 50.16% of the variance. The Principal Component 2 (PC2; “Indoor Green”) represented Green Corridor Area and Indoor Plants and explained 21.13% of the variance. We calculated the scores of each principal factor for each school based on the regression method, and we used these scores to replace the original school physical environment in the following analysis (Table 2). Table 2. Component matrix of principal component analysis of environmental variables Variable PC1 (Green Environment) PC2 (Indoor Green) GSA 0.92 0.22 GSA/stu 0.799 −0.047 PD 0.887 0.14 VSD 0.79 0.079 LP 0.877 0.236 RF 0.538 0.014 GCA −0.033 0.908 Inplant 0.244 0.856 Proportion explained 50.16 21.125 Cumulative proportion 50.16 71.285 We used regression models to determine whether green environment (PC1) and indoor green (PC2) affected students’ perception of the school environment. The results showed that PC1 had a significantly positive effect on students’ perception of school environment, with a standard estimation β = 0.46 (p = 0.04, R2 = 0.17). As Figure 2 shows, the greener the school environment, the higher students’ perception of the environment. However, PC1 has not had any significant effect on the students’ interaction with natural elements (β = –0.05, p = 0.86, R2 = –0.04). PC2, meanwhile, had no significant effect on either students’ perception of the school environment (β = 0.15, p = 0.53, R2 = –0.03) or interaction with natural elements (β = –0.14, p = 0.56, R2 = –0.04). Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 68 Figure 2. The influence of green environment on students’ perception of the school environment and preservation score (n = 20) Note: every black spot represents one school with school code shown on the right; the error bar represents standard error; the estimation in the equation is nonstandard). School Green Activity Our interviews with teachers provided a list of 223 activities related to environmental protection held in schools over the last two years. Each school had held 2–25 green activities (see Appendix B for more details). The most popular type of activity was talking (26% of total activities), followed by hand-written activities (15%). Cleaning was least popular, only accounting for 6% of the total activities. The proportion of each activity type is shown in Figure 5 in Appendix B. We built three regression models to determine whether the number of green activities at a school affected students’ perception of the school environment and environmental attitudes. The results showed that green activities did not have any significant effect on the perception of the school environment, with β = 0.07 (p = 0.78, R2 = –0.05). Green activities also did not significantly affect the preservation score, with β = 0.03 (p = 0.91, R2 = –0.05). However, these activities significantly affected the utilization score, with β = –0.51 (p = 0.02, R2 = 0.22). In other words, the more green activities, the lower the perception of nature as something to utilize (Figure 3). Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 69 Figure 3. The influence of green activities on students’ utilization value (n = 20) Note: every black spot represents one school with school code shown on the right; the error bar represents standard error; the estimation in the equation is nonstandard) Students’ Environmental Attitudes and Pro-Environmental Behavior We developed a pathway model to understand how perceptions of the school environment and interaction with natural elements at school affect students’ environmental attitudes and PEB. The fit indices for the pathway model are χ2 = 0.193, df = 3, χ2/df = 0.064 (p = 0.979), RMSEA = 0.000, CFI = 1, NNFI = 1, and indicate a good model fit. As Figure 4 shows, perception of the school environment has a direct positive influence on students’ preservation score (β = 0.14, p < .01), which in turn significantly increases PEB (β = 0.22, p < .01). Meanwhile, perception of the school environment has a direct effect on PEB (β = 0.08, p < 0.01), as does interaction with natural elements at (β = 0.37, p < 0.01; Figure 4). Neither perception of school environment nor interaction with natural elements in school have a significant influence on the utilization score. The overall model explains 14% of the variance in preservation, 1% of variance in utilization, and 40% of variance in PEB. Tables 6-10 in Appendix D show the mean, standard deviation, correlations between constructs, standardized regression weights (direct effects), correlations, squared multiple correlations, and standardized regression weights (indirect effects) for the pathway model. Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 70 Figure 4. Pathway model showing the influence of students’ perceptions of school green space, natural activity within school, and EE activity out of school on environmental values and pro-environmental behavior (n = 1,597; number on the pathway represents path coefficient) Discussion This study explores the influence of school green space on students’ environmental attitudes and pro-environmental behaviors. The results show that students’ perception of their school environment is positively correlated with the objective attributes of the school’s green environment. Students’ perception of the school environment and their interaction with natural elements at school have a direct, significant positive effect on their preservation environmental attitude and PEB. Among the four hypotheses we proposed, Hypothesis 1 (a green school environment has a direct and positive influence on students’ PSE), Hypothesis 2 (PSE and INE have a direct positive effect on preservation), and Hypothesis 4 (PSE and INE have a direct positive effect on PEB) have been supported, whereas Hypothesis 3 (PSE and INE have a direct negative effect on utilization) has been rejected. The research thus has indicated that the green spaces in Chinese schools enhance children’s environmental attitudes and pro-environmental behavior. The study indicated that the school environment is perceptible. This result is consistent with some previous studies, which found that students can perceive the quality of a school’s physical environment (Akpinar, 2016; Collado & Corraliza, 2012; Samborski, 2010). Children can learn through the physical environment: it sharpens students’ perception, enhances their sensitivity, and assists them in developing an aesthetic sense (Demirbaş & Demirkan, 2003). This study further demonstrates that children’s perception of the school environment has a positive relation with PEB. In other words, students who attend a school with more green space, plant diversity, and landscape possibility will exhibit more pro-environmental behavior. Previous studies also showed that an Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 71 adult’s perceptions of a natural element’s restorative qualities can predict the individual’s environmental behavior (Hartig, Kaiser, & Bowler, 2001). Although the positive correlation does not necessarily mean a cause-effect relation, this study suggests that school managers and designers could construct the school green environment in ways that positively impact students’ environmental orientation and behavior (Kong, Yaacob, & Ariffin, 2015). Our results also indicated that students’ interaction with natural elements at school was an important factor in cultivating students’ environmental attitudes and PEB. Because school is the place where children spend the majority of their time when not at home, opportunities to experience and learn from nature provided at school can promote students’ affinity with nature (Collado, Evans, Corraliza, & Sorrel, 2015). Therefore, teachers should consider how to promote students’ interaction with the outdoor environment at school. As learning does not only happen in the classroom, teachers should try to encourage students’ free-choice learning on the school ground, providing them with various opportunities to interact with natural elements while at school. Psychological studies show that children have different learning modalities (Smith, 1998). Children of different genders, ages, and personalities have different preferences for how they engage with different environmental characteristics and activities (Barbour, 1999; Bell et al., 2008). Schools must therefore seek to meet the needs of all children; the richer the environmental structure, the stronger its affordances, which can enable more children to enjoy the school green environment. This study found that students’ PSE and INE can explain their attitude toward preservation of the environment, but cannot explain their attitude toward utilization of the environment. The logic behind this might be that students’ preservation attitude originated from their appreciation of the beauty of their school green space, and this appreciation stimulated them to have more interaction with the environment. People tend to protect what they find beautiful or perceive to be good. Regarding students’ attitude toward utilization, PSE and INE cannot explain the variations in it, but the school’s green activities can (with a R2 = 0.22). This “contradictory” result could be attributed to the lack of one variable in the pathway model: students’ self-reported gains from the green activities held in school. Previous research on the pedagogical approach of eco-schools found that they had an impact on students’ attitude toward utilization (Boeve-de Pauw & Van Petegem, 2013b), which is consistent with our results. Eco-schools’ pedagogical approach involves school-based educational projects or activities that are related to the sustainable development of the school environment, which are similar to the green activities in this study. Together with the results of previous studies, this study has surfaced some interesting implications: to improve the preservation dimension of students’ environmental attitude, an appreciation activity such as experiencing nature might be effective. At the same time, to improve the utilization dimension of students’ environmental attitude, more pedagogical activities should be added, such as lectures, hands-on activities, and eco-school-like projects. We can perhaps better understand this from the psychological essence of the two environmental attitudes: Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 72 the preservation attitude is a positive mental activity, often conveyed as “make the good things stay”—that is, because nature is beautiful, we should protect it. On the other hand, utilization of nature is an antidromic mental activity that requires denying primitive human desires—“We should not eat wild animals,” or “we should not use too much of the woods.” To overcome these, more mental strength is needed, so more intensive, pedagogical interventions must be applied. Preservation and utilization are not contrary to each other (Johnson & Manoli, 2008; Liu & Chen, 2020; Wiseman & Bogner, 2003), so we cannot judge environmental education’s success solely on the basis of students having improved preservation attitudes. To some degree, beautiful school environments and educational approaches are equally important for cultivating students’ environmental literacy. This study did not find any strong relationship between green activities and students’ perception of the school environment or interaction with nature at school. This presumably indicates problems with the current green activities at schools: our interviews show that most of the green activities were not based on school grounds, such as lectures or hand-written activities. Instead, the activities were disconnected from the school environment. Previous studies have shown that school grounds and classrooms are powerful tools that influence teaching and learning (Orr, 1997). The physical environment and environmental education should not be two separate identities; teachers and school managers may need more training on how to use the school environment as a teaching tool. Despite many interesting findings in our research, this study also has some limitations. First, we used principal factor analysis to reduce the school physical environment variables and obtain the principal component factor of the “green environment.” This included many interesting variables, such as landscape possibility, plants species diversity, plant structural diversity, etc. However, this method does not indicate the potential influence of any individual attribute on students. Future studies could dive into these individual attributes, and we expect such research will lead to more suggestions about the design and construction of school environments. Second, the reliability of the OEE and 2-MEV scale used in this study has not met the often cited standard (< 0.7) (Nunnally, 1978), but they were very close to 0.7, hence still tolerable; however, we assume that those two scales need further improvement with reference to the reliability coefficient in future studies. Third, the 20 primary schools randomly sampled do not represent the current trends in school environment innovation. Most studies have focused on innovative school design, including participatory (Wake & Eames, 2013) and permaculture design (Holmgren, 2002; Skanavis & Manolas, 2015). Such studies find that these designs may provide a better environment for children (Macnab et al., 2014). Future studies should engage in more detailed comparisons of schools with different design philosophies and explore the roles of different spaces in the school environment. In addition, some schools in our study were more restrictive of students’ activities due to safety considerations, which greatly affects the role of the school environment (Wang, 2013). Future studies could pay more attention to this factor. Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 73 Despite these limitations, our findings indicate that a school’s physical environment does have a positive influence on students’ environmental attitudes and PEB via their perceptions, as does students’ interaction with natural elements at school. These findings highlight the importance of the school physical environment and students’ interactions with the environment. School designers, principals, and teachers should consciously include the school physical environment in their pedagogy to better cultivate students’ environmental literacy. Acknowledgments We greatly appreciate the help and support of the schools, children, and their families. Wanlu Liu earned a master’s degree in environmental education from the Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences. Her research interest is how informal education integrated with formal education enhances students’ environmental attitudes and behaviors. Jin Chen is a professor and the director of Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences, and principal investigator of Ecology and Evolution of Plant Animal Interaction (EEPAI). His research interests are frugivore- plant interactions and environmental education. He focuses on fundamental ecological research and practical activities of links between biological and social problems, and makes flora conservation and education more effective by implementing multidisciplinary methods. References Akoumianaki-Ioannidou, A., Paraskevopoulou, A. T., & Tachou, V. (2016). School grounds as a resource of green space to increase child-plant contact. Urban Forestry & Urban Greening, 20, 375-386. Akpinar, A. (2016). How is high school greenness related to students’ restoration and health? Urban Forestry & Urban Greening, 16, 1-8. Albuquerque, U. P., & Alves, R. R. N. (Eds.). (2016). Introduction to ethnobiology. New York: Springer. Atran, S., Medin, D., & Ross, N. (2004). Evolution and devolution of knowledge: A tale of two biologies. Journal of the Royal Anthropological Institute, 10(2), 395-420. doi:10.1111/j.1467-9655.2004.00195.x Baines, E. & Blatchford, P. (2011). Children’s games and playground activities in school and their role in development. In A. D. Pellegrini (Ed.), The Oxford Handbook of the Development of Play. New York: Oxford University Press. Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 74 Bamberg, S., & Möser, G. (2007). Twenty years after Hines, Hungerford, and Tomera: A new meta-analysis of psycho-social determinants of pro- environmental behaviour. Journal of Environmental Psychology, 27(1), 14-25. Barbour, A. C. (1999). The impact of playground design on the play behaviors of children with differing levels of physical competence. Early Childhood Research Quarterly, 14(1), 75-98. Bell, J. F., Wilson, J. S., & Liu, G. C. (2008). Neighborhood greenness and 2-year changes in body mass index of children and youth. American Journal of Preventive Medicine, 35(6), 547-553. doi:10.1016/j.amepre.2008.07.006 Blair, D. (2009). The child in the garden: An evaluative review of the benefits of school gardening. The Journal of Environmental Education, 40(2), 15-38. doi:10.3200/JOEE.40.2.15-38 Boeve-de Pauw, J., & Van Petegem, P. (2013a). A cross-cultural study of environmental values and their effect on the environmental behavior of children. Environment and Behavior, 45(5), 551-583. doi:10.1177/0013916511429819 Boeve-de Pauw, J., & Van Petegem, P. (2013b). The effect of eco-schools on children’s environmental values and behaviour. Journal of Biological Education, 47(2), 96-103. doi:10.1080/00219266.2013.764342 Bogner, F. X., & Wiseman, M. (2006). Adolescents’ attitudes towards nature and environment: Quantifying the 2-MEV model. Environmentalist, 26(4), 247- 254. doi:10.1007/s10669-006-8660-9 Boldemann, C., Blennow, M., Dal, H., Mårtensson, F., Raustorp, A., Yuen, K., & Wester, U. (2006). Impact of preschool environment upon children's physical activity and sun exposure. Preventive Medicine, 42(4), 301-308. doi:10.1016/j.ypmed.2005.12.006 Braun, T., Cottrell, R., & Dierkes, P. (2018). Fostering changes in attitude, knowledge and behavior: Demographic variation in environmental education effects. Environmental Education Research, 24(6), 899-920. doi:10.1080/13504622.2017.1343279 Burns, H. (2010). The good school playground guide: Developing school playgrounds to support the curriculum and nurture happy, healthy children. Retrieved from https://www.ltl.org.uk/resources/results.php?id=712 Carrus, G., Scopelliti, M., Lafortezza, R., Colangelo, G., Ferrini, F., Salbitano, F., ... & Sanesi, G. (2015). Go greener, feel better? The positive effects of biodiversity on the well-being of individuals visiting urban and peri-urban green areas. Landscape and Urban Planning, 134, 221-228. https://www.ltl.org.uk/resources/results.php?id=712 Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 75 Castéra, J., Clément, P., Munoz, F., & Bogner, F. X. (2018). How teachers' attitudes on GMO relate to their environmental values. Journal of Environmental Psychology, 57, 1-9. doi:10.1016/j.jenvp.2018.04.002 Chawla, L. (2015). Benefits of nature contact for children. Journal of Planning Literature, 30(4), 1-20. doi: 10.1177/0885412215595441 Chawla, L., & Cushing, D. F. (2007). Education for strategic environmental behavior. Environmental Education Research, 13(4), 437-452. doi:10.1080/13504620701581539 Chawla, L., Keena, K., Pevec, I., & Stanley, E. (2014). Green schoolyards as havens from stress and resources for resilience in childhood and adolescence. Health & Place, 28, 1-13. China Green Building Council (2013). Green campus evaluation criteria CSUS/GBC 04-2013. Retrieved from http://www.jianbiaoku.com/webarbs/book/50165/921502.shtml Chongqing Statistics Bureau (2018). Chongqing statistical yearbook 2018. China Statistics Press. Retrieved from http://tjj.cq.gov.cn/zwgk_233/tjnj/2018/indexch.htm Collado, S., & Corraliza, J. A. (2012). Perceived restoration and environmental orientation in a sample of Spanish children. Procedia-Social and Behavioral Sciences, 38, 264-274. doi:10.1016/j.sbspro.2012.03.348 Collado, S., Evans, G. W., Corraliza, J. A., & Sorrel, M. A. (2015). The role played by age on children's pro-ecological behaviors: An exploratory analysis. Journal of Environmental Psychology, 44, 85-94. doi:10.1016/j.jenvp.2015.09.006 Collado, S., Staats, H., & Corraliza, J. A. (2013). Experiencing nature in children's summer camps: Affective, cognitive and behavioural consequences. Journal of Environmental Psychology, 33, 37-44. doi:10.1016/j.jenvp.2012.08.002 Dadvand, P., Nieuwenhuijsen, M. J., Esnaola, M., Forns, J., Basagaña, X., Alvarez- Pedrerol, M., . . . Su, J. (2015). Green spaces and cognitive development in primary schoolchildren. Proceedings of the National Academy of Sciences, 112(26), 7937-7942. doi:10.1073/pnas.1503402112 Demirbaş, O. O., & Demirkan, H. (2003). Focus on architectural design process through learning styles. Design Studies, 24(5), 437-456. doi:10.1016/S0142-694X(03)00013-9 Dyment, J. E. (2005). Green school grounds as sites for outdoor learning: Barriers and opportunities. International Research in Geographical & Environmental Education, 14(1), 28-45. doi:10.1080/09500790508668328 http://www.jianbiaoku.com/webarbs/book/50165/921502.shtml http://tjj.cq.gov.cn/zwgk_233/tjnj/2018/indexch.htm Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 76 Evans, J. (2001). In search of peaceful playgrounds. Education Research and Perspectives, 28(1), 45-56. Feng, J., Xiao, X., Zhou, J., & Wu, X. (2018). Nature education industry survey report in 2018. Retrieved from https://new.qq.com/omn/20190310/20190310A0WA6J.html Fjørtoft, I. (2004). Landscape as playscape: The effects of natural environments on children's play and motor development. Children, Youth and Environments, 14(2), 21-44. Gärling, T., & Golledge, R. G. (1989). Environmental perception and cognition. In E. H. Zube & G. T. Moore (Eds.), Advances in environment, behavior, and design (pp. 203-236). New York: Springer. Gifford, R., & Sussman, R. (2012). Environmental attitudes. In S. D. Clayton (Ed.), The Oxford handbook of environmental and conservation psychology. Oxford: Oxford University Press. doi:10.1093/oxfordhb/9780199733026.013.0004 Gill, T. (2014). The benefits of children's engagement with nature: A systematic literature review. Children, Youth and Environments, 24(2), 10-34. doi:10.7721/chilyoutenvi.24.2.0010 Giusti, M., Barthel, S., & Marcus, L. (2014). ‘Nature, routines and affinity with the biosphere.’ Children, Youth and Environments, 24(3): 16–42. Guo, H., & Yang, Z. (2008). Beijing shi xiao xuesheng shijian anpai yu shengcun zhuangtai yanjiu [Study on time arrangement and living status of primary school students in Beijing]. Educational Scientific Research, 11, 31-32. Hart, R. (1979). Children’s experience of place. New York: Irvington. Hartig, T., Kaiser, F. G., & Bowler, P. A. (2001). Psychological restoration in nature as a positive motivation for ecological behavior. Environment and Behavior, 33(4), 590-607. doi:10.1177/00139160121973142 Hinds, J., & Sparks, P. (2008). Engaging with the natural environment: The role of affective connection and identity. Journal of Environmental Psychology, 28(2), 109-120. doi:10.1016/j.jenvp.2007.11.001 Holmgren, D. (2002). Permaculture: Principles & pathways beyond sustainability. Holmgren Design Services. Hooper, D., Coughlan, J., & Mullen, M. R. (2008). Structural equation modelling: Guidelines for determining model fit. Electronic Journal of Business Research Methods, 6(1), 53-60. https://new.qq.com/omn/20190310/20190310A0WA6J.html Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 77 Huh, S. Y., & Gordon, C. M. (2008). Vitamin D deficiency in children and adolescents: Epidemiology, impact and treatment. Reviews in Endocrine and Metabolic Disorders, 9(2), 161-170. doi:10.1007/s11154-007-9072-y Ittelson, W. H. (1976). Environment perception and contemporary perceptual theory. In H. M. Proshansky, W. H. Ittelson, & L. G. Rivlin (Eds.), Environmental psychology: People and their physical settings, 2nd ed. (pp. 141-154). Holt, Rinehart and Winston. Jansson, M., Gunnarsson, A., Mårtensson, F., & Andersson, S. (2014). Children's perspectives on vegetation establishment: Implications for school ground greening. Urban Forestry & Urban Greening, 13(1), 166-174. Johnson, B., & Činčera, J. (2015). Examining the relationship between environmental attitudes and behaviour in education programmes. Sociální Studia, 3, 97-111. doi:10.5817/SOC2015-3-97 Johnson, B., & Manoli, C. C. (2008). Using Bogner and Wiseman’s Model of Ecological Values to measure the impact of an earth education programme on children’s environmental perceptions. Environmental Education Research, 14(2), 115-127. doi:10.1080/13504620801951673 Johnson, B., & Manoli, C. C. (2010). The 2-MEV scale in the United States: a measure of children's environmental attitudes based on the theory of ecological attitude. The Journal of Environmental Education, 42(2), 84-97. doi:10.1080/00958964.2010.503716 Kaiser, F. G., Hartig, T., Brügger, A., & Duvier, C. (2013). Environmental protection and nature as distinct attitudinal objects: An application of the Campbell paradigm. Environment and Behavior, 45(3), 369-398. doi:10.1177/0013916511422444 Karimian, Z., Samiei, L., & Kazemi, F. (2017). Assessment of user preferences of campus green space at Ferdowsi University of Mashhad-Iran. Iranian Journal of Horticultural Science, 48, 1-11. doi:10.22059/ijhs.2017.63642 Kong, S. Y., Yaacob, N. M., & Ariffin, A. R. M. (2015). Physical environment as a 3- D textbook: Design and development of a prototype. Asia Pacific Journal of Education, 35(2), 241-258. doi:10.1080/02188791.2014.906384 Kweon, B.-S., Ellis, C. D., Lee, J., & Jacobs, K. (2017). The link between school environments and student academic performance. Urban Forestry & Urban Greening, 23, 35-43. doi:10.1016/j.ufug.2017.02.002 Kyttä, M. (2002). Affordances of children's environments in the context of cities, small towns, suburbs and rural villages in Finland and Belarus. Journal of Environmental Psychology, 22(1-2), 109-123. Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 78 Li, D., & Chen, J. (2015). Significant life experiences on the formation of environmental action among Chinese college students. Environmental Education Research, 21(4), 612-630. doi:10.1080/13504622.2014.927830 Li, D., Chiang, Y. C., Sang, H., & Sullivan, W. C. (2019). Beyond the school grounds: Links between density of tree cover in school surroundings and high school academic performance. Urban Forestry & Urban Greening, 38, 42-53. Li, X., Ni, G., & Dewancker, B. (2019). Improving the attractiveness and accessibility of campus green space for developing a sustainable university environment. Environmental Science and Pollution Research, 26(32), 33399- 33415. Lindemann-Matthies, P., Junge, X., & Matthies, D. (2010). The influence of plant diversity on people’s perception and aesthetic appreciation of grassland vegetation. Biological Conservation, 143(1), 195-202. https://doi.org/10.1016/j.biocon.2009.10.003 Lindemann-Matthies, P., & Köhler, K. (2019). Naturalized versus traditional school grounds: Which elements do students prefer and why? Urban Forestry & Urban Greening, 46. https://doi.org/10.1016/j.ufug.2019.126475 Liu, W., & Chen, J. (2020). Modified Two Major Environmental Values scale for measuring Chinese children’s environmental attitudes. Environmental Education Research, 26(1), 130-147. http://dx.doi.org/10.1080/13504622.2019.1697431 Louv, R. (2008). Last child in the woods: Saving our children from nature-deficit disorder. Chapel Hill, NC: Algonquin Books. Lysgaard, J. G., Larsen, N., & Læssøe, J. (2015). Green flag eco-schools and the challenge of moving forward. In V. Thoresen, D. Doyle, J. Klein, & R. Didham (Eds.), Responsible Living (pp. 135-150). Cham: Springer. Määttä, S., Gubbels, J., Ray, C., Koivusilta, L., Nislin, M., & Sajaniemi, N., et al. (2019). Children's physical activity and the preschool physical environment: the moderating role of gender. Early Childhood Research Quarterly, 47, 39- 48. Macnab, A. J., Stewart, D., Gagnon, F. A., Beery, M., Adatia, R., Segantin, O., & Skaer, C.-F. (2014). School food gardens: Fertile ground for education. Health Education, 114(4), 281-292. doi:10.1108/HE-05-2013-0019 Mahdjoubi, L., & Akplotsyi, R. (2012). The impact of sensory learning modalities on children’s sensitivity to sensory cues in the perception of their school environment. Journal of Environmental Psychology, 32(3), 208-215. doi:10.1016/j.jenvp.2012.02.002 https://doi.org/10.1016/j.biocon.2009.10.003 https://doi.org/10.1016/j.ufug.2019.126475 http://dx.doi.org/10.1080/13504622.2019.1697431 Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 79 Maller, C., & Townsend, M. (2006). Children's mental health and wellbeing and hands-on contact with nature. International Journal of Learning, 12(4), 359- 372. Malone, K., & Tranter, P. J. (2003). School grounds as sites for learning: Making the most of environmental opportunities. Environmental Education Research, 9(3), 283-303. doi:10.1080/13504620303459 Meiboudi, H., Lahijanian, A., Shobeiri, S. M., Jozi, S. A., & Azizinezhad, R. (2017). Development and validation of sustainability criteria of administrative green schools in Iran. Journal of Environmental Management, 197, 605-609. doi:10.1016/j.jenvman.2017.04.045 Milfont, T. L., & Duckitt, J. (2004). The structure of environmental attitudes: A first- and second-order confirmatory factor analysis. Journal of Environmental Psychology, 24(3), 289-303. doi:10.1016/j.jenvp.2004.09.001 Milfont, T. L., & Duckitt, J. (2010). The environmental attitudes inventory: A valid and reliable measure to assess the structure of environmental attitudes. Journal of Environmental Psychology, 30(1), 80-94. doi:10.1016/j.jenvp.2009.09.001 Miller, J. R. (2005). Biodiversity conservation and the extinction of experience. Trends in Ecology & Evolution, 20(8), 430-434. doi:10.1016/j.tree.2005.05.013 Moore, R. C. (1996). Outdoor settings for playing and learning: Designing school grounds to meet the needs of the whole child and whole curriculum. North American Montessori Teacher's Association Journal, 21(3), 97-120. Müderrisoglu, H., & Altanlar, A. (2011). Attitudes and behaviors of undergraduate students toward environmental issues. International Journal of Environmental Science & Technology, 8(1), 159-168. Nabhan, G. P., & Trimble, S. (1994). The geography of childhood: Why children need wild places. Boston: Beacon Press. Nunnally, J. C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill. Nussbaum, M. C. (2011). Creating capabilities. Cambridge, MA: Harvard University Press. Orr, D. W. (1997). Architecture as pedagogy II. Conservation Biology, 11(3), 597- 600. Pergams, O. R., & Zaradic, P. A. (2008). Evidence for a fundamental and pervasive shift away from nature-based recreation. Proceedings of the National Academy of Sciences, 105(7), 2295-2300. doi:10.1073/pnas.0709893105 Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 80 Pooley, J. A., & o’Connor, M. (2000). Environmental education and attitudes: Emotions and beliefs are what is needed. Environment and Behavior, 32(5), 711-723. doi:10.1177/0013916500325007 Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40(3), 879-891. doi:10.3758/BRM.40.3.879 Samborski, S. (2010). Barren or biodiverse schoolgrounds: Their effects on children. Children, Youth and Environments, 20(2), 67-115. https://www.jstor.org/stable/10.7721/chilyoutenvi.20.2.0067?seq=1 Sando, O. J., & Sandseter, E. B. H. (2020). Affordances for physical activity and well-being in the ECEC outdoor environment. Journal of Environmental Psychology, 69. https://doi.org/10.1016/j.jenvp.2020.101430 Scholl, K. G., & Gulwadi, G. B. (2018). College campus landscapes within a learning ecosystem. Planning for Higher Education, 46(2), 50-64. Sivarajah, S., Smith, S. M., & Thomas, S. C. (2018). Tree cover and species composition effects on academic performance of primary school students. Plos One, 13(2), e0193254. https://doi:10.1371/journal.pone.0193254 Skanavis, C., & Manolas, E. (2015). School gardens and ecovillages: Innovative civic ecology educational approaches at schools and universities. In W. L. Filho (Ed.), Transformative approaches to sustainable development at universities: Working across disciplines (pp. 559-570). Cham: Springer. Smith, A. (1998). Accelerated learning in practice. London: A&C Black. Soga, M., & Gaston, K. J. (2016). Extinction of experience: The loss of human– nature interactions. Frontiers in Ecology and the Environment, 14(2), 94- 101. doi:10.1002/fee.1225 Stokols, D. (1978). Environmental psychology. Annual Review of Psychology, 29(1), 253-295. Tan, H., Chen, S., Shi, Q., & Wang, L. (2014). Development of green campus in China. Journal of Cleaner Production, 64, 646-653. doi:10.1016/j.jclepro.2013.10.019 Taylor, A. (1993). The learning environment as a three-dimensional textbook. Children's Environments, 10(2), 170-179. https://www.jstor.org/stable/41514891 Tian, Y., & Wang, C. (2016). Environmental education in China: Development, difficulties and recommendations. Journal of Social Science Studies, 3(1), 31- 43. https://www.jstor.org/stable/10.7721/chilyoutenvi.20.2.0067?seq=1 https://doi.org/10.1016/j.jenvp.2020.101430 https://doi.org/10.1371/journal.pone.0193254 Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 81 Titman, W. (1994). Special places, special people: The hidden curriculum of school grounds. Retrieved from https://eric.ed.gov/?id=ED430384 Tucker, R., & Izadpanahi, P. (2017). Live green, think green: Sustainable school architecture and children’s environmental attitudes and behaviors. Journal of Environmental Psychology, 51, 209-216. doi:10.1016/j.jenvp.2017.04.003 Ulrich, R. S., Zimring, C., Zhu, X., DuBose, J., Seo, H.-B., Choi, Y.-S., . . . Joseph, A. (2008). A review of the research literature on evidence-based healthcare design. HERD: Health Environments Research & Design Journal, 1(3), 61- 125. doi:10.1177/193758670800100306 van den Berg, A. E., Wesselius, J. E., Maas, J., & Tanja-Dijkstra, K. (2017). Green walls for a restorative classroom environment: A controlled evaluation study. Environment and Behavior, 49(7), 791-813. Voigt, A., Kabisch, N., Wurster, D., Haase, D., & Breuste, J. (2014). Structural diversity: A multi-dimensional approach to assess recreational services in urban parks. Ambio, 43(4), 480-491. doi:10.1007/s13280-014-0508-9 Wake, S. J., & Eames, C. (2013). Developing an “ecology of learning” within a school sustainability co-design project with children in New Zealand. Local Environment, 18(3), 305-322. doi:10.1080/13549839.2012.748723 Wang, T. (2013). Chinese school principals’ behavioral intentions in relation to green school practices. (Doctoral dissertation). Auburn University, Auburn, Alabama. Wee, C. (2017). Assessing the perception of campus green space and stress levels among students at Michigan State University. Michigan State University. Environmental Design. Whitburn, J., Linklater, W., & Abrahamse, W. (2020). Meta‐analysis of human connection to nature and proenvironmental behavior. Conservation Biology, 34(1), 180-193. doi:10.1016/j.jenvp.2010.01.003 Wiseman, M., & Bogner, F. X. (2003). A higher-order model of ecological values and its relationship to personality. Personality and Individual Differences, 34(5), 783-794. doi:10.1016/S0191-8869(02)00071-5 Wu, Z. (2002). Green schools in China. The Journal of Environmental Education, 34(1), 21-25. doi:10.1080/00958960209603478 Zamani, Z., & Moore, R. (2013). The cognitive play behavior affordances of natural and manufactured elements within outdoor preschool settings. Landscape Research, 1, 268-278. Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 82 Zhang, W., Zhao, J., & Chen, J. (2019). Nature club programs promote adolescents’ conservation behavior: A case study in China’s biodiversity hotspot. The Journal of Environmental Education, 50(3), 192-207. doi:10.1080/00958964.2019.1604480 Zhao, D.-X., He, B.-J., & Meng, F.-Q. (2015). The green school project: A means of speeding up sustainable development? Geoforum, 65, 310-313. doi:10.1016/j.geoforum.2015.08.012 Zhao, J., Wu, J., & Wang, H. (2019). Characteristics of urban streets in relation to perceived restorativeness. Journal of Exposure Science & Environmental Epidemiology, 30(2), 309-319. Zhiyan, J., Hongying, Z., & Xuhong, S. (2004). An overview of "green school" development in China in 2001. Chinese Education & Society, 37(3), 49-54. doi:10.1080/10611932.2004.11031632 Zube, E. H. (1999). Environmental perception. In C. W. Finkl (Ed.) Encyclopedia of Earth science (pp. 214–216). New York: Springer. Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 83 Appendix A. Table 3. Basic information about selected schools School Code District Funding Year* Number of Students Number of Classes School area (m2) 1 Shapingba 1966 570 12 4,071 2 Jiangbei 2008 1200 36 23,847 3 Shapingba 1936 (2010) 1150 24 12,446 4 Yubei 2005 1665 37 32,966 5 Shapingba 1970 (2011) 1170 23 18,182 6 Yuzhong 1957 1158 47 7,072 7 Jiulongpo 2008 1600 27 14,540 8 Jiulongpo 1998 1552 32 13,120 9 Jiulongpo 1988 1398 27 10,755 10 Dadukou 1963 (2013) 2098 42 11,294 11 Jiangbei 2002 721 17 5,503 12 Yubei 1993 1442 26 13,306 13 Banan 2015 1055 25 11,959 14 Nanan 2006 3100 74 30,091 15 Banan 1941 (2007) 934 20 5,035 16 Yubei 2013 1198 33 21,716 17 Nanan 1999 572 23 3,597 18 Yuzhong 1911 1090 26 7,055 19 Yubei 1922 (2010) 1500 30 19,308 20 Beibei 2007 480 12 10,589 *Renovation year shown in parentheses Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 84 Appendix B. Table 4. Environmental variables and green activities (GA) held in the last two years at each school School code GSA (m2) GSA/st u (m2) GR (%) PD VSD LP GCA (m2) RF Inplant GA 1 1,155 2.03 28.4 21 7 1 0 1 3.5 5 2 11,192 9.33 46.9 74 10 4 0 2 6.5 11 3 4,129 3.59 33.2 76 6 6 0 2 2.5 10 4 17,677 10.62 53.6 135 9 7 230 8 17.5 8 5 6,726 5.75 37.0 88 9 6 0 8 3 14 6 2,775 2.40 39.2 50 5 0 35 1 0 10 7 4,375 2.73 30.1 43 9 3 15 4 0 7 8 4,714 3.04 35.9 61 7 4 50 4 22.5 25 9 2,811 2.01 26.1 80 4 2 55 3 4.5 2 10 2,623 1.25 23.2 56 7 4 156 7 2.5 12 11 2,211 3.07 40.2 52 7 2 102 2 12.5 6 12 3,480 2.41 26.2 55 7 2 15 3 4 9 13 3,165 3.00 26.5 47 7 3 473 0 24 4 14 19,080 6.15 63.4 113 11 10 203 3 17.5 23 15 1,462 1.57 29.0 41 6 1 15 0 1 4 16 6,760 5.64 31.1 86 7 4 64 3 6 9 17 1,855 3.24 51.6 67 8 3 105 4 8 20 18 2,450 2.25 34.7 47 6 3 313 5 1 25 19 8,246 5.50 42.7 92 9 6 9 4 2.5 9 20 4,490 9.35 42.4 59 7 3 0 0 0.5 10 Figure 5. Proportion of each type of green activity held in schools Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 85 Appendix C. Table 5. Correlation coefficients of environmental variables (the upper triangle represents the coefficient of significance) GSA GSA/stu PD VSD LP RF GCA Inplant GSA 1 <0.001 <0.001 <0.001 <0.001 0.06 0.25 0.04 GSA/stu 0.75 1 0.001 0.003 0.007 0.23 0.43 0.21 PD 0.85 0.68 1 0.015 <0.001 0.01 0.35 0.08 VSD 0.73 0.59 0.48 1 <0.001 0.07 0.45 0.11 LP 0.83 0.55 0.81 0.70 1 0.01 0.22 0.05 RF 0.35 0.17 0.52 0.34 0.49 1 0.34 0.431 GCA 0.16 -0.04 0.10 0.03 0.18 0.10 1 0.003 Inplant 0.41 0.19 0.33 0.28 0.37 0.04 0.59 1 Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 86 Appendix D. Table 6. Means, standard deviation, and correlations of constructs in the pathway model Construct Mean SD PSE INE OEE PRE UT PSE 3.89 0.73 1 INE 2.75 0.93 0.26** 1 OEE 2.98 0.84 0.29** 0.40** 1 PRE 4.23 0.55 0.23** 0.34** 0.22** 1 UT 2.28 0.80 0.02 −0.01 0.09** −0.18** 1 PEB 3.02 0.66 0.28** 0.53** 0.39** 0.43** −0.16** ** means <0.01 Table 7. Standardized regression weights (direct effects) for the pathway model, with 95% confidence intervals (CI) and p-values Independent variable Dependent variable Estimate Lower 95% CI Upper 95% CI p-value PRE <--- PSE 0.142 0.092 0.193 0.004 PRE <--- INE 0.272 0.228 0.319 0.004 UT <--- INE −0.053 −0.112 0.003 0.075 UT <--- OUT 0.112 0.052 0.169 0.004 UT <--- PSE −0.003 −0.069 0.05 0.814 PRE <--- OUT 0.07 0.014 0.119 0.013 PEB <--- PSE 0.081 0.032 0.123 0.004 PEB <--- INE 0.368 0.336 0.399 0.004 PEB <--- OUT 0.179 0.136 0.226 0.004 PEB <--- PRE 0.217 0.199 0.234 0.004 PEB <--- UT −0.135 −0.179 −0.095 0.004 Table 8. Correlations for the pathway model, with 95% confidence intervals (CI), and p-values Parameter Estimate Lower 95% CI Upper 95% CI p-value PSE <--> INE 0.259 0.221 0.296 0.004 INE <--> OUT 0.403 0.361 0.445 0.004 PSE <--> OUT 0.288 0.247 0.328 0.004 e1 <--> e2 −0.196 −0.245 −0.146 0.004 Green Spaces in Chinese Schools Enhance Children’s Environmental Attitudes… 87 Table 9. Squared multiple correlations for the pathway model, with 95% confidence intervals (CI) and p-values Parameter Estimate Lower 95% CI Upper 95% CI p-value UT 0.011 0.003 0.024 0.004 PRE 0.14 0.113 0.173 0.004 PEB 0.401 0.365 0.44 0.004 Table 10. Standardized regression weights (indirect effects) for the pathway model, with 95% confidence intervals (CI) and p-values Independent variable Dependent variable Mediator Standardized regression weight Lower 95% CI Upper 95% CI p-value OUT PEB PRE 0 -0.017 0.015 0.978 UT INE PEB PRE 0.066 0.052 0.081 0.004 UT PSE PEB PRE 0.031 0.017 0.046 0.004 UT