The Mechanism of Young People’s Perceptions of Urban Parks on Their Recommendation Intentions

Abstract

Urban parks, as ubiquitous spaces, are vital in enhancing people’s quality of life owing to their significant ecological benefits. However, urban parks have not achieved full age inclusivity, failing in particular to prioritize the youth demographic. Against this background, this study focused on college students as a representative youth group. Grounded in Symbolic Interactionism theory and employing structural equation modeling (SEM) for data analysis, it evaluated how young people’s perceptions of urban parks correlate with recommendation intentions through place attachment. The results showed the following: (1) beyond the natural perceptions of a park, perceptions about activities and facilities are significantly positively associated with recommendation intentions; (2) perceptions of nature, activities, and facilities correlate markedly and positively with place dependence, and perceptions of nature, activities, and facilities correlate markedly and positively with place identification; and (3) place attachment and place identification exerted chained mediating effects between perceptions of urban parks and recommendation intentions. These results provide implications for planners and designers to upgrade the internal symbols of public facilities, such as urban parks, and offer practical value for deepening interactions with the youth group.

1. Introduction

Urban parks exist both as physical landscapes and as emotional constructs within the public consciousness. As a common public facility, their functional positioning has gradually transitioned from public infrastructure to public service facilities [1]. At the same time, China’s urban governance has shifted from extensive, sprawling development to intensive, quality-driven growth, and public needs have become increasingly diverse and multilayered.

In China, the notion of a people-centered city is reflected in broad urban development strategies and in the intricate details of urban parks [2]. Driven by a people-centered value system, the public’s sense of fulfillment, happiness, and security has also become a crucial benchmark, reflecting higher-level needs for greater care and refinement in urban parks [3,4]. However, urban parks currently fall short of the ideal public facilities expected by the public, being hindered by homogenization [5], the uneven distribution of green spaces, and notable disparities in usage rates [6]. Beyond internal design, these common public facilities do not adequately respond to the different needs of various groups, so they are not truly age-inclusive. While children [7] and the elderly [8] are the main user groups for urban parks, the needs of people with disabilities are gaining more attention [9]. Still, the youth demographic has not received the attention it needs, leading young people, particularly college students, to lack a comprehensive and objective understanding of urban parks. Their low reliance on and low identification with these spaces weaken their recommendation effect, thus preventing them from driving more young people to enjoy the ecological benefits of urban parks.

The concept of recommendation intention derives from the positive behavioral intentions of tourists during travel, which encompasses revisit intention and preference intention [10]. Current research on recommendation intention mainly explores two dimensions: From the demand side, most studies have focused on the recommender’s experience, satisfaction, and word-of-mouth recommendations, examining their generation mechanisms using various recommended destinations as examples. On the supply side, the recommended destinations, as hybrids of natural and cultural symbols, have been found to be mainly influenced by their natural attributes [11]. Existing studies using structural equation modeling have identified the environment [12] and spatial physical environment [13] as key factors, with cultural attributes such as social experience [14] now gaining more attention. Factors from both perspectives inform this study’s extraction of urban park symbols. Within the specific domain of urban parks, research on recommendation intention has largely focused on influencing factors and enhancement strategies. Moreover, results on influencing factors align with those of prior research. Regarding enhancement strategies, studies have advocated for optimizing natural aspects such as biodiversity [15] and environmental adaptability [16] in design while proposing the integration of virtual reality (VR) technology [17] and augmented reality (AR) technology [18] into the renovation and upgrading of urban parks on the human level.

However, research on recommendation intention has exhibited three key limitations: First, while studies on three behavioral intents have been done, few have examined recommendation intention, with revisit intent dominating research [19]. This trend reflects a prevailing pattern of personal, self-serving repeat consumption, which reduces a location’s appeal to new or first-time visitors. Second, regarding recommended destinations, academic research has largely focused on tourist attractions, reflecting a deep exploration of commercial value. However, studies on the recommendation effects for public spaces such as parks have been insufficient, leaving their deeper value unexplored. Third, from the recommender perspective, research has primarily targeted universal visitor groups, leading to a dearth of specialized studies on specific demographics, thus hindering the achievement of all-age inclusive public services. Moreover, in practice, recommendation intention is much stronger in the youth demographic. This is because, after encountering visual or video content about a location, this group not only develops a strong sense of immersion and impulsive desire but also experiences a heightened motivation to share and recommend after experiencing it, thus exerting a greater promotional impact [20]. This distinctiveness has not been adequately recognized in the context of urban parks and has not been leveraged as a core driver of sustainable development.

The uniqueness of youth groups is also evident in the realm of place attachment, which can be analyzed through the lenses of developmental psychology and intergenerational perspectives. From a developmental psychology standpoint, research indicates that place attachments originate from childhood experiences and evolve throughout an individual’s life [21]. Childhood serves as the foundational stage for greenfield attachment, with subsequent life stages, such as adulthood, reshaping spatial meanings and usage frequency [22]. Notably, existing studies provide empirical evidence that the determinants of place identity vary between adolescents and adults. Younger individuals tend to rely more on the perceived quality of activities (functional dependence) rather than on emotional connections [23]. This finding suggests that adolescence represents a critical transitional phase during which place attachments shift from memories rooted in childhood to functional dependence and identification.

The intergenerational perspective indicates that this developmental transition is further influenced by social and historical contexts. For instance, place attachment is markedly lower in Generation Z compared to Generation Y, a trend attributed to increased mobility and digital lifestyles [24]. Concurrently, research has demonstrated that in green space environments, macrospatial environmental quality significantly predicts place dependence among young individuals, whereas this correlation does not extend to older adults [25]. Therefore, the uniqueness of the youth formation mechanism does not lie in a single factor, but in the collaborative interaction between its specific development stage and intergenerational characteristics. These discoveries are also applicable to the study of urban parks. Meanwhile, in the research of urban parks, the existing studies have not fully explored the uniqueness of the youth group, but have paid more attention to the needs of children and the elderly [7,8].

Therefore, this study focused on the research domain of recommendation intentions, overcoming the constraints of previous investigations into popular tourist destinations. Instead, it centered on parks—ubiquitous public spaces that better align with public needs and embody the human touch of a city. Simultaneously, young people were selected as the recommending group to study their perceptions of urban parks. Furthermore, this study used place attachment as a mediating variable to refine the investigation of young people’s recommendation intentions, thereby elucidating the underlying influence pathways at the micro level. To explore the relationship between young people and internal symbols within urban parks, the study employed Symbolic Interactionism as the theoretical framework and sought to extend the concept of traditional physical symbols to include social symbolic meanings. Ultimately, this study theoretically extended the body of research on the influence mechanisms of recommendation intentions from young people’s perspectives and provided practical references for building age-inclusive urban parks and improving public facility provision in modern people-oriented cities. To that end, this study aimed to answer the following questions:

(1) From the perspective of young people, what perceptual factors influence their willingness to recommend urban parks?

(2) How do the perceptual factors of young people influence their recommendation intentions?

2. Theoretical Framework and Research Hypotheses

2.1. Theoretical Foundation

Blumer first formally proposed “Symbolic Interactionism” and systematically elaborated on the theory [26]. Initially applied to sociological research, this theory focused on urban sociology and examined the behaviors of groups such as youth [27] and immigrants [28], thus establishing itself as a classic sociological theory. In summary, this theoretical framework emphasizes the role of symbols as mediators that innovatively connect humans and society. Individuals, based on their cognitive capacities and social experiences, assign new meanings to objects, continually refining these interpretations through interaction to realize social construction [26,29]. In recent research, scholars have applied this theory to public spaces such as parks, expanding the scope from human-to-human interactions to human-to-park interactions [30,31,32]. However, a review of the literature showed that current studies largely focus on either the symbolic construction process and outcomes of parks [33] or the social interactions between people [30], seldom correlating these two aspects or extending the analysis to interactions between other entities.

Even so, scholars have extracted symbols from tourist destinations to study loyalty metrics such as recommendation intentions. The three core independent variables in this research revolve around symbols: Self-consistency reflects the degree of correspondence between tourists’ self-symbols and destination brand symbols, brand identity differentiates the destination’s symbols from other places, and lifestyle consistency enables the destination to become a symbolic carrier of tourists’ value orientations. This approach shows that, in the context of symbolic consumption in tourism, a place serves not only as a brand symbol with symbolic meaning but also as an important anchor for people to define their social positioning and express their life philosophies [34]. While this approach offers insights for selecting symbols in this study, certain limitations persist: First, the interpretation of symbols tends toward static and unidirectional perspectives, thus neglecting the dynamic interactions between tourists and places. Second, the research has largely focused on matching tourists with destinations, thus overlooking tourists’ own role in reproducing symbolic value.

Later, scholars studied Wuyishan National Park in China using a mixed-methods approach combining qualitative and quantitative research to examine the formation mechanisms of its recreational appeal. They found that understanding and acceptance of symbols within urban parks influenced people’s dependence on and identification with them [35], thus providing a valuable reference for the variable selection in this study. However, this line of study remains incomplete, lacking in-depth exploration of the bidirectional nature of symbolic interaction. Specifically, it has focused on the one-way transmission of symbols from visitors to the park, so it has not clearly articulated the diverse, conflicting phenomena inherent in the process.

2.1.1. Research Subject Selection

In addition, it is necessary to clarify why we have chosen urban parks as the focus of our study. First, existing research indicates that a significant number of young people in China frequent parks, as demonstrated by various factors including usage patterns, frequency, duration, and behaviors. Specifically, 41.81% of the surveyed youth visit parks one to two times per week. In terms of duration, 42.67% spend between half an hour and one hour at the park during each visit. Among all observed behaviors, young people predominantly engage in activities focused on relaxation and connecting with nature, highlighting their strong preference for parks as venues for rest and recreation [36]. In addition, more than half of young people engage in physical activities (such as walking and moderate-to-vigorous exercise) in urban parks [37].

Second, urban parks are distinguished by their non-consumer nature, fundamentally differentiating them from commercial tourist destinations, which are often outdoor venues. Compared to commercial tourist destinations, urban parks are integrated into the urban landscape and serve as convenient green spaces that young people can visit regularly without requiring meticulous planning [38]. Additionally, considering the economic pressures faced by young individuals [39], it is evident that economic factors may drive young people to prefer urban parks. Second, in terms of psychological motivation, individuals’ motivations for tourism spending may arise from a desire for novelty and exploration or a need to escape the pressures of daily life. In the realm of environmental psychology, urban parks fulfill functions such as restoring attention—an aspect that commercial tourist destinations often struggle to provide [40].

Third, beyond commercial tourist destinations, various other spaces serve as important reference points, underscoring the significance of urban parks in fulfilling the recreational and leisure preferences of young individuals. Compared to crowded indoor entertainment venues, urban parks provide young people with a significantly higher sense of rejuvenation and enjoyment when their goal for leisure is mental recovery and relaxation [41]. Furthermore, taking movie theaters as an example, price is also an important factor young people consider when choosing leisure and entertainment venues [42]. In non-commercial indoor venues, such as libraries, young people prioritize the educational functions of libraries, perceiving them as essential spaces for academic pursuits. While social elements exist, they are predominantly associated with technology-driven collaborative efforts, such as group project discussions, rather than with unstructured leisure and entertainment activities [43]. Furthermore, when compared to other types of green spaces, such as national or rural parks, which offer greater ecological benefits but possess relatively low accessibility, urban parks are better positioned to address the daily relaxation needs of young people [44].

The above comparison, which covers commercial tourist destinations, indoor entertainment venues, and various types of green spaces, further confirms the appeal of urban parks for young people seeking leisure and entertainment. The preferences of young individuals are undeniably significant. This study employs Symbolic Interactionism to illustrate that young people’s perceptions of symbols such as natural landscapes, activities, and facilities may be positively correlated with their willingness to recommend these parks. However, we contend that this is not the only factor justifying the necessity of this research. Even when certain preferences are acknowledged, actual behavior may still be influenced by various factors, such as place attachment. In essence, as young people cultivate a greater functional dependence on and emotional connection with urban parks, they are more likely to visit and endorse them. Moreover, preference is a dynamic and variable cognitive process. Although young individuals may temporarily lose interest in parks due to external factors during specific periods, this does not diminish the necessity of this research; rather, it highlights its urgency—specifically, the need to enhance the attractiveness of parks in a dynamic environment by optimizing their symbolic elements.

2.1.2. The Logic of Symbolic Interactionism

According to Peirce’s semiotics, any element within the park—whether natural or man-made—can be considered a signifier that indicates a particular signified. When this referential relationship is recognized and shared within a social culture, the element transforms into a symbol [45]. This principle underpins the interpretation of elements such as nature, activities, and facilities as symbols in this study.

Consequently, the analysis begins with the logical relationships among perception, place attachment, and recommendation intention within the youth demographic. Specifically, this study starts from the theory of symbolic interaction. Elements such as nature, activities, and facilities in urban parks can all be regarded as symbols. When young people receive these symbols, they will develop an attachment to them over time and offer new interpretations. Then, they will recommend them to other groups, thereby achieving the reuse and dissemination of symbols. Therefore, symbols have achieved a transformation from physical meaning to real meaning, and at the same time, they are not confined to existing groups, and potential groups can be transformed into true visitors.

The initial step involves conveying the natural elements, activities, and facilities of urban parks as physical symbols to this group. In the second step, following the reception of physical symbols, the young group will not only develop perceptions and attachments to these symbols but also interpret them. Natural symbols, beyond their fundamental ecological significance, can be imbued with historical and cultural meanings [46]. Regarding activity and facility symbols, existing research, through the analysis of social media data, has demonstrated that activities and facilities in parks serve as symbolic expressions at both individual and cultural or group levels [47]. From the perspective of interpretation methods, young people mainly endow it with meaning from three levels. The first level is the sensory level. The young group mainly starts from the perspectives of sensory stimulation and aesthetic pleasure, and recognizes and transforms some unique or distinctive symbols into some visual symbols [48]. The second level is the experience level or behavioral level, with the core being an interactive approach, endowing urban parks with more functional symbols [49]. The third level is the reflective level, which involves interpreting the identity or emotional symbols contained in urban parks with the goal of resonance, thereby deepening the attachment to them [50].

The third step is for the youth group to recommend the perceived physical symbols of urban parks and the redefined symbols to other groups. Before making a recommendation, the young group will also assess the value of the recommendation and the acceptance level of the other party [51].

The fourth step involves the recommended party developing a strong interest in visiting the park, thereby returning to the first step and creating a cycle. The specific pathways include word-of-mouth communication within familiar groups and posting relevant information on social media and other platforms to attract new audiences [47].

In summary, by choosing college students as the primary agents of symbolic interaction with urban parks among the youth demographic, this study better reflected the creative value of symbols while avoiding the misguided tendency to center them, thus facilitating a transition from human–park interaction to human–human interaction. As shown in Figure 1, this process focuses on the dynamic evolution of young people’s attitudes toward symbols, which progresses through stages of reception, perception, attachment, and creation. Moreover, the innovation of this study was in its starting point: the natural, activity, and facility symbols provided by urban parks to young people. This process culminates in this group recommending ontological symbols to other populations, thus transcending the limitations of interaction between two subjects and establishing a symbolic recommendation effect.

2.2. Research Questions and Hypotheses

2.2.1. Perception and Place Attachment to Urban Parks

Earlier research on perceived place attachment did not differentiate between specific location types. Furthermore, scholars observed that people’s cognition significantly explains place attachment and proposed that perceiving a place as under threat can also enhance attachment to it [52]. As research progressed, scholars began to subdivide both perception and place types, yet the findings turned out to be complex and sometimes contradictory.

On one hand, a majority of studies support a positive relationship. For example, in rural settings, research demonstrated that perceptions of environmental elements observed incidentally, daily, or permanently all positively influence place attachment [53]. In urban parks, a study showed that among three independent variables—objective exposure indicators (NDWI and NDVI) and subjective perception characteristics (PC) of urban green spaces—seniors’ subjective perceptions of green space quality (e.g., facilities) were the most critical factor shaping their sense of place (including identification and reliance) [54].

On the other hand, a counterexample emerges in the context of community parks. Research has found that positive perceptions of functional facilities such as restrooms negatively affect place attachment. This finding indicates that people prioritizing utility exhibit stronger task-oriented visits, rather than seeking emotional connections with the park [55]. Furthermore, urban park symbols have been categorized into biotic natural elements, abiotic environmental features, and anthropogenic infrastructure [56]. This study similarly extracted attributes such as natural elements and facilities. From an inclusivity perspective, urban park design facilitates activities including resting, sightseeing, and contemplation, reflecting how park symbols encourage people to engage in such behaviors [57]. Consequently, this study also incorporates the attribute of activities.

For young adults, who are in a transitional phase between childhood and adulthood, childhood-formed attachments to parks coexist with newly emerging place perceptions driven by maturing cognition (e.g., aesthetic, social, and self-expressive needs). Urban parks, unlike community parks, offer richer emotional and experiential affordances, and young adults rarely visit them with purely utilitarian goals. Simultaneously, a psychological study showed that perceptions of facilities with specific characteristics and memorable experiences could influence place attachment [58]. Therefore, based on the above comparative analysis, this study focuses on the urban park context for young adults and proposes the following hypothesis:

H1a. 

Young people’s perceptions of urban park natural attributes have a positive association with place dependence.

H1b. 

Young people’s perceptions of the activity attributes of urban parks have a positive association with place dependence.

H1c. 

Young people’s perceptions of the facility attributes of urban parks have a positive association with place dependence.

H2a. 

Young people’s perceptions of the natural attributes of urban parks have a positive association with place identification.

H2b. 

Young people’s perceptions of the facility attributes of urban parks have a positive association with place identification.

H2c. 

Young people’s perceptions of the activity attributes of urban parks have a positive association with place identification.

2.2.2. Perception and Recommendation Intentions of Urban Parks

At a general level, research on tourist destinations has largely confirmed a positive relationship between perception and behavioral intentions. For instance, a study of tourists visiting two destinations in Spain employed structural equation modeling to reveal that the perceived image of the destination was the primary factor influencing behavioral intentions, such as recommendation intentions, thereby fully demonstrating the reference value of perception as an independent variable [59]. However, further evidence suggests that the strength of this relationship may vary across user groups. For both domestic and international tourists in central Portugal, a study showed that their perceived evaluations exerted a positive and significant influence on recommendation intention, although the impact was greater among international tourists [60].

Turning to park-specific settings, a study revealed that the attributes of environment, landscape, facilities, and culture all exhibited significant positive correlations with recommendation intention, with the natural environment attribute having the most noteworthy influence [13]. Field observations further showed that young people preferred role-playing activities in parks, which partly reflects the significance of the activity attribute in parks [12].

Furthermore, as a segment of the non-economically active population, young people are often influenced by expenditure factors when traveling [39], leading them to choose urban parks as low-cost recreational destinations. Moreover, they are characterized by a strong desire for exploration, a preference for new things, and a willingness to evaluate and share their feelings, all of which drive their recommendation intentions. Therefore, the perceived environmental quality of urban parks among college students may increase their willingness to recommend these spaces. Specifically, from the youth-friendly perspective, their perceptions imbue the symbols inside the city park with certain social meanings, thus transferring symbolic meanings, which reflects the increase in recommendation intention. Based on this notion, the following hypotheses are proposed:

H3a. 

Young people’s perceptions of the natural attributes of urban parks have a positive association with their recommendation intention.

H3b. 

Young people’s perceptions of the activity attributes of urban parks have a positive association with their recommendation intention.

H3c. 

Young people’s perceptions of the facility attributes of urban parks have a positive association with their recommendation intention.

2.2.3. Place Attachment and Recommendation Intentions to Urban Parks

A substantial body of literature within tourism research has established a robust link between place attachment and recommendation intention. Specifically, place attachment, as a positive psychological phenomenon, has been found to derive from a location’s appeal and manifests in two dimensions: place dependence and place identification. Both factors foster and enhance tourists’ recommendation intention [61]—a perspective that has gained widespread attention. Furthermore, existing research using meta-analysis methods has shown that both place dependence and place identification exert positive influences on tourists’ recommendation intention, with identification exerting a stronger effect.

This effect may stem from recommenders extending beyond the physical attributes of the destination to embrace a value orientation of self-sharing, thus fostering interaction [62].

However, a closer examination of context-specific studies reveals that the relative strength of place attachment versus place identity varies systematically across location types, including suburban nature parks, rural tourism destinations, golf tourism sites, and winery tourism locations [63,64,65,66]. In suburban natural parks, functional place dependence has a stronger influence than affective place identification [63]. In contrast, in golf tourism destinations, this notable effect is limited exclusively to place dependence, with place identification exerting no significant impact on recommendation intention [65]. This inconsistency suggests that the influence of place attachment on recommendation intentions is not uniform but rather moderated by the type of destination and the nature of the activities it affords.

Against this backdrop, young people represent a distinctive user group. Those who are more knowledgeable about and attached to urban parks can better recognize the parks’ value. This enhanced value recognition reduces the perceived psychological risk of engaging in inappropriate or misinterpreted recommendations, ultimately increasing their willingness to actively share. Given that urban parks offer a blend of functional and affective affordances, and given young adults’ transitional life stage and explorative tendencies, both dimensions of place attachment are expected to positively influence their recommendation intentions. Thus, the following hypotheses are proposed:

H4. 

Place dependence toward urban parks has a positive association with young people’s recommendation intentions.

H5. 

Place identification toward urban parks has a positive association with young people’s recommendation intentions.

2.2.4. Place Dependence and Place Identity on Urban Parks

In place attachment, the relationship between the dimensions of place dependence and place identification has attracted scholarly attention. Specifically, using structural equation modeling, studies conducted on long-distance runners at West Lake in Hangzhou [67], divers in Taiwan [68], eco-tourists on Hainan Island [69], and visitors to Canadian national parks [70] revealed that place dependence positively affected place identification as a universally established pathway. These results indicate that individuals’ attachment to a place starts with functional use.

While this pathway has been well established in general tourist and recreational populations, it remains an open question whether the same holds true for youth groups in urban park settings. However, given that adolescents gradually ascribe greater social significance to a place as the frequency and depth of their interactions with it increase, integrating themselves into it and attributing emotional value to it, young people’s repeated use of urban parks for practical purposes (such as exercise, social gatherings, and free leisure) may also foster a deeper sense of place identity through their growing familiarity and engagement. Based on this, the following hypothesis is proposed:

H6. 

Young people’s place dependence on urban parks has a positive association with place identification.

2.2.5. The Mediating Role of Place Attachment to Urban Parks

The idea that place attachment is a mediating factor in tourism studies has been widely recognized within academia. In fact, this idea has been concretely demonstrated across diverse contextual settings, such as traditional villages in rural areas [71] and coastal cities [72], where place attachment often functions as a mediating variable influencing destination loyalty. In this context, visitor loyalty is defined as the complex psychological process through which tourists—through perception, evaluation, and emotional engagement—develop intentions to revisit and recommend the destination [73].

However, closer inspection reveals that the strength and nature of this mediation vary by context. Taking traditional villages in rural areas as an example, a study conducted at a Chinese farm stay confirmed that place identity played a stronger mediating role in this context [71]. In contrast, research on urban characteristics and behavioral intentions (including recommendation intentions) found that place attachment mediates the relationship, but with an important nuance: emotional symbols and human-established interactions exert a greater driving influence than purely functional attributes [74]. This implies that in more socially dense and symbol-laden environments, the mechanisms of place attachment’s mediation shift toward interpersonal and affective pathways.

Urban parks differ from both traditional villages and generic urban settings. Yet they share a hybrid nature: they contain natural elements, facilities, and opportunities for social interaction. Therefore, these results can be extended to urban parks, where young people’s place attachment can enhance their recommendation intention. As such, we posit that their place attachment mediates the effect of place attachment on recommendation intention. Based on this, the following hypotheses are proposed:

H7a. 

Place attachment mediates the relationship between young people’s perceptions of urban park nature and their recommendation intentions.

H7b. 

Place attachment mediates the relationship between young people’s perceptions of urban park activities and their recommendation intentions.

H7c. 

Place attachment mediates the relationship between young people’s perceptions of urban park facilities and their recommendation intentions.

In summary, this study constructed a mediation model from a youth-friendly perspective. Specifically, university students’ perceptions of urban parks were the independent variable, their recommendation intention was the dependent variable, and the independent variable influenced the dependent variable through the mediating variable of place attachment (Figure 2).

3. Materials and Methods

3.1. Questionnaire Design

The questionnaire used to collect data in this study contained four sections (

Table 1. The main variables designed in this study.

Construct		Item (Please Comment on the Park You Often Visit)
Urban park perception	UPP1: Natural perception	UPP1-1: Number of trees
		UPP1-2: Lawn coverage
		UPP1-3: Plant species
		UPP1-4: Landscape richness
		UPP1-5: Environmental tranquility
		UPP1-6: Air quality
	UPP2: Activity perception	UPP2-1: Bazaar activities
		UPP2-2: Competitive activities
		UPP2-3: Sports activities
		UPP2-4: Immersive activities
		UPP2-5: Music performance activities
	UPP3: Facility perception	UPP3-1: Rest service facilities
		UPP3-2: Barrier-Free Facilities
		UPP3-3: Sports facilities
		UPP3-4: Social facilities
		UPP3-5: Interactive facilities
		UPP3-6: Safety facilities
		UPP3-7: Digital and intelligent facilities
		UPP3-8: Supporting facilities for commercial services
Place attachment	PA1: Place dependence	PA1-1: I need to rely on the park to do what I want to do.
		PA1-2: I get more satisfaction in the park than in other public spaces.
		PA1-3: My activities in the park cannot be replaced by other public spaces.
		PA1-4: The park is the best place to do what I want to do.
		PA1-5: It is difficult for me to get the same quality experience in other public spaces.
	PA2: Place identity	PA2-1: I like to go to the park.
		PA2-2: I have good memories in the park.
		PA2-3: After leaving the park, I will have nostalgia.
		PA2-4: The park means a lot to me.
		PA2-5: The park has become a part of my life.
Recommendation intention		RI1: I will give a positive evaluation of the park to others.
		RI2: I will share my good memories of the park with others.
		RI3: I will encourage others to visit the park.

). The first section included participants’ demographic characteristics, such as gender, age, ethnic groups, education level, major, average monthly consumption expenditure, and visit frequency. We adhered to the industry standard “Standard for Classification of Urban Green Space” (CJJ/T 85-2017) [75], issued by the Ministry of Housing and Urban-Rural Development of the People’s Republic of China, to categorize park types. In this standard, park green spaces are further subdivided into four categories: comprehensive parks, community parks, specialized parks, and pocket parks. The second section contained a perception scale for urban parks among college students. Drawing from scales developed by Xie et al. [76] and Kaltenborn et al. [77], it measured the perceptions of urban parks across three dimensions: nature, activities, and facilities, covering a total of 19 items. The third section featured 10 items encompassing two place attachment components, i.e., place dependence and place identification, referencing Williams et al.’s scale [78]. The fourth section captured recommendation intention, having three items based on Zeithaml’s measurement framework [79]. Moreover, all questionnaire items used a five-point Likert scale, where 1 indicates strongly disagree, and 5 indicates strongly agree.

The theoretical foundation of this study, Symbolic Interactionism, aligns with its variable structure. The ontological representation of signs and individual interpretation of signs in Symbolic Interactionism correspond to the Urban Park Perception (UPP) and Place Attachment (PA) components in this study’s design, namely, the perceptions of nature, activities, and facilities, as well as place dependence and identity, as measured in the questionnaire. The output of the symbolic meaning component on the theoretical end corresponds to the recommendation intention (RI) in this study. Thus, Symbolic Interactionism offers a micro-level explanation of meaning generation for the mediation pathway, rather than merely providing a common-sense description of correlation.

3.2. Data Collection

For the youth population, this study took college students as the research subjects. Employing the voluntary sampling method within non-probability sampling, and drawing on previous research, we distributed the questionnaire via the Wenjuanxing platform (https://www.wjx.cn, accessed on 18 August 2026) to collect data [80]. To ensure alignment between the survey population and research subjects, we ensured that the questionnaire design included questions on educational background, age, and university type. The survey was distributed via campus communities and other channels. To guarantee the questionnaire’s reliability and validity, we conducted a pre-survey before the formal study, distributing 80 questionnaires. The questionnaire was revised based on the results of this pre-survey. During the formal survey period from 28 August to 8 September 2025, 452 questionnaires were gathered. These questionnaires were then validated by examining attention–distraction questions, response times, and answer consistency to confirm that respondents were genuine university students. A total of 452 questionnaires were initially collected, and after systematic screening, 435 valid responses were retained. Due to the logic design of the questionnaire, respondents who indicated that they “seldom or never visit parks” skipped the questions regarding visit types and accompanying persons, resulting in 414 valid responses for these two items. Based on matching IP addresses and highly similar response patterns, and after removing a small number of duplicate cases, the final analytical sample for the main analysis was confirmed to be N = 411, with a sample validity rate of 90.9%.

3.3. Data Analysis

Using SPSS 27.0 (IBM Corp., Armonk, NY, USA) and Mplus 28.0 (Muthén & Muthén, Los Angeles, CA, USA), this study used several statistical methods to measure outcomes (Figure 3). Descriptive statistics were first used to characterize the demographic features of the sample. Then, Cronbach’s α and Kaiser-Meyer-Olkin (KMO) were employed to assess reliability and validity [81,82]. Confirmatory factor analysis (CFA) was used to examine item structure and dimensions, with the convergent validity for each dimension being tested via average variance extracted (AVE) and composite reliability (CR) [83].

Since the measurements of respondents’ perceptions of urban parks, place attachment, and willingness to recommend were conducted simultaneously, there was a possibility of common response bias when they completed the questionnaire, leading to common method bias (CMB) and thereby affecting the accuracy of the results. Therefore, this study employed Harman’s one-factor test to conduct an exploratory factor analysis on all variables in the scale. The results showed that there were six factors with eigenvalues greater than 1, among which the first factor explained 39.98% of the variance—below the 40% threshold [84]. Consequently, this study does not exhibit a serious common-method bias issue, and the empirical analysis is not affected.

To further rule out the potential influence of common method bias, this study employed the Unmeasured Latent Method Factor Control to construct a two-factor model. The results showed that adding the common method factor did not significantly improve model fit (ΔCFI < 0.01, ΔTLI < 0.01, ΔRMSEA < 0.015), further corroborating that this study does not suffer from severe common method bias [85].

After completing the test for outliers in the common method effect, to further ensure the stability of subsequent path coefficient estimates in the structural model, this study will also perform multicollinearity diagnostics (i.e., the variance inflation factor, VIF) on all latent variables to rule out statistical interference that may result from significant correlations among variables.

Then, to test the proposed research hypotheses, we employed structural equation modeling (SEM) to explore relationships between perceptions of urban parks, between perceptions of urban parks and place attachment, between place attachment and recommendation intention, and the direct link between perceptions of urban parks and recommendation intention. In addition, for the mediation effect analysis, the PROCESS procedure in SPSS 27.0 software was used [86]. The sample underwent 5000 bootstrap samples, with a 95% confidence interval set.

In addition to the main hypothesis testing, we conducted four supplementary analyses: (1) We conducted a correlation analysis to examine the bivariate relationships among the core variables, providing preliminary evidence for the hypothesized associations. (2) We included age and monthly consumption as control variables in the structural equation model. (3) Through model comparison, we compared partial mediation models, full mediation models, and models with only direct effects based on the AIC, BIC, and R² metrics. (4) In addition to SEM, we employed fuzzy-set Qualitative Comparative Analysis (fsQCA) as a supplementary analytical tool. Whereas SEM estimates the average net effect of each predictor under linear and symmetric assumptions, fsQCA adopts a set-theoretic logic that captures asymmetric causal relations and allows for equifinality—the possibility that the same outcome may arise from different combinations of conditions across respondent subgroups [87].

We combined these two analytical strategies for the following reason: SEM tests whether the proposed mediated structure holds on average across the full sample, while fsQCA probes whether alternative configurational paths exist for specific subgroups that deviate from the aggregate pattern [88]. This dual approach allows us to assess both the generalizability of the mediated model and the potential heterogeneity that a linear model may otherwise average out.

4. Results

4.1. Demographic Background of the Respondents

Based on the descriptive statistics of the sample (

Table 2. Summary statistics (N = 411).

Measures	Measure Types	Number	Proportion (%)
Gender	Male	211	51.3
	Female	200	48.7
Age	19	82	20.0
	20	90	21.9
	21	67	16.3
	22	84	20.4
	23	88	21.4
Ethnic groups	Han ethnicity	373	90.8
	Ethnic minority	38	9.2
Education level	Junior college	107	26.0
	Undergraduate	304	74.0
Major	Natural science	206	50.1
	Humanities and social sciences	205	49.9
Average monthly consumption expenditure	1000 yuan and below	38	9.2
	1001–2000 yuan	211	51.3
	2001–3000 yuan	78	19.0
	3001–5000 yuan	69	16.8
	Over 5000 yuan	15	3.6
Probability of going to the park	Occasionally go	207	50.4
	Often go	158	38.4
	Frequently	46	11.2
Types of parks frequently visited *	Comprehensive park	195	47.8
	Community park	187	45.9
	Specialized park	125	30.2
	Pocket park	117	28.3
Followers to the park	None	150	36.2
	Friend	108	26.1
	Family	83	20.1
	Lover	61	14.7
	Clubs/Organizations	12	2.9

Note: * This was a multiple-response question; respondents could select more than one park type. Therefore, the sum of percentages exceeds 100%.

), all respondents were current university students aged 19–23, with the majority holding bachelor’s degrees, accounting for 74.0% of the total. In terms of academic disciplines, respondents were evenly distributed across the natural sciences (50.1%) and the humanities/social sciences (49.9%). Gender distribution was balanced, with there being 211 male and 200 female participants. Ethnically, the majority of respondents were Han Chinese (90.8%), with the remainder belonging to minority groups. Regarding average monthly expenditure, most respondents reported spending between 1001 and 2000 RMB (51.3%), indicating moderate disposable income among young adults, consistent with mainstream consumption patterns for this age group. Regarding urban park-related questions, 50.4% of respondents visited parks occasionally, whereas 11.2% visited frequently. Moreover, based on the Urban Green Space Classification Standards issued by the Ministry of Housing and Urban-Rural Development of the People’s Republic of China [75], the questionnaire categorized urban parks into four types, with comprehensive parks and community parks being the most frequently selected by respondents. This highlights their relatively larger spatial scale and broader range of services [89]. Concerning companions, most respondents (63.8%) preferred visiting with others. This preliminary result indicates that the dependent variable of recommendation intention holds significant importance in urban park research contexts, thus providing direct empirical support for further analysis of its underlying mechanisms.

To further examine whether demographic characteristics (e.g., age, gender, education level, monthly consumption expenditure) were associated with the key behavioral outcomes (i.e., probability of going to the park, types of parks frequently visited, and followers to the park), we conducted chi-square tests. The results showed that most demographic variables did not exhibit significant associations with these outcome variables (see Appendix A,

Table A1. Chi-square test results.

	Probability of Going to the Park	Types of Parks Frequently Visited	Followers to the Park
Gender	0.12	0.98	0.23
Age	0.82	0.39	0.05
Ethnic groups	0.92	0.43	0.29
Education level	0.15	0.10	0.58
Major	0.04 *	0.46	0.26
Average monthly consumption expenditure	0.33	0.61	0.25

Note: * p < 0.05.

for detailed results). This suggests that the observed patterns in our main analysis are not substantially driven by demographic differences, which further supports the robustness of our findings.

4.2. Reliability and Validity Testing

To determine the consistency and stability of measurement issues, as well as the degree of discrepancy between measurement results and true values, this study next examined the reliability and validity of the scale and performed confirmatory factor analysis. The results are presented in

Table 3. Reliability and validity test of the scale.

Dimension	Items	Unstd.	S.E.	z-Value	Std.	Cronbach’s α	AVE	CR
Natural perception	NP1	1			0.991	0.932	0.705	0.934
	NP2	0.782	0.031	25.465	0.794
	NP3	0.821	0.032	25.483	0.795
	NP4	0.816	0.028	29.078	0.834
	NP5	0.812	0.033	24.946	0.788
	NP6	0.802	0.029	27.438	0.817
Activity perception	AP1	1			1	0.933	0.744	0.935
	AP2	0.810	0.031	26.233	0.798
	AP3	0.788	0.032	24.948	0.782
	AP4	0.851	0.027	31.639	0.85
	AP5	0.869	0.026	33.763	0.866
Facility perception	FP1	1			0.976	0.941	0.673	0.942
	FP2	0.853	0.035	24.703	0.792
	FP3	0.814	0.035	23.524	0.776
	FP4	0.822	0.032	25.447	0.802
	FP5	0.826	0.032	26.039	0.809
	FP6	0.872	0.031	27.97	0.831
	FP7	0.856	0.034	24.864	0.794
	FP8	0.782	0.035	22.475	0.76
Place dependence	PD1	1			0.994	0.915	0.695	0.919
	PD2	0.816	0.031	25.952	0.804
	PD3	0.837	0.031	26.843	0.814
	PD4	0.791	0.034	23.084	0.765
	PD5	0.795	0.034	23.495	0.771
Place identity	PI1	1			0.963	0.917	0.696	0.919
	PI2	0.899	0.036	24.937	0.814
	PI3	0.761	0.038	20.288	0.738
	PI4	0.849	0.035	24.365	0.805
	PI5	0.901	0.034	26.701	0.837
Recommendation intention	RI1	1			0.998	0.904	0.773	0.910
	RI2	0.836	0.032	25.818	0.832
	RI3	0.800	0.034	23.382	0.795

Note: Unstd. = Unstandardized estimate, Std. = Standardized estimate (β), S.E. = Standard error, z-value = z statistic, CR = Composite Reliability, AVE = Average Variance Extracted. All reported coefficients are statistically significant at the 0.001 level.

. The overall reliability of the complete questionnaire (32 items) was high (Cronbach’s α = 0.950), with a KMO value of 0.928. The Cronbach’s α coefficients for each subscale ranged from 0.904 to 0.941, and the KMO values were all greater than 0.6. This result indicates sufficient inter-item correlation for factor analysis. Furthermore, standardized factor loadings all surpassed the 0.5 threshold, and the AVE values for each dimension ranged from 0.673 to 0.773, with all values being above 0.5. This result confirms that the scale possesses strong convergent validity.

If variables lack discriminant validity or exhibit high correlations, issues such as multicollinearity and aggregation validity distortion may arise. Therefore, this study measured discriminant validity by comparing the square root of AVE with the correlation coefficients between variables. According to the criterion proposed by Fornell and Larcker [83], the square root of AVE for a construct needs to exceed the correlation coefficient between that construct and any other one. As shown in

Table 4. Results of the discriminant validity.

	NP	AP	FP	PD	PI	RI
NP	0.839
AP	0.436	0.863
FP	0.441	0.434	0.820
PD	0.443	0.371	0.377	0.879
PI	0.448	0.434	0.466	0.470	0.834
RI	0.327	0.423	0.400	0.383	0.405	0.835

Note: The bold numbers on the diagonal represent the square root of the average variance extracted (AVE). The lower triangle shows the Pearson correlation coefficients between constructs. All reported coefficients are statistically significant at the 0.001 level.

, all constructs met this criterion. The results indicate that the measurement model has adequate discriminant validity.

To ensure that the aforementioned correlations do not affect the estimation of path coefficients in subsequent structural models due to multicollinearity issues, this study further conducted a Variance Inflation Factor (VIF) diagnostic analysis on the predictor variables. VIF is commonly used as an indicator of multicollinearity; a VIF value greater than 10 indicates a severe multicollinearity problem [90], while stricter standards recommend that VIF should not exceed 5 [91]. As shown in

Table 5. Multicollinearity diagnosis results.

Variables	Tolerance	VIF
Natural perception	0.688	1.454
Activity perception	0.667	1.500
Facility perception	0.658	1.519
Place dependence	0.651	1.536
Place identity	0.732	1.367

, the VIF values for all latent variables range from 1.367 to 1.536, and the tolerance values range from 0.651 to 0.732 (all greater than 0.1). All indicators are well below the recommended thresholds mentioned above, indicating that although there is significant correlation among the variables, the degree of multicollinearity is extremely low and insufficient to cause substantial bias in the estimation of path coefficients in the structural model.

4.3. Goodness-of-Fit Testing of the Measurement Model

To verify the model-data fit, this study examined model fit using multiple dimensions, including CMIN/df and GFI, referencing the criteria proposed by Hayduk [92], Bagozzi and Yi [93], Hu and Bentler [94,95], and Hair et al. [96]. As shown in

Table 6. Goodness of fit for measurement model.

Fit Indices	Model Indicator Values	Standard	Conclusion	Source
CMIN/df	1.319	<5 Acceptable; <3 Good fit	Good fit	Hayduk [92]
GFI	0.914	>0.8 Acceptable; >0.9 Good fit	Good fit	Bagozzi and Yi [93]
CFI	0.988	>0.8 Acceptable; >0.9 Good fit	Good fit	Bagozzi and Yi [93]
NFI	0.951	>0.9 Acceptable; >0.95 Good fit	Good fit	Hu and Bentler [95]
TLI (NNFI)	0.986	>0.8 Acceptable; >0.9 Good fit	Good fit	Hair et al. [96]
RMSEA	0.028	<0.08	Good fit	Hair et al. [96]
SRMR	0.029	<0.08	Good fit	Hu and Bentler [94]
Model Summary	Acceptable Model Fit

Note: CMIN/DF = The ratio of the chi-square value to the degree of freedom, GFI = Goodness-of-fit index, CFI = Comparative fit index, NFI = Normed fit index, TLI = Tucker–Lewis index, RMSEA = Root mean square error of approximation, SRMR = Standardized root mean square residual.

, all model indices met the fit criteria, indicating a satisfactory model fit. This result indicates a high degree of alignment between the model and the collected data.

4.4. Correlation Analysis

This study conducted a Pearson correlation analysis on six key variables; the results are shown in Table 4. All variables showed significant positive correlations (r = 0.327–0.470, p < 0.001). This indicates that a change in one variable is accompanied by a change in the same direction in the others; that is, higher scores on one variable tend to be associated with higher scores on the others [97]. These results provide preliminary support for subsequent hypothesis testing.

Although the correlation coefficients between all latent variables reached statistical significance, to avoid the potential impact of multicollinearity on the estimation of the structural model due to minor correlations in a large sample, this study further conducted a variance inflation factor (VIF) diagnostic analysis in the subsequent measurement model evaluation (see Table 5 for details).

4.5. Hypothesis Testing of the Model

4.5.1. Direct Effect Test of Model

This study next tested the structural equation model and hypotheses, with results presented in Figure 4. Regarding the relationship between perceptions of urban parks and place dependence, perceptions at the levels of nature (β = 0.289, p < 0.001), activities (β = 0.175, p < 0.01), and facilities (β = 0.179, p < 0.01) all had positive and statistically significant associations with place dependence. These results support H1a, H1b, and H1c. In the path from perception to place identity, perceptions of nature (β = 0.163, p < 0. 01), activities (β = 0.173, p < 0.001) and facilities (β = 0.228, p < 0.001) were positively and significantly associated with place identity, thus supporting H2a, H2b and H2c.

The pathways associated with recommendation intention comprise two distinct routes: one originating from perceptions of the urban park and the other from place attachment. Regarding the relationship between perceptions of urban parks and recommendation intention, as hypothesized, perceptions at both the activity and facility levels were positively related to recommendation intention, thus supporting H3b and H3c. However, perceptions at the nature level were not significantly associated with recommendation intention (β = 0.020, p > 0.05), thus failing to support H3a. Regarding the relationship between place attachment and recommendation intention in urban parks, the p-values for paths H4 and H5 were both under 0.001. Moreover, place dependence was positively and significantly related to recommendation intention (β = 0.174, p < 0.01), and place identification was also positively and significantly associated with recommendation intention (β = 0.151, p < 0.01). Thus, the results support H4 and H5.

Regarding the relationship between place dependence and place identification, the findings are consistent with existing research: Place dependence was significantly and positively associated with place identification (β = 0.244, p < 0.001). This finding suggests that people’s attachment to urban parks extends beyond functional dependence to emotional support, thus supporting H6.

4.5.2. Mediating Effect Test Based on Bootstrap

Table 7. Results of the mediation effect test.

Hypothesis	Path	Effect	Percentile 95%CI		Test Results
			Lower	Upper
H7a	NP → PD → PI → RI	0.010	0.004	0.021	Supported
H7b	AP → PD → PI → RI	0.006	0.002	0.015	Supported
H7c	FP → PD → PI → RI	0.006	0.002	0.016	Supported

shows that place dependence and place identification had significant positive indirect relationships across multiple pathways, with bootstrap confidence intervals not containing zero. The results indicated that perceived naturalness has a significant positive indirect association with recommendation intention through place dependence and place identification. Moreover, activity perception was positively associated with recommendation intention through place attachment. The indirect relationship between facility perception and recommendation intention via place attachment was also significant. These results support H7a, H7b, and H7c.

4.6. Additional Analyses

4.6.1. Robustness Tests

To address potential confounding effects of individual differences, we added age and monthly consumption as control variables in the SEM, pointing to place dependence (PD), place identity (PI), and recommendation intention (RI). As shown in Appendix B,

Table A2. Comparison of standardized path coefficients (STDYX) between no-control and control models.

Hypothesis	Path	No-Control Model	With Control Model	Change
H1a	NP → PD	0.289 ***	0.299 ***	0.010
H1b	AP → PD	0.175 **	0.169 **	−0.006
H1c	FP → PD	0.179 **	0.180 ***	0.001
H2a	NP → PI	0.163 **	0.161 **	−0.002
H2b	AP → PI	0.173 ***	0.175 ***	0.002
H2c	FP → PI	0.228 ***	0.228 ***	0.000
H3	PD → PI	0.244 ***	0.245 ***	0.001
H4	PI → RI	0.151 **	0.152 **	0.001
H5a	NP → RI	0.02	0.031	0.011
H5b	AP → RI	0.209 ***	0.199 ***	−0.010
H5c	FP → RI	0.162 **	0.163 ***	0.001
H6	PD → RI	0.174 **	0.167 **	−0.007

Note: ** p < 0.01, *** p < 0.001.

, after including these controls, all core path coefficients remained virtually unchanged (all changes < 0.02) and retained the same significance levels. The chain mediation effects also remained stable (see Appendix B,

Table A3. Bootstrap analysis of chain mediation effects between no-control and control models.

Hypothesis	Path	No-Control Model	With Control Model
		Percentile 95%CI
H7a	NP → PD → PI → RI	[0.004, 0.021]	[0.004, 0.022]
H7b	AP → PD → PI → RI	[0.002, 0.015]	[0.002, 0.014]
H7c	FP → PD → PI → RI	[0.002, 0.016]	[0.002, 0.016]

). These results indicate that individual differences do not substantively affect our conclusions.

4.6.2. Model Comparison Tests

To address concerns regarding alternative causal explanations for our findings, we compared our hypothesized partial mediation model (M0) with two alternative specifications: a full mediation model (M3), which constrains all direct paths to zero, and a direct-only model (M2), which omits the mediators. As shown in Appendix C,

Table A4. Model comparison tests for alternative explanations.

Model	Δχ2	AIC	BIC	R2
Partial mediation model (M0)	—	3272.105	3344.44	0.289
Full mediation model (M3)	38.383 ***	3304.488	3364.767	0.22

Note: M0 = hypothesized model retaining all direct and indirect paths; M3 = full mediation model with all direct paths from NP, AP, FP to RI constrained to zero. Δχ² = chi-square difference between M3 and M0 (M0 is the reference model). AIC = Akaike Information Criterion; BIC = Bayesian Information Criterion. CFI and RMSEA are not reported for M0 due to saturation (df = 0). *** p < 0.001.

, the partial mediation model (M0) fitted the data significantly better than the full mediation model (M3), with a chi-square difference of 38.383 on 3 degrees of freedom (p < 0.001). Moreover, M0 yielded lower AIC and BIC values (3272.105 and 3344.440, respectively) than M3 (3304.488 and 3364.767). These findings effectively rule out full mediation as a viable alternative explanation and support our proposed partial mediation structure. To further evaluate the mediation model against a model devoid of mediators, we assessed the variance explained in the outcome variable (RI). The partial mediation model (M0) accounts for 28.9% of the variance in RI, in contrast to 24.2% in the direct-only model (M2). This enhancement in R² signifies that the proposed mediators provide substantial explanatory power beyond the direct effects of the independent variables. Collectively, these model comparisons substantiate the superiority of our hypothesized partial mediation model over plausible alternative explanations.

4.6.3. fsQCA Analysis

To examine whether the mediated relationships identified above hold uniformly across all respondents, or whether alternative configurational paths exist for certain subgroups, we conducted an fsQCA to identify sufficient combinations of conditions that lead to the outcome.

The results reveal multiple equifinal pathways (solution consistency = 0.817). These configurational paths do not contradict the SEM findings; rather, they show that while the partial mediation model best represents the average structure, there exist alternative combinations for specific cases. Results are presented in Appendix D,

Table A5. Demonstration of fsQCA configurations obtained in the sample study.

Condition	Configuration 1	Configuration 2	Configuration 3	Configuration 4	Configuration 5	Configuration 6	Configuration 7
NP			⊗		●	⊗	⊗
AP	●		●	⊗	●
FP	●	●		●	⊗	●	●
PD		●		⊗	●	●
PI	⊗	⊗	●	●			●
Raw coverage	0.367	0.374	0.34	0.247	0.331	0.351	0.346
Unique coverage	0.021	0.016	0.013	0.012	0.052	0.004	0.000
Consistency	0.884	0.872	0.899	0.883	0.873	0.884	0.893
Overall consistency	0.817
Overall coverage	0.616

Note: Black circles (“●”) indicate the presence of a condition, and crossed-out circles (“⊗”) indicate the absence of a condition. Blank spaces indicate a “do not care” situation, in which the causal condition may be either present or absent. Configuration 7 has zero unique coverage, indicating that its empirical instances are fully covered by other configurations; it is retained for its theoretical relevance and high consistency (0.893).

.

Taken together, the fsQCA results serve two complementary purposes relative to the SEM analysis. First, they corroborate the main SEM findings: the key predictors that were significant in the net-effect model also appear as core conditions across the sufficient configurations, supporting the robustness of the mediated structure. Second, they reveal that the average mediated model does not exhaust all possible causal mechanisms—for certain subgroups, the outcome can be achieved through alternative combinations of conditions. This suggests that while the partial mediation model best captures the overall sample pattern, the underlying causal structure is more heterogeneous across respondents than the linear model alone would indicate.

5. Discussion

5.1. Key Findings and Comparisons with Previous Studies

The key study results are as follows. First, young people’s perceptions of urban parks were all positively and statistically significantly associated with place attachment, consistent with previous research findings. For example, previous studies indicate that in urban parks, older adults’ subjective perceptions of green space quality, including the attributes of park facilities, significantly influence their sense of place, which encompasses their feelings of belonging and dependence on the space [54]. This finding suggests that similar conclusions also apply to younger individuals in urban green spaces, such as city parks. However, with respect to perceptions of facilities, scholars have observed that in urban community parks, negative perceptions of restroom facilities adversely affect place attachment. In contrast, other recreational amenities, such as benches and water fountains, exhibit a positive correlation with place attachment [55]. In this study, perceptions at the facility level actively engendered the formation of place attachment. This may stem from younger generations’ heightened attention to detail, thus interpreting facility quality as a symbol of human care. Encountering facilities that blend design aesthetics with user-friendly features optimizes their interpretation of these symbols, thereby strengthening attachment and identification [98,99].

Second, both dependence and identification with urban parks were positively and significantly related to young people’s willingness to recommend them, aligning with prior research [63]. These findings suggest that, in addition to their subjective personal appreciation of the symbols in urban parks, this group also values the specific functional benefits provided by the parks, based on the meanings they ascribe to them. This also reduces the risk associated with sharing information with others, thus extending the interaction between parks and young people to encompass interactions between parks and other entities—a distinctive feature of this study.

Third, in addition to perceptions of the natural environment, perceptions of facilities and activities are both significantly and positively associated with the willingness to recommend. This finding contradicts the conclusions of existing studies; for example, previous research has shown that in the specific context of parks, attributes such as the environment, landscape, facilities, and culture are all significantly positively correlated with recommendation intentions, with natural environmental attributes having the most significant impact [13]. The lack of a significant correlation between natural perception and recommendation intention may stem from several factors. First, from a symbolic perspective, natural symbols are more easily replaceable than activity and facility symbols. Their ubiquity and low functional value in the lives of young people diminish their appeal, rendering them easily substituted [100]. Second, young individuals often infuse activity and facility symbols in parks with new meanings based on their creativity, thereby revitalizing these spaces [50]. Furthermore, leisure activities that require effort tend to provide deeper psychological satisfaction compared to those that are merely relaxing [101]. Consequently, young people are more inclined to select activity and facility symbols as vehicles for expression, as these symbols not only fulfill their desire for personal gain but also allow for the reimagining of park symbols. Lastly, regarding their personal habits, young individuals tend to engage in activities they deem worthy of recommendation, which are more likely to be accepted by those receiving the recommendations [51]. The significant correlation between young people’s perception of activities and facilities and their willingness to recommend. For activities, social activities are the core driving force for young people to use parks, thus highlighting their demand for social spaces in urban parks [49]. This further indicates that urban parks can optimize supply from the perspective of promoting interaction and other aspects to better encourage young people to have a willingness to recommend. For facilities, their multi-functionality and flexibility can meet the diverse needs of the youth group and be adapted to various types of activity scenarios [102]. At the same time, unlike other groups, demands such as technology and safety are also the core needs that the current youth group hopes urban park spaces will possess [103,104].

Fourth, place dependence and place identification played a chain-like mediating role between young people’s perceptions of urban parks and their willingness to recommend them. Furthermore, when young people developed dependence on urban parks through their interactions, it promoted the formation of identification. In this study, the group’s instrumental attachment, emotional connection, and value creation toward recommended locations formed a holistic chain process. Coupled with their enthusiasm for sharing and willingness to transmit information and value, their perceptions could not directly influence recommendation willingness. Instead, symbolic transmission through place attachment was needed to achieve the symbiosis of authentic and symbolic representations.

5.2. Policy Implications

(1) Urban parks should focus on the intrinsic needs of young people in their planning and design.

Urban parks must not only continue to leverage their core strengths in areas such as ecological conservation, recreation, and emergency shelter, but also make improvements in enhancing their appeal and fostering a sense of belonging among young people.

Recommended measures are specifically outlined in the following three dimensions. From a natural perspective, the diversity of plant life should be increased to create ecologically balanced spaces with varied topography. At the same time, consideration should be given to enriching landscape diversity to achieve an aesthetic experience that engages both sight and sound. Furthermore, creating urban park spaces with fresh air and a tranquil environment should be included as key evaluation factors, thereby attracting young people to actively visit the parks to relax. From an activities perspective, urban parks can build upon the growing popularity of competitive sports (such as frisbee and trail running) by introducing a range of emerging activities, including music performances (such as music festivals), immersive experiences (such as camping and picnics), and market events (such as flea markets), thereby meeting young people’s needs for leisure and enjoyment. From a facilities perspective, in addition to further improving existing elements such as rest areas and safety facilities, urban parks should also embody greater youth appeal. By focusing on digital, social, and commercial amenities, parks can bridge the gap with young people and rejuvenate the park experience. Consequently, the hybrid park model —comprising these various elements—infuses urban parks with greater vitality, thereby driving the development of public facilities toward a more people-centered approach.

(2) The formation of young people’s willingness to recommend urban parks must follow the principles of gradual development.

This study indicates that young people’s perceptions of urban parks—in terms of nature, activities, and facilities—are indirectly positively correlated with their willingness to recommend them through place attachment. In reality, however, young people currently show limited enthusiasm for visiting parks; therefore, in their minds, urban parks still require time to transition from ubiquitous green patches to accessible life-support systems. In addition to urban parks needing to evolve their internal symbolism, young people should also further clarify the functional and emotional fulfillment that urban parks provide them, thereby deepening their reliance on and sense of identification with these spaces. Furthermore, young people should promptly update their comprehensive understanding of urban parks and leverage the ripple effect among groups to drive promotion, thereby participating in the sustainable development of urban parks.

(3) The symbolic interactive relationship between urban parks and young people deserves attention.

While urban parks provide high-quality, youth-appealing material symbols (such as nature, activities, and facilities), young people should also leverage their creative strengths to imbue these symbols with new meanings. At the same time, they can participate in the process of symbol implementation through various means, such as volunteer service, thereby promoting the reuse and dissemination of these symbols. In summary, paying attention to the push-and-pull dynamics between the two is the new solution for enhancing the appeal of public facilities such as urban parks.

5.3. Contributions

The study made two major contributions. First, at the research perspective level, existing studies have predominantly focused on general respondents while neglecting specific demographic segments. Drawing inspiration from the age-friendly perspective, this study highlighted the neglect of young people’s needs in urban park settings. Yet, driven by pressures and feelings of loneliness, they urgently seek recreational spaces for relaxation. Therefore, adopting a youth-friendly perspective centered on this specific demographic, this study examined the symbolic connections between young people and urban parks to investigate the mechanisms through which their perceptions influence recommendation intentions. This approach provided a fresh angle for refining recommendation effect research while overcoming the limitations of existing micro-level studies in urban parks, which have largely focused on elderly and child demographics. Furthermore, on the theoretical level, this study, grounded in Symbolic Interactionism, constructed a model and presented results that enrich and expand research in the recommendation intention field. Moreover, while existing research on urban park symbols has largely focused on the park’s intrinsic symbols, this study examined symbols self-constructed by the youth group. It further expanded the perspective from human–park interactions to human–human interactions mediated by the park, thus achieving the co-construction of symbols.

5.4. Limitations

Certain limitations remained that require addressing in future research. First, a more comprehensive set of factors influencing recommendation intentions warrants examination. This study concentrated on three perceptual dimensions as independent variables. We recognize that other perceptual attributes, such as cultural and heritage values, were not included in our analysis. However, these dimensions were adequate to test our core hypotheses, and multiple robustness checks confirm that our conclusions remain reliable. Future research could encompass a wider array of natural and cultural attributes and develop an evaluation indicator system to furnish diagnostic tools for policymakers and practical design guidelines for park planners and designers. Furthermore, we did not consider other recreational venues (such as libraries and movie theaters) as comparison variables. Consequently, we cannot directly compare differences in young people’s preferences across various recreational venues at the variable level. To compensate for this limitation to some extent, we have drawn upon existing literature on other recreational venues to provide a comparative context for our findings. Future research could employ a multiple-choice scoring comparison design, which might help us gain a more comprehensive understanding of young people’s preferences. Second, this cross-sectional design does not allow us to capture how perceptions and intentions change over time. Nevertheless, this limitation does not bias the estimated associations at the single time point measured, because the relationships we report are based on concurrent data. The absence of longitudinal follow-up does not invalidate the cross-sectional correlations we observed. In future research, we plan to adopt a longitudinal design to analyze how young people’s perceptions of urban parks affect their recommendation intentions at different time points, thereby revealing the dynamic relationships among these variables. Third, since our analysis was conducted solely for descriptive purposes, we adopted a binary classification: humanities and social sciences, and natural sciences. We acknowledge that this simplified approach may not fully reflect the diversity within disciplines or the compatibility between them. However, this classification was used solely to outline the composition of the sample. In future research, we will employ more diverse disciplinary classifications to investigate potential heterogeneity. Fourth, this study used a voluntary online sampling method, which may limit generalizability to the entire population. However, our primary goal was to explore the theoretical relationships among park symbol perceptions, place attachment, and recommendation intentions among young people in urban parks. The conclusions are valid for this target population and for the theoretical mechanisms examined. Future research should employ probability sampling across more diverse geographic regions and demographic groups to enhance external validity.

6. Conclusions

In urban parks, college students, as part of the youth demographic, exhibit perceptions that largely exert a positive and significant influence on place attachment. However, their perceptions of the natural aspects of urban parks do not markedly influence place identification, which indicates that designers should better understand the significance of youth-friendliness (i.e., the benefits of urban parks) and satisfy their higher-level needs in terms of botanical diversity, landscape aesthetics, and other symbols.

Both dimensions of place attachment, i.e., dependence and identification, have a significant positive influence on recommendation intention. This result suggests that this demographic should sufficiently value the emotional bonds of attachment and recognition formed through interactions with urban parks, leveraging their social networks to extend these interactions to broader groups.

At the same time, since there is no significant positive correlation between the perception of nature and the willingness to recommend, while activity and facility attributes show a significant positive relationship with the willingness to recommend, planners adopting a youth-friendly perspective should place greater emphasis on the diversity and novelty of activities and facilities without neglecting natural attributes. This not only imbues them with meaning and drives the optimization, renewal, and communication of symbolic value but also contributes to the high-quality development of urban parks.

Through the chain-like mediating effects of place dependence and place identification, young people’s perceptions of urban parks influence their recommendation willingness via three mediating pathways: Perceptions of nature, activities, and facilities significantly affect recommendation willingness through place dependence and place identification. These pathways can guide their deepened understanding and the creation of new symbolic meanings, offering insights for policymakers pursuing integrated innovation in age-inclusive urban development and high-quality city growth.