The Effect of Group Activity Spaces in Community Parks on Social Interactions

Abstract

Social interaction plays a vital role in fostering interpersonal relationships and building social cohesion. As primary venues for residents' daily recreation and interactions, unlocking its potential for promoting social activities is essential for enhancing place attachment and advancing social well-being. This research selects typical community parks in Harbin as sample sites. Combining behavioral annotation and environmental observation, it visualizes group activities through digital behavior maps and evaluates social interaction levels through three indicators: social activeness, interpersonal distance, and social density. Kernel density estimation, correlation analysis, and variance analysis are employed to identify the environmental features that promote social interaction within community parks. The results indicate that the level of social interaction is significantly influenced by spaces and facilities: the addition of recreational facilities increases social activeness; adding paved surfaces and lighting tend to extend interpersonal distance, while a higher number of enclosure interfaces shortens it; and the increasing of convenience facilities contributes to higher social density. Moreover, discrete leisure spaces and linear boundary spaces are identified as primary hotspots of group activity. Based on these findings, this research proposes spatial optimization strategies that promote social interaction in community parks, to revitalize community parks through micro-scale spatial improvements, thereby encouraging broader social participation and enhancing residents' physical and mental well-being.

Graphical abstract

Keywords

Community Park / Social Interaction / Interpersonal Distance / Social Density / Urban Spatial Optimization

Highlight

	· Evaluates social interaction levels using social activeness, interpersonal distance, and social density
	· Reveals that space, boundary, and facility in community parks significantly affect social interaction
	· Finds differences in social activeness, distance, and density across types of group activity spaces

1 Background

The rapid urbanization, coupled with the rise of digital networks, has reshaped residents' daily habits and modes of social interaction. These changes have led to the weakening of interpersonal relationships, and in extreme cases, to social withdrawal and isolation. Urban public spaces play a critical role in fostering social engagement and strengthening social bonds, thereby helping alleviate psychological distress^[1]. As cities shift toward qualitative development, community parks—typical small-scale urban green spaces—have gained prominence due to their high accessibility^[2]. They not only function as vital sites for everyday social interaction, but also serve as key spatial resources for urban micro-renewal initiatives^[3][4]. Accordingly, examining the relationship between social behaviors and the spatial characteristics of community parks at a micro scale, and developing activity space design strategies that foster social interaction, are essential for enhancing residents' social well-being and mental health and for informing urban design and regeneration strategies.

1.1 Methods for Assessing Social Interaction in Outdoor Spaces

Social interaction refers to communicative actions rooted in social bonds, typically measured by the level of connection and interpersonal contact^[6]. It is commonly evaluated at four dimensions: direction (relationships between interacting individuals), depth (degree of mutual dependence), breadth (spatial extent of interaction), and frequency (how often interaction occurs)^[7]. Psychologists and sociologists have employed surveys, structured interviews, and behavioral observation to assess social interaction level based on factors such as group size, intensity of interpersonal contact, emotional connectedness, and duration and frequency of interaction (Tab.1)^[8]~^[18]. Among these, the social participation observation scale developed by Mildred B. Parten and colleagues categorizes children's interactions into six types—unoccupied behavior, solitary independent play, onlooker, parallel activity, associative play, and organized supplementary play^[15]. This typology has become a classic paradigm in assessing interaction levels^[19]. Nevertheless, survey- and interview-based methods, which rely heavily on subjective perceptions, fall short in capturing the direct relationship between social behaviors and the surrounding environment objectively, thereby lacking the quantitative rigor to guide evidence-based spatial design decisions^[20].

With the advancement of research, scholars have begun to view social interaction as a result of the interplay between spatial relationships and relational attributes. Urban Design studies suggest that social density can serve as an indicator of social interaction level, as higher occupancy often correlates with greater opportunities for interaction^[16]. In addition, research in Proxemics has highlighted the critical role of interpersonal distance in shaping interpersonal interactions^[21]. American anthropologist Edward T. Hall conceptualized interpersonal distance as the physical proximity between individuals in social contexts, categorizing it into four types: intimate, personal, social, and public distances^[17]. Subsequent studies have largely adopted Hall's distance categories directly. However, due to cultural differences and evolving relationships between social behavior and space, there remain gaps and misinterpretations in understanding appropriate interpersonal distance scales in small public spaces^[18]. This underscores the need to re-examine the micro-scale dimensions of social interaction spaces.

Overall, it is essential to integrate interdisciplinary perspectives from Psychology, Sociology, and Proxemics to assess social interaction in public spaces. Such assessment should be grounded in a comprehensive understanding of interpersonal dynamics and spatial usage patterns. By incorporating spatial indicators such as group size, contact intensity, interpersonal distance, activity density, and activity frequency, it is possible to systematically evaluate the direction, depth, breadth, and frequency of social interaction. This, in turn, supports the reconstruction of an evaluation framework for social interaction in public spaces, which can guide the design of spaces for social activities.

1.2 The Impact of Community Parks on Social Interaction

Community parks refer to independently designated green spaces equipped with basic recreational and service facilities, designed to meet the everyday leisure needs of nearby residents. These spaces embody both the economic characteristics of public goods and the sociological attributes of the public realm^[22]. From an economic perspective, existing research has primarily focused on the supply–demand relationships of park services and diverse interactions, analyzing multiple demographic groups' preferences and needs for activity spaces and facilities^[23], as well as examining how the equity of park siting and layout^[24], facility allocation, and service^[25][26] influence the frequency and diversity of group activities. It has been found that proximity to restaurants, commercial areas, and office buildings can significantly increase the intensity of visiting parks^[27]. In addition, existing research has evaluated the physical activity levels of group activities based on health benefit objectives and explored how community park spaces and facilities affect the type, duration, and frequency of activities^[28]~^[32]. For instance, higher shape index of community park and tree diversity have been proved to be negatively correlated with activity diversity^[28]; areas with higher vegetation coverage and richer recreational or convenience facilities tend to support a wider variety of activity types^[29][30]; furthermore, good lighting, hygiene, and safety conditions can increase the duration and frequency of activities^[31][32].

Overall, community park space and facility serve not only as key points influencing residents' activity behaviors but also as essential criteria in planning and design guidance^[33]. However, existing research remains limited in examining how spatial and facility characteristics influence the levels of social interaction^[34], lacking evaluation of the sociological attributes of community parks from an integrated perspective of spatial design and behavioral interaction^[35], making it difficult to clarify the mechanisms by which physical park features stimulate social interaction. Therefore, it is urgent to determine whether interventions in community park space and facility can alter the levels of social interaction within group activities. This is essential for improving the precision of social demand assessments in community parks and for meeting the finer-granular requirements of urban micro-regeneration.

To re-conceptualize micro-scale social interaction and to uncover the potential and value of group activity spaces in community parks to promote social interactions, and to provide a basis for the renewal and construction of community parks, this study takes selected community parks in Harbin as a case study. By conducting behavioral annotation with GPS tracking, this research generated digital behavioral maps of group activities to explore the following questions: 1) How can group activities be evaluated from the perspective of promoting social interaction? 2) Do spatial and facility characteristics of activity spaces influence social interaction level? 3) What are the characteristics of different types of group activity spaces in community parks, and are there differences in their social interaction levels? And 4) which spaces and facilities in community parks can guide and encourage diverse forms of interaction?

2 Research Methods

2.1 Sample Site Selection

This research followed the definition and criteria for community parks outlined in the Standard for Classification of Urban Green Space (CJJ/T 85–2017), screening green spaces with an area of 0.5 ~ 2 hm² and a service radius of less than 1 km around the residential areas in the central city area of Harbin as sample sites. All selected community parks should be located within a 15-min living circle of residential communities. Internally, the parks should be equipped with adequate service and recreational facilities and provide ample activity spaces to meet the social needs of users. A total of 11 community parks meeting these criteria were finally selected as sample sites (Fig.1).

The observation period extended from September 1 to November 7, 2021. During this time, each community park was surveyed on one weekday and one weekend day, with observations conducted across five time slots: 7:00–9:00, 9:00–11:00, 13:00–15:00, 15:00–17:00, and 17:00–19:00. All observation days were sunny, with average temperatures ranging from 10℃ to 13℃. Activity data were collected using a quadrat method. Based on Yoshinobu Ashihara's external modulus theory, 20 ~ 25 m is considered the threshold distance for facial recognition, which facilitates interpersonal communication while maintaining strong spatial rhythm and variation^[36]. Accordingly, the quadrat size was set at 20 m × 20 m (Fig.1). Only quadrats in which group activities involving two or more individuals occurred were included as units for analysis. A total of 293 such units were recorded.

2.2 Data Collection

Behavioral annotation and environmental observation are effective tools for assessing human–place interplay and are widely used to extract activity characteristics and spatial demands^[37]. This study adopted these tools and incorporated the Social Interaction Scale (SIS) to evaluate the social interaction of group activity units in community parks. The degrees of social interaction were classified as onlooker, parallel, associative, and cooperative (Tab.2)^[15][20]. A team of three trained observers conducted on-site behavioral annotation during the observation periods. Following the SIS observation protocol, when sampling high-density activity areas, the observers divided the park into zones and conducted instantaneous sampling within each. To avoid misrecording due to the activity variation, only activities without an interaction type change within a 15-min period were recorded as annotation events. Observers used the 2bulu app as the GPS tracker to log the latitude and longitude coordinates of each activity event, while also annotated additional information in the app, including the time, the activity type, the level of social interaction as measured by SIS, and the size of the activity group. In total, 110 on-site observations were conducted across all sample parks, yielding 466 recorded group activity annotation events. Finally, the annotated activities were visualized using ArcGIS to generate digital behavior maps.

This scale integrates evaluation items from established international environmental quality assessment tools for community parks, including Community Park Audit Tool (CPAT), Bedimo-Run Assessment Tool-Direct Observation (BRAT-DO), and Neighborhood Green Space Tool (NGST)^[38]. Items that were not applicable to the selected sites (e.g., restroom, sink, water feature) were excluded (Tab.3). A total of 14 indicators across six categories were selected, covering site metrics (area, perimeter), number of enclosure interfaces, amenity facilities (bench, picnic table, shade structure, paved surface), recreational facilities (fitness equipment, playground, sports court), natural quality (tree), and safety and maintenance (lighting, warning sign, entrance). These indicators were quantified through on-site observation. Among them, the number of enclosure interfaces was scored as 0, 1, 2, 3, or 4 according to the number of side boundaries. Paved surface was coded as a binary variable: presence as 1, absence as 0.

2.3 Data Analysis

This research used indicators of social activeness, interpersonal distance, and social density to evaluate the level of social interaction in group activities. Social activeness (V) was measured by combining activity frequency, degree of social interaction, and group size. The calculation is as follows:

(1)V=\sum (F\times SIS\times G),

where F refers to activity frequency, representing the number of observation time slots when the activity occurs. If an activity appears in all five time slots, then F = 5; if only in one, F = 1. SIS represents the degree of social interaction, with values defined in Tab.2; G denotes group size, i.e., the number of participants in the group activity. For example, in one unit, observers recorded three people conversing during three time slots, and six people playing table tennis throughout all five time slots. The interaction degree of the conversation is associative (SIS = 3), and that of table tennis activity is cooperative (SIS = 4), then the social activeness score for conversation is 3 × 3 × 3, and for table tennis is 5 × 4 × 6. The total social activeness for the unit is the sum of them, i.e., 147.

Interpersonal distance was measured using the ArcGIS distance measuring tool, representing the physical distance between annotated points. Social density was defined as the number of participants per unit area and was calculated with quadrats as follows:

(2)D=P/A,

where D is the density, P is the number of participants, and A is the area of the corresponding quadrat (m²).

Finally, SPSS was utilized to statistically analyze the correlations between spatial/facility characteristics and social interaction levels across all parks and within each activity unit. Meanwhile, this research applied Kernel Density Estimation (KDE) to visualize the spatial distribution and clustering patterns of group activities, and overlaid the KDE results with maps of activity spaces and facilities in the sample parks to validate the analytical results of the correlations.

3 Analysis of Social Interaction Levels

3.1 Social Activeness

Through on-site observation, this research identified ten types of social activities across the sample parks, and categorized their participants into groups and subgroups based on group size and the social characteristics of participants (Tab.4, Fig.2). Group activities typically involve larger numbers of participants. In contrast, subgroups—often composed of dyads or triads—exhibit more direct, frequent, and intense interpersonal interactions^[7]. By comparing each activity's degree of social interaction, activity frequency, and group size, and calculating the average social activeness for each type of activity (Tab.5), the results show that group activities such as board and card games and square dancing are cooperative interactions. These activities not only recorded the highest frequencies and participant numbers but also scored the highest in terms of social activeness. Activities with moderate activeness include group-based ball games with cooperative relationships, as well as unorganized casual conversing by subgroups, both show relatively high degrees of social interaction and frequent participation with larger group sizes. Conversely, walking in dyads and child supervision in triads exhibit lower frequencies and social activeness. Although whipping tops is generally played in groups, it falls into the category of parallel activity with minimal interaction among players, which registers the lowest scores across degrees of social interaction, activity frequency, and group size—making it the least socially active behavior observed.

3.2 Interpersonal Distance

The results (Fig.3) reveal significant differences in interpersonal distance across various group activities in community parks. Activities characterized by strong social bonds and conducted in groups typically keep intimate distances of around 0.45 m. For instance, in board and card games, both core participants and onlookers tend to maintain a proximity within the range of 0.45 m. Interactions among close friends and family members, often occurring in dyads or triads, usually with the personal distance of 0.45 ~ 1.2 m. Relevant activities include fitness exercises, casual conversing, walking, and child supervision. Specifically, the internal distances for child supervision range from 0.5 m to 1.2 m, while inter-group distances can extend from 1 m to 10 m. Activities such as Tai Chi and square dancing maintain social distances of approximately 1.5 ~ 2.5 m, which helps preserve comfortable spatial arrangements and avoids the discomfort caused by excessive proximity. Ball games, including badminton and table tennis, exhibit relatively small variations in interpersonal distance (3 ~ 4 m). According to proxemics theory, individuals in formal or spatially constrained public settings tend to maintain a public distance of 3.7 ~ 7.5 m. However, the physical constraints and interactive nature of specific activities can significantly alter this normative range. In this research, whipping top shows a marked deviation: due to the limited site size and the inherent susceptibility to interference and risk of conflict during gameplay, participants frequently maintain inter-group distances exceeding 10 m, while intra-group interactions typically occur within 2 ~ 5 m.

3.3 Social Density

High-density activities in community parks include board and card games (0.88 persons/m²), fitness exercises (0.65 persons/m²), and casual conversing (0.6 persons/m²), which involve stationary engagement. Moderate-density activities include ball games, square dancing, and child supervision, while low-density activities consist of walking, whipping top, and playing musical instrument (Fig.4). By applying the least squares method to fit the relationship between interpersonal distance and social density, an L-shaped curve emerges (Fig.4). When social density is below 0.5 persons/m², activity types are relatively limited, spatial distribution is loose, and interpersonal distances typically range from 1.2 m to 3 m, with minimal influence from surrounding spatial features or facilities. When the density reaches 0.5 ~ 1 persons/m², the average interpersonal distance falls within the personal distance of 0.45 ~ 1.20 m. People have the greatest spontaneity in choosing how to use the space, and the probability of changing activity is also the highest. However, when density exceeds 1 persons/m², interpersonal distances tend to converge to the intimate distance of 0.45 m. In such conditions, high social density suppresses variability in activities and reduces the likelihood of other types of activities emerging.

4 Analysis of the Influence of Community Park Environmental Features on Social Interaction Levels

4.1 Correlations Between Community Park Environmental Features and Social Interaction

According to the correlation analysis between social activity indicators and environmental features (Tab.6), the spatial and facility characteristics of community parks have a significant impact on social activeness, social distance, and social density. Specifically, social distance shows a significant positive correlation with the number of bench, picnic table, paved surface, shade structure, and lighting. In addition, the perimeter, area, and number of fitness equipment have a significant positive correlation with social activeness. The perimeter and area reflect the maximum carrying capacity for the number of participants in social activities: spaces with greater capacity are more likely to attract organized group activities, thereby enhancing the overall social activeness of the site. Meanwhile, social density is positively correlated with the number of bench and picnic table, indicating that high-density activities are more dependent on the availability of leisure amenities.

Through the KDE analysis of the spatial distribution and clustering patterns of group activities, it is evident that spatial form and facilities in community parks have a certain influence on social density. Discrete recreational spaces tend to promote the level of activity clustering, while linear spaces along site boundaries show moderate clustering of activities. From morning to evening, activity hotspots consistently concentrate around amenity facilities including picnic tables and seating areas, as well as recreational facilities such as fitness equipment zones, indicating that the presence and spatial arrangement of such facilities significantly affect both the number of participants and the degree of spatial clustering in activities.

4.2 Differences in Activity Levels Across Various Social Activity Spaces

Referring to international classification standards for spatial zones in community parks^[39], this research categorized group activity spaces in the study area into five types based on activity patterns and spatial functions: central gathering spaces, peripheral leisure spaces, anchored recreational spaces, equipment-based fitness spaces, and facility-based play spaces (Fig.5). Central gathering spaces are open areas accommodating multiple coexisting activities, characterized by high spatial extensibility. These spaces support a variety of activities such as square dancing, whipping top, walking, and child supervision, and are often equipped with lighting facilities to enable nighttime use. Peripheral leisure spaces are shaded linear zones along park edges, often featuring shade structures and attracting more leisure activities such as casual conversing. Anchored recreational spaces rely on the presence of convenience and recreational facilities such as picnic tables and benches to support activities like board and card games and ball games. They are typically enclosed with a strong sense of territoriality. Equipment-based fitness spaces and facility-based play spaces are fewer in number and show a strong dependence on specific equipment and site conditions, often located near park entrances.

This research applied Brown-Forsythe ANOVA to further compare differences in social activity indicators among the five types of social activity spaces (Tab.7). Results show significant differences across social activeness, interpersonal distance, and social density (p < 0.05). Central gathering spaces exhibit the highest level of social activeness but the lowest social density and relatively longer interpersonal distances. Peripheral leisure spaces show the closest interpersonal distances, indicating more intimate social interactions. Anchored recreational spaces are characterized by high social density and activeness, along with relatively close social distances. Equipment-based fitness spaces and facility-based play spaces, due to their specific functions, exhibit lower levels of social activeness.

4.3 Effects of Community Park Environmental Features on Social Interaction

Among the five types of group activity spaces, this research focused on central gathering spaces, peripheral leisure spaces, anchored recreational spaces, and equipment-based fitness spaces. It is worth noting that although facility-based play spaces (i.e., areas equipped with fixed children's play equipment) are essential components of community parks, the number of units in sample parks that clearly featured such facilities was limited (N < 30). As such, the sample size did not meet the minimum data requirements and statistical validity for reliable correlation analysis and inference. Therefore, facility-based play spaces were excluded from the statistical analysis.

The correlation analysis between social park environmental features and social activity indicators across various space types shows that in central gathering spaces, interpersonal distance has a significant negative correlation with the number of enclosure interface (r = –0.262, Sig. = 0.013), and positively correlated with area (r = 0.692, Sig. = 0.001). This indicates that the interpersonal distance is constrained by the spatial scale and degree of enclosure. Central gathering spaces mainly accommodate organized activities such as Tai Chi and square dancing. When bounded by trees or other strongly defined side interfaces and boundaries, activities tend to follow a dual-row pattern, maintaining a consistent distance of 1.5 ~ 2 m. In contrast, in fully open spaces, fitness exercises and square dancing tend to form layered concentric patterns, with a core of 3 ~ 5 rows of participants, a 2 ~ 5 m buffer zone between dancers and onlookers, and an outer ring maintaining an approximate 3 m safety distance from the boundary. In addition, social activeness and social density in these gathering spaces are significantly correlated with area and the number of lighting, suggesting that nighttime lighting functions as a catalyst for activating the space use and enhancing participation.

Periphery leisure spaces primarily support casual conversing. The variation in interpersonal distance within these spaces is significantly correlated with the number of enclosure interface (r = –0.202, Sig. = 0.027). People who gather near boundaries generally maintain intimate to personal distances (0.45 ~ 1.2 m), which aligns with the boundary effect described in environmental psychology^[17].

Anchored recreational spaces exhibit a high dependency on facilities. Interpersonal distance in these areas has a significant negative correlation with the number of enclosure interface (r = –0.394, Sig. = 0.037), and significant positive correlations with the number of bench and fitness equipment (r = 0.284, Sig. = 0.013; r = 0.362, Sig. = 0.025). Benches tend to attract casual conversing at personal distance (0.45 ~ 1.2 m), whereas four-seat picnic tables accommodate intimate-distance interactions such as board and card games (within 0.45 m). In addition, bench is a key factor influencing social density in anchored recreational spaces. Arranging benches along boundaries and providing a sense of prospect and shelter are more effective at encouraging social interaction.

In equipment-based fitness spaces, social density is positively correlated with the number of bench, picnic table, shade structur, and sports court, suggesting that functional integration in such spaces helped attract users' clustering. Benches and picnic tables likely provide places for rest and conversation, while shading structures may extend the duration of stay, thereby further increasing social cohesion.

5 Conclusions and Discussion

5.1 The Relationships Between Social Interaction and Space

This research, from the perspective of promoting diverse interactions, conducted a detailed investigation of social interactions in community parks and proposed an innovative evaluation method that used social activeness, interpersonal distance, and social density as key indicators. Compared with the interpersonal distance categories proposed and widely validated by Hall, the results of this research indicate that group activities in community parks broaden both social and public distances. People tend to enter a model of safe and sharing space, maintaining a more close social distance (1.5 ~ 2.5 m). Public distance exceeds the conventional threshold of 7.5 m to a critical range of 10 m. Personal and intimate distances show more flexible variations, often presenting as more compressed or closely clustered configurations. Relevant research also confirms that social relationships can be reshaped in small-scale public spaces, and their spatial patterns do not strictly conform to Hall's classification—social behaviors can also occur at much closer distances^[18]. Meanwhile, this research identifies facility-related factors that influence the levels of social interaction, finding that social activeness in community park is positively associated with the presence of fitness and recreational facilities, which aligns with the findings of Tinghong Guo et al., who suggested that recreational amenities (e.g., playground) enhance park vitality by supporting social interaction behaviors^[40]. Furthermore, this research reveals that interpersonal distance has a significant positive correlation with the number of lighting and a significant negative correlation with the number of enclosure interface. Related research similarly confirms that low lighting levels are negatively correlated with pedestrian mobility^[41], and relatively low spatial permeability tends to promote social interactions and reduce distances^[42]. The number and distribution of amenity facilities emerge as key factors influencing social density, discrete recreational spaces and linear boundary spaces are identified as the primary zones for activities and clustering. However, these findings contrast with those of Shuolei Chen et al., who concluded—based on measures of contact intensity and group size—that safety and maintenance features are the primary factors promoting social interaction in Nanjing's community parks, while amenity facilities have only limited effectiveness in attracting social groups^[34]. A comparative study of community parks in Belgium also finds that increasing amenity and recreational facilities have no significant impact on the number of people participating social interactions per hectare^[43]. These differing conclusions suggest that cultural, geographic, and climatic contexts may influence social interaction patterns and spatial preferences, and further research with a broader geographic scope and larger sample size is needed to verify the generalizability of the relationship between facilities and social catalysts.

5.2 Spatial Optimization Strategies to Promote Social Interaction

Based on systematic observational data, this research analyzed how spaces and facilities in community parks contribute to social interaction, and proposed the following recommendations for spatial optimization in community park design.

1) To optimize the spatial layout and functional zoning for social activities, a nested spatial configuration of "central activity hub + functional clusters" can be adopted to construct an integrated social space network. In this layout, the central activity hub serves as a gathering space, while functional clusters—such as fitness, recreational, and leisure modules—are distributed around the periphery, stimulating diverse patterns of social behavior among residents.

2) Quantify spatial configuration standards for social activities to enhance multi-activity capacity. Based on the level of dependence on facilities, spatial standards can be categorized into three types. First, for activities in gathering spaces by groups, site demand can be estimated based on social density. For example, 4 ~ 8 m²/person for square dancing, 9 m²/person for whipping top, and 4 ~ 6 m²/person for Tai Chi. Second, for activities taking place in leisure spaces by sub-groups (e.g., casual conversing, child supervision), activity capacity can be calculated based on interpersonal distance and the number of participants. For instance, the radius of interaction for casual conversing is 0.5 ~ 0.8 m, and 0.5 ~ 1.2 m for child supervision. Third, for activities in play, fitness, or recreational spaces that are highly dependent on specific facilities, spatial scale should be assessed according to the size of facilities. For example, 8 ~ 16 m² per board and card game table unit, and 3 ~ 6 m² per fitness equipment unit.

3) Leverage environmental amenities to create diverse and interactive social activity spaces. In recreational areas, increase the setting of resting facilities including physical structures, benches, and potential seating elements. In gathering areas, adjust the coverage of lighting to activate nighttime vitality and extend the duration of active use. Along spatial boundaries, install shade structures to create comfortable leisure zones that encourage lingering and interaction. In fitness areas, incorporate layered and diverse planting designs, and increase seating and shade structures to create multifunctional spaces that further strengthen social cohesion.

5.3 Limitations and Prospects

This research employed a combined method of behavioral annotation and environmental observation tools, supplemented by precisely geolocated GIS mapping to create digital behavior maps and analyze the level of social interaction and its influencing factors within small-scale spatial contexts. This method provides fundamental information during the early stages of urban planning and design, assisting designers in developing evidence-based social spaces. However, direct observations based on a small sample size may involve case-specific biases. Future research could expand the sample size to improve the generalizability of the findings. In addition, as the limited accuracy of GPS positioning may result in slight measurement errors, the application of digital technologies should be further developed to more accurately track interaction trajectories, exploring the dynamic spatial distribution of social interactions and informing the layout of functional and spatial planning.

References

[1] Hanlon, P. , Wightman, H. , Politis, M. , Kirkpatrick, S. , Jones, C. , Andrew, M. K. , ..., , & Hoogendijk, E. O. (2024) The relationship between frailty and social vulnerability: A systematic review. The Lancet Healthy Longevity, 5 ( 3), e214– e226.

[2] Wu, W. , Wang, Z. , Liu, R. , & Gu, C. (2021) On community park design under the normalization of the epidemic—Taking Yuekang Park as an example. Chinese Landscape Architecture, 37 ( S1), 74– 79.

[3] Chen, J. , & Zhang, J. (2022) Research on the relationship between community park space and characteristics of outdoor activities of the elderly. Chinese Landscape Architecture, 38 ( 4), 86– 91.

[4] Rosso, F. , Pioppi, B. , & Pisello, A. L. (2024) Tactical urban pocket parks (TUPPs) for subjective and objective multi-domain comfort enhancement. Journal of Environmental Management, 349 , 119447– .

[5] & Kaźmierczak, A. (2013) The contribution of local parks to neighbourhood social ties. Landscape and Urban Planning, 109 ( 1), 31– 44.

[6] Rasidi, M. H. , Jamirsah, N. , & Said, I. (2012) Urban green space design affects urban residents' social interaction. Procedia-Social and Behavioral Sciences, 68 , 464– 480.

[7] Zheng, H. S. (2003). New Introduction to Sociology. China Renmin University Press.

[8] Xu, Y. , Lee, J. H. , & Matarrita-Cascante, D. (2021) Profiling attached residents in an urban community in the U.S.: An empirical study of social–landscape interactions within a park. Social Sciences, 11 ( 1), 5– .

[9] Peters, K. , Elands, B. , & Buijs, A. (2010) Social interactions in urban parks: Stimulating social cohesion?. Urban Forestry & Urban Greening, 9 ( 2), 93– 100.

[10] Yuan, Q. , Zhao, J. , & Leng, H. (2022) Research on the impact of green space activity behavior in winter residential districts on the mental health of the elderly. Chinese Landscape Architecture, 38 ( 3), 45– 50.

[11] Moulay, A. , Ujang, N. , & Said, I. (2017) Legibility of neighborhood parks as a predicator for enhanced social interaction towards social sustainability. Cities, 61 , 58– 64.

[12] Maas, J. , Van Dillen, S. M. E. , Verheij, R. A. , & Groenewegen, P. P. (2009) Social contacts as a possible mechanism behind the relation between green space and health. Health & Place, 15 ( 2), 586– 595.

[13] Dadvand, P. , Hariri, S. , Abbasi, B. , Heshmat, R. , Qorbani, M. , Motlagh, M. E. , ..., , & Kelishadi, R. (2019) Use of green spaces, self-satisfaction and social contacts in adolescents: A population-based CASPIAN-Ⅴ study. Environmental Research, 168 , 171– 177.

[14] Tao, Y. , Yang, J. , & Chai, Y. (2020) The anatomy of health-supportive neighborhoods: A multilevel analysis of built environment, perceived disorder, social interaction and mental health in Beijing. International Journal of Environmental Research and Public Health, 17 ( 1), 13– .

[15] & Parten, M. B. (1932) Social participation among pre-school children. Journal of Abnormal and Social Psychology, 27 ( 3), 243– 269.

[16] Guo, X. , Yang, Y. , Cheng, Z. , Wu, Q. , Li, C. , Lo, T. , & Chen, F. (2022) Spatial social interaction: An explanatory framework of urban space vitality and its preliminary verification. Cities, 121 , 103487– .

[17] Hall, E. T. (1969). The Hidden Dimension. The Bodley Head Ltd.

[18] Song, S. , Xiao, M. , & Lu, X. (2024) Body, behavior and space: Revisiting the concept of "small scale" in public spaces. Urban Planning International, 39 ( 1), 60– 66.

[19] Bakeman, R. , & Brownlee, J. R. (1980) The strategic use of parallel play: A sequential analysis. Child Development, 51 ( 3), 873– 878.

[20] Chen, S. , Sleipness, O. , Christensen, K. , Yang, B. , & Wang, H. (2023) Developing and testing a protocol to systematically assess social interaction with urban outdoor environment. Journal of Environmental Psychology, 88 , 102008– .

[21] Su, J. , Huang, J. , Qing, L. , He, X. , & Chen, H. (2022) A new approach for social group detection based on spatio-temporal interpersonal distance measurement. Heliyon, 8 ( 10), e11038– .

[22] Qiu, B. , Zhang, F. , & Wan, Z. (2019) Main problems of study on community parks. Modern Urban Research, ( 3), 35– 41.

[23] Luo, T. , Fu, W. , & Xia, L. (2017) How to provide recreational services in local parks: A case study in Shanghai, China. Chinese Landscape Architecture, 33 ( 2), 113– 117.

[24] Zhou, C. , & Zhang, Y. (2021) Mini-park layout formation method in high-density cities. Chinese Landscape Architecture, 37 ( 10), 60– 65.

[25] Ye, S. , Zhen, D. , & Chen, B. (2020) Spatial accessibility of urban park green space based on service radius. Geospatial Information, 18 ( 4), 65– 69.

[26] Hu, Y. , Yu, Y. , & Zhang, Q. (2021) Urban green space feature and its impacts on elderly people's social life: Evidence of GPS tracking and social investigation in Shanghai. Shanghai Urban Planning Review, ( 2), 96– 103.

[27] Zhou, C. , An, Y. , Zhao, J. , Xue, Y. , & Fu, L. (2022) How do mini-parks serve in groups? A visit analysis of mini-park groups in the neighbourhoods of Nanjing. Cities, 129 , 103804– .

[28] Wang, M. , Qiu, M. , Chen, M. , Zhang, Y. , Zhang, S. , & Wang, L. (2021) How does urban green space feature influence physical activity diversity in high-density built environment? An on-site observational study. Urban Forestry & Urban Greening, 62 , 127129– .

[29] Van Dillen, L. , Ghekiere, A. , Van Cauwenberg, J. , Veitch, J. , De Bourdeaudhuij, I. , Van Dyck, D. , ..., , & Deforche, B. (2018) Park characteristics preferred for adolescent park visitation and physical activity: A choice-based conjoint analysis using manipulated photographs. Landscape and Urban Planning, 178 , 144– 155.

[30] Mu, B. , Liu, C. , Mu, T. , Xu, X. , Tian, G. , Zhang, Y. , & Kim, G. (2021) Spatiotemporal fluctuations in urban park spatial vitality determined by on-site observation and behavior mapping: A case study of three parks in Zhengzhou City, China. Urban Forestry & Urban Greening, 64 , e127246– .

[31] Song, J. , Sun, Y. , & Xie, Y. (2017) Spatial planning and design strategies based on the characteristics of the elderly society activity in neighborhood: A case study of outdoor activity spaces in typical communities of Shenzhen. City Planning Review, 41 ( 5), 27– 36.

[32] Li, J. , Fan, C. , Tian, L. , Ouyang, W. , & Miao, W. (2021) Study on the impact of neighborhood built environment on social activities of elderly. Human Geography, 36 ( 1), 56– 65.

[33] Wei, W. , & Yao, N. (2023) Research on comprehensive health spaces and facilities: Connotation, classification and research framework. Urban Problems, 331 ( 2), 26– 37.

[34] Chen, S. , Sleipness, O. , Christensen, K. , Yang, B. , Park, K. , Knowles, R. , ..., , & Wang, H. (2024) Exploring associations between social interaction and urban park attributes: Design guideline for both overall and separate park quality enhancement. Cities, 145 , 104714– .

[35] Li, J. , & Yang, Z. (2023) From spatial design to social design: Epistemological and methodological discussions on micro-renewal of leftover spaces. Landscape Architecture, 30 ( 8), 44– 50.

[36] Yoshinobu Ashihara. (1985). Exterior Design in Architecture. China Architecture & Building Press.

[37] Vidal, D. G. , Teixeira, C. P. , Fernandes, C. O. , Olszewska-Guizzo, A. , Dias, R. C. , Vilaça, H. , ..., , & Maia, R. L. (2022) Patterns of human behaviour in public urban green spaces: On the influence of users' profiles, surrounding environment, and space design. Urban Forestry & Urban Greening, 74 , 127668– .

[38] Zhu, X. , Zhang, Y. , & Zhao, W. (2021) Systematic review of environmental quality assessment tools for urban green spaces. Landscape Architecture, 28 ( 9), 90– 95.

[39] & Qu, Z. (1999) The planning and design quality of residential park green space should be further improved—A brief introduction to the planning and design of Japanese residential park foundation. Beijing City Planning & Construction Review, ( 4), 28– 32.

[40] Guo, T. , Dong, L. , & Liu, C. (2020) Identifying features influencing the use of small-scale urban park under the health perspective. Chinese Landscape Architecture, 36 ( 4), 78– 82.

[41] Jens, K. , & Gregg, J. S. (2021) How design shapes space choice behaviors in public urban and shared indoor spaces—A review. Sustainable Cities and Society, 65 , 102592– .

[42] Zerouati, W. , & Bellal, T. (2020) Evaluating the impact of mass housings' in-between spaces' spatial configuration on users' social interaction. Frontiers of Architectural Research, 9 ( 1), 34– 53.

[43] Poppe, L. , Van Dyck, D. , De Keyser, E. , Van Puyvelde, A. , Veitch, J. , & Deforche, B. (2023) The impact of renewal of an urban park in Belgium on park use, park-based physical activity, and social interaction: A natural experiment. Cities, 140 , 104428– .

RIGHTS & PERMISSIONS

© Higher Education Press 2025