Research on the Factors Influencing the Perception of Urban Park Recreational Behavior Based on the “Homo Urbanicus” Theory

Abstract

The improvement of the quality of urban parks plays a vital role in the construction and development of urban society. This study examined the factors influencing recreational satisfaction with urban parks, providing a reference for measures to improve the quality of urban parks. Based on the theory of “Homo Urbanicus”, we conducted a factor analysis of recreation satisfaction with urban parks and constructed a structural equation model. It can be seen from the analysis results produced by the causal model that “supporting facilities” and “functional facilities” have the highest contributions to the factor load of the model, which indicates that individual recreational satisfaction with urban parks is mainly based on the relationship between individuals and space conditions. In other words, these elements are necessary to meet the requirement of the “physical” in the “Homo Urbanicus” theory and to supplement the “physical–group–rational” method structure in the “Homo Urbanicus” theory. This conclusion indicates that the improvement of the “functional facilities” and “supporting facilities” of urban parks will play an important role in improving recreational satisfaction with urban parks. The results of this study can help researchers to build an urban park quality evaluation system from the perspective of recreation quality.

1. Introduction

The Third United Nations Conference on Housing and Urban Sustainable Development, which was held in 2016, clarified the relationship between urban life quality and individual behavior perception. In February 2018, the Chinese government put forward the concept of the “park city” [1], which clarified that the person is at the core of a city, and the sustainable development of a city should be person-oriented. Therefore, Chinese urban parks and “person-oriented” cities fall into a research context that is related to actual developmental needs in Chinese national conditions.

As an urban green infrastructure, green space in parks plays an important role in the maintenance of the urban natural structure and ecological sustainability. In terms of scale, China’s urban park construction has led to great achievements in recent years, but the recreational resources in urban parks are limited in terms of their detail and quality, mainly manifested in the following: Firstly, the social experience mechanism of urban parks is not perfect. Recreational activities in urban parks are mostly based on communication between families and friends, while social communication between different groups is less significant. Social interaction in urban parks is, to some extent, important for promoting social cohesion [2]. Moreover, the characteristics of urban park recreational experience are not well-defined. Most urban parks are affected by the phenomenon of “one side of a thousand cities”, and urban parks lack a unique label in terms of recreational experience [3]. Finally, there are problems related to regional identification and regional culture. Many urban parks do not play the role of regional cultural heritage, which leads to bias regarding the recognition of urban areas in which urban parks are located and can easily cause regional recognition obstacles [4,5].

According to the above three points, the problems affecting urban parks in terms of recreational social experience, recreational characteristics, and regional recognition are mainly reflected in three aspects: recreational satisfaction, place dependence, and recreational motivation in urban parks. Therefore, the analysis of these various elements, recreation satisfaction, place dependence, and recreation motivation, will help us to further explore the experience mechanism of urban park recreation and study the relationships between various elements in the structure of urban park recreational experience, which is also essential for improving the quality of urban parks.

With the continuous development and improvement of cities, the quality of urban parks has become a comprehensive symbol of the developmental level of a city, and research on recreation in urban parks has become an important part of urban research [6,7]. Recreational behavior has the characteristics of general dispersion and randomness in terms of spatial distribution [8]. William E. Hammitt followed the argument of Cyrus W. Young that recreational behavior in urban parks is actually a local preference generated by the coupling of behavior and place [9]. Changes in space, time, and place elements can follow changes in the subject’s perception and experience of recreational behavior, a notion which is also the objective basis for the study of the subject’s perception and experience of entertainment. Later, this perception difference was also explained according to cultural cognitive structure and objective event factors [10,11,12]. The study of recreational behavior is mainly reflected in the relationship of the “spatial individual” with the “subject structure”. This is largely reflected in the research on recreational behavior satisfaction, preference, or motivation and behavior-related decision-making mechanisms [13,14]. Therefore, recreational behavior can also be interpreted as a whole-system mechanism developed from a single element [15].

Alan D. Bright argued that the key factor affecting recreational behavior experience is the importance of the individual and advocated for the establishment of the absolute evaluative status of the individual’s subjective perception in environmental experience [16] so as to design external conditions that could change their behavioral choices according to their subjective incentive needs [17,18]. Joan Vitesso was influenced by Alan D. Bright. Using place attachment to study the emotional feedback system of recreational behavior, Joan provided an improved model of human–land emotional connection between objective entertainment behavior and subjective experience [19]. This kind of human–land emotional connection mainly reflects recreational behavior expressed through the environmental values of the subproject [20,21]. Daniel R. Williams argued that the experience generated by recreational behavior has a significant impact on local attachment sensitivity [22]. Based on Bright and Joan Vitesso, William E. divided Bright’s personal subjective perception into four dimensions, namely familiarity, belonging, identity, and dependence, through the structural equation model, in which identity and familiarity have significant impacts on dependence and belonging [23].

The study of recreational behavior in space is mainly reflected in the feedback structure relationship between subjective perception and objective factors [24]. Joan Vitesso argued that this feedback structure relationship is mainly manifested in a type of recreational behavior emotional feedback called “local attachment” [25]. In addition, the study of recreational behavior includes cultural and service perception of recreational behavior [26], psychological perception and behavior [27], subjective wellbeing [28], the stage value of recreational behavior [29], the cultural background and physical and mental performance of recreational behavior subjects [30], the narrative nature of the spatial structure, and many other behavioral perception indicators that are difficult to quantify [31]. There are also multi-level and multi-structural differences in the research objects of recreational behavior, mainly manifested in the subjective factors of causal structures such as individual occupation, gender, age, race, and recreational behavior [32,33,34,35,36,37,38,39].

Improving the recreational quality of parks is key to improving the quality of urban parks. Cheng Siya used the method of importance–performance analysis to study recreational satisfaction with 50 urban parks in Beijing [40] and found that the park space facilities, activity types, crowd characteristics, and other factors have different degrees of influence on recreational activities in urban parks [41,42]. This conclusion was confirmed in the research on urban parks conducted by Huang Songyao and Liang Huilin using social media data and survey data. In the study of urban parks, Heo Seulkee pointed out the importance of the relationship between the convenience experienced by individuals in life, social economy, age, gender, and other indicators, as well as access to park green spaces for recreation opportunities [43]. Dai Daixin argued that in urban parks, a variety of sports facilities and more inclusive design, planning, and management strategies are conducive to promoting recreational participation among young people [44]. In addition, Shams Khadija argued that the physiological perception of individuals has an impact on their psychological perception, which is mainly reflected in the happiness brought about by health satisfaction in the case of urban parks [45]. This sense of happiness helps individuals to further strengthen their local attachment to urban parks. This kind of local attachment also enables psychological repair for the individual [46]. Liu Qunyue found that higher familiarity can help to improve the local identity of the recreational person and enhance the individual’s location connection [47]. In fact, the main idea of recreational behavior research is to take the person as the main body among the elements of environmental space and study the influences of individual behavior and space through the relationship between elements in the structure.

The “Homo Urbanicus” theory was first proposed by Mr. Hok Lin Leung (2012). According to the theoretical point of view expressed in Mr. Hok-Lin Leung’s book, Old Concepts and New Environment, “Homo Urbanicus” is defined as a “person who pursue space contact through rational choice of settlement” [48]. It is a unified urban research methodology recognized by the academic community and is also a relatively new methodology. Therefore, it has a strong practical application and reference value for Chinese cities with a generally rapid development of urbanization. The “Homo Urbanicus” theory is a methodological theory of urban planning and urban design with the measurement of the noumenon value of urban person as the core. The “Homo Urbanicus” theory focuses on the value measurement of “rational choice and settlement”, holding that man is a rational animal [49]. “Rationality” is a subjective control factor; that is, the method of “pursuing the balance of self-existence/coexistence with minimum effort to achieve the optimization of space opportunities” [50]. “Inhabitation” (representing the space–time element structure of individuals) is an objective factor, and the subjective and objective systems jointly affect the measurement of “value” content. The “Homo Urbanicus” theory holds that the principle of measuring all problems in cities must be “person-oriented” [51,52].

The “Homo Urbanicus” theory recognizes the unity of the monism of “rationality” and “coexistence/self-existence” in terms of value, emphasizes the “human nature”; “physical nature”, and “rationality” that individuals experience in places in theory; and advocates the adoption of “person”, “event”, “opportunity”, and “space” as the four dimensional elements of the structure of space research [53]. In this study, we aimed to apply the “Homo Urbanicus” theory to explore the following two questions: What is the relationship between the elements in the structure of urban park recreation experience? How can the quality of urban parks be improved in terms of recreation experience? This study took city parks as a carrier and combined the factors of recreation behavior in city parks with the “city people” theory to obtain the structure of four-dimensional elements of recreation behavior from three aspects: social experience, recreation experience, and place identification. Factor analysis was conducted to obtain the influence of spatial condition elements on individual spatial perception and then analyze the influence of urban parks on individual spatial perception through structural equation modeling from the four-dimensional elements of the “Homo Urbanicus” theory (Figure 1).

2. Materials and Methods

2.1. Research Methods

According to the four-dimensional structure of the “Homo Urbanicus” theory, the recreational behavior of the participants, the process of the occurrence of recreational behavior, the opportunity to engage in recreational behavior, and the spatial carrier of recreational behavior in the balance of the relationship between self-existence and coexistence correspond to the “participants or decision-makers in the balance of the relationship between self-existence and coexistence”, “spatial contact behaviors such as going to work and commuting”, and the “timing of decision-making”. The four-dimensional structural relationship of the “spatial scope of decision-making events” between “person”, “events”, “opportunity”, and “space” actually comprises the four basic dimensions of recreational contact between the person and spatial environment.

Deconstructing this structure, we can obtain three individual identities, namely recreation subject participants (purposeful participants), collaborative participants, and continuous participants, from the “recreational behavior participants in the balance between self-existence and coexistence”. From the “recreational behavior occurrence process”, we can obtain the four basic elements of “time of occurrence”, “place of occurrence”, “experience results”, and “preference feedback”, which correspond to the experience and feedback obtained according to the time sequence of events and individual behavior. From the “space carrier of recreational behavior”, we can obtain the five basic elements of “landscape”, “convenience”, “regularity”, “facility”, and “place”. From the dimension of “recreational behavior opportunity”, we can obtain the three basic elements of “spatial perception”, “perception of relationships with others”, “self-behavior perception”, and “recreational behavior opportunities”, which correspond to the triple relationship between “thing–I” (self in space), “other–I”, and “self–I” (Figure 2).

The causal model analysis of the structural equation model requires dependent variables and independent variables; thus, it is necessary to deconstruct the causal logic of the occurrence of events for the dimensions of “person”, “event”, “opportunity”, and “space” (Figure 3).

Based on the structural elements corresponding to the four dimensions of “person”, “event”, “opportunity”, and “space”, in the study, we designed a questionnaire to obtain three levels of scale information: the basic personal information of participants in the sample, the sample evaluation scale of park space resources, and the spatial perception evaluation of the sample (Appendix A).

In this study, a structural equation model (SEM) [54,55,56] was used to investigate the four-dimensional structural elements of the “urban human” theory of urban park recreation behavior and to explore the relationship between the spatial elements of urban parks and the spatial perception of individual recreation behavior. We obtained the basic data based on the questionnaire survey and imported the data into SPSS 23.0 software for exploratory factor analysis (EFA) and conducted KOM and Bartlett’s sphericity tests: KMO values > 0.9 indicate very suitable, 0.8 indicates suitable, 0.7 is suitable, 0.6 is not very suitable, and 0.5 or less indicates extremely unsuitable [56,57]; Bartlett’s sphericity test was considered significant at the level of p < 0.001. Factor models were constructed separately for both spatial elements of urban parks and individual spatial recreation perceptions, the factor models that met a certain explanation ratio were screened according to the total variance explanation, and after completing the Cronbach’s alpha coefficient reliability test of the factor models, the AMOS software was used to construct structural equation models based on the results of the factor model data from SPSS 23.0. The theoretical model generated by AMOS was imported into the sav data of SPSS 23.0 software, and the RMR and GFI models were tested for goodness of fit to generate the causal full model, which is the structure of the influence of spatial elements of urban parks on the spatial perception of individual recreation behavior.

2.2. Data Collection

According to the basic information obtained through the questionnaire, as shown in Appendix A and Appendix B, the questionnaire was created, and the field survey was conducted.

The survey lasted from 3 July 2021 to 10 March 2022, and lasted for nine months. During this time, there were three special periods of epidemic prevention and control in Zhengzhou City and 20 devastating floods in July. A total of 68 people were included in two groups. Nineteen parks within the Fourth Ring Road of Zhengzhou were selected for the field distribution of the questionnaire. The questionnaire was distributed among the groups from 17:00 p.m. to 18:00 p.m. (spring and autumn), from 16:00 p.m. to 17:40 p.m. (winter), and from 17:30 p.m. to 19:00 p.m. (summer). The questionnaire was distributed, filled in, and returned on site. A total of 2626 questionnaires were distributed, and 2626 were returned, with a recovery rate of 100% (

Table 1. Sample description statistics.

Item	Class	Frequency	Rate (%)	Item	Class	Frequency	Rate (%)
Gender	Male	1201	45.7	Political outlook	Party member	719	27.4
	Female	1425	54.3		The masses	1907	72.6
	total	2626	100		total	2626	100
Park type	Street center park	527	20.0	Degree of education	High school	353	13.4
	City square	955	36.3		Junior college	913	34.8
	Urban landscape walkway	187	7.1		Undergraduate	1198	45.6
	Large green park	957	36.4		Master	145	5.5
					Doctor	17	0.6
	Total	2626	100		Total	2626	100
Marital status	Unmarried	524	20	Residence	Native	1231	46.9
	Married	2102	80		Non-resident	630	24
					Foreigner who has settled	765	29.1
	Total	2626	100		Total	2626	100
Family economic status	Very difficult	46	1.8	Monthly income level	No income	114	4.3
	Fair enough	1130	43		Below 3000	534	20.3
	Basically well-off	1376	524		3000–6000	1676	63.8
	Rich	71	2.7		6000–10,000	231	8.8
	Very rich	3	0.1		10,000+	71	2.7
	Total	2626	100		Total	2626	100
Age	Under 30	486	18.5	Number of family members in the community	Single	254	9.7
	30–40	698	26.2		Couple	389	14.8
	40–50	620	23.6		Family of three	869	33.1
	50–60	468	17.8		A large family	818	31.2
	Over 60	363	13.8		Multi generation family	296	11.3
	total	2626	100		Total	2626	100

).

According to the descriptive statistics of the sample, the gender distribution of the sample tended to be balanced, with slightly more women than men, and the number of the masses far exceeded the number of party members in terms of the political situation, reaching 72.6%. Among the four types of urban parks, the sample size of large-scale green space parks was the largest, with 957 samples, accounting for 36.4%, while the percentage of undergraduates was the highest (45.6%) in terms of the educational level. In terms of the subjective definition of family economic status, the residents tended to be “passable” and “basically well-off”. In terms of the “age” variable, the numerical difference in the sample size at each interval was not large, and the participants’ ages were mainly concentrated in the ranges of 30 to 40 and 40 to 50. As for the number of family members, three-member families accounted for 33.1% of the total population in the first instance, indicating that this is still the dominant pattern of family size in China’s urban areas. It is also worth noting that the sample of “many families in one family” closely followed, with a sample size of 31.2%; this shows that the “two-child policy” has gradually been affecting the size of urban families. It is worth mentioning that on the “living situation” scale, “local residents” accounted for only 46.9%, while the “non-resident” category comprised more than 50% of the total sample, reaching 53.1%, indicating that Zhengzhou is a city open to the migrant population. In this context of space contact, outsiders need to be more considered more inclusively in regard to city park humanistic care, which is an important factor in the design of urban public spaces.

We carried out research on the four-dimensional structure of the “Homo Urbanicus” theory; sorted and screened the variables, including “health and safety”, “function and use”, “park space elements”, “park landscape elements”, “human needs of park activities”, “park vitality and openness”, “heterogeneity of use objects”, “park supporting facilities”, and other aspects; and designed a Likert scale. The evaluation scale (X1~X30) and the evaluation (Y1~Y22) of the perception of park space resources were determined with 52 variables in total (

Table A1. Distribution of survey results for the evaluation scale of park space resources.

Proportion of 1 to 5 Points
Question	1 Point	2 Points	3 Points	4 Points	5 Points
X1. What is your assessment of the scale of green plants in the park?	0.2%	1.4%	51.8%	44.1%	2.4%
X2. Do you think there are many kinds of green plants in the park?	0.0%	2.1%	51.8%	44.6%	1.5%
X3. What do you think of the level of plant pruning and maintenance in the park?	0.0%	3.8%	44.5%	48.7%	3.0%
X4. How satisfied are you with the selection of plants in the park?	0.0%	3.9%	47.1%	45.5%	3.4%
X5. What do you think of the coverage of trees in the park?	0.1%	3.2%	58.7%	34.5%	3.5%
X6. Do you think the transportation to the park is convenient?	0.2%	3.2%	38.7%	42.8%	15.2%
X7. Do you think it is convenient to consume food in the park?	1.0%	7.8%	38.1%	47.3%	5.9%
X8. How do you think the park is doing in terms of vehicle management?	0.6%	6.4%	44.4%	45.8%	2.8%
X9. Do you think the park is far from your home?	0.8%	7.1%	38.2%	46.5%	7.4%
X10. What do you think of the rainwater accumulation in the park?	0.2%	6.7%	37.6%	50.4%	5.1%
X11. What do you think of the overall design of the park?	0.0%	23.5%	23.5%	71.2%	4.2%
X12. What do you think of the footpath planning in the park?	0.0%	1.6%	28.6%	67.7%	2.1%
X13. What do you think of the parking space in the park?	2.1%	25.2%	38.2%	33.0%	1.6%
X14. What do you think of the public health of the park?	0.2%	2.4%	48.4%	46.0%	3.0%
X15. What do you think of the monitoring quality of the park?	0.9%	9.8%	59.5%	27.0%	2.8%
X16. Do you think the park can be used for seating?	0.3%	13.1%	47.7%	35.0%	3.9%
X17. What do you think of the fitness facilities in the park?	1.7%	12.4%	43.1%	39.8%	3.1%
X18. What do you think of the rain shelter in the park?	2.4%	15.1%	52.1%	27.7%	2.7%
X19. Do you think the square area of the park is sufficient?	0.4%	6.5%	47.1%	41.1%	4.9%
X20. What do you think of the parent–child amusement facilities in the park?	1.8%	11.6%	41.9%	41.0%	3.8%
X21. Do you think the illumination of the park at night is sufficient?	0.7%	8.0%	55.2%	32.1%	4.0%
X22. Do you think the elderly in the park are safe?	0.2%	2.8%	43.0%	48.7%	5.3%
X23. Do you think the children engaged in activities in the park are safe?	0.2%	3.1%	42.1%	50.1%	4.5%
X24. Do you think there are many kinds of personal activities in the park?	0.3%	3.8%	49.1%	42.4%	4.5%
X25. Do you think there is enough space for children’s activities in the park?	0.2%	4.2%	40.4%	51.2%	4.0%
X26. Please comment on the number of public hand basins in the park.	0.4%	5.0%	42.2%	48.4%	4.0%
X27. Please comment on the number of public toilets in the park.	0.2%	4.2%	32.1%	58.7%	4.8%
X28. Please comment on the number of trash cans in the park.	0.2%	2.1%	29.7%	50.0%	18.0%
X29. Do you think the park is usually lively?	0.1%	0.7%	33.9%	60.4%	4.9%
X30. Do you think you can buy food and drink in the park?	1.6%	3.7%	33.2%	51.6%	9.8%

and

Table A2. Distribution of survey results for the park space perception evaluation scale.

Proportion of 1 to 5 Points
Question	1 Point	2 Points	3 Points	4 Points	5 Points
Y1. How long do you stay in the park on average?	2.4%	29.1%	45.4%	18.9%	4.1%
Y2. Who will accompany you to the park?	5.8%	22.1%	35.9%	25.9%	10.2%
Y3. Do you think the park has made you gain knowledge?	1.2%	6.7%	57.2%	32.8%	2.1%
Y4. Do you think you have gained confidence in the park?	1.0%	6.1%	62.7%	28.3%	1.8%
Y5. How is your willingness to communicate with others in the park?	0.9%	7.4%	56.8%	32.2%	2.7%
Y6. Do you think the park is indispensable in your life?	0.8%	4.4%	48.4%	41.2%	5.2%
Y7. Are you willing to help others in the park?	0.8%	2.7%	12.4%	79.9%	4.2%
Y8. Do you expect to get help from others in the park?	1.0%	4.0%	27.7%	64.3%	3.0%
Y9. Do you have the intention to carry out plant education for children in the park?	1.3%	4.1%	47.8%	42.3%	4.5%
Y10. Do you think you are a part of nature in the park?	0.4%	4.7%	51.5%	37.1%	6.3%
Y11. Do you think the park has made you healthier?	0.2%	4.6%	57.6%	32.5%	5.1%
Y12. Do you think the children in the park are active?	0.1%	2.9%	48.5%	43.8%	4.7%
Y13. Do you want others in the park to notice you?	1.5%	10.1%	52.7%	34.4%	1.2%
Y14. What kind of public impression do you want to leave in the park?	0.6%	9.9%	37.8%	48.6%	3.0%
Y15. What do you think of the collective activity culture in the park?	1.8%	3.2%	39.6%	53.1%	2.4%
Y16. What is your willingness to participate in the collective activities of the park?	0.8%	4.4%	43.7%	48.1%	3.0%
Y17. Do you think the park can be used to make friends?	1.6%	2.6%	40.3%	51.9%	3.6%
Y18. Does the park provide you with more personal interests?	0.7%	13.0%	30.5%	51.4%	4.3%
Y19. What do you think of the overall historical and cultural nature of the park?	1.1%	13.8%	14.0%	66.3%	4.8%
Y20. How do you rate the management level of the park?	0.1%	1.1%	10.7%	75.6%	12.5%
Y21. Do you think you will be threatened by strangers in the park?	0.2%	1.3%	16.1%	49.9%	32.6%
Y22. Do you think you belong to Zhengzhou in this park?	0.4%	2.4%	13.2%	58.0%	26.0%

in Appendix A).

According to the answers to the questionnaire, based on the evaluation scale X1–X30 of the park’s space resources, the highest proportion of the five points included in the X28 “evaluation of the number of garbage cans in the park” was 18.0%, which reflected a general recognition of the park’s space sanitation management level. Secondly, X6, “Do you think the traffic to the park is convenient?”, reached 15.2%, indicating that the current traffic accessibility of the city is satisfactory, and people are generally satisfied with the distribution mode of the park in the traffic road grid. With the lowest score of 1 point, the highest-rating answer was that to X18, “Do you think the rain shelter facilities in the park are sufficient?”, reaching 2.4. In addition, for X13, “park parking spaces”, it reached 2.1%; for X20, “personal amusement facilities”, it reached 1.8%; and for X17, “fitness facilities in the park”, it reached 1.7%. With similarly low scores of 2 points, X13, X11, and X18 accounted for 25.2%, 23.5%, and 15.1%, respectively. In fact, the problem descriptions of X13, X11, X18, X17, and X20 concern the functions and design contents of the parks’ internal facilities, which shows that, at present, users are not particularly satisfied with their experiences of some detailed functions of the parks.

In the survey results for the scale of individual perception and evaluation of the park space, it can be seen that the variable with the lowest score and the largest proportion is Y2, “Who do you like to accompany you to the park?”. The proportion of this variable, with a score of 1, reaches 5.8%, but the choice of answers for this question ranges from the “self” to “other” social relations, thus falling under the problem variable of factual description, as the question is not “intuitive”. Therefore, it is not comparable with Y1, Y15, Y17, and Y13, which have the same high proportion, with 1 point. For the values of 2.4%, 1.8%, 1.6%, and 1.5% generated by Y1, Y15, Y17, and Y13, respectively, these four variables actually relate to the sense of integration between the individual and other individuals in the environment of the park. Among the five zones, the highest proportions are those of Y21, Y22, and Y20. The three variables are “Do you think there is a security threat from strangers in the park?”, “Do you think you are a part of Zhengzhou in the park?”, and “How do you rate the management level of the park?”, indicating that the interviewees generally believed that the park offered a full sense of security and that newcomers to the place were mainly subject to security management. Moreover, the interviewees generally had strong feelings of human–land connection towards Zhengzhou.

2.3. Data Processing and Analysis

SPSS 23.0 software was used to conduct KOM and Bartlett sphericity tests on the 30 variables of X1–X30 (Table A1 in Appendix A and

Table 2. KOM and Bartlett sphericity test.

KOM and Bartlett Test
Kaiser–Meyer–Olkin Measure of Sampling Adequacy		0.938
Bartlett’s Test of Sphericity	Approx. Chi-Square	27,341.778
	df	435
	Sig.	0.000

). The factor analysis with SPSS 23.0 software was used to extract four-factor, five-factor, and six-factor models, respectively. Finally, we chose to implement the five-factor model. According to the Kaiser normalized maximum variance method, convergence was achieved after 11 iterations. We then obtained a total variance interpretation table and rotated component matrix table (

Table A3. Total variance explained of X1–X30.

Total Variance Explained
Group	Initial Characteristic Value			Extracted Sum of Squares of the Loadings
	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
1	8.427	28.089	28.089	8.427	28.089	28.089
2	2.582	8.606	36.696	2.582	8.606	36.696
3	2.039	6.797	43.492	2.039	6.797	43.492
4	1.415	4.717	48.210	1.415	4.717	48.210
5	1.084	3.614	51.824	1.084	3.614	51.824
6	1.061	3.537	55.361
7	0.877	2.923	58.284
8	0.825	2.749	61.033
9	0.747	2.490	63.523
10	0.697	2.322	65.845
11	0.672	2.242	68.087
12	0.663	2.210	70.297
13	0.652	2.175	72.471
14	0.630	2.099	74.570
15	0.602	2.008	76.578
16	0.587	1.958	78.536
17	0.566	1.886	80.422
18	0.542	1.806	82.228
19	0.526	1.755	83.983
20	0.514	1.713	85.696
21	0.499	1.664	87.360
22	0.491	1.636	88.996
23	0.476	1.586	90.582
24	0.469	1.564	92.146
25	0.429	1.429	93.574
26	0.418	1.394	94.968
27	0.407	1.357	96.325
28	0.405	1.352	97.677
29	0.357	1.191	98.868
30	0.340	1.132	100.000

Extraction method: principal component analysis.

in Appendix B and

Table A5. Rotated component matrix of X1–X30.

Rotated Component Matrix a
	Component
	1	2	3	4	5
X1	0.002	0.125	0.733	0.047	−0.060
X2	0.119	0.008	0.746	0.131	0.139
X3	0.218	−0.019	0.690	0.113	0.238
X4	0.244	0.112	0.517	0.095	0.336
X5	0.135	0.368	0.381	−0.251	0.383
X6	0.397	−0.238	0.236	0.216	0.461
X7	0.238	0.063	−0.054	0.471	0.558
X8	0.078	0.349	0.210	0.127	0.585
X9	0.343	0.059	−0.031	0.252	0.513
X10	0.202	0.048	0.280	−0.110	0.623
X11	0.013	0.044	0.451	0.341	0.265
X12	−0.099	0.168	0.146	−0.012	0.349
X13	−0.078	0.759	−0.035	−0.088	0.144
X14	0.446	−0.001	0.428	0.145	0.222
X15	0.345	0.615	0.095	0.062	0.178
X16	0.508	0.414	0.047	−0.040	0.142
X17	0.658	0.286	−0.091	0.264	0.225
X18	0.427	0.589	−0.156	0.189	0.213
X19	0.639	0.148	0.223	0.043	0.193
X20	0.628	0.301	−0.088	0.297	0.213
X21	0.370	0.586	0.085	0.054	0.161
X22	0.639	0.067	0.240	0.178	0.008
X23	0.652	0.118	0.190	0.215	−0.012
X24	0.213	0.598	0.159	0.214	−0.028
X25	0.517	0.202	0.185	0.312	0.017
X26	0.037	0.576	0.159	0.434	−0.078
X27	0.203	0.236	0.207	0.626	0.070
X28	0.443	−0.219	0.264	0.515	0.115
X29	0.198	0.143	0.332	0.477	0.015
X30	0.303	0.035	0.017	0.706	0.091

Extraction method: principal component analysis. Rotation method: Kaiser standardized maximum variance method. ^a. The rotation converged after 14 iterations. The blue number is the factor with larger load selected from the factor component matrix.

in Appendix C).

The Kaiser–Meyer–Olkin (KMO) test statistic is an index that compares simple variables with partial correlation coefficients. It is mostly used for factor analysis of multivariate statistics.

The KMO value is between 0 and 1. The closer the KMO value is to 1, the stronger the correlation between variables is, and the more suitable the data are for factor analysis. The closer the KMO value is to 0, the weaker the correlation between variables is, and the data are not suitable for factor analysis. It is generally believed that a KMO value > 0.9 is very suitable, 0.8 is suitable, 0.7 is relatively suitable, 0.6 is not very suitable, and a value below 0.5 is extremely unsuitable.

The statistical value of Bartlett’s sphericity is 0.938, indicating that Bartlett’s ball test is significant at the level of p < 0.001. The approximate chi-square value of the test is 27,341.778, and the degree of freedom, 435, can indicate that the test content is relatively consistent with the measurement results, and that the test validity is high, meaning that the data are suitable for factor analysis.

After determining the number of factors, the public factors were named according to the principle of “naming by referring to the load value of the subject factor” and the meanings of the subjects with high load values and the most subjects: Factor 1 includes seven questions, X16, X17, X19, X20, X22, X23, and X25, which mainly concern the completeness and evaluation of the use of public activity facilities in the park; thus, it is named as “functional facilities”. Factor 2 includes six questions, X13, X15, X18, X21, X24, and X26, which mainly concern the individual’s experience of supporting facilities in the process of using the park. Although X24 does not directly evaluate this metric, “the rich variety of crowd activities” can be regarded as an indirect manifestation of complete supporting facilities; hence, it is named as “supporting facilities”. Factor 3 includes four questions, X1, X2, X3, and X4, which mainly concern the experience of the recreational person in regard to park vegetation greening and the green landscape; thus, it is named as “landscape greening”. Factor 4 includes three questions, X27, X28, and X30, which mainly concern the conditions necessary for the recreational person to conduct demand-related behavior in the park; thus, it is named as “demand facilities”. Factor 5 includes three questions, X7, X8, and X10, which are mainly related to the evaluation of the park space management level by visitors; thus, it is named as “space management level”.

The exploratory factor analysis (EFA) showed that the explanatory degree of the five-factor model in regard to the total variance was 51.824%. Through the analysis of this scale, the relevant variables can be divided into “functional facilities”, “supporting facilities”, “landscape greening”, “demand facilities”, and “space management level” according to the five-factor model based on the park space conditions, which are considered as the five potential variables necessary for recreational participants to come into space contact with the park.

Cronbach’s α, the coefficient method, which is generally used for factor testing, was used to test the reliability of the five-factor model.

The summary of the case processing stage shows that there were 2626 data points in this study, with no missing values, and the total sample size was 2626 cases (

Table 3. Case processing summary.

Case Processing Summary
		number	%
Case	Valid	2626	100.0
	Exclusion a	0	0.0
	Total	2626	100.0

^a. Based on the column deletion of all variables in the procedure.

). In the table, Cronbach’s α coefficient value is 0.907 (

Table 4. Cronbach’s α reliability statistics.

Reliability Statistics
Cronbach’s α	Cronbach’s coefficient based on the standardization project	Number of items
0.907	0.907	30

), indicating that the 30 variables have fairly high internal consistency. Generally speaking, for Cronbach’s α, the degree of consistency is related to the measurement content. The larger the coefficient is, the stronger the consistency is. It is generally believed that for Cronbach’s α, a coefficient value greater than 0.7 is an acceptable test result, indicating good consistency.

Using the SPSS 23.0 software for a factor analysis of neighborhood social capital, we conducted a test of three-factor, four-factor, and five-factor models. Finally, in the model selection process, we chose to adopt the three-factor model. That is, we selected “human attraction”, “space belonging”, and “space trust” as the three potential variables of individual perception of the park space (

Table A6. Rotated component matrix of Y1–Y22.

Rotated Component Matrix a
	Component
	1	2	3
Y1	−0.215	0.472	0.204
Y2	−0.220	0.414	0.171
Y3	0.338	0.476	0.094
Y4	0.368	0.553	0.056
Y5	0.397	0.524	0.014
Y6	0.264	0.613	0.036
Y7	−0.029	0.416	0.283
Y8	0.220	0.301	0.386
Y9	0.359	0.546	0.022
Y10	0.340	0.639	−0.085
Y11	0.232	0.569	0.003
Y12	0.319	0.551	−0.062
Y13	0.677	0.117	0.127
Y14	0.713	0.128	0.080
Y15	0.555	0.182	0.187
Y16	0.555	0.303	0.117
Y17	0.726	0.163	0.083
Y18	0.784	0.167	−0.118
Y19	0.744	0.110	−0.041
Y20	0.383	0.226	0.374
Y21	−0.002	−0.084	0.813
Y22	0.096	0.060	0.792

Extraction method: principal component analysis. Rotation method: Kaiser standardized maximum variance method. ^a The rotation converges after six iterations. The blue number is the factor with larger load selected from the factor component matrix.

in Appendix C). Factor 1 is “human attraction”, including seven questions, Y13, Y14, Y15, Y16, Y17, Y18, and Y19; Factor 2 is “spatial sense of belonging”, including Y4, Y6, Y10, and Y11; and Factor 3 is “space trust”, including Y8, Y21, and Y22. It can be seen from

Table A4. Total variance explained of Y1–Y22.

Total Variance Explained
Group	Initial Eigenvalues			Extracted Sum of Squares of the Loadings
	Total	% of Variance	Cumulative %	Total	Variance Percent	Cumulative %
1	6.316	28.708	28.708	6.316	28.708	28.708
2	1.866	8.483	37.191	1.866	8.483	37.191
3	1.576	7.163	44.354	1.576	7.163	44.354
4	1.309	5.949	50.303
5	1.192	5.419	55.722
6	1.032	4.691	60.412
7	0.820	3.729	64.141
8	0.759	3.452	67.593
9	0.711	3.230	70.823
10	0.647	2.941	73.764
11	0.601	2.733	76.497
12	0.588	2.674	79.171
13	0.570	2.591	81.762
14	0.541	2.458	84.220
15	0.513	2.331	86.551
16	0.487	2.215	88.766
17	0.456	2.071	90.837
18	0.445	2.021	92.859
19	0.443	2.013	94.872
20	0.418	1.901	96.773
21	0.393	1.787	98.560
22	0.317	1.440	100.000

Extraction method: principal component analysis.

in Appendix B that the explanatory power of the four-factor model in regard to the total variance is 44.354%.

3. Results

3.1. Model Results and Verification

3.1.1. Theoretical Model

In the “Homo Urbanicus” theory, the individual’s perception of the park space reflects non-quantitative psychological feedback on the space expressed by the individual in terms of space contact behavior. This feedback directly affects the individual’s motivation to engage in space contact behavior in the future, forming the reproductive role of space elements. We sought to analyze the factors influencing the individual’s sense of regional security, human attraction, and spatial belonging; explore the objective constraints of the park space resources; and design an overall plan for this study corresponding to the structural, causal model of the recreational behavior of people in contact with space. Through factor analysis of the 22 variables of park space perception in the scale design, three common factors were extracted. According to the research plan, a structural equation model, the causal model, was built for the test, and three potential variables were derived. The five dimensions of the model, “functional facilities”, “supporting facilities”, “landscape greening”, “demanding facilities”, and “management level”, were taken as the independent variables of the causal model of the recreational behavior structure, according to the “Homo Urbanicus” theory, and their effects on the dependent variables of “human attraction” and “space belonging” and the role of “space trust” were explored. The theoretical model is shown in Figure 4.

3.1.2. Complete Causal Model

According to the theoretical model basis generated by AMOS, the SAV data of SPSS were imported to generate a full causal model (Figure 5) and passed the goodness-of-fit test using the AMOS RMR, GFI, AGFI, and PGFI models (

Table 5. Goodness-of-fit test of RMR and GFI model.

Model	RMR	GFI	AGFI	PGFI
Default model	0.060	0.926	0.912	0.783
Saturated model	0.000	1.000
Independence model	0.138	0.866	0.858	0.818
Zero model	0.172	0.000	0.000	0.000
Model	RMR	GFI	AGFI	PGFI

), indicating that “functional facilities”, “supporting facilities”, “landscape greening”, “demand facilities”, “space management level”, “human attraction”, and “space belonging” were valid. The causal model, composed of “spatial trust”, had a high level of reliability and validity.

3.2. Analysis of Model Results

3.2.1. Measurement Equation and Independent Variable Factor Results

The independent variable factor results are as follows:

(1)

X17 and X20 have the greatest impacts on the potential variable “functional facilities”, that is, “fitness facilities in the park” and “parent–child recreation facilities in the park”, both reaching a correlation coefficient of 0.74.

(2)

X28, “Whether the rain shelter facilities in the park are sufficient”, has the greatest impact on the potential variable “supporting facilities”, and the correlation coefficient is 0.71.

(3)

X2, “Is there a rich variety of greening plants in the park?”, and X3, “Plant maintenance and pruning level in the park”, have the greatest impacts on the potential variable “landscape greening”, with correlation coefficients of 0.74.

(4)

X18, “Whether the number of trash cans in the park is enough”, has the greatest impact on the potential variable “demand facilities”, and the correlation coefficient is 0.65.

(5)

The most influential factor for the potential variable “space management level” is X8, “How well the park manages motor vehicles”, with a correlation coefficient of 0.68.

3.2.2. Analysis of the Structural Equation Results

The structural equation results are as follows:

(1)

According to the full causal model analysis, the potential variables that have the greatest impacts on the three variables “human attraction”, “sense of space belonging”, and “sense of space trust” are, respectively, “supporting facilities”, “landscape greening”, and “functional facilities”, and their related factor loads are 0.78037 and 0.68, respectively. Although “landscape greening” has the highest contribution value for “sense of space belonging”, its value is only 0.37. Therefore, it is judged to be relevant but not significant.

(2)

In the case of the full causal model, the load coefficients of “human attraction” for the relevant factors of “sense of space belonging” and “sense of space trust” are high, reaching 0.53 and 0.54, respectively. Thus, it can be used as the intermediary variable in the structural equation model, playing a mediating role in the causal structure analysis, which ranges from the five elements of park space resources to the three factors of park space perception. That is, to obtain “space belonging” and “space trust”, we need to enhance the “human attraction”. However, “functional facilities” are the only variable that can directly cross the intermediary variable in order to realize the most significant and relevant impact on “space trust”, indicating that “functional facilities” play a key role in the causal model. “Supporting facilities” had the highest relevant factor load of 0.78 for “human attraction” and also played an indirect role in “space trust”.

(3)

Regarding the correlations of the other potential variables, the second highest correlation coefficient of “space trust” is “demand facilities”, reaching 0.54, which shows that “demand facilities” also have a significant positive impact on “space trust”.

(4)

For the “social interaction function” and “humanistic value” of urban parks to be strengthened, this model offers a design and planning strategy direction for “supporting facilities” and “functional facilities”. That is, “improving and upgrading supporting facilities and paying attention to the functional facilities of urban parks will help to increase social interaction functions and enhance the humanistic value of parks”.

4. Discussion

According to the rotated factor analysis results, the three latent variables of “human attraction”, “sense of spatial belonging”, and “sense of spatial trust” reached a value of 44.354% in the total variance interpretation. The theoretical model passed the goodness of fit test using the AMOS RMR, GFI, AGFI, and PGFI models. Based on the definition of “Homo Urbanicus” in the “Homo Urbanicus” theory,” people who pursue space contact through reasonable choice of residence” engage in selective contact with space. This contact behavior is essentially a mechanism for the selection of objective factors by the subject. This mechanism mainly originates from the subjective cognitive system composed of the purpose needs, behavioral experience, and process feedback of the subject. It can also be understood as the attraction of local culture. Because this attraction fosters human–land attachment, the local recognition brought about by behavioral feedback in response to space contact causes the behavioral subject to further improve their sense of trust in space. Therefore, “human attraction” can be constructed as the intermediary variable of “space belonging” and “space trust”.

Through the results of the causal model of the structural equation model, it is shown that “functional facilities” and “supporting facilities” are the two factors that have the greatest impacts on individual park space perception.

Li Fangzheng conducted an investigation on urban parks in 11 cities in the Beijing–Tianjin–Hebei region, discussed the relationship between the building density and green space areas of large cities and their impacts of the flow of people, and argued that the urban green space area is positively related to the service radius of urban parks. In this paper, this can be understood as the impact of “landscape greening” on individual recreational perception, but from the perspective of the causal model described in this paper, compared with the other four independent variables, “landscape greening” has no significant impact on the three latent variables of “human attraction”, “space belonging”, and “space trust”. This may be due to conceptual differences between the “service radius of urban parks” and the three latent variables mentioned in this paper [58].

Liu Ruixue’s research on satisfaction with urban parks among recreational users highlighted that the park area, air quality, vegetation, mosquitoes, recreational facilities, sign system, landscape visual effects, maintenance of facilities and plants, and environmental cleanliness, as well as nine other predictive variables, significantly affect a person’s satisfaction with urban parks. Among these factors, it was found that the sign system has the greatest impact, followed by recreational facilities. In this paper, we show that the sign system has certain correlations with “functional facilities” and “supporting facilities”. Although, in Liu Ruixue’s research, “human attraction”, “space trust”, and “satisfaction” have different meanings, and the factor load data and components of their interactions are different, these latent variables refer to the subjective judgment of individuals based on local perception; hence, the conclusion of these researchers is consistent with the findings of this paper [59].

In addition, Bao Yu argued that children are a very important participant group in urban parks; thus, children’s recreational behavior and environmental and psychological perception are also crucial for improving the quality of urban parks. In the study in question, Bao Yu found that recreational facilities are the most significant independent variable in the impact model of social perception and regional security perception. This conclusion is consistent with the conclusion that “functional facilities” have a strong correlation with “space trust” in this study and shows that this conclusion is also applicable to children [60].

5. Conclusions

The structural equation model is a causal model of recreational behavior based on the “Homo Urbanicus” theory, creating a theoretical model of this theory to explore recreational behavior in urban parks. This model shows that, based on the basic structure of “people”, “events”, “time”, and “space”, through the experience of, and feedback on, individual recreational behavior, a causal structure model of individual perception of park space resources, in regard to individual park space, can be obtained. According to the interpretation of “self-existence” and “coexistence” in the “Homo Urbanicus” theory, recreational behavior is actually a kind of contact behavior between people and space. From the perspective of the causal model explored in this study, this contact behavior is reflected in the five independent variables of space resources, namely “functional facilities”, “supporting facilities”, “landscape greening”, “demand facilities”, and “space management level”, which, respectively, reflect goal demands, equipment support, and natural contact demands. Basic physiological needs and space service needs form the basic elements of the five factors. Therefore, the independent variable “park space resources” is divided into five factors in the model: “functional facilities”, “supporting facilities”, “landscape greening”, “demand facilities”, and “space management level”. Meanwhile, “individual park space perception” is divided into three factors in the model: “human attraction”, “sense of space belonging”, and “sense of space trust”. This is in line with the interpretation of the basic conditions and elements of individual contact with space in the “Homo Urbanicus” theory (Figure 6).

The research is based on the causal model of the “Homo Urbanicus” theory, which was used to create a person-focused theoretical model of the “Homo Urbanicus” theory of urban park recreational behavior. Through the results of the causal model of the structural equation model, it is shown that “functional facilities” and “supporting facilities” are the two factors that have the greatest impacts on individual park recreation perception, and they are also the basis of the individualized space perception path of “space contact–space attraction–space trust”. To improve the recreational quality of urban parks, “functional facilities” and “supporting facilities” should be taken into account as the main reference criteria.

The main innovation of this study lies in its use of the “Homo Urbanicus” theory to study the impacts of elements of recreational behavior, marking an exploratory application of this theory. Individual recreational experience is crucial for spatial participation in urban parks. In terms of methodology, according to the scale of “self-existence” and “coexistence” that constructs the influence relationship in “Homo Urbanicus” theory, recreational behavior is also an expression of a causal model of the interaction between “person” and “space” structural elements. This study was the first to use the “Homo Urbanicus” theory as a research method for urban park recreational behavior in the academic field; thus, it has inevitable limitations, such as the changes in the content and demand of “self-existence” and “coexistence” among individuals in different seasons and the cross-changes in individual environmental perception between different ages and different time nodes, which are elements to be improved and refined in the future.