Research on the Influencing Factors of Urban Community Residents’ Willingness to Segregate Waste Based on Structural Equation Model

Abstract

Segregation of household waste is an important means of achieving resource recovery, minimization and harmlessness of waste, which is of great significance in addressing the dilemma of the “rubbish siege”. However, at present, urban community residents still face many challenges in the practice of household waste classification, such as lack of classification knowledge, imperfect classification facilities and weak and persistent classification behavior, which seriously restrict the effective promotion of garbage classification work. In this paper, a model of the factors influencing community residents’ willingness to separate household waste was developed based on the theory of planned behavior and tested by using structural equation modeling (SEM) with a sample of 218 surveys conducted among residents of community X in Nanchang, Jiangxi Province. It was found that urban community residents were generally willing to sort their household waste subjectively. The five factors of waste sorting recognition, intrinsic moral constraints, group behavioral incentives, time and space factors and waste sorting facilities positively influenced urban community residents’ willingness to sort household waste. Government job satisfaction and legal and regulatory constraints had no significant influence on urban community residents’ willingness to sort household waste and did not reach a statistically significant level. Based on this, in the future, we should strengthen public education, enhance group behavioral incentives, improve supporting infrastructure, standardize and improve laws and regulations to improve residents’ willingness to separate household waste and promote the process of urban household waste segregation in China.

1. Introduction

With the progress of China’s reform and opening up and the continuous rise in the level of economic development, more and more people are flocking to cities, leading to a rising urban population in China that is still growing at a high rate. At the same time, this is accompanied by a large amount of household waste, which seriously hinders the sustainable development of cities. According to a survey, urban waste in China is growing at an annual rate of 9%, and nearly two-thirds of cities are being inundated by rubbish [1]. Rubbish is a misplaced resource. Every year, Chinese residents can reuse 60 million tons of household waste, with an economic value of CNY 25 billion [2]. Separating waste is the most effective way to turn waste into treasure while centralizing harmful waste, reducing it, making it resourceful and harmless, which is of great significance in solving the dilemma of waste management and promoting urban ecological civilization. Since 2000, China has been actively learning from the advanced experience of waste management at home and abroad and has formulated and improved relevant laws and regulations. Relevant departments have issued the Implementation Plan of the Household Waste Segregation System, clarifying the objectives, tasks and requirements for garbage segregation. The importance and specific classification methods of garbage segregation are widely publicized through the media, internet and community, so as to improve the public’s awareness and attention to garbage segregation. And waste segregation has gradually spread and achieved some success. However, it is undeniable that the implementation of household waste segregation policies in domestic cities is still subject to multiple resistance, and the actual results are not obvious [3]. As residents are the generators and disposers of household waste, it is natural that the community should become the most important factor in the management of municipal household waste. The participation willingness and behavior of individual residents have become the key to solving the dilemma of garbage management.

In the field of waste management, domestic academics have conducted research relatively late compared to foreign countries, but after decades of development, fruitful results have been achieved, involving multiple fields and multiple perspectives, including the main directions of realistic dilemmas, influencing factors and improvement mechanisms, multiple subjects such as government, market and individuals and various elements such as subjective attitudes, laws and regulations, technology and equipment and group behavior. At present, the focus of research in this field has tended to be on the microscopic subject of residents, mainly exploring the psychology and behavior of individual residents in waste segregation. Some studies show that at the present stage, the problems of urban community household waste segregation mainly focus on the lack of waste segregation awareness, lack of relative knowledge, lack of community identity and responsibility awareness, etc. [4,5]. Therefore, the core reasons need to be further explored, and the corresponding countermeasures and suggestions should be put forward. However, in general, there are more behavioral studies and fewer willingness studies, more qualitative studies and fewer quantitative studies, and most scholars use logistic regression analysis and SPSS correlation analysis to study the influencing factors, but these methods can only study the influence of individual factors on the dependent variable, and there are measurement errors and limitations. In this context, this paper takes the theory of planned behavior as the basis, takes the residents of community X in Nanchang as the survey object, uses the structural equation model analysis method in SPSS 26.0 software and combines SPSS data analysis with structural equation modeling validation to conduct a study on the current situation of residents’ household waste segregation and its influencing factors. It is hoped that this thesis can enrich the relevant research results in China to a certain extent and provide the theoretical basis and practical guidance for promoting the process of urban household waste segregation in China.

2. Materials and Research Methods

2.1. Study Subjects

In June 2019, the Ministry of Housing and Construction, together with a number of relevant departments, issued the Notice on the Household Waste Segregation in Cities at and above the prefecture level and decided that from 2019 onwards, cities at prefecture level and above across the country would be comprehensively promoted to carry out waste classification work, while waste classification and treatment systems in the 46 designated pilot cities would need to be essentially completed by the end of 2020. Nanchang is one of these 46 key cities, but so far, its household waste segregation and management is still in progress, facing common problems encountered by many cities. In addition, compared with the eight pilot cities, such as Beijing and Shanghai, Nanchang’s urban development level and the status quo of household waste classification and management are more similar to many other cities in China, especially for these 46 cities and other cities in China that are in the process of starting household waste segregation management; there are more common problems and measures to learn from. Therefore, this study is based on a simple random sampling method, using the residents of community X in Nanchang, Jiangxi Province, as the target population for sample selection.

2.2. Theoretical Basis and Research Hypothesis

2.2.1. Theory of Planned Behavior

The theory of planned behavior (TPB), introduced by Ajzen, is a theory in psychology that describes the relationship between individual attitudes and behavior by constructing a model theory to predict and explain human behavior in a variety of different situations. The theory suggests that behavioral outcomes are directly influenced by behavioral intentions, which are determined by a combination of three factors: perceived behavioral control, behavioral attitudes and subjective norms. Perceptual behavioral control reflects the conditions of control in the actual situation and therefore directly predicts the probability that the behavior will eventually occur, which in turn has a direct impact on the actual behavior. In addition, the theory of planned behavior suggests that individual characteristics, such as personality, intelligence, experience, age, gender, cultural background and employment status, may also influence behavioral attitudes, subjective norms and perceptual behavioral control, ultimately influencing behavioral intentions and leading to behavioral outcomes [6]. With its strong explanatory power, TPB has been widely used in many fields, such as psychology, education, information science and environmental science, and has also provided a proven theoretical framework for domestic and international scholars in the field of waste sorting. Taylor, a foreign scholar, has constructed a model of household waste management based on the theory of planned behavior [7], and Chen Jian, Liu Qiao and Shi Shi Ying, domestic scholars, have analyzed and argued the influencing factors of residents’ waste sorting behavior based on the theory of planned behavior [8,9,10]. However, intention does not always lead to behavioral results. Ajzen himself acknowledges that situational factors, opportunities such as convenience and resource factors are more likely to combine to determine waste segregation behavior than a sense of subjective control [11]. Prat and Somsak incorporate external environmental factors into the model of influencing factors, arguing that intrinsic subjective factors and external environmental factors determine the behavior of household waste classification together [12]. Some scholars, such as Tonglet, also suggest the addition of elements such as community concern, ethical norms and situational factors to the model to improve the explanatory power of the theory [13]. While these scholars have affirmed the theory of planned behavior in their empirical studies, they have also suggested that other factor variables should be incorporated into the explanatory model to enhance its explanatory power. However, it is undeniable that the theory of planned behavior still occupies an important position as the most fundamental theoretical basis and conceptual framework in the field of waste management. The theoretical model of planned behavior is shown in Figure 1.

2.2.2. Research Hypothesis and Conceptual Model

In this paper, through literature research, on the basis of the theory of planned behavior, some significant influencing factors are summarized and sorted into three major dimensions and seven major factors. Many scholars in domestic and international academia have already explored the influence of behavioral attitudes, subjective norms and perceptual behavioral control on residents’ willingness to separate household waste. For example, Wang Feng et al. introduced the qualitative comparative analysis fuzzy set method and, based on the theoretical interpretation framework of planned behavior proposed by Ajzen, explored the “joint effect” of three factors, such as environmental attitude, social trust and perceived behavior control, on the behavior strategy of household garbage classification [14]. Fernanda, based on the theory of planned behavior, used the partial least squares structural equation modeling approach to emphasize the significant impact of perceived cost and benefit factors on consumer participation in household waste sorting [15]. We will not continue to explore these works here, but we further explore the influential role of some specific factors under these dimensions.

(1)

Behavioral and attitudinal dimensions: recognition of waste separation/segregation, satisfaction with government work

Behavioral attitude is an individual’s stable and overall psychological tendency, consisting of the emotional reaction of liking or disliking a specific object. Here, it refers to the degree of liking or disliking, acceptance or rejection of the behavior of individual residents toward household waste segregation, and it encompasses the degree of personal approval of household waste segregation and satisfaction with the work of relevant government departments. The more residents understand the importance and value of waste segregation, the more likely they are to separate their waste [16]. At the same time, if the government departments are doing a good job in separating household waste, residents will be more satisfied with their work, and they will be more willing to separate household waste [17].

Therefore, we propose the hypothesis that

H1a. 

The recognition of waste segregation positively influences the willingness of urban community residents to separate their household waste.

H1b. 

Government job satisfaction positively influences residents’ willingness to separate household waste in urban communities.

(2)

Subjective normative dimensions: intrinsic moral constraints, legal and regulatory constraints, group behavioral incentives

Subjective norms are the social pressures that individuals feel when performing a specific act, which may come from other individuals or social groups, etc. This paper refers to the social moral requirements, legal and regulatory constraints and group behavioral influences felt by individual residents when carrying out the act of separating household waste. Residents’ sorting behavior is more likely to occur if they see it as a social responsibility and incorporate it into their own moral system [17]. In addition, the power of laws and regulations can act as an external constraint, greatly facilitating the realization of separate household waste recycling and disposal. Finally, as people are social in nature, their psychology and behavior are inevitably influenced by other members of society. When family members, friends or neighbors are willing to separate household waste and strictly follow the requirements, individuals will be influenced by them to change their mentality and behavior in order to maintain a good image or to save themselves from exclusion, and the behavior of the group will also provide a learning experience for individuals to follow.

Therefore, we propose the hypothesis that

H2a. 

Intrinsic moral constraints positively influence the willingness of urban community residents to separate their household waste.

H2b. 

Legal and regulatory constraints positively influence the willingness of urban community residents to separate household waste.

H2c. 

Group behavioral incentives positively influence the willingness of urban community residents to separate their household waste.

(3)

Perceptual behavioral control dimensions: temporal–spatial factors, waste segregation facilities

Perceptual behavioral control refers to how easy or difficult an individual perceives a behavior to be when performing it. In this case, it refers to the individual’s perception of the ease or difficulty of the act of sorting, including the perceived ease of time and space and the perceived convenience of waste sorting facilities. Residents who work long hours and have tight leisure time often choose to dispose of their household waste hastily due to lack of time and energy. Wan Azlina Wan Ab et al. suggest that the convenience and facilities provided by the government will have a direct impact on people’s motivation to separate and recycle their waste [18]. If government departments fail to provide appropriate waste segregation facilities, this may also weaken residents’ willingness to separate waste.

Therefore, we propose the hypothesis that

H3a. 

Spatial and temporal factors positively influence the willingness of urban community residents to separate their household waste.

H3b. 

The condition of waste segregation facilities positively influences the willingness of urban community residents to separate their household waste.

(4)

Socio-demographic variables

In practice, residents’ willingness to separate household waste is often the result of a combination of factors, involving a complex and diverse range of influences. The theory of planned behavior suggests that sociological demographic characteristics, such as gender, age and educational background, also play an important role in the behavioral intentions of individuals. Some scholars have also shown through their research that residents’ willingness to participate in household waste sorting is influenced by various demographic and sociological characteristics. However, because the variability brought about by residents’ demographic sociological attributes is reflected in the attitudes of each of the observed questions, we will not continue to explore them here.

Based on the above research hypotheses, we construct here a conceptual model of the factors influencing residents’ willingness to separate household waste in urban communities, as shown in Figure 2.

2.3. Research Methodology

2.3.1. Scale Design

There are no established scales that can be used directly in the field of waste management; therefore, each question in this questionnaire was set up with reference to previous scholars’ research scales as far as possible to enhance the quality of the questionnaire. The questionnaire consists of two parts: the first part is related to the sociological demographic characteristics of the respondents, and the second part is the Influencing Factors Scale, which is the main part of this questionnaire. The Influencing Factors Scale consists of 7 dimensions and 18 measurement questions. The questions are set up through a 5-point Likert scale design method and are quantified into scores of 1, 2, 3, 4 and 5 according to strongly disagree, disagree, uncertain, agree and strongly agree. Each respondent chose their answer according to their personal situation, and the score for each question reflected their attitude toward the issue, with the final total score indicating their overall attitude toward the content of the scale.

2.3.2. Data Collection

A total of 235 questionnaires were collected, and some incomplete and non-responsive questionnaires were excluded, resulting in 218 valid questionnaires, with a 92.8% return rate. It is generally accepted that the sample size should be 5 times the number of items, and to make the data more reliable, the sample size should be more than 10 times the number of items. As the scale for this study included 18 questions, it was desirable to maintain a sample size of 180 or more according to this principle, and a sample size of 218 was returned for this study, which was in line with the original sample size setting.

2.3.3. Data Analysis Methods

The data were processed using SPSS 25.0 and AMOS 24.0 software. (1) The questionnaire data were analyzed for basic sample characteristics and descriptive statistics and correlation analysis; (2) the questionnaire scale was tested for reliability; (3) the SEM was constructed, and model fit and path coefficient analysis were carried out.

3. Results and Analysis

3.1. Basic Sample Characteristics and Descriptive Statistical Analysis

3.1.1. Basic Sample Characteristics

This paper uses frequency and percentage to conduct descriptive statistical analysis of five sociological demographic characteristics of the survey respondents: gender, age, education, monthly household income and occupation. In terms of gender, 51.4% of the respondents were male, and 48.6% were female, with a relatively balanced ratio of men to women. In terms of age, the majority of respondents were in the 20–39 age group, accounting for 58.3% of the respondents, more than half of the sample size. In terms of educational background, the majority of respondents were college-educated or above, accounting for 66.1%, while 33.9% had less than high school education. In terms of monthly household income, the group with a monthly household income of CNY 5000–10,000 accounted for a relatively high proportion, while the number of people with a monthly income of CNY 20,000 or above was relatively small, accounting for 6.4%. The occupational background was dominated by student groups and enterprise workers, with the fewest being employees of government offices and public institutions.

3.1.2. Descriptive Statistical Analysis

In order to have a clearer understanding of the willingness of residents in community X of Nanchang to separate household waste and the scores of the influencing factors, we calculated the mean scores by means of a mean analysis in descriptive statistics and divided them by the number of items to obtain the mean of each item (

Table 1. Study variable scores.

Observed Variables	Average	Title Item	Average Score per Item
Waste segregation recognition	14.13	3	4.71
Government job satisfaction	7.59	2	3.795
Intrinsic moral constraints	12.61	3	4.20
Legal and regulatory constraints	7.71	2	3.855
Group behavioral incentives	13.13	3	4.38
Time/space factor	8.03	2	4.015
Waste segregation facilities	11.38	3	3.793
Willingness to separate waste	15.8	4	3.95

). The results show that the mean value of the survey respondents’ willingness to separate household waste is 3.95, which is close to the “agree” attitude, and the mean value of each dimension of the factors influencing the willingness to separate household waste is above 3.9.

3.2. Test of Reliability and Validity

The data recovered from the questionnaire scale must have reliability and validity to ensure that the results of the later data analysis have credibility. Therefore, we need to conduct a reliability and validity analysis of the scale.

Reliability of the scale was tested using reliability analysis in SPSS 25.0 software (

Table 2. Results of reliability analysis.

Dimensionality	Measurement Questions	Cronbach’s α Value Deleted from Item	Cronbach’s α Value
Waste segregation recognition	1. Segregation of household waste has a protective effect on land resources	0.867	0.871
	2. Segregation of household waste helps to recycle and reuse resources	0.868
	3. Segregation of household waste is of great value and relevance	0.867
Government job satisfaction	4. Your city has enacted a system or regulation for separating household waste	0.875	0.728
	5. The government department in your city attaches great importance to the work related to segregation of household waste	0.870
Intrinsic moral constraints	6. You feel very guilty if you do not separate your household waste as required	0.864	0.857
	7. Segregation of household waste can reflect individual social responsibility	0.865
	8. Segregation of household waste can reflect the quality of individual cultivation	0.866
Legal and regulatory constraints	9. Laws and regulations governing the segregation of household waste can regulate and guide your behavior	0.871	0.942
	10. Community codes of practice for sorting household waste can regulate and guide your behavior	0.870
Group behavioral incentives	11. Your family’s household waste sorting behavior affects your willingness to participate	0.863	0.933
	12. The behavior of friends or community members in separating household waste affects your willingness to participate	0.862
	13. Advocacy by the public media or relevant authorities can influence your willingness to participate	0.861
Time and space factors	14. Does it take too much time and energy to sort your waste?	0.868	0.831
	15. Does taking up too much space in your home affect your willingness to separate your household waste?	0.869
Waste segregation facilities	16. Your city has very good facilities for separating household waste, and it is very convenient to separate household waste	0.867	0.829
	17. Your community has a number of waste segregation facilities, so it is easy to separate your waste	0.869
	18. Your community’s waste segregation facilities are clearly marked	0.866
Willingness to separate waste	19. You have the knowledge needed to be able to accurately sort your household waste	0.848	0.867
	20. You have enough time and energy to separate your household waste	0.799
	21. You can provide sufficient space in your home for the segregation of household waste	0.816
	22. In your daily life, you are very willing to separate your household waste	0.848

), and the results showed that the Cronbach α value for each variable was greater than 0.7, indicating that the reliability of the questionnaire was good.

The results of the KMO and Bartlett’s test using factor analysis in SPSS25.0 software (

Table 3. KMO and Bartlett’s test results.

		Independent Variable	Dependent Variable
Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy		0.784	0.811
Bartlett’s sphericity test	Approximate cardinality	2564.493	426.661
	Freedom	153	6
	Significance	0.000	0.000

) showed that the KMO values of both the independent and dependent variables of the scale were greater than 0.6, and Bartlett’s sphere test corresponded to a significance of 0.000, which was less than 0.05, indicating that the scale data were suitable for further factor analysis. We then carried out an exploratory factor analysis using factor orthogonal rotation (

Table 4. Factor loading matrix.

		Factor
		1	2	3	4	5	6	7	8
Waste segregation recognition	Q1		0.855
	Q2		0.849
	Q3		0.765
Government job satisfaction	Q4							0.876
	Q5							0.832
Intrinsic moral constraints	Q6			0.735
	Q7			0.851
	Q8			0.897
Legal and regulatory constraints	Q9					0.935
	Q10					0.934
Group code of conduct	Q11	0.888
	Q12	0.888
	Q13	0.823
Time/space factor	Q14						0.876
	Q15						0.874
Waste segregation facilities	Q16				0.738
	Q17				0.886
	Q18				0.833
Willingness to separate waste	Q19								0.814
	Q20								0.889
	Q21								0.867
	Q22								0.816

), and the scale was divided into eight dimensions according to the aggregation of each question item after factor orthogonal rotation, which was consistent with the initial division of dimensions in this study, and the factor loadings of each dimension were all greater than 0.5, indicating that the scale had good validity.

3.3. Correlation Analysis

In this paper, the Pearson analysis in SPSS 25.0 software was used to measure the correlation between all variables in the model (

Table 5. Correlation analysis between variables.

Variables	Waste Separation Recognition	Government Job Satisfaction	Intrinsic Moral Constraints	Legal and Regulatory Constraints	Group Behavioral Incentives	Time/Space Factor	Waste Separation Facilities	Willingness to Separate Waste
Waste segregation recognition	1
Government job satisfaction	0.210 **	1
Intrinsic moral constraints	0.503 **	0.258 **	1
Legal and regulatory constraints	0.177 **	0.267 **	0.164 *	1
Group behavioral incentives	0.488 **	0.243 **	0.405 **	0.433 **	1
Time/space factor	0.359 **	0.110	0.333 **	0.120	0.317 **	1
Waste segregation facilities	0.322 **	0.278 **	0.358 **	0.147 *	0.276 **	0.396 **	1
Willingness to separate waste	0.479 **	0.184 **	0.484 **	0.232 **	0.470 **	0.511 **	0.496 **	1

Note: * indicates significant correlation at 0.05 level; ** indicates significant correlation at 0.01 level.

). The results show that the p-values of the seven independent variables and the dependent variable are all less than 0.01, and the coefficients are greater than 0, indicating a significant positive correlation between the two sides. Waste segregation recognition is weakly positively correlated with government job satisfaction and legal and regulatory constraints and moderately positively correlated with group behavioral incentives, time/space factor, waste segregation facilities and willingness to separate waste and strongly positively correlated with intrinsic moral constraints. Government job satisfaction is weakly positively correlated with the other seven variables. Intrinsic moral constraints are weakly positively correlated with legal and regulatory constraints and moderately positively correlated with group behavioral incentives, time/space factor, waste segregation facilities and willingness to separate waste. Legal and regulatory constraints are moderately positively correlated with group behavioral incentives and weakly positively correlated with the time/space factor, waste segregation facilities and willingness to separate waste. Group behavioral incentives are weakly positively correlated with waste segregation facilities and moderately positively correlated with the time/space factor and willingness to separate waste. The time/space factor is moderately positively correlated with waste segregation facilities and strongly positively correlated with willingness to separate waste. Waste segregation facilities are moderately positively correlated with willingness to separate waste, which is consistent with the hypothesis. Hypotheses H1a, H1b, H2a, H2b, H2c, H3a and H3b are all tentatively supported, but the causal relationship between the variables cannot be accounted for yet. The next step in this paper is to explore the causal relationships that the variables have with each other by constructing structural equation models.

3.4. Structural Equation Model Testing

3.4.1. Initial Structural Equation Model

In this paper, we used the AMOS 24.0 software tool to construct the initial structural equation model I for the study of the influencing factors of household waste segregation of urban community residents, and the data were imported into the hypothetical path diagram to derive the standardized path coefficient, as shown in Figure 3.

(1)

Model goodness-of-fit test

The maximum likelihood estimation method was chosen to test the fit of model I. The degree of fit of the model to the sample data was tested by comparing the values of the seven main indicators with the reference values (

Table 6. Table of parameters for the overall fit of model I.

Fitted Indicators	CMIN/DF	CFI	NFI	IFI	GFI	RMR	RMSEA
Fitting criteria	<3	>0.9	>0.9	>0.9	>0.9	<0.05	<0.08
Indicator values	2.056	0.938	0.887	0.939	0.866	0.061	0.070
Fitting results	Conformity	Conformity	General conformity	Conformity	General conformity	General conformity	Conformity

). As can be seen in the table, the chi-squared freedom ratio (CMIN/DF) of the model in this study is 2.056, which is less than 3, indicating that the model has parsimonious fitness; the RMSEA is 0.070, which is less than 0.08, indicating a reasonable fit. Although the NFI value of the model is 0.887, and the GFI value is 0.866, both less than 0.9, and the RMR is 0.061, greater than 0.05, not reaching the reference value, the difference is not significant, and the CFI value is 0.938, and the IFI value is 0.939, both greater than 0.9, indicating that the model fits relatively well, and the model is acceptable and needs to be further revised.

(2)

Path coefficient test

According to the path diagram I of the influencing factors of waste segregation of urban community residents, we can analyze the path coefficients of the independent and dependent variables. As can be seen in

Table 7. Table of the impact factor path coefficients of model I.

Paths	Unstd. Estimate	Std. Estimate	S.E.	C.R.	p	Is It Established
Willingness to separate waste ←Waste segregation approval	0.246	0.161	0.125	1.972	0.049	Yes
Willingness to separate garbage←Government job satisfaction	0.021	0.044	0.022	0.962	0.336	No
Willingness to separate waste←Intrinsic moral restraint	0.172	0.184	0.061	1.271	0.024	Yes
Willingness to separate waste←Constraints of laws and regulations	0.026	0.038	0.042	0.614	0.539	No
Willingness to sort waste←Group behavioral incentives	0.156	0.170	0.070	2.230	0.026	Yes
Willingness to separate waste←Spatial and temporal factors	0.227	0.313	0.057	3.970	***	Yes
Willingness to separate waste←Waste segregation facilities	0.214	0.272	0.059	3.611	***	Yes

Note: *** indicates significant correlation at 0.001.

, the standardized path coefficients between waste segregation recognition, intrinsic moral constraints, group behavioral incentives, temporal and spatial factors, waste segregation facilities and waste segregation intention are 0.161, 0.184, 0.170, 0.313 and 0.272, respectively, and the p-values are all less than 0.05, indicating that these five factors positively influence urban community residents’ willingness to separate their household waste, and the influence is significant. Hypotheses H1a, H2a, H2c, H3a and H3b are supported. The standardized path coefficients between government job satisfaction, legal and regulatory constraints and willingness to separate rubbish are 0.044 and 0.038, with p-values greater than 0.05, indicating that these two factors positively influence urban community residents’ willingness to separate rubbish but are not significant; therefore, the hypotheses are rejected, and Hypotheses H1b and H2b are not valid.

3.4.2. Modified Structural Equation Model

According to the results of the above analysis, the two independent variables of government job satisfaction and legal and regulatory constraints have a low and poorly significant effect on the dependent variable of willingness to separate rubbish, with no socio-statistical characteristics. The former may be due to the fact that residents have inherent prejudices and believe that the main role of government departments in the introduction and implementation of waste management policies, publicity and supporting facilities is their necessary obligation and responsibility, while their attitude toward the work of government departments and whether they are satisfied or not does not stimulate their own willingness and behavior to separate waste. The latter may be due to the fact that in the field of waste management, most initiatives of government departments are still at the stage of advocacy and encouragement, and the laws and regulations introduced have not yet formed a compulsory binding force on residents, which does not have a significant impact on residents’ own willingness to participate in household waste segregation. Therefore, by synthesizing the above model analysis theoretically and practically, the path between government job satisfaction and willingness to classify rubbish/legal and regulatory constraints and willingness to classify rubbish is considered to be removed, and the path diagram II of the influencing factors of domestic rubbish classification of urban community residents is obtained, as shown in Figure 4.

(1)

Model goodness-of-fit test

Model II of the corrected influencing factors was tested again for goodness of fit, and the results are presented in

Table 8. Table of parameters for the overall fit of model II.

Fitted Indicators	CMIN/DF	CFI	NFI	IFI	GFI	RMR	RMSEA
Fitting criteria	<3	>0.9	>0.9	>0.9	>0.9	<0.05	<0.08
Indicator values	2.034	0.951	0.909	0.952	0.888	0.047	0.069
Fitting results	Conformity	Conformity	Conformity	Conformity	General conformity	Conformity	Conformity

. Comparing the fit indices of the model before and after the adjustment, we can see that the CFI is 0.951, the NFI is 0.909, and the IFI is 0.952, and the fit indices are better than those before the adjustment, all of which are greater than 0.90. The fit effect is obviously improved, and this shows that the secondary constructed model II of the influencing factors of household waste classification of urban community residents fits very well and has strong explanatory power.

(2)

Path coefficient test

According to the results of the standardized coefficients of all paths in the revised model II, the original research hypotheses H1a, H2a, H2c, H3a and H3b are still supported, with the temporal–spatial factor (β = 0.309) having the greatest influence on the intention of urban community residents to separate household waste, followed again by waste segregation facilities (β = 0.279), closely followed by intrinsic moral constraints (β = 0.242), group behavioral incentives (β = 0.185) and waste segregation identification (β = 0.158). As shown in

Table 9. Table of the impact path coefficients of model II.

Paths	Unstd. Estimate	Std. Estimate	S.E.	C.R.	p	Is It Established
Willingness to separate waste←Waste segregation approval	0.243	0.158	0.125	1.941	0.047	Yes
Willingness to separate waste ←Intrinsic moral restraint	0.196	0.242	0.082	2.216	0.024	Yes
Willingness to sort waste←Group behavioral incentives	0.169	0.185	0.064	2.664	0.008	Yes
Willingness to separate waste ←Spatial and temporal factors	0.227	0.309	0.058	3.938	***	Yes
Willingness to separate waste ←Waste segregation facilities	0.216	0.279	0.057	3.784	***	Yes

Note: *** indicates significant correlation at 0.001.

.

4. Conclusions and Policy Implications

4.1. Conclusions

This paper conducted an SPSS empirical analysis and AMOS structural equation modeling test on 218 scales of residents’ willingness to separate household waste and their influencing factors in community X, Nanchang, Jiangxi Province. After the theoretical analysis and empirical study, the following research conclusions were drawn:

(1)

With regard to the current situation of urban community residents’ willingness to participate in household waste segregation, we found that the mean score of the respondents’ household waste segregation intention is high (3.95), close to the “agree” attitude. If we disregard the actual occurrence of household waste segregation behavior, urban community residents are generally willing to separate household waste in terms of subjective willingness.

(2)

The five factors of waste segregation recognition, intrinsic moral constraints, group behavioral incentives, time and space factors and waste segregation facilities positively influence residents’ willingness to separate household waste in urban communities. With a high degree of recognition, strong moral binding force, strong group incentive, abundant time and perfect space and facilities, residents have stronger willingness to classify garbage.

(3)

There is no significant effect of government job satisfaction and legal and regulatory constraints on residents’ willingness to separate household waste in urban communities, which do not reach a statistically significant level. This may be due to the cognitive bias of residents, who believe that government departments have the main obligation and responsibility in the area of waste management and do not take the initiative to understand the work of government departments and that the effectiveness of their work and whether they are satisfied or not do not stimulate their own sense of participation and do not stimulate their own willingness and behavior in waste segregation. In addition, most government initiatives are still in the advocacy and encouragement stage, and the laws and regulations in place are not yet binding on residents; therefore, there is an urgent need to improve the power of the legal system.

In the construction of the influence factor model, despite Ajzen’s planned behavior theory and the theories of other scholars, there are some important influencing factors that are not included, such as personal economic status, environmental awareness and knowledge of garbage classification. In addition, the present study uses cross-sectional data for empirical analysis, which can only reflect the situation at a certain point in time and cannot reveal the changing trend of residents’ household waste segregation intention over time. Therefore, in future studies, it is necessary to further improve the variable system, and longitudinal data can be considered to continuously follow up the investigation of residents’ household waste segregation willingness, understand its change process and provide scientific basis for the formulation of more effective garbage classification policies.

4.2. Policy Insights

(1)

To strengthen public education and publicity and to combine all channels and resources as far as possible to help residents correctly understand household waste sorting, the sorting system and the sorting standards, so that they can truly understand why “waste is a misplaced resource”, thus increasing their acceptance of household waste sorting behavior and internalizing it to guide them to carry out waste segregation.

(2)

To enhance group behavioral incentives, focusing on creating small role models around us, publicizing typical examples of waste segregation and commending advanced individuals and communities to form a social demonstration force and create a good atmosphere of household waste segregation. To cultivate new-age moral qualities and civic responsibilities and to internalize the awareness and behavior of household waste segregation into citizens’ intrinsic sense of moral responsibility.

(3)

To improve the supporting infrastructure, reasonably allocate rubbish sorting equipment and scientifically plan the throwing distance to facilitate residents to sort and put out their rubbish as much as possible and to save time for sorting. To guide enterprises to optimize production and advocate households to set up simple household waste segregation bins to save space for sorting.

(4)

To standardize and improve laws and regulations, to do a good job in top-level design, to clarify the standards of household waste classification and to introduce a special waste management system, so as to standardize the system and unify the standards. To scientifically divide the subjects responsible for rubbish classification, establish a permanent supervision mechanism, urge the subjects to consciously fulfill their responsibilities and obligations and ensure the unity of rights and responsibilities.