Environmental Risk Perception, Risk Culture, and Pro-Environmental Behavior

International Journal of Environmental Research and Public Health Article Environmental Risk Perception, Risk Culture, and Pro-Environmental Behavior Jingjing Zeng 1 , Meiquan Jiang 1 and Meng Yuan 2, * 1 Zhongnan University of Economics and Law, Wuhan 430073, China; jjzeng@zuel.edu.cn (J.Z.); jiangmeiquan 3@gmail.com (M.J.) 2 Department of Political Science, Northern Illinois University, DeKalb, IL 60115, USA * Correspondence: Z 1710495@students.niu.edu Received: 21 January 2020; Accepted: 3 March 2020; Published: 7 March 2020 Abstract: Mixed evidence exists regarding the relationship between environmental risk perception and pro-environmental behavior. This study uses an existing online survey conducted by the Center of Ecological Civilization (CEC) of China University of Geosciences from December 2015 to March 2016 and examines how cultural bias influences environmental risk perception and behavior. We found that an individual’s pro-environmental behavior is not only influenced by environmental risk perception, but also by his or her cultural worldviews. Built on culture theory (CT), our empirical results suggest that young Chinese people are more located in “high-group” culture, where egalitarian culture and hierarchical culture dominate. The higher scores of hierarchical and egalitarian cultures of Chinese youth, the more likely they are to protect the environment. Moreover, the relationship between cultural worldviews and pro-environmental behaviors are mediated by perceived environmental risks Keywords: environmental risk perception; cultural theory of risk; pro-environmental behavior 1. Introduction Environmental problems have become a serious social risk that need to be urgently solved all over the world. The public’s voluntary pro-environmental behavior, including that of youths, is critical to achieve environmental sustainability [ 1 – 6 ]. Moreover, because college students will be the leaders of society and decision makers of public institutions, understanding their environmental concerns and behavior has great value for policymakers who want to improve environmental policy compliance and governance [ 7 ]. Environmental risk perception stimulates people’s sense of urgency and responsibility to protect the environment and encourages more environmental protection behaviors [ 1 , 2 , 8 ]. However, other scholars have found that there is a discrepancy between environmental risk perception and environmental protection behaviors [ 9 , 10 ]. Individuals who possess the same level of perceived environmental risk can have inconsistent behaviors in environmental protection [ 11 , 12 ]. In addition, existing research mainly built on the rational choice model (RCM) or relied on individual cost-benefit analysis [ 12 , 13 ]. Like other cultural scholars, this paper tries to look to what RCM cannot explain by using the culture theory [ 14 – 16 ]. According to culture theory (CT), people have di ff erent preferences of how the society should be organized, which then a ff ects how they evaluate and respond to risk [ 17 ]. Existing studies have confirmed the relationship between cultural worldviews and risk perception suggested by CT [ 18 – 21 ]. Other scholars further found that risk perception, as a mediator, can influence the relationship between culture and pro-environmental behavior [ 11 , 22 ]. However, existing studies did not answer whether culture can influence pro-environmental behavior directly. This study fills this gap by investigating how Int. J. Environ. Res. Public Health 2020 , 17 , 1750; doi:10.3390 / ijerph 17051750 www.mdpi.com / journal / ijerph

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 2 of 18 culture influences pro-environmental behavior directly. This study also tries to test the generalizability of CT by applying CT to non-Western countries This paper uses an existing online survey conducted by the Center for Ecological Civilization of China University of Geosciences (CEC) from December 2015 to March 2016. Our sample includes 8084 students from 152 universities in 30 provinces across China, who were surveyed on the Internet This research makes contributions in four aspects. First, this large-scale data allows us to map the distribution of environmental risk perception and pro-environmental behavior among young people in China. Second, this paper is the first research testing the influence of individuals’ cultural worldviews on pro-environmental behavior in China. More specifically, we test the e ff ects of culture worldviews, environmental risk perception and other control variables, including knowledge and demographics, on pro-environmental behavior. Third, this paper analyzes how the relationship between risk culture and pro-environmental behavior is mediated by an individual’s cultural bias. Fourth, Douglas (1982) used the “grid-group” framework to analyze how the relations between individuals and society determine people’s perception of risk. CT was then introduced by Wildavsky to study politics in the United States and was applied to many western countries [ 23 ]. However, there is only a few pieces of research on China [ 21 , 22 ]. This study will contribute to improve CT’s generalizability by applying CT to Chinese context and providing some implications for measuring culture based on existing research The paper starts with a review of CT and its application in China. Based on the theory and existing research on the relationships among cultures, environmental risk perception, and pro-environmental behavior, we suggest three sets of research hypotheses in Section 2 . Section 3 introduces data, variables, and methods used to test our hypotheses. Section 4 describes the result of our data testing and examines its robustness. Section 5 is the conclusion of the paper 2. Theoretical Background and Hypotheses 2.1. Cultural Theory of Risk and Di ff erences of Chinese Youth Culture Types “Grid-group” cultural theory of risk (Figure 1 ) was proposed by Douglas (1970), a religious anthropologist. Douglas’s (1970) “grid-group” framework helps us to understand how individuals’ perception and behavior are influenced by four di ff erent social types with di ff erent core values Influenced by “grid-group” framework, subsequent researchers further revised and clarified four types of culture. For example, Dake (1992) improved Douglas’s (1970) model and combined the dimensions of grid and group to develop four types of risk culture (Figure 2 ): fatalism, hierarchy, individualism, and egalitarianism Int. J. Environ. Res. Public Health 2020 , 17 , x 2 of 18 investigating how culture influences pro-environmental behavior directly. This study also tries to test the generalizability of CT by applying CT to non-Western countries. This paper uses an existing online survey conducted by the Center for Ecological Civilization of China University of Geosciences (CEC) from December 2015 to March 2016. Our sample includes 8084 students from 152 universities in 30 provinces across China, who were surveyed on the Internet. This research makes contributions in four aspects. First, this large-scale data allows us to map the distribution of environmental risk perception and pro-environmental behavior among young people in China. Second, this paper is the first research testing the influence of individuals’ cultural worldviews on pro-environmental behavior in China. More specifically, we test the effects of culture worldviews, environmental risk perception and other control variables, including knowledge and demographics, on pro-environmental behavior. Third, this paper analyzes how the relationship between risk culture and pro-environmental behavior is mediated by an individual’s cultural bias. Fourth , Douglas (1982) used the “grid-group” framework to analyze how the relations between individuals and society determine people’s perception of risk. CT was then introduced by Wildavsky to study politics in the United States and was applied to many western countries [23]. However, there is only a few pieces of research on China [21,22]. This study will contribute to improve CT’s generalizability by applying CT to Chinese context and providing some implications for measuring culture based on existing research. The paper starts with a review of CT and its application in China. Based on the theory and existing research on the relationships among cultures, environmental risk perception, and proenvironmental behavior, we suggest three sets of research hypotheses in Section 2. Section 3 introduces data, variables, and methods used to test our hypotheses. Section 4 describes the result of our data testing and examines its robustness. Section 5 is the conclusion of the paper. 2. Theoretical Background and Hypotheses 2.1. Cultural Theory of Risk and Differences of Chinese Youth Culture Types “Grid-group” cultural theory of risk (Figure 1) was proposed by Douglas (1970), a religious anthropologist. Douglas’s (1970) “grid-group” framework helps us to understand how individuals’ perception and behavior are influenced by four different social types with different core values. Influenced by “grid-group” framework, subsequent researchers further revised and clarified four types of culture. For example, Dake (1992) improved Douglas’s (1970) model and combined the dimensions of grid and group to develop four types of risk culture (Figure 2): fatalism, hierarchy, individualism, and egalitarianism. Figure 1. Douglas’ (1970) “grid-group” social type quadrant. Figure 1. Douglas’ (1970) “grid-group” social type quadrant.

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 3 of 18 Int. J. Environ. Res. Public Health 2020 , 17 , x 3 of 18 Figure 2. Dake’s (1992) “grid-group” risk culture type quadrant. In culture theory, “grid” dimension refers to rules that connect an individual with another on an ego-centered basis [24]. Grid is the preference of the social structures that rely on roleor classbased rules (high grid) versus those emphasizing equal opportunity and status among individuals (low grid) [22,25]. “Group” dimension refers to the extent that one person is bound by a group. Group is reflected in the preference of social structures that encourage cooperation, collectivism, and collective interests (high group) versus those that support competition and personal interests (low group) [24,25]. However, unlike other psychometric scales, cultural theory suggests that grid and group dimensions vary separately and independently on their own [26–28]. In Chinese social and cultural systems, the collective thinking makes Chinese people emphasize collective interests over personal interests and connect with each other closely. In addition, China is a country in which Han ethnicity (Han Zu) accounts for a majority of the overall Chinese population. Studies have shown that countries with a single ethnicity tend to prefer collectivism culture [25]. In the political context, China’s one-party system and centralized government can increase people’s preference for bureaucracy and social system that emphasizes role or class stratification. In fact, not much literature has analyzed and compared the cultural characteristics of different countries in terms of grid and group dimension. As an exception, Chai et al. used the World Values Survey and found that the culture of China is placed in the high group and high grid quadrant: hierarchy [29]. Our research population is Chinese youth with college education or above. Some existing research found that the Chinese younger generation tends to be more individualistic than the older generation [30]. Others have suggested that younger generations are more likely to participate in or support social environmental movements that are consistent with egalitarian values [25]. Therefore, we expect that Chinese young people are more likely to be located in the high-group quadrant. Hypothesis 1. Compared with other culture types, the culture of Chinese youth is more likely to be Hierarchy and Egalitarianism. 2.2. Cultural Worldviews and Pro-Environmental Behavior According to the CT, people have different preferences of how the society should be organized, which influences how they evaluate and respond to risk [17]. Different from the traditional rational model which assumes that people rationalize their choices based upon cost benefit analysis, culture theory assumes that people think and act in certain ways that are consistent with their culture [15,16,25,31]. In other words, individuals will view the issue that opposes their preferred lifestyles as dangerous, and act on their perceived danger from the issue [28,32]. Values applied by each of the four cultures are constraining decisions and behaviors and are difficult to change [33]. Therefore, we will analyze four different cultural types separately to infer their guiding roles in behavior. Because individuals who prefer an egalitarian culture believe that everyone in society is equal [13–15], they are more likely to accept ideas supporting equal and fair treatment between human and nature [34,35] and are more willing to seek common interests for both the development of human society and the natural environment. Therefore, they tend to worry about environmental risks and be more active in environmental protection [36–40]. In contrast, individuals who prefer an Figure 2. Dake’s (1992) “grid-group” risk culture type quadrant In culture theory, “grid” dimension refers to rules that connect an individual with another on an ego-centered basis [ 24 ]. Grid is the preference of the social structures that rely on roleor class-based rules (high grid) versus those emphasizing equal opportunity and status among individuals (low grid) [ 22 , 25 ]. “Group” dimension refers to the extent that one person is bound by a group. Group is reflected in the preference of social structures that encourage cooperation, collectivism, and collective interests (high group) versus those that support competition and personal interests (low group) [ 24 , 25 ]. However, unlike other psychometric scales, cultural theory suggests that grid and group dimensions vary separately and independently on their own [ 26 – 28 ]. In Chinese social and cultural systems, the collective thinking makes Chinese people emphasize collective interests over personal interests and connect with each other closely. In addition, China is a country in which Han ethnicity (Han Zu) accounts for a majority of the overall Chinese population. Studies have shown that countries with a single ethnicity tend to prefer collectivism culture [ 25 ]. In the political context, China’s one-party system and centralized government can increase people’s preference for bureaucracy and social system that emphasizes role or class stratification. In fact, not much literature has analyzed and compared the cultural characteristics of di ff erent countries in terms of grid and group dimension. As an exception, Chai et al. used the World Values Survey and found that the culture of China is placed in the high group and high grid quadrant: hierarchy [ 29 ]. Our research population is Chinese youth with college education or above. Some existing research found that the Chinese younger generation tends to be more individualistic than the older generation [ 30 ]. Others have suggested that younger generations are more likely to participate in or support social environmental movements that are consistent with egalitarian values [ 25 ]. Therefore, we expect that Chinese young people are more likely to be located in the high-group quadrant Hypothesis 1. Compared with other culture types, the culture of Chinese youth is more likely to be Hierarchy and Egalitarianism 2.2. Cultural Worldviews and Pro-Environmental Behavior According to the CT, people have di ff erent preferences of how the society should be organized, which influences how they evaluate and respond to risk [ 17 ]. Di ff erent from the traditional rational model which assumes that people rationalize their choices based upon cost benefit analysis, culture theory assumes that people think and act in certain ways that are consistent with their culture [ 15 , 16 , 25 , 31 ]. In other words, individuals will view the issue that opposes their preferred lifestyles as dangerous, and act on their perceived danger from the issue [ 28 , 32 ]. Values applied by each of the four cultures are constraining decisions and behaviors and are di ffi cult to change [ 33 ]. Therefore, we will analyze four di ff erent cultural types separately to infer their guiding roles in behavior Because individuals who prefer an egalitarian culture believe that everyone in society is equal [ 13 – 15 ], they are more likely to accept ideas supporting equal and fair treatment between human and nature [ 34 , 35 ] and are more willing to seek common interests for both the development of human society and the natural environment. Therefore, they tend to worry about environmental risks and be

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 4 of 18 more active in environmental protection [ 36 – 40 ]. In contrast, individuals who prefer an individualistic culture are more concerned about freedom than equality. They are more tolerant of environmental risks [ 41 ]. Because they believe that regulations protecting the environment will simultaneously have restrictions on the development of business and industry, they view environmental protection as threats to freedom, and this view will further decrease their pro-environmental behavior [ 41 ]. Individuals who prefer a hierarchical culture believe that everyone in society has their own role and status [ 42 ], and environmental risk is seen as a “hidden symbolic cultural model of social elite status and authority” [ 29 ]. They trust authority and believe that over-worrying about environmental risks is challenging the status quo and authorities. Therefore, they are more likely to devalue environmental protection, perceive environmental risk as low, and less likely to demonstrate pro-environmental behavior. However, some research found that hierarchists can closely align with the egalitarians in some environmental risk perceptions such as climate change [ 37 ]. Since the hierarchists respect experts and authority, they are more likely to accept information from environmental experts or government authorities [ 32 ]. As the Chinese government has been increasing environmental risk propaganda in recent years, it may send signals to hierarchists that environmental protection is important and they should take actions to protect the environment. Fatalism is a culture that favors unpredictable actions. Because fatalists are unlikely to maintain a stable position in life, it has been proven that they have little to do with environmental problems [ 8 , 43 , 44 ]. Consequently, fatalism is often excluded from many studies that predict environmental protection behaviors Any individual can be viewed as a hybrid of the four types of culture [ 25 ], and an individual’s environmental perception and behavior are influenced by the mixture of di ff erent cultural types [ 21 ]. Therefore, we created four indicators for each of the four types of respondents’ risk culture [ 45 ]. We expect: Hypothesis 2 a. Cultural worldviews are significantly associated with individual pro-environmental behavior Hierarchical and egalitarian cultures will increase people’s pro-environmental behavior Hypothesis 2 b. Individualistic culture will decrease people’s pro-environmental behavior Hypothesis 2 c. Fatalist culture will not influence people’s pro-environmental behavior 2.3. Cultural Worldviews, Environmental Risk Perception, and Pro-Environmental Behavior Environmental risk perception is an individual’s understanding of the importance and urgency of environmental protection and the relationship between people and the environment. Individuals’ propensities to protect the environment can be dependent on their perceived environmental risk For example, they may decide to change their behavioral habits and lifestyles to protect the environment when they perceive a high level of environmental risks. Pro-environmental behavior is a human behavior that reflects a relatively consistent eco-friendly propensity to buy, use or dispose of a particular product [ 46 ]. Pro-environmental behavior also refers to some general environmental behavior such as taking the initiative to understand the relevant information of ecological civilization, participating in ecological civilization activities, turning o ff the lights when not in use, and paying attention to garbage classification [ 47 ]. In addition to the finding that risk perception can directly influence pro-environmental behaviors, previous studies have also shown that the influence of cultural worldviews on pro-environmental behavior can be mediated by environmental risk perception [ 8 , 11 ]. According to culture theory, specific behaviors result from specific attitudes that are constrained by culture [ 8 , 11 ]. Therefore, each of the four types of cultures will influence the levels of environmental risk perception, which will further influence the behavioral choices Prior studies on the relationship between cultural worldview and risk perception have found that individuals with fatalistic culture and individualistic culture rate risk lower than those with

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 5 of 18 hierarchical and egalitarian culture [ 31 , 43 ]. In addition, research has found that environmental risk perception can explain the varieties of pro-environmental behavior [ 1 , 2 , 45 , 48 , 49 ], and individuals with higher awareness of environmental risk will demonstrate more pro-environmental behavior than those with lower awareness of environmental risk [ 50 , 51 ]. However, many scholars have found that there is a discrepancy between environmental risk perception and participation in environmental protection activities, which is referred to as attitude-behavior gap or value-action gap [ 9 – 11 ]. In this study, we expect: Hypothesis 3 a. Risk perception plays a mediating role in the relationship between risk culture and pro-environmental behavior Hypothesis 3 b. The higher the environmental risk perception of an individual is, the more pro-environmental behavior he or she has 3. Data, Variables and Methods 3.1. Data Data source: our paper uses an existing online survey conducted by the Center of Ecological Civilization (CEC) of China University of Geosciences from December 2015 to March 2016. China University of Geosciences is a comprehensive university with diverse majors including earth system science, applied science and many interdisciplinary research areas. To collect a large-scale sample across di ff erent provinces, the Center for Ecological Civilization first set up an ecological survey team consisting of 50 students from di ff erent majors to ensure the diversity of respondents. Next, a “snowball sampling” approach was used. The students from the survey team recruited 250 college students from di ff erent provinces as the initial respondents. These 250 students were asked to answer the survey and send the survey to their middle school classmates who were studying in di ff erent regions or universities. This process was repeated multiple times until there were very few responses received. In order to ensure the randomness of sample distribution, each respondent was required to recruit no more than five participants, and these five participants should not have studied in the same province. The survey covered 152 universities in 30 provinces (excluding Tibet, Hong Kong, Macau, and Taiwan) and received a total of 14,097 responses. Among the participants in the survey, 66.14% are college students, 29.69% are graduate students and 4.5% are doctoral students. The survey was distributed by using the most popular social software in China: Tencent and Wechat Data screening: To ensure high quality of the data, the data screening of this paper includes three steps: first, we removed the observations with missing values; second, we removed the cases in which respondents answered the survey in less than three minutes; third, the illogical answers were removed. For example, those who chose government when they were asked “Who is responsible for the construction of ecological civilization?” but ranked government as lower than enterprises and individuals when they were asked to rank “Who should be responsible for construction of ecological civilization?” were removed. As a result, our final sample size included 8084 valid cases 3.2. Methods and Variables To investigate the determinants of pro-environmental behaviors of college students and the mediating e ff ect of environmental risk on the relationship between culture and pro-environmental behavior, we constructed three multiple regression models: Behavior i = α 0 + α 1 Culture i + n X k = 1 α 2 Control i , k + µ 1 (1)

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 6 of 18 Perception i = β 0 + β 1 Culture i + n X k = 1 β 2 Control i , k + µ 2 (2) Behavior i = δ 0 + δ 1 Culture i + δ 2 Perception i + n X k = 1 δ 3 Control i , k + µ 3 (3) where Behaviror i is the sum of scores of seven types of environmental actions. The respondents were asked to rate separately, using a 5-point scale, the level and frequency of pro-environmental behaviors. This research selected seven types of behavior from the survey, and these behaviors have also appeared in previous research on pro-environmental behaviors [ 46 , 47 , 49 ]. For example, we selected questions regarding respondents’ attention to the classification of the trash and environmental protection signs to measure purchasing and using of green product [ 46 ]. We also selected questions regarding turning o ff the lights when you leave the room and questions regarding their participation in environmental activities, including donations for eco-public activities and ecological civilization activities, to measure general pro-environmental behaviors [ 47 , 49 ]. Table 1 summarized our selected seven types of pro-environmental behavior (see Table 1 ). Perception i denotes the perceived worries of environmental issues. Students were asked to answer whether they worry about the current overall ecological environment in China (5-point scale: 1: Definitely not; 2: Probably not; 3: Neutral; 4: Probably worry; 5: Definitely very worry) Table 1. Measuring Pro-environmental Behavior Key Themes Description Mean Std. Min Max Have you taken the initiative to learn about the ecological civilization? 5-point scale (1: never; 2: hardly; 3: occasionally; 4: often; 5: always) 2.943 0.738 1 5 Have you ever participated in the ecological civilization activities organized by the school? 4-point scale (1: have not participated and consider it unnecessary; 2: have not participated but consider it necessary; 3: have participated but thought it was not very useful; 4: have participated and thought it makes sense) 2.632 0.878 1 4 Will you make donations for eco-public activities? 4-point scale (1: never considered doing this; 2: depending on economic condition; 3: if there is such an activity around, I am very happy to participate; 4: already done this) 2.841 0.712 1 4 Have you ever promoted the knowledge of ecological civilization to the people around you? 5-points scale (1: never; 2: hardly; 3: occasionally; 4: often; 5: always) 3.044 0.782 1 5 Do you turn o ff the lights or air conditioners when you leave your bedroom or empty classroom? 5-points scale (1: never; 2: hardly; 3: occasionally; 4: often; 5: always) 4.549 0.687 1 5 Do you pay attention to the classification mark of the trash cans when you throw away garbage? 5-points scale (1: never; 2: hardly; 3: occasionally; 4: often; 5: always) 3.928 0.941 1 5 Do you pay attention to energy saving and environmental protection signs when buying goods? 5-points scale (1: never; 2: hardly; 3: occasionally; 4: often; 5: always) 3.056 1.014 1 5 Culture i is the individual’s risk culture type (including fatalism, hierarchy, individualism and egalitarianism). CT scholars have been debating whether individual cultural bias should be classified as a dominant culture in four quadrants or measured by four indices indicating di ff erent levels for each of the four types of culture. This paper argues that each individual should be treated as a hybrid of the four types of culture and uses four cultural indices separately to measure each of the cultures [ 16 , 17 ]. CT scholars have noticed that worldview is not an ideal measurement for culture and suggested that a combination of worldview and relational statement measures have better validity [ 52 ]. In this study, we select survey items that best approximated the four types of worldviews for two reasons (in Appendix A ). First, existing research justified the value of selecting worldview survey items in existing research to measure cultures [ 29 , 53 – 55 ]. Second, research on Asian countries have found limitations and challenges by using measures developed for research in the United States [ 8 , 56 ].

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 7 of 18 Therefore, this research tries to select survey items that have better face validity. Questions regarding the role of authority and government and preference for self-interest over other values should have face validity [ 52 ]. In addition, while questions regarding attitude to economic activity are viewed as maybe having face validity in Swedlow et al.’s (2019) research, these questions have been used in some existing cultural research [ 53 , 57 ]. Therefore, we selected four survey questions to measure four types of culture: (1) What motivates you to protect the environment, (2) who should take the responsibility for environmental protection, (3) attitude toward commerce and industry, and (4) ranking of responsibility in environmental protection (Table 2 ) [ 14 , 21 , 28 ]. For each question, the respondent was assigned di ff erent scores for each question for each of the four types of cultures according to their answers (see more detail in Table 2 ). The scores for each question were then summed to create four cultural indices [ 45 ]. Because we use an existing research that was not designed to test culture, the respondents were not told explicitly that these questions will be used to measure their cultures. Therefore, we believe that response bias, such as social desirability bias, should be minimal Table 2. Measuring Culture Questions Cultural Bias Scores What motivates you to protect the environment? Fatalism: No motivation and irrelevant Yes = 1, No = 0 Individualism: Self interest Hierarchy: Laws and rules Egalitarianism: Respect for nature as human being Who should take the responsibility for environmental protection? Fatalism: Others Yes = 1, No = 0 Individualism: Businesses enterprises Hierarchy: Government Egalitarianism: Every individual Attitude to commerce and industry? Fatalism: Irrelevant Individualism: Unrestricting Hierarchy: Regulation Egalitarianism: Anti-commerce Ranking of responsibility in environmental protection Fatalism: Others Rank in 1 st = 5 Rank in 2 nd = 4 Rank in 3 rd = 3 Rank in 4 th = 2 Excluded by the respondents = 1 Individualism: Business Hierarchy: Government Egalitarianism: Everyone X is the vector of controls, including knowledge, gender, grade, school, and the province in which the respondent attended her or his school. Age is not included in this questionnaire. However, this survey has a question of grade. Although grade and age are not the same thing, to some extent, grade should be correlated with age for university students in China. Consistent with previous cultural surveys in public opinion research [ 58 , 59 ], our study used two variables to measure knowledge. First, we used a master’s degree or doctorate as a proxy for “policy elite” or “intellectual.” Second, we created a dummy variable to indicate whether the respondents were majoring in an ecological environment related area The statistical analysis process of the findings is performed in the following way: first, we regressed pro-environmental behavior on culture to test the main e ff ect of culture (Equation (1)). If α 1 is significantly positive, Hypothesis 2 set for this study is valid and we can continue to carry out the next test. Second, we regressed pro-environmental behavior on environmental risk perception to check whether the β 1 is significant (Equation (2)). Third, we regressed pro-environmental behavior on both culture and environmental risk perception (Equation (3)). In this test, we will focus on the size and significance of δ 1 in Equation (3). We compare the coe ffi cient in di ff erent steps: if both β 1 and δ 2 are significant and the value of δ 1 becomes smaller, it means that risk perception plays a mediating role in which culture predict pro-environmental behavior (Hypothesis 3 a). If δ 2 is significantly positive, it means that Hypothesis 3 b is valid, meaning that risk perception can increase the pro-environmental behavior.

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 8 of 18 4. Analysis and Results 4.1. Summary Statistics of Related Variables Table 3 reports the descriptive statistics of the variables, from which the overall characteristics of the sample data can be obtained: the score of pro-environmental behavior ranges from 7 to 28 and mean value is 22.995, which indicates that the mean value (average level) of seven types of pro-environmental behavior of college students is relatively high. However, the standard deviation is large, indicating that the di ff erences in behavior among college students are large. In terms of the distribution of cultural bias, overall, Chinese youth has a highest mean value of egalitarian culture (5.409), followed by hierarchical culture (5.094), individualistic culture (3.627), and fatalism (2.128). This shows that Chinese youth are more likely to be hierarchists and egalitarians. Figure 3 further explains the distribution of di ff erent scores in four cultural bias among young Chinese people. X-axis is the score of individuals’ four cultural types and ranges from 1 to 8. Y-axis is the percentage of each score in di ff erent cultures. The overall distribution of hierarchical cultural scores are very similar with the egalitarian culture, and most of people have higher scores in egalitarian and hierarchical culture while most of people have lower scores in individualism and fatalism. There are 4876 people having a score of 5 or above for egalitarian culture, which accounted for 60.32% of respondents. Similarly, there are 7990 people having a score of 5 or more for hierarchical culture, which accounted for 61.73% of responses. This means that Chinese youth are more likely to be hierarchists and egalitarians Table 3. Summary of Variables Variables Description Mean Std. Min Max Obs. Dependent Variable Pro-environmental behavior Statistics on the degree and frequency of participation in environmental protection activities, such as participating in ecological civilization activities, turning o ff lights actively, etc 22.995 3.398 7 28 8084 Independent Variables Environmental risk perception Statistics on answers concerned about the current overall ecological environment in China 4.145 0.705 1 5 8084 Risk Culture Fatalism 2.218 0.682 1 8 8084 Individualism 3.627 1.383 1 8 8084 Hierarchy 5.093 1.730 1 8 8084 Egalitarianism 5.409 1.925 1 8 8084 Control Variables Knowledge Education Completion of a master’s degree and above 0.0468 0.211 0 1 8084 Major Related Majoring in ecological environment related area 2.527 0.7456 1 4 8084 Demographic factors Gender Male = 1, Female = 0 0.356 0.746 0 1 8084 Grade Level of grade (1–4: from first year of university to fourth year undergraduate student; 5,6: from first year to second year graduate student; 7–9: from first year to third year Ph.D. student) 5.195 1.312 1 9 7980 School Categorical variables indicating the school where students attended 49.093 42.269 1 153 8084 Province Categorical variables indicating the provinces where students study at 13.801 7.652 1 30 8084 Then, we used one-way ANOVA test to analyze whether there were differences in pro-environmental behaviors among four di ff erent cultures. The culture type was rated highest among the four cultures was coded as an individual’s dominant culture. Since the behavior distribution of individuals dominated by fatalistic culture did not follow the normal distribution (by analyzing its distribution), we only conducted variance analysis on the behavior among other three types of culture, and the results were as follows (Table 4 ):

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 9 of 18 Table 4. Analysis of Variance Source SS Df MS F Prob > F Between groups 394.81626 2 197.40813 17.37 0.0000 Within groups 89,080.4379 7838 11.3652 Bartlett’s test for equal variances: chi 2(2) = 1.2849, Prob > chi 2 = 0.526 Int. J. Environ. Res. Public Health 2020 , 17 , x 9 of 18 Grade Level of grade (1–4: from first year of university to fourth year undergraduate student; 5,6: from first year to second year graduate student; 7–9: from first year to third year Ph.D. student) 5.195 1.312 1 9 7980 School Categorical variables indicating the school where students attended 49.093 42.269 1 153 8084 Province Categorical variables indicating the provinces where students study at 13.801 7.652 1 30 8084 Figure 3. Distribution of Risk Cultural Biases (Average Frequency). Then , we used one-way ANOVA test to analyze whether there were differences in proenvironmental behaviors among four different cultures. The culture type was rated highest among the four cultures was coded as an individual’s dominant culture. Since the behavior distribution of individuals dominated by fatalistic culture did not follow the normal distribution (by analyzing its distribution) , we only conducted variance analysis on the behavior among other three types of culture , and the results were as follows (Table 4): Table 4. Analysis of Variance. Source SS Df MS F Prob > F Between groups 394.81626 2 197.40813 17.37 0.0000 Within groups 89,080.4379 7838 11.3652 Bartlett’s test for equal variances: chi 2(2) = 1.2849, Prob > chi 2 = 0.526 The results suggest significant differences in pro-environmental behaviors among three cultural types. Figure 4 describes the score of pro-environmental behavior among four types of dominant cultures. Individuals who are affiliated with egalitarian culture and hierarchical culture have higher pro-environmental behavior score. In addition, Figure 5 shows the relationship between risk perception level and pro-environmental behavior. The level of pro-environmental behavior increases as environmental risk perception increases. Figure 3. Distribution of Risk Cultural Biases (Average Frequency) The results suggest significant di ff erences in pro-environmental behaviors among three cultural types. Figure 4 describes the score of pro-environmental behavior among four types of dominant cultures. Individuals who are a ffi liated with egalitarian culture and hierarchical culture have higher pro-environmental behavior score. In addition, Figure 5 shows the relationship between risk perception level and pro-environmental behavior. The level of pro-environmental behavior increases as environmental risk perception increases Int. J. Environ. Res. Public Health 2020 , 17 , x 10 of 18 Figure 4. Risk Culture and Pro-environmental Behavior. Figure 5. Risk perception and Pro-environmental Behavior. Table 5 reports correlations among our key variables. Most of correlation coefficients are between − 0.5 and 0.5 ( Although egalitarianism and hierarchy are negatively correlated at moderate level ( − 0.611), we would argue that it is reasonable given that hierarchy and egalitarianism share the attribute of high group.) It shows that there is no significant multicollinearity. In the preliminary analysis of the relationship between environmental risk perception and pro-environmental behavior , there is a positive correlation at the 1% level , which makes us more confident to test our hypothesis that there is a correlation between risk culture and pro-environmental behavior. At the same time , we can see the varieties of associations between four risk cultures and pro-environmental behaviors , with the egalitarian culture showing a significant positive correlation. However , it is worth noting that although fatalism has a significant negative correlation , it is not suitable for explanation due to the small sample size. Table 5. Correlation matrix of main variables. Variables Behavior Perception Fatalism Individuali-Sm Hierarchy Egalitaria-Nism Behavior 1.00 perception 0.232 *** 1.00 Fatalism − 0.032 *** − 0.044 *** 1.00 Individualism − 0.010 0.041 *** − 0.070 *** 1.00 Hierarchy − 0.015 0.030 *** − 0.043 *** 0.075 *** 1.000 Egalitarianism 0.087 *** 0.090 *** − 0.125 *** − 0.301 *** − 0.611*** 1.000 Notes: *** denotes significance level 1%. 4.2. Results Figure 4. Risk Culture and Pro-environmental Behavior.

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 10 of 18 Int. J. Environ. Res. Public Health 2020 , 17 , x 10 of 18 Figure 4. Risk Culture and Pro-environmental Behavior. Figure 5. Risk perception and Pro-environmental Behavior. Table 5 reports correlations among our key variables. Most of correlation coefficients are between − 0.5 and 0.5 ( Although egalitarianism and hierarchy are negatively correlated at moderate level ( − 0.611), we would argue that it is reasonable given that hierarchy and egalitarianism share the attribute of high group.) It shows that there is no significant multicollinearity. In the preliminary analysis of the relationship between environmental risk perception and pro-environmental behavior , there is a positive correlation at the 1% level , which makes us more confident to test our hypothesis that there is a correlation between risk culture and pro-environmental behavior. At the same time , we can see the varieties of associations between four risk cultures and pro-environmental behaviors , with the egalitarian culture showing a significant positive correlation. However , it is worth noting that although fatalism has a significant negative correlation , it is not suitable for explanation due to the small sample size. Table 5. Correlation matrix of main variables. Variables Behavior Perception Fatalism Individuali-Sm Hierarchy Egalitaria-Nism Behavior 1.00 perception 0.232 *** 1.00 Fatalism − 0.032 *** − 0.044 *** 1.00 Individualism − 0.010 0.041 *** − 0.070 *** 1.00 Hierarchy − 0.015 0.030 *** − 0.043 *** 0.075 *** 1.000 Egalitarianism 0.087 *** 0.090 *** − 0.125 *** − 0.301 *** − 0.611*** 1.000 Notes: *** denotes significance level 1%. 4.2. Results Figure 5. Risk perception and Pro-environmental Behavior Table 5 reports correlations among our key variables. Most of correlation coe ffi cients are between − 0.5 and 0.5 (Although egalitarianism and hierarchy are negatively correlated at moderate level ( − 0.611), we would argue that it is reasonable given that hierarchy and egalitarianism share the attribute of high group.) It shows that there is no significant multicollinearity. In the preliminary analysis of the relationship between environmental risk perception and pro-environmental behavior, there is a positive correlation at the 1% level, which makes us more confident to test our hypothesis that there is a correlation between risk culture and pro-environmental behavior. At the same time, we can see the varieties of associations between four risk cultures and pro-environmental behaviors, with the egalitarian culture showing a significant positive correlation. However, it is worth noting that although fatalism has a significant negative correlation, it is not suitable for explanation due to the small sample size Table 5. Correlation matrix of main variables Variables Behavior Perception Fatalism Individuali-Sm Hierarchy Egalitaria-Nism Behavior 1.00 perception 0.232 *** 1.00 Fatalism − 0.032 *** − 0.044 *** 1.00 Individualism − 0.010 0.041 *** − 0.070 *** 1.00 Hierarchy − 0.015 0.030 *** − 0.043 *** 0.075 *** 1.000 Egalitarianism 0.087 *** 0.090 *** − 0.125 *** − 0.301 *** − 0.611*** 1.000 Notes: *** denotes significance level 1% 4.2. Results In this paper, multiple linear regression methods were used to explore the relationship between individual environmental risk perception and pro-environmental behavior. In order to ensure the validity of the regression results, we tested the assumptions of OLS regression, including linear correlation between independent variables and dependent variables (From the significance test of sample correlation coe ffi cient, r = P ( X i − X )( Y − Y ) q P ( X i − X ) 2 q P ( Y i − Y ) 2 , p < 0.01) and the normally distributed error term (by testing the distribution of the residuals). We also used heteroscedasticity-robust standard error [ 60 – 62 ]. The regression results are shown in Table 6 . The dependent variable in model 1 is risk perception, and model 1 was used to test the e ff ect of culture worldviews on environmental risk perception. The dependent variable of Models 2–4 is pro-environmental behavior. Risk perception, culture and both of them were entered sequentially to investigate their impacts on pro-environmental behavior. More specifically, Model 2 tested the e ff ect of environmental risk perception on pro-environmental behavior. Model 3 tested the e ff ect of culture on pro-environmental

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 11 of 18 behaviors. Model 4 tested the mediating e ff ect of risk perception on the relationship between culture and pro-environmental behavior Table 6. Regression results (Risk Cultural Biases) Variables Perception Behavior Model 1 Model 2 Model 3 Model 4 Risk perception 1.008 *** 0.970 *** (0.058) (0.057) Fatalism − 0.003 *** − 0.048 − 0.045 (0.014) (0.066) (0.062) Individualism 0.046 *** − 0.046 0.002 (0.006) (0.029) (0.029) Hierarchy 0.059 *** 0.106 *** 0.048 *** (0.006) (0.029) (0.029) Egalitarianism 0.0738 *** 0.207 *** 0.135 *** (0.006) (0.028) (0.028) Education 0.172 *** 0.555 *** 0.757 *** 0.590 *** (0.041) (0.215) (0.217) (0.213) Major Related 0.079 *** 0.923 *** 0.994 *** 0.917 *** (0.011) (0.051) (0.052) (0.051) Grade − 0.001 − 0.013 *** -0.022 0.023 (0.017) (0.078) (0.080) (0.078) Gender − 0.032 *** − 0.134 *** − 0.154 *** − 0.123 *** (0.006) (0.030) (0.031) (0.030) School − 0.000 − 0.000 − 0.000 − 0.000 (0.000) (0.000) (0.000) (0.000) Province − 0.002 0.004 0.001 0.004 (0.001) (0.005) (0.005) (0.004) Constant 3.282 *** 17.117 *** 19.507 *** 16.320 *** (0.095) (0.316) (0.427) (0.469) Observations 7980 7980 7980 7980 R 2 0.037 0.095 0.063 0.099 F 23.40 99.25 44.7 66.31 Notes: robust standard errors in parentheses; *** denotes significance level of 1% Hypothesis 2, that cultural worldviews are significantly associated with individual pro-environmental behaviors, was supported in Model 3. The results suggest that Chinese youth who have the higher scores of hierarchical and egalitarian are more likely to demonstrate pro-environmental behavior. Although the coe ffi cients of individualistic culture and fatalist culture were not significant, the coe ffi cients were negative as we expected Hypothesis 3 a expected that risk perception plays a mediating role for the relationship between culture and pro-environmental behavior. First, Model 1 shows that di ff erent cultural worldviews have di ff erent impacts on perception. While fatalistic culture has a significantly negative correlation with environmental risk perception, the other three cultures are positively correlated with environmental risk perception In addition, the results in Model 2 suggest a significant positive correlation between risk perception and pro-environmental behavior, which is consistent with existing studies [ 8 ]. It supports Hypothesis 3 b that the increase in environmental risk perception will increase individuals’ pro-environmental behavior. Finally, culture and environmental risk perception were included in the Model 4 and the results suggested that environmental risk perception is positively correlated with pro-environmental behavior at a significant level. However, the correlations between all cultural types and pro-environmental behaviors were significantly reduced when compared with Model 3. It suggests that that risk perception plays a mediating role for the relationship between culture and pro-environment behavior.

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 12 of 18 Path analysis is displayed in Figure 6 . Hierarchy and egalitarianism are not only directly related to pro-environmental behavior, but also may increase pro-environmental behavior by increased environmental risk perception, showing that risk perception plays partial mediating role in this process (Figure 6 ). Int. J. Environ. Res. Public Health 2020 , 17 , x 12 of 18 correlation between risk perception and pro-environmental behavior , which is consistent with existing studies [8]. It supports Hypothesis 3 b that the increase in environmental risk perception will increase individuals’ pro-environmental behavior. Finally , culture and environmental risk perception were included in the Model 4 and the results suggested that environmental risk perception is positively correlated with pro-environmental behavior at a significant level. However, the correlations between all cultural types and pro-environmental behaviors were significantly reduced when compared with Model 3. It suggests that that risk perception plays a mediating role for the relationship between culture and pro-environment behavior. Path analysis is displayed in Figure 6. Hierarchy and egalitarianism are not only directly related to pro-environmental behavior , but also may increase pro-environmental behavior by increased environmental risk perception, showing that risk perception plays partial mediating role in this process (Figure 6). Figure 6. Path model explains the mediating effect of risk perception on the relationship between risk culture and pro-environmental behavior,*** denotes significance level of 1% In addition , the effects of control variables, including knowledge level on pro-environmental behavior are summarized in Table 6. Knowledge has a positive influence at a significant level. Young people in higher education are more likely to have a higher level of environmental risk perception and pro-environmental behavior. Also , young people who are majoring in environment related areas are more likely to be engaged in environmental protection. Therefore , we suggest that the increase of knowledge level of young people on environmental issues will facilitate the development of proenvironmental behavior. The gender variable is significantly negative , suggesting that women are more likely to participate in pro-environmental behavior than men , which is consistent with previous research findings [63,64]. 4.3. Robust Check In order to ensure the robustness of the research findings , this study used four dominant culture types to replace the cultural indices for regression analysis, and the dominant culture was the cultural type for which the cultural index was reported as highest. Therefore , dominant cultures are measured by categorical variables. The results are shown in Table 7 , which is very similar to the results in Section 4.2. In particular , in model 8 , the coefficient of risk perception is still significantly positive , and risk perception works as a mediator variable, explaining the positive correlation between egalitarian culture and pro-environmental behavior. Figure 6. Path model explains the mediating e ff ect of risk perception on the relationship between risk culture and pro-environmental behavior,*** denotes significance level of 1% In addition, the e ff ects of control variables, including knowledge level on pro-environmental behavior are summarized in Table 6 . Knowledge has a positive influence at a significant level. Young people in higher education are more likely to have a higher level of environmental risk perception and pro-environmental behavior. Also, young people who are majoring in environment related areas are more likely to be engaged in environmental protection. Therefore, we suggest that the increase of knowledge level of young people on environmental issues will facilitate the development of pro-environmental behavior. The gender variable is significantly negative, suggesting that women are more likely to participate in pro-environmental behavior than men, which is consistent with previous research findings [ 63 , 64 ]. 4.3. Robust Check In order to ensure the robustness of the research findings, this study used four dominant culture types to replace the cultural indices for regression analysis, and the dominant culture was the cultural type for which the cultural index was reported as highest. Therefore, dominant cultures are measured by categorical variables. The results are shown in Table 7 , which is very similar to the results in Section 4.2 . In particular, in model 8, the coe ffi cient of risk perception is still significantly positive, and risk perception works as a mediator variable, explaining the positive correlation between egalitarian culture and pro-environmental behavior.

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 13 of 18 Table 7. Regression results (Dominant Risk Culture) Variables Perception Behavior Model 5 Model 6 Model 7 Model 8 Risk perception 1.008 *** 0.998 *** (0.058) (0.057) Individualism 0.185 *** 0.305 0.121 (0.053) (0.275) (0.268) Hierarchy 0.192 *** 0.308 0.116 (0.046) (0.250) (0.242) Egalitarianism 0.222 *** 0.753 *** 0.531 *** (0.046) (0.250) (0.243) Education 0.174 *** 0.555 *** 0.764 *** 0.590 *** (0.044) (0.215) (0.220) (0.215) Major Related 0.085 *** 0.923 *** 1.008 *** 0.923 *** (0.010) (0.051) (0.052) (0.051) Grade − 0.014 − 0.013 *** − 0.033 0.010 (0.016) (0.078) (0.080) (0.078) Gender − 0.032 *** − 0.134 *** − 0.158 *** − 0.126 *** (0.007) (0.030) (0.031) (0.030) School − 0.000 − 0.000 − 0.000 − 0.000 (0.000) (0.000) (0.000) (0.000) Province − 0.002 0.004 0.001 0.003 (0.001) (0.005) (0.005) (0.004) Constant 3.941 *** 17.117 *** 20.750 *** 16.817 *** (0.064) (0.316) (0.331) (0.404) Observations 7980 7980 7980 7980 R 2 0.015 0.095 0.056 0.098 F 13.79 99.25 47.50 73.41 Notes: robust standard errors in parentheses; *** denotes significance level of 1% 5. Brief Conclusion and Future Research There are some limitations in this study. First of all, this research uses attitudinal items in an existing survey that approximate worldviews to measure culture, which decreases the construct validity of measurement. Future research should try to use the survey items directly measuring relations or cultural statement. However, some of consistent results with existing research on China [ 21 , 22 ] provide evidence that our measurements have predictive validity. Considering the condition that China does not have a large-scale survey using cultural measurements developed in the United States, our paper suggests that future research seeking for measuring cultures based on existing survey should select survey items regarding role of authority and government and preference for self-interest over other values. Second, there are some other important factors, such as age, were not tested. We have used grade instead of age because we believe that grade should be related with age in Chinese context. However, we also suspect that age and grade are not the same thing. Therefore, future research should include this demographic factor. Finally, future research should ask whether the influence of the cultural biases and risk perception on pro-environmental behaviors vary with di ff erent environmental policy areas such as energy, water, and wastes management To explain the attitude-action lag between risk perception and pro-environment behavior, prior research focused on the rational choice model or benefits and costs analysis. Unlike other studies, our paper used cultural biases to fill the gap of the previous research. Our research suggests that people think and act in such a way that cultural biases function as an orienting mechanism that helps people navigate a world full of uncertainties and risks [ 32 ]. In the case of young people’s pro-environmental behavior, culture constraints individuals’ core values and behavioral preferences in the face of current environmental issues [ 65 – 67 ].

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 14 of 18 Additionally, the results show that Chinese youth are more likely to be a ffi liated with hierarchical and egalitarian culture. Both of these two cultures belong to the high-group culture type in the grid-group cultural type quadrant. High group cultures emphasize a close connection between individuals and through collective identity, i.e., they foster a high level of collective thinking. This could be explained by China’s national sociocultural conditions. Both egalitarian and hierarchical cultures have a positive e ff ect on the pro-environmental behavior among young people. However, contrary to our expectation, we found no influence of individualism on pro-environmental behaviors. We suggest one possible explanation is that individualists act on self-interests which is very context dependent. Compared with other cultures, they are more willing to act when their personal interests exceed social interests or their private costs are less than social costs. For example, people living in poverty could be less likely to act in protecting environment while people having better income could be more likely to support environmental protection. By contrast, the e ff ect of self-interest might not influence egalitarians’ and hierarchists’ choices as much as individualism. In addition, there are many other factors, such as media coverage on environmental protection, family influences and pressures, religious and spiritual ideals, identities etc., that may influence environmental behaviors and mediate environmental value. Future research should test how other di ff erent factors influence environmental perception and behavior Our research results confirmed that risk perception plays a mediating role in the correlation between egalitarian culture and pro-environmental behavior. Although this has been studied in previous studies, we focused on Chinese young people by using a much larger sample size. Moreover, we found that environmental risk perception is significantly related to pro-environmental behavior. Therefore, raising the perception of environmental risks among young people can help young people develop pro-environmental behaviors Among control variables, the knowledge level of young people is also significantly related to pro-environmental behaviors. Therefore, we suggest that providing better education and improving environmental knowledge among youth may help them to understand environmental risks more comprehensively and objectively, and thus enhancing their pro-environmental behavior For the policy makers, we suggest that, individuals will understand and accept policy recommendations only when they recognize that policies are compatible with their cultural beliefs and in line with their cultural rationality. Therefore, to reach a consensus on environmental policies, the government should consider the varieties of the public’s environmental risk perceptions and their environmental awareness when formulating policies. The cultural attributes of environmental policies and governance should not be ignored Author Contributions: Conceptualization, J.Z. and M.Y.; methodology, M.Y. and M.J.; software, M.J.; validation, J.Z. and M.Y.; formal analysis, M.J. and M.Y.; investigation, J.Z. and M.Y.; resources, J.Z. and M.Y; data curation, J.Z.; writing-original draft preparation, J.Z. and M.J.; writing-review and editing, M.Y.; visualization, M.Y. and M.J.; supervision, J.Z. and M.Y.; project administration, J.Z. and M.Y.; funding acquisition, J.Z. All authors have read and agreed to the published version of the manuscript Funding: The authors acknowledge financial support from the fund project of the National Natural Science Foundation of China (Grant No: 71974203), Zhongnan university of economics and law graduate education achievement award cultivation project (Grant No: CGPY 201904), Interdisciplinary innovation research project (Grant No:2722019 JX 002), and Zhongnan University of Economics and Law Fundamental Research Funds for the Central Universities (Grant No: 2722020 JCT 026) Conflicts of Interest: The authors declare no conflict of interest.

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 15 of 18 Appendix A. Appendix Table A 1. Survey Items for Measuring Pro-environmental Behavior Survey Items in English Survey Items in Chinese Have you taken the initiative to learn about the ecological civilization? (1) Never (2) Hardly (3) Occasionally (4) Often (5) Always 你会 主 动 了 解 有 关 生 态 文 明 的相 关 信 息 吗 ？ （ 1 ） 从 来 不 会 （ 2 ） 几 乎不 会 （ 3 ） 偶 尔 会 （ 4 ） 经 常 会 （ 5 ） 总 是 会 Have you ever participated in the ecological civilization activities organized by the school? (1) I have not participated and consider it unnecessary. (2) I have not participated but consider it necessary. (3) I have participated but thought it was not very useful. (4) I have participated and thought it makes sense 你 是 否 参 加 过 学 校 组织 的 生 态 文 明 活 动 ？ （ 1 ） 没 有 参 加 过 但 认 为 是 必 要 的 。 （ 2 ） 没 有 参 加 过 但 认 为 是 必 要 的 。 （ 3 ） 参 加 过 但 是 认 为 没 有 太大 用 处 。 （ 4 ） 参 加 过 且 认 为 意 义 重 大 。 Will you make donations for eco-public activities? (1) I never considered doing this. (2) It depends on economic condition. (3) If there is such an activity around, I am very happy to participate. (4) I have already done this 你 是 否 为一些 生 态 公 益 活 动 进 行 爱 心 捐 赠 ？ （ 1 ） 从 未 考 虑 过这 样 做 。 （ 2 ） 视 经 济 情 况决 定 。 （ 3 ） 如 果 身 边 有 这 样 的 活 动 ， 非 常 乐 意 参 加 。 （ 4 ） 已 经 在 这 样 做 了 。 Have you ever promoted the knowledge of ecological civilization to the people around you? (1) Never (2) Hardly (3) Occasionally (4) Often (5) Always 你会 主 动 向 身 边 人 宣 传 生 态 文 明 的 知 识 吗 ？ （ 1 ） 从 来 不 会 （ 2 ） 几 乎不 会 （ 3 ） 偶 尔 会 （ 4 ） 经 常 会 （ 5 ） 总 是 会 Do you turn o ff the lights or air conditioners when you leave your bedroom or empty classroom? (1) Never (2) Hardly (3) Occasionally (4) Often (5) Always 你 在 离 开 寝室 、 空 教 室 时 会 主 动 关 闭 电 灯 或 空 调 吗 ？ （ 1 ） 从 来 不 会 （ 2 ） 几 乎不 会 （ 3 ） 偶 尔 会 （ 4 ） 经 常 会 （ 5 ） 总 是 会 Do you pay attention to the classification mark of the trash cans when you throw away garbage? (1) Never (2) Hardly (3) Occasionally (4) Often (5) Always 在 路 边 丢 弃 垃圾 时 ， 你会 主义 垃圾 桶 的 分 类 标 志 吗 ？ （ 1 ） 从 来 不 会 （ 2 ） 几 乎不 会 （ 3 ） 偶 尔 会 （ 4 ） 经 常 会 （ 5 ） 总 是 会 Do you pay attention to energy saving and environmental protection signs when buying goods? (1) Never (2) Hardly (3) Occasionally (4) Often (5) Always 购 买 商 品 时 ， 你会 注 意 商 品 的 节 能 环 保 标 志 吗 ？ （ 1 ） 从 来 不 会 （ 2 ） 几 乎不 会 （ 3 ） 偶 尔 会 （ 4 ） 经 常 会 （ 5 ） 总 是 会 Table A 2. Survey Items for Measuring Culture Survey Items in English Survey Items in Chinese What motivates you to protect the environment? (1) No motivation and irrelevant (2) Self interest (3) Laws and rules (4) Respect for nature as human being 当 你 做 出 保 护 生 态 的 行 为 时 出 发 点 是 ？ （ 1 ） 没 有 兴 趣 且与 我 无 关 （ 2 ） 个人 利 益 （ 3 ） 法 律 规 定 （ 4 ） 尊 重 自 然 Who should take the responsibility for environmental protection? (1) Others (2) Business enterprises (3) Government (4) Every individual 你 认 为 保 护 生 态 环 境 是 谁 的 责 任 ？ （ 1 ） 其 他 （ 2 ） 企 业 （ 3 ） 政 府 （ 4 ） 个人 What is your attitude towards industry and commerce? (1) Irrelevant (2) Unrestricting (3) Regulation (4) Anti-commerce 你 对 工 商 业 的 态 度 是 什么 ？ （ 1 ） 如 我 无 关 （ 2 ） 减 少 限 制 （ 3 ） 规 则 约 束 （ 4 ） 反 对 商 业 Ranking of responsibility in environmental protection. Government / Business enterprises / Every individual / Others 对 于 环 境 保 护 中 的 责 任 进 行 排 序 。 政 府 ／ 企 业 ／ 个人 ／ 其 他 References 1 Lin, T.; Cao, X.; Huang, N.; Xu, L.; Li, X.; Zhao, Y.; Lin, J. Social cognition of climate change in coastal community: A case study in Xiamen City, China Ocean Coastal Manag 2018 . [ CrossRef ] 2 Lieske, D.J.; Wade, T.; Roness, L.A. Climate change awareness and strategies for communicating the risk of coastal flooding: A Canadian Maritime case example Estuarine Coast. Shelf Sci 2014 , 140 , 83–94. [ CrossRef ] 3 Chen, X.Y.; Li, F.; Zhang, J.D.; Zhou, W.; Wang, X.Y. Spatiotemporally mapping and multiple driving forces identifying of PM 2.5 variation and its joint management strategies across China J. Clean. Prod 2019 [ CrossRef ] 4 Li, L.; Xia, X.; Chen, B.; Sun, L. Public participation in achieving sustainable development goals in China: Evidence from the practice of air pollution control J. Clean. Prod 2018 , 201 , 499–506. [ CrossRef ] 5 Kates, N. Attention deficit disorder in adults. Management in primary care Can. Fam. Phys. M é d. De Fam Can 2005 , 51 , 53–59.

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 16 of 18 6 Leeuwis, C. Reconceptualizing Participation for Sustainable Rural Development: Towards a Negotiation Approach Dev. Chang 2000 , 31 , 931–959. [ CrossRef ] 7 Duan, H.X.; Rosanne, W.F. A Cross-Cultural Study on Environmental Risk Perception and Educational Strategies: Implications for Environmental Education in China Int. Electron. J. Environ. Educ 2010 , 1 , 1–19 8 Xue, W.; Zhao, S.Y. The Environmental Worldviews and Climate Change Mitigation Behaviors: Testing the New Ecological Scale in the Smallest Space Analysis for Chinese Samples Int. J. Environ. Sci. Dev 2015 , 6 , 547–550. [ CrossRef ] 9 Lorenzoni, I.; Nicholson-Cole, S.; Whitmarsh, L. Barriers perceived to engaging with climate change among the UK public and their policy implications Glob. Environ. Chang 2007 , 17 , 459. [ CrossRef ] 10 Fisk, D. Engaging the public with climate change: Behaviour change and communication, edited by Lorraine Whitmarsh, Sa ff ron O’Neill and Irene Lorenzoni Int. J. Ambient Energy 2011 , 32 , 112. [ CrossRef ] 11 Lacroix, K.; Gi ff ord, R. Psychological Barriers to Energy Conservation Behavior: The Role of Worldviews and Climate Change Risk Perception Environ. Behav 2017 , 50 , 749–780. [ CrossRef ] 12 Deng, F.; Zhou, C.; Zheng, Y Exploring the Impact of Risk-Benefit Perception on Public Acceptance to Nuclear Power—Based on a Survey at Guangzhou University City ; University of South China: Hengyang, China, 2016 13 Chen, D.; Cheng, C.Y.; Urpelainen, J. Support for renewable energy in China: A survey experiment with internet users J. Clean. Prod 2016 , 112 , 3750–3758. [ CrossRef ] 14 Swedlow, B. A Cultural Theory of Politics PS Political Sci. Politics 2011 , 44 , 703–710. [ CrossRef ] 15 Swedlow, B. Advancing Policy Theory with Cultural Theory: An Introduction to the Special Issue Policy Stud. J 2014 , 42 , 465–483. [ CrossRef ] 16 Swedlow, B.; Ripberger, J.T.; Liu, L.Y.; Silva, C.L. Assessing the Validity of Di ff erent Approaches to Operationalizing Cultural Theory in Survey Research Soc. Sci. Electron. Publ 2016 . [ CrossRef ] 17 Agar, M.; Douglas, M.; Wildavsky, A. Risk and Culture Anthropol. Q 1983 , 56 , 102. [ CrossRef ] 18 Brenot, J.; Bonnefous, S.; Marris, C. Testing the Cultural Theory of Risk in France Risk Anal 1998 , 18 , 729–739 [ CrossRef ] 19 Marris, C.; Langford, I.H.; O’Riordan, T.A. Quantitative Test of the Cultural Theory of Risk Perceptions: Comparison with the Psychometric Paradigm Risk Anal 1998 , 18 , 635–647. [ CrossRef ] 20 Bouyer, M.; Bagdassarian, S.; Chaabanne, S. Personality Correlates of Risk Perception Risk Anal 2001 , 21 , 457–465. [ CrossRef ] 21 Yuan, M.; Zeng, J.J.; Swedlow, B.; Qi, R. Environmental Concern among Chinese Youth: The Roles of Knowledge and Culture Social Science Quarterly 2019 . under review 22 Xue, W.; Hine, D.; Marks, A.; Phillips, W.; Zhao, S. Cultural worldviews and climate change: A view from China Asian J. Soc. Psychol 2016 , 19 , 134–144. [ CrossRef ] 23 Ellis, R.J.; Thompson, F. Culture and the Environment in the Pacific Northwest Am. Political Sci. Rev 1997 , 91 , 885–897. [ CrossRef ] 24 Hoppe, R. Guest co-editor’s introduction: Cultural theory’s gift for policy analysis J. Comp. Policy Anal. Res Pract 2002 , 4 , 235–241. [ CrossRef ] 25 Chai, S.-K.; Dorj, D.; Hampton, K.; Liu, M. The Role of Culture in Public Goods and Other Experiments PS Political Sci. Politics 2011 , 44 , 740–744. [ CrossRef ] 26 Kahan, D.M.; Braman, D.; Slovic, P.; Gastil, J.; Cohen, G. Cultural Cognition of the Risks and Benefits of Nanotechnology Soc. Sci. Electron. Publ 2009 , 4 , 87–90. [ CrossRef ] [ PubMed ] 27 Kahan, D.M.; Jamieson, K.H.; Landrum, A.R.; Winneg, K. Culturally antagonistic memes and the Zika virus: An experimental test J. Risk Res 2017 , 20 , 1–40. [ CrossRef ] 28 Kahan, D.M. Ideology, Motivated Reasoning, and Cognitive Reflection: An Experimental Study Judgm. Dec Mak 2013 , 8 , 407–424. [ CrossRef ] 29 Chai, S.K.; Kim, M.S. Cultural Comparisons of Beliefs and Values: Applying the Grid-Group Approach to the World Values Survey Beliefs Values 2009 , 1 , 193–208. [ CrossRef ] 30 Zeng, R.; Greenfield, P.M. Cultural evolution over the last 40 years in China: Using the Google Ngram Viewer to study implications of social and political change for cultural values Int. J. Psychol 2015 , 50 , 47–55 [ CrossRef ] 31 Dake, K.; Wildavsky, A. Individual Di ff erences in Risk Perception and Risk-Taking Preferences. In The Analysis, Communication, and Perception of Risk ; Springer: Boston, MA, USA, 1991; pp. 15–24.

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 17 of 18 32 Bordat-Chauvin, E. Cultural Policies and Change: Mexico and Argentina after the Neoliberal Turn (1983–2012) Lat. Am. Policy 2016 , 7 , 147–162. [ CrossRef ] 33 Schwartz, S.H. A Theory of Cultural Values and Some Implications for Work Appl. Psychol 1992 , 48 , 23–47 [ CrossRef ] 34 Schalock, R.L.; Verdugo, M.A.; Jenaro, C.; Wang, M.; Wehmeyer, M.; Jiancheng, X.; Lachapelle, Y. Cross-Cultural Study of Quality of Life Indicators Am. J. Ment. Retard 2005 , 110 , 298–311. [ CrossRef ] 35 Jenaro, C.; Verdugo, M.A.; Caballo, C. Cross-cultural study of person-centred quality of life domains and indicators: A replication J. Intellect. Disabil. Res 2005 , 49 , 734–739. [ CrossRef ] [ PubMed ] 36 Jones, J.C. How Safe is Safe Enough? Risk Manag 2005 , 7 , 73–74. [ CrossRef ] 37 McEvoy, J.W.; Diamond, G.A.; Detrano, R.C.; Kaul, S.; Blaha, M.J.; Blumenthal, R.S.; Jones, S.R. Risk and the Physics of Clinical Prediction Am. J. Cardiol 2014 , 113 , 1429–1435. [ CrossRef ] [ PubMed ] 38 Zeng, J.J.; Liu, T.; Feiock, R.; Li, F. The impacts of China’s provincial energy policies on major air pollutants: A spatial econometric analysis Energy Policy 2019 , 132 , 392–403. [ CrossRef ] 39 Zeng, J.J.; Yuan, M.; Feiock, R. What Drives People to Complain about Environmental Issues? An Analysis Based on Panel Data Crossing Provinces of China Sustainability 2019 , 11 , 1147. [ CrossRef ] 40 Zeng, J.J.; Liu, D.J.; Yi, H.T. Agglomeration, Structural Embeddedness, and Enterprises’ Innovation Performance: An Empirical Study of Wuhan Biopharmaceutical Industrial Cluster Network Sustainability 2019 , 11 , 3922. [ CrossRef ] 41 Hermand, D.; Karsenty, S.; Py, Y.; Guillet, L.; Chauvin, B.; Simeone, A.; Teresa, M.; Sastre, M.; Mullet, E. Risk Target: An Interactive Context Factor in Risk Perception Risk Anal 2003 , 23 , 821–828. [ CrossRef ] 42 Lachapelle, P.R.; McCool, S.F. Claiming Ownership: A Response to Manning and Ginger Soc. Nat. Resour 2007 , 20 , 193–197. [ CrossRef ] 43 Thompson, M.; Ellis, R.; Wildavsky, A Cultural Theory ; Westview Press: Boulder, CO, USA, 1990; pp. 49–50 44 Gi ff ord, R.; Nilsson, A. Personal and social factors that influence pro-environmental concern and behaviour: A review Int. J. Psycholo 2014 , 49 , 141–157. [ CrossRef ] [ PubMed ] 45 Herron, K.G.; Jenkins–Smith, H.C. U.S. Perceptions of Nuclear Security in the Wake of the Cold War: Comparing Public and Elite Belief Systems Int. Stud. Q 2002 , 46 , 451–479. [ CrossRef ] 46 Rohini, S. The Influence of Cultural Values and Environmental Attitudes on Green Consumer Behaviour Int J. Behav. Sci 2012 , 7 , 83–98 47 Kilbourne, W.; Pickett, G. How materialism a ff ects environmental beliefs, concern, and environmentally responsible behavior J. Bus. Res 2008 , 61 , 885–893. [ CrossRef ] 48 Li, F.; Qiu, Z.Z.; Zhang, J.D.; Liu, C.Y.; Cai, Y.; Xiao, M.S.; Zhu, L.Y. Temporal variation of major nutrients and probabilistic eutrophication evaluation based on stochastic-fuzzy method in Honghu Lake, Middle China Sci. China Ser. E Technol. Sci 2019 , 62 , 417–426. [ CrossRef ] 49 Li, F.; Zhang, J.D.; Liu, W.C.; Liu, J.A.; Huang, J.H.; Zeng, G.M. An exploration of an integrated stochastic-fuzzy pollution assessment for heavy metals in urban topsoil based on metal enrichment and bioaccessibility Sci Total Environ 2018 , 644 , 649–660. [ CrossRef ] 50 Slovic, P.; McDaniels, T.; Lawrence, J.; Axelrod, L.J.; Giulio, R.T Perceptions of Risk to Humans and to Nature: A Research Plan ; Springer: Dordrecht, The Netherlands, 1996; pp. 199–209 51 Fern á ndez-Contreras, M.-E.; Sarri á , C.; Nieto, S.; Lazo, P.A. Amplification of human genomic sequences by human papillomaviruses universal consensus primers J. Virol. Methods 2000 , 87 , 171–175. [ CrossRef ] 52 Swedlow, B.; Johnson, B.B. The Value-Added by Cultural Theories of Political Values: Comparing Ideology, Partisanship, and Two Cultural Value Explanations. 2019. Available online: https: // ssrn.com / abstract = 3426221 (accessed on 7 May 2019) 53 Boyle, R.P.; Coughlin, R.M. Conceptualizing and Operationalizing Culture Theory Variables. In Politics Culture Polic ; Westview Press: Boulder, CO, USA, 1994; pp. 191–218 54 West, J.; Bailey, I.; Winter, M. Renewable energy policy and public perceptions of renewable energy: A cultural theory approach Energy Policy 2010 , 38 , 5739–5748. [ CrossRef ] 55 Simmons, R. Improvement and Public Service Relationships: Cultural Theory and Institutional Work Public Adm 2016 , 94 , 933–952. [ CrossRef ] 56 Liu, L.-Y. How Radical Is Too Radical? Public Perception of Taiwanese Environmental Nonprofit Organizations’ Activism Soc. Sci. Q 2019 , 99 , 1426–1445. [ CrossRef ]

Int. J. Environ. Res. Public Health 2020 , 17 , 1750 18 of 18 57 Coughlin, R.M.; Lockhart, C. Grid-Group Theory and Political Ideology J. Theor. Politics 1998 , 10 , 33–58 [ CrossRef ] 58 Stimson, G.V.; Fitch, C.; Jarlais, D.D. Rapid Assessment and Response Studies of Injection Drug Use: Knowledge Gain, Capacity Building, and Intervention Development in a Multisite Study Am. J. Public Health 1975 , 96 , 288–295. [ CrossRef ] [ PubMed ] 59 Welzel, M.; Roth, W. Do interviews really assess students’ knowledge? Int. J. Sci. Educ 2004 , 20 , 25–44 [ CrossRef ] 60 Huber, P.J. Robust Estimation of a Location Parameter Ann. Math. Stat 1964 , 35 , 73–101. [ CrossRef ] 61 Eicker, F. Limit theorems for regressions with unequal and dependent errors. In Berkeley Symposium on Mathematical Statistics & Probability ; The Regents of the University of California: Oakland, CA, USA, 1967 62 White, H.A. Heteroscedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroscedasticity Econometrica 1980 , 48 , 817–838. [ CrossRef ] 63 Alexandra, G.; Delhomme, P. Marie Line FELONNEAU. Peer. Pressure and risk taking in young drivers’ speeding behavior Transp. Res. Part F Tra ffi c Psychol. Behav 2015 , 35 , 101–111 64 Feygina, I.; Tom, R.T. Procedural Justice and System-Justifying Motivations. In Social and Psychological Bases of Ideology and System Justification ; Oxford University Press: Oxford, UK, 2009; pp. 351–370 65 Harris, P.G. Environmental Perspectives and Behavior in China: Synopsis and Bibliography Environ. Behav 2006 , 38 , 5–21. [ CrossRef ] 66 Johnson, B.B.; Swedlow, B. Cultural Theory’s Contributions to Risk Analysis: A Thematic Review with Directions and Resources for Further Research Risk Anal 2019 . [ CrossRef ] 67 Kim, H.K.; Kim, Y. Risk Information Seeking and Processing About Particulate Air Pollution in South Korea: The Roles of Cultural Worldview Risk Anal 2018 , 39 , 1071–1087. [ CrossRef ] © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http: // creativecommons.org / licenses / by / 4.0 / ).