Next Article in Journal
Features of the Microalgae and Cyanobacteria Growth in the Flue Gas Atmosphere with Different CO2 Concentrations
Previous Article in Journal
Foreign Trade as a Channel of Pandemic Transmission to the Agricultural Sector in Poland
Previous Article in Special Issue
“Power to” for High Street Sustainable Development: Emerging Efforts in Warsaw, Poland
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 16
Issue 16
10.3390/su16167074
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Correlation of the Walk Score and Environmental Perceptions with Perceived Neighborhood Walkability: The Quantile Regression Model Approach

by
Suin Jin
1 and
Eun Jung Kim
2,*
1
Spatial Information Research Institute, Jeonbuk-do 55365, Republic of Korea
2
Department of Urban Planning, Keimyung University, Daegu 42601, Republic of Korea
*
Author to whom correspondence should be addressed.
Sustainability 2024, 16(16), 7074; https://doi.org/10.3390/su16167074
Submission received: 15 July 2024 / Revised: 14 August 2024 / Accepted: 16 August 2024 / Published: 18 August 2024
(This article belongs to the Collection Towards More Walkable and Liveable Cities: Perceptions, Attitudes, Methods, Technologies and Policies)
Browse Figures
Versions Notes

Abstract

:
The walk score, which is widely used as an index of walkability, does not include pedestrian’s perception, so there is a limit to explaining the level of perceived walkability in a neighborhood. The purpose of this study is to examine how an objectively measured walk score and subjectively measured environmental perceptions correlate with perceived neighborhood walkability. This study conducted a survey on 371 participants aged 18 or older living in Daegu, South Korea to examine perceived neighborhood walkability and perception of the built environment. We measured the walk score based on participants’ location using a geographic information system. We used the quantile regression model, whereby we investigated the effects of explanatory variables (e.g., the walk score, perceptions of the built environment) by classifying perceived neighborhood walkability by quantile into Q10, Q25, Q50, Q75, and Q90. The walk score had a positive association with people with low perceived neighborhood walkability (Q10), but a negative association with people with high perceived neighborhood walkability (Q90). Regarding views of the built environment, in most quantiles, people perceived the environment as walkable if there were abundant green spaces and diverse alternative routes. Conversely, odors, smoke, hills, and stairs impeded walkability. This indicates that along with an objective walkability index, perceptions of the built environment play an important role in determining perceived neighborhood walkability. This implies that our results can help identify appropriate policies to promote walkability for citizens.

1. Introduction

Physical activity not only prevents cardiovascular diseases such as obesity, hypertension, and type 2 diabetes mellitus, but also helps solve problems of depression and social isolation [1,2,3,4]. Although the physical and mental health benefits of physical activity are well known, in practice, people do not engage in sufficient physical activity [5,6]. Physical inactivity has become a serious public health problem worldwide, and the public health sector is aiming to encourage physical activity to promote citizens’ health [7].
Walking and cycling are physical activities that people can easily access and have the advantages of promoting physical and mental health [8]. Recently, personal and shared means of transport such as e-scooters and shared bicycles have been activated to improve people’s mobility [9]. These modes of transport help individuals easily move to their desired destinations, which can encourage people to be more active [10,11].
Despite advances in various personal and shared transport, walking is still the most common physical activity that people can choose and is an inexpensive and environmentally friendly means of transport [12,13,14]. However, walking has decreased with the increasing supply of automobiles [15,16,17,18]. Various studies have been conducted to encourage walking in terms of urban planning, traffic engineering, and public health, and many scholars have discovered that the built environment has a significant effect on walking [19,20]. Thus, it is essential to understand which characteristics of the built environment foster walking [21,22].
Diverse indices such as the walkability index, the walk score, and the pedestrian index of the environment have been developed to quantitatively measure the overall walkability of a neighborhood [23]. The walk score is a walkability index used worldwide; it is available for free at the link and is highly accessible [24]. The walk score is available for the US, Canada, Australia, and New Zealand, but not for European or Asian countries. For this reason, some researchers have applied the walk score methodology to measure the walk score in Asian cities [25,26].
In addition, since cities in Asia (such as South Korea) have high density, unlike Western nations, there could be limitations in explaining walkability using the walk score [27]. The walk score is a walkability index based on the access to amenities needed for daily life (e.g., restaurants, coffee shops, libraries, etc.). Therefore, high-density cities such as Seoul in South Korea tend to have a higher walk score than other cities in the US and Canada, except for a few high-density cities, due to their high access to amenities. In response, Kim and colleagues tried to verify the validity and reliability of the score value after measuring the walk score in Seoul, South Korea [26,28]. Hence, it is necessary to determine whether the walk score can explain walkability as actually perceived by pedestrians in the Asian context.
The walk score is based on a geographic information system (GIS) for access to a destination. Towne et al. [29] found that these walk scores did not completely explain the neighborhood walkability because the conditions of walking infrastructure were not considered. On the other hand, pedestrians’ perceptions of the quality of the built environment play an important role in their choice on whether to walk [30,31]. As such, to create a walkable city, it is necessary to examine pedestrians’ perceptions of the built environment [32]. Specifically, research must be conducted to explore environments in which pedestrians prefer to walk and continue walking for a long time [14,33]. Adkins et al. [34] argued that it is necessary to focus on the quality of walking experienced by pedestrians and suggested that walking can be improved by understanding people’s perceptions of the built environment.
As discussed above, walking is a healthy means of transport that promotes physical activity and is affected by the built environment. Therefore, it is important to understand the aspects of the environment that can promote walking. The walk score is an objectively measured index. However, since the walk score does not include people’s subjective perceptions, it has limitations in explaining the level of walkability that people perceive when walking. Therefore, we aimed to study the relationship between the walk score and perceived neighborhood walkability, and to further investigate whether perceptions of the built environment are associated with perceived neighborhood walkability.
This study is organized as follows. Section 2 provides a review of literature on walk score, elements of walkable neighborhood environment, and the usefulness of the quantile regression model. Section 3 describes the study area and participant recruitment, measure, and analytical methods. Section 4 shows the results, and Section 5 includes discussions. Finally, Section 6 summarizes the main findings of the study.

2. Literature Review

2.1. The Walk Score

The walk score gauges the walkability of a neighborhood and is used in numerous fields such as real estate, urban planning, public health, and finance [24]. The walk score is determined by access to nine types of amenities (grocery stores, restaurants, shopping, coffee, banks, parks, schools, books, entertainment) and two kinds of pedestrian friendliness (intersection density and average block length). The walk score has been used previously to promote walkability and has shown a significant relationship with physical activity and health [35,36,37]. For example, Brown et al. [38] investigated the relationship between the walk score and purposive walking and found that a 10-point increase in the walk score led to a 19% increase in the possibility of walking to one’s destination. Similarly, Cole et al. [37] examined the association between the walk score and walking for transport to and from one’s home among adults in Australia. They revealed that residents living in areas with relatively higher walk scores (≥50 points) were more likely to walk 30 min a day compared to residents living in areas with lower walk scores (≤24 points). As such, the walk score has a positive effect on promoting walking and is an efficient indicator for determining the walkability of a neighborhood. However, most prior studies have been conducted in Western countries, and while some have looked at Asian countries, there is still a need to review whether the walk score is applicable to the Asian context [39].

2.2. Elements of Walkable Neighborhood Environment

This section explores studies that review the elements of creating a walkable environment. Many studies have been conducted on standards for improving walking quality and evaluating walkability in the built environment; we reviewed some representative articles published in the past 10 years that present elements for creating a walkable environment (Table 1). Khder et al. [40] identified comfort, safety, accessibility, and connectivity as key factors in encouraging walking. Moura et al. [41] presented the framework of the seven Cs (connected, convenient, comfortable, convivial, conspicuous, coexistence, commitment) for a pedestrian-friendly context, and Papageorgiou et al. [42] proposed mobility, connectivity, comfort, safety, and convenience as five elements related to walkability. Arisar et al. [43] suggested connectivity, convenience, safety, and ease as requirements for creating a walkable campus, and Harun et al. [44] presented comfort, connectivity, safety, and accessibility as factors that enable an interesting walking experience. Naharudin et al. [45] and Manzolli et al. [46] mentioned five criteria (comfort/comfortability, connectivity, convenience, conviviality, and conspicuous/conspicuousness) that are most commonly used as important walkable elements. Darmawan and Rahmi [47] argued that a walkable environment must be well-connected, convenient, comfortable, safe, convivial, and conspicuous. Loo [48] proposed three walkability principles (convenience, comfort, and safety) as a person-centered, place-based methodological framework to promote walking. Fonseca et al. [49] stated that pedestrian networks must be convenient, comfortable, connected, safe, and attractive. Many articles examined in this study suggest that comfort, connectivity, convenience, and safety can create a walkable environment. In at least seven of the ten studies reviewed here, these four factors were included as elements of the walkable neighborhood environment.
These four attributes (comfort, connectivity, convenience, and safety) are important as the elements of walkable neighborhood environments for the following reasons. First, comfort is associated with the emotional reaction of pedestrians, and this attribute is mainly affected by a pedestrian environment [49]. A comfortable walkaway is an environment where people want to walk, and it can increase walking time by making pedestrians feel shorter on the route [50,51]. Second, connectivity is a factor that assesses how well-connected the load network is and whether walking is uninterrupted [52]. Well-connected streets are characterized by the absence of pedestrian obstacles, various alternative routes to the destination, and more [53]. Third, convenience means being able to easily reach a destination and places of daily amenities within walking distance [54]. Convenience is associated with ease and efficiency of walking, which encourages people to walk [55,56]. Lastly, safety is one of the important factors of walkable environments. It should consider whether pedestrians are safe from exposure to traffic accidents and crime [30,52]. High rates of traffic collisions and crimes reduce willingness to walk and increase dependence on vehicles. Kim and Jin [52] proposed these four elements: the 3Cs (comfort, connectivity, and convenience) and 1S (safety) as aspects of a walkable environment.

2.3. Usefulness of the Quantile Regression Model

Most studies generally use the ordinary least squares (OLS) regression to analyze the correlation between their dependent and independent variables [57]. The OLS model has the disadvantage of ignoring subsamples that deviate from the average [58]. OLS analysis can be conducted by dividing the samples, but this may cause sample selection bias [57,59]. Accordingly, a quantile regression model has been proposed as an alternative to the OLS model [60,61]. Since the quantile regression model uses all samples without dividing them, there is no sample selection bias, and it is helpful in explaining the effect of independent variables at any point in the distribution because it can explain the conditional distribution of the dependent variable [62]. Moreover, because the quantile regression model provides estimated regression coefficients at various points along the distribution of the dependent variable, it can explain the distribution located at the extremes of the dependent variable [63,64].
Prior studies on walkability have primarily scrutinized only the relationship between dependent and independent variables [34,65,66,67]. However, an individual’s walking activity varies depending on the aspects of the built environment, and not all factors of the built environment have an equal impact on perceived walkability. To improve walkability, it is necessary to study the full distribution instead of the average effects of perceived neighborhood walkability. In other words, since we measured perceived neighborhood walkability (the dependent variable in our study) subjectively, we must focus on the full distribution, including people at both extremes. Hence, we performed an analysis using the quantile regression model.

3. Materials and Methods

3.1. Study Area and Participant Recruitment

The research site was the city of Daegu in South Korea. The city of Daegu comprises an area of approximately 883.7 km2 and consists of eight districts and counties. Daegu had a population of approximately 2,385,412 people as of 2021, making it the fourth most populous city in South Korea. To transform itself into a walkable city, Daegu has been making efforts to encourage walking activity among citizens through myriad initiatives such as the Transit Mall Project and the Safe Walking Environment Project [68].
We examined perceived neighborhood walkability and perceptions of the built environment through a survey, which we administered from 12 October to 8 November 2022. The survey was conducted by three experienced researchers face-to-face, and citizens aged 18 or older living in Daegu were selected as subjects for the study. We collected 487 samples, of which 371 questionnaires (100% answered the questions) available in this study were considered valid responses and were used for analysis. This sample size of 371 out of the total population, aged 18 or older, produced a 95% confidence interval of accuracy and a 5.09% margin of error ratio. Participants were randomly selected from residents.
The locations of the research site and the respondents’ homes are shown in Figure 1. This study was approved by the research team’s Institutional Review Board. The survey was conducted after obtaining informed consent from the participants.

3.2. Measures

3.2.1. The Dependent Variable

We examined perceived neighborhood walkability, the dependent variable in our study, through the abovementioned survey. We asked the respondents to rate the walkability of the neighborhood in which they lived on a scale of 0–100 points. One of the purposes of this study is to examines the association between the objectively measured variable of walk score and subjectively measured variables of neighborhood walkability. Therefore, we aligned the perceived neighborhood walkability used in this study with the walk score ranging from 0 to 100. Moreover, we also added the descriptions of the walkability levels in the questionnaire as those of the walk score, so that survey participants can evaluate the perceived walkability of their neighborhood well. The following are five degrees of the walk score level provided on its official website. Communities with a walk score of 0–24 are ‘car-dependent: almost all errands require a car’, 25–49 are ‘car-dependent: most errands require a car’, 50–69 are ‘somewhat walkable: some errands can be accomplished on foot’, 70–89 are ‘very walkable: most errands can be accomplished on foot’, and 90–100 are ‘walkers’ paradise: daily errands do not require a car’ [69]. The mean perceived neighborhood walkability was 76.10 (Standard Deviation (SD) = 17.57).

3.2.2. The Independent Variables

We calculated the walk score, the key independent variable in this study, based on the addresses of the respondents using ArcGIS 10.5 (ESRI, Redlands, CA, USA). The walk score is computed via access to the nine types of amenities, and the basic score of a relevant point is provided by the location, number, and weight of amenities. In Table 2, the total count and weight refer to the number of amenities and the assigned weight. If the total count is greater than 1, it means that multiple amenities are included, and the nth-nearest amenity of that type will receive the nth-assigned weight. Amenities are weighted higher in order of importance to the level of walkability: grocery stores and restaurants have a total weight of 3, shopping and coffee have a total weight of 2, and the rest have a total weight of 1. For example, a convenience store would include only the closest one and is weighted by 3 points. For restaurants, up to ten are included in the order of proximity, with weighted points of 0.75, 0.45, 0.25, …, 0.2, depending on the proximity. A penalty may be applied to the basic score for the two types of pedestrian friendliness: intersection density and average block length.
More specifically, access to amenities applies the distance decay function, and amenities located within 400 m (0.25 miles) receive the maximum score; the farther the distance, the lower the score. For pedestrian friendliness, a lower intersection density and longer average block length led to greater point deduction, and a penalty of up to 10% can be applied. The walk score is determined in the range of 0–100 points.
To calculate walk scores for the addresses of the survey respondents, we collected data on the nine types of amenities and the two kinds of pedestrian friendliness from the D-Data Hub [70], the Financial Supervisory Service of Korea [71], Road Name Address [72], BigData MarketC [73], and the Spatial Information Open Platform [74]. As shown in Table 3, the mean of the walk scores of the survey respondents (N = 371) was 77.73 (SD = 9.24).
As another independent variable, we conducted the survey to examine the respondents’ perceptions of the built environment of the neighborhoods where they lived; a total of 17 items for 3Cs+S were mentioned previously in the literature review section. We selected these items based on previous studies. The details are as follows: First, the items corresponding to comfort include green spaces, natural scenery, street cleanliness, odors and smoke, and noise level. Second, connectivity involves multiple alternative routes, sidewalk connections, and pedestrian obstacles. Third, for the convenience factor, there are numerous facilities, sidewalk widths, sidewalk conditions, hills and stairs, and pedestrian shelters. Fourth, safety factors include pedestrian segregation, crosswalks and traffic lights, traffic volume, and security facilities. As an example, in an item about green spaces, the question asked, “Do you think there are enough green spaces when walking in your neighborhood?” For the items that negatively affected walking, such as odors and smoke, noise level, pedestrian obstacles, and hills and stairs, the survey asked if the respondents were uncomfortable when walking due to each item. For example, regarding hills and stairs, it asked, “Do you feel uncomfortable with the slope and stairs of the street when walking in your neighborhood?” These variables on perceptions of the built environment were rated on a 5-point Likert scale (1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree).

3.2.3. Control Variables

The control variables used in this study included the respondents’ individual traits such as gender, age, perceived distance of neighborhood unit, and car ownership, which we collected through the survey. As for the descriptive statistics on individual traits, 37.5% were male and 62.5% were female, and the average age was 34.80 (SD = 14.46). Regarding car ownership, 57.4% said they did not own a car and 42.6% said they did.

3.3. Methodology

We used a quantile regression model to examine the factors that affect perceived neighborhood walkability. The quantile regression model, first proposed by Koenker and Bassett [75], can estimate different regression coefficients according to the quantile of the dependent variable, which is especially useful when investigating the quantile-specific characteristics of the dependent variable in more detail. Accordingly, we studied how the effects of the independent variables differed between quantiles with low and high perceived neighborhood walkability. The quantile regression model does not respond sensitively to outliers of the dependent variable and has the advantage of being able to scrutinize the relationship with independent variables according to the conditions of the dependent variable without losing data [75]. We performed the analysis by applying the quantile regression model. We used R software (version 4.2.0) for the analysis.

4. Results

Table 4 displays the results of the quantile regression model for perceived neighborhood walkability. We estimated the regression coefficients to be quantiles at 0.10, 0.25, 0.50, 0.75, and 0.90. The quantile regression model portrays the difference in the regression coefficients according to the distribution of the dependent variable for each quantile. Figure 2 depicts a graph of the quantile regression estimates.

4.1. Effect of the Walk Score on Perceived Neighborhood Walkability

The walk score demonstrates a statistically significant effect on perceived neighborhood walkability at the extremes of the distribution (Q10, Q90), but the direction was the opposite. In Q10 (b = 0.412), respondents with a high walk score claimed that their neighborhood was more walkable, whereas in Q90 (b = −0.092), those with a high walk score stated that their neighborhood was not walkable. Figure 2a indicates that the estimates of the quantile regression coefficients decline from positive to negative as the quantile increases. Interestingly, this implies that the walk score has different effects on determining neighborhood walkability depending on the quantile. In other words, for those who rated their neighborhood walkability in the bottom 10% (Q10), a higher walk score (better access to amenities) played a positive role in determining neighborhood walkability. For those who rated their neighborhood walkability in the top 10% (Q90), this had the reverse effect.

4.2. How Perceptions of the Built Environment Affect Perceived Neighborhood Walkability

This section explores the association between perceptions of the built environment and perceived neighborhood walkability. Prior to the main analysis, we estimated internal consistency reliability for a 3Cs+S framework by Cronbach’s alpha coefficient. Cronbach’s alpha coefficient of comfort, connectivity, convenience, and safety were 0.72, 0.50, 0.74, and 0.64, respectively. Of course, we reversed code items of odors and smoke, noise level, pedestrian obstacles, and hill and stairs, which negatively affect walking. Although the threshold of the Cronbach’s alpha considered is more than 0.6, some studies considered the value of 0.5 as still acceptable [76,77,78]. So, Cronbach’s alpha in all four dimensions could be accepted.
We investigated perceptions on 17 built environments, and finally used eight items as variables, excluding variables with multicollinearity problems. The results are as follows: First, among the variables corresponding to the comfort factor, green spaces and odors and smoke were important variables (i.e., sub-items of the comfort factor) that determined perceived neighborhood walkability. Green spaces served as a statistically significant variable in all quantiles of the quantile regression model, and the effect of green spaces on the dependent variable was about three times greater in Q10 (b = 5.371) than in Q90 (b = 1.743). This indicates that the effect of green spaces is at least three times greater for Q10 (those who rated their neighborhood walkability in the bottom 10%) than for Q90 (those who rated their neighborhood walkability in the top 10%). The regression coefficient of green spaces generally decreased as the quantile of the dependent variable increased (Q10: 5.371 → Q90: 1.743). Figure 2b shows that the graph of the quantile regression estimates of green spaces exhibits a decline as the perceived neighborhood walkability increases. In contrast, odors and smoke showed a negative regression coefficient, which was statistically significant at Q50 and above. In other words, odors and smoke decreased perceived neighborhood walkability. In particular, the effect of Q50 (b = −2.480) was greater than that of the other quantiles, and the effect gradually declined from Q50 and above. Street cleanliness did not show a statistically significant effect on perceived neighborhood walkability in the quantile regression model.
Second, multiple alternative routes, used as a variable of connectivity, had a great effect on perceived neighborhood walkability. Further, multiple alternative routes were statistically significant in all quantiles. In other words, multiple alternative routes improved perceived neighborhood walkability, and the effect of multiple alternative routes decreased as the quantile increased (Q10: 6.458→Q90: 2.625), which is also indicated in Figure 2d.
Third, among the variables corresponding to convenience, hills and stairs were the most noticeable. This had a negative impact on perceived neighborhood walkability. However, we noted statistically significant outcomes for Q10 (b = −2.826) through Q50 (b = −1.546), but the results were not statistically significant in Q75 and Q90. People who experienced inconveniences in walking due to hills and stairs tended to perceive the area as not walkable. The sidewalk width was found to have a positive correlation with perceived neighborhood walkability, but this was statistically significant only for Q10 (b = 3.664).
Fourth, traffic volume and security facilities corresponding to safety did not have a significant influence on perceived neighborhood walkability; traffic volume and security facilities were statistically significant in Q10 and Q50, respectively, but not in other quantiles. In the quantile regression model, we found statistically significant results in some quantiles. As for traffic volume, people experiencing inconveniences in walking due to traffic volume tended to perceive their neighborhood as not walkable in Q10 (b = −3.872). In other quantiles, there were no statistically significant outcomes. For security facilities, we only found statistically significant results for Q50 (b = 1.794), which indicates that people perceive their neighborhood as walkable when security facilities are properly installed.

4.3. How Individual Characteristics Affect Perceived Neighborhood Walkability

For individual traits, the results were not statistically significant for most of the variables. Gender, age, and car ownership did not have a significant effect on the dependent variable in the quantile regression models. However, perceived distance of neighborhood unit showed a significant outcome only in Q10 (b = 3.723), indicating that those who rated their neighborhood walkability in the bottom 10% perceived their neighborhood as more walkable if they thought the perceived distance of the neighborhood unit was longer.

5. Discussion

As walking has become increasingly important, research has been conducted to encourage it. We examined the correlation between the walk score and perceived neighborhood walkability using a quantile regression model, and we scrutinized the effect of perceptions of the built environment on perceived neighborhood walkability. The implications of our results are as follows. First, the effect of the walk score and perceptions of the built environment changes according to the distribution of perceived neighborhood walkability in the quantile regression model. The quantile regression model showed that the effects of the walk score and perceptions of the built environment vary depending on the level of perceived neighborhood walkability, which contributes to an in-depth understanding of the relationship between perceived neighborhood walkability and its determinants. Furthermore, the quantile regression model can help policymakers decide which specific groups to focus on to improve walkability. This signals that the quantile regression model is useful for selecting an effective policy according to the lower or upper quantiles of perceived walkability.
Second, the walk score revealed interesting results depending on perceived neighborhood walkability. Those who rated their neighborhood as not walkable and those who rated it as walkable at the extremes of the distribution exhibited conflicting results. Specifically, those who rated their neighborhood walkability in the bottom 10% perceived their neighborhood as walkable if the walk score was high, whereas those who rated their neighborhood walkability in the top 10% perceived their neighborhood as not walkable if the walk score was high. This finding must be interpreted considering the characteristics of the walk score calculated based on access to amenities. In other words, access to amenities can promote walking for those who rated their neighborhood walkability as low, but it had the reverse effect for those who rated their neighborhood walkability as high. Many amenities can foster neighborhood walkability to a certain extent, but too many amenities may cause urban problems such as parking issues, traffic congestion, and noise, which reduce neighborhood walkability.
Third, perceptions of the built environment had a substantial effect on determining perceived neighborhood walkability. The discussion by four elements of the built environmental perceptions is summarized as follows.
  • Comfort
For people in the full distribution of perceived neighborhood walkability, more abundant green spaces while walking led to higher perceived walkability. This is similar to the findings of a previous study showing that an aesthetically satisfying environment has a positive correlation with walking activity [21]. Green spaces are walking-inducing factors that reduce stress and provide enjoyment while walking, and they play an important role in determining whether people choose to walk in their neighborhoods [23,79,80]. Hence, an adequate arrangement of green spaces is essential for creating a walkable street, and a walking environment must be designed such that visually rich greenness can be viewed when walking. In particular, the effect of green spaces was greatest for those who rated their perceived neighborhood walkability in the bottom 10%, which shows that green spaces have a substantial effect on determining walkability for those who perceive their neighborhood walkability as low. In the case of street cleanliness, litter on the streets was considered to be a factor that hindered walking by reducing aesthetics, but this study did not find a statistically significant association between street cleanliness and the perceived neighborhood walkability. Meanwhile, people who feel uncomfortable while walking due to odors and smoke tend to perceive their neighborhood as not walkable. In particular, odors and smoke were important factors for those who perceived their neighborhood as walkable (Q50, Q75, and Q90). This implies that it is necessary to consider improving the quality of walking rather than enhancing the physical walking environment for people with high perceived walkability. This outcome is similar to that of Villanueva et al. [81] who demonstrated that people who live in a walkable neighborhood walk more frequently than those who do not.
  • Connectivity
The existence of multiple alternative routes increased perceived neighborhood walkability. This indicates that people see their neighborhood as walkable when there are diverse alternative routes for pedestrians to choose from to get to their destination. Multiple alternative routes had a positive effect on perceived neighborhood walkability in all quantiles, and the effect was greatest in quantile (Q10) where perceived neighborhood walkability was low. Alternative routes can divert pedestrian traffic concentrated on congested streets and provide alternatives in unexpected situations such as construction and accidents. Furthermore, walkability is expected to improve even further as pedestrians can choose their preferred route when walking. As such, it is necessary to consider providing alternative routes in terms of connectivity as a street design plan to promote walkability in the future.
  • Convenience
Hills and stairs were found to be factors impeding perceived neighborhood walkability. This is similar to the results of previous studies in which people walked less on slopes or streets with stairs [19,82]. Hills and stairs cause physical difficulties in walking, which is why they are seen as not walkable. This suggests that hills and stairs must be considered first for those who rated their neighborhood walkability as low (Q10, Q25, Q50) when improving the walking environment. Sidewalk width had a significant relationship with the perceived neighborhood walkability only in the lowest quartile (Q10). This means that sidewalk width plays an important role in determining walkability for those who rate their perceived neighborhood walkability low. Traffic volume was the only significant result in Q10, indicating that traffic volume was an important factor in walking for those who perceived their neighborhood as less walkable.
  • Safety
Traffic volume was negatively associated with the perceived neighborhood walkability only in Q10. Pedestrians can feel fear of traffic collisions on high-traffic roads. Meanwhile, security facilities were found to have a positive association with perceived neighborhood walkability only in the middle quartile (Q50). Security facilities can reduce the fear of crime while walking, and further promote neighborhood safety. These results can be used to improve street design and walking conditions according to the level of walkability in the neighborhood.
In summary, the effect of perceptions of the built environment on perceived neighborhood walkability varies depending on the quantile, but generally the same direction is shown in all quantiles. In other words, the results of the quantile regression model revealed that green spaces and multiple alternative routes had a positive effect on perceived neighborhood walkability, whereas odors, smoke, hills, and stairs had a negative effect. However, we also confirmed that the effect of perceptions of the built environment varied depending on the level of neighborhood walkability. These results can offer suggestions for policymakers to improve walkability for citizens with low perceived walkability.
The limitations of this study and directions for future research are as follows. We used individual traits such as gender, age, perceived distance of neighborhood unit, and car ownership as control variables. In addition, various personal variables such as walking duration, walking frequency, and perceptions of health may show a significant relationship with perceived neighborhood walkability. Thus, it is necessary to consider variables related to health and walking as control variables in future studies. We examined Daegu, a large city in South Korea. However, more abundant policy implications can be derived by conducting further research that compares large cities, such as Seoul and Busan, or different types of cities such as big, small, and medium-sized cities.
Despite these limitations, this study has several implications. We performed the analysis using the quantile regression model to focus on the full distribution of perceived neighborhood walkability. This is significant in that we investigated the different effects of the walk score and perceptions of the built environment according to the perceived neighborhood walkability quantiles. Furthermore, the walk score is a quantitative walkability index based on access to amenities. This study is significant in exploring the relationship between the walk score and perceived neighborhood walkability to determine whether the walk score can explain walkability in a large city in Asia. Moreover, it is important to study the effects of subjective factors on perceived neighborhood walkability.

6. Conclusions

Walking is a sustainable means of transport and the most accessible physical activity for people. We examined the effect of the walk score and perceptions of the built environment on perceived neighborhood walkability using the quantile regression model. We confirmed that the effect of the walk score varies depending on the level of perceived neighborhood walkability and discovered that perceptions of the built environment have a significant effect on how people perceive neighborhood walkability. This indicates that there is difficulty in improving walkability for everyone using only the walk score. Furthermore, understanding citizens’ perceptions of the built environment can be effective in promoting individual walkability. Built environmental factors that have a positive relationship with perceived neighborhood walkability can be effectively used to design walkable communities. Moreover, we need to pay attention to the areas with lower level of perceived neighborhood walkability, especially to provide better conditions of walking environment for everyone. This study suggested that it is important to consider more green spaces, more alternative routes, wider sidewalk width, less hills and stairs, and lower traffic volume in the pedestrian environment. The results of this study will contribute to establishing urban policies related to the walkable environment from the perspective of urban planning and design.

Author Contributions

Conceptualization, E.J.K.; methodology, E.J.K. and S.J.; software, E.J.K. and S.J.; validation, E.J.K. and S.J.; formal analysis, E.J.K. and S.J.; investigation, E.J.K. and S.J.; resources, E.J.K. and S.J.; data curation, E.J.K. and S.J.; writing—original draft preparation, E.J.K. and S.J.; writing—review and editing, E.J.K. and S.J.; visualization, S.J.; supervision, E.J.K.; project administration, E.J.K.; funding acquisition, E.J.K. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (No. NRF-2021R1A2B5B01002628).

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board of Keimyung University (IRB No. 40525-202207-HR-036-02, 12 October 2022).

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

The data presented in this study are available upon request from the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Oyeyemi, A.L.; Kolo, S.M.; Rufai, A.A.; Oyeyemi, A.Y.; Omotara, B.A.; Sallis, J.F. Associations of Neighborhood Walkability with Sedentary Time in Nigerian Older Adults. Int. J. Environ. Res. Public Health 2019, 16, 1879. [Google Scholar] [CrossRef] [PubMed]
  2. Oreskovic, N.M.; Charles, P.R.; Shepherd, D.T.; Nelson, K.P.; Bar, M. Attributes of form in the built environment that influence perceived walkability. J. Archit. Plann. Res. 2014, 31, 218–232. [Google Scholar] [PubMed]
  3. Yu, J.; Yang, C.; Zhao, X.; Zhou, Z.; Zhang, S.; Zhai, D.; Li, J. The associations of built environment with older people recreational walking and physical activity in a Chinese Small-Scale City of Yiwu. Int. J. Environ. Res. Public Health 2021, 18, 2699. [Google Scholar] [CrossRef] [PubMed]
  4. Zhang, Y.; van Dijk, T.; Wagenaar, C. How the Built Environment Promotes Residents’ Physical Activity: The Importance of a Holistic People-Centered Perspective. Int. J. Environ. Res. Public Health 2022, 19, 5595. [Google Scholar] [CrossRef] [PubMed]
  5. Reisi, M.; Ahmadi Nadoushan, M.; Aye, L. Local walkability index: Assessing built environment influence on walking. Bull. Geogr. Socioecon. Ser. 2019, 46, 7–21. [Google Scholar] [CrossRef]
  6. Dyck, D.V.; Cardon, G.; Deforche, B.; De Bourdeaudhuij, I. Do adults like living in high-walkable neighborhoods? Associations of walkability parameters with neighborhood satisfaction and possible mediators. Health Place 2011, 17, 971–977. [Google Scholar] [CrossRef] [PubMed]
  7. Jensen, W.A.; Brown, B.B.; Smith, K.R.; Brewer, S.C.; Amburgey, J.W.; McIff, B. Active Transportation on a Complete Street: Perceived and Audited Walkability Correlates. Int. J. Environ. Res. Public Health 2017, 14, 1014. [Google Scholar] [CrossRef] [PubMed]
  8. Ton, D.; Duives, D.C.; Cats, O.; Hoogendoorn-Lanser, S.; Hoogendoorn, S.P. Cycling or walking? Determinants of mode choice in the Netherlands. Transp. Res. Part A Policy Pract. 2019, 123, 7–23. [Google Scholar] [CrossRef]
  9. Latinopoulos, C.; Patrier, A.; Sivakumar, A. Planning for e-scooter use in metropolitan cities: A case study for Paris. Transp. Res. Part D Transp. Environ. 2021, 100, 103037. [Google Scholar] [CrossRef]
  10. Zagorskas, J.; Burinskienė, M. Challenges Caused by Increased Use of E-Powered Personal Mobility Vehicles in European Cities. Sustainability 2020, 12, 273. [Google Scholar] [CrossRef]
  11. Jie, F.; Standing, C.; Biermann, S.; Standing, S.; Le, T. Factors affecting the adoption of shared mobility systems: Evidence from Australia. Res. Transp. Bus. Manag. 2021, 41, 100651. [Google Scholar] [CrossRef]
  12. Rafiemanzelat, R.; Emadi, M.I.; Kamali, A.J. City sustainability: The influence of walkability on built environments. Transp. Res. Procedia. 2017, 24, 97–104. [Google Scholar] [CrossRef]
  13. Liu, J.; Zhou, J.; Xiao, L. Built environment correlates of walking for transportation: Differences between commuting and non-commuting trips. J. Transp. Land Use 2021, 14, 1129–1148. [Google Scholar] [CrossRef]
  14. Fonseca, F.; Papageorgiou, G.; Tondelli, S.; Ribeiro, P.; Conticelli, E.; Jabbari, M.; Ramos, R. Perceived walkability and respective urban determinants: Insights from Bologna and Porto. Sustainability 2022, 14, 9089. [Google Scholar] [CrossRef]
  15. Xiao, L.; Yang, L.; Liu, J.; Yang, H. Built Environment Correlates of the Propensity of Walking and Cycling. Sustainability 2020, 12, 8752. [Google Scholar] [CrossRef]
  16. Zeng, F.; Shen, Z. Study on the Impact of Historic District Built Environment and Its Influence on Residents’ Walking Trips: A Case Study of Zhangzhou Ancient City’s Historic District. Int. J. Environ. Res. Public Health 2020, 17, 4367. [Google Scholar] [CrossRef] [PubMed]
  17. Lee, H.-S.; Shepley, M.M. Perceived neighborhood environments and leisure-time walking among Korean adults: An application of the theory of planned behavior. Health Environ. Res. Des. J. 2012, 5, 99–110. [Google Scholar] [CrossRef] [PubMed]
  18. Vale, D.S.; Pereira, M. Influence on pedestrian commuting behavior of the built environment surrounding destinations: A structural equations modeling approach. Int. J. Sustain. Transp. 2016, 10, 730–741. [Google Scholar] [CrossRef]
  19. Moniruzzaman, M.; Páez, A. An investigation of the attributes of walkable environments from the perspective of seniors in Montreal. J. Transp. Geogr. 2016, 51, 85–96. [Google Scholar] [CrossRef]
  20. Gerike, R.; Koszowski, C.; Schröter, B.; Buehler, R.; Schepers, P.; Weber, J.; Wittwer, R.; Jones, P. Built Environment Determinants of Pedestrian Activities and Their Consideration in Urban Street Design. Sustainability 2021, 13, 9362. [Google Scholar] [CrossRef]
  21. Su, M.; Tan, Y.-y.; Liu, Q.-m.; Ren, Y.-j.; Kawachi, I.; Li, L.-m.; Lv, J. Association between perceived urban built environment attributes and leisure-time physical activity among adults in Hangzhou, China. Prev. Med. 2014, 66, 60–64. [Google Scholar] [CrossRef] [PubMed]
  22. Saadi, I.; Aganze, R.; Moeinaddini, M.; Asadi-Shekari, Z.; Cools, M. A Participatory Assessment of Perceived Neighbourhood Walkability in a Small Urban Environment. Sustainability 2022, 14, 206. [Google Scholar] [CrossRef]
  23. Arellana, J.; Saltarín, M.; Larrañaga, A.M.; Alvarez, V.; Henao, C.A. Urban walkability considering pedestrians’ perceptions of the built environment: A 10-year review and a case study in a medium-sized city in Latin America. Transp. Rev. 2020, 40, 183–203. [Google Scholar] [CrossRef]
  24. Walk Score. Walk Score Professional. Available online: https://www.walkscore.com/professional/why-walkscore.php (accessed on 15 December 2022).
  25. Zhang, J.; Tan, P.Y.; Zeng, H.; Zhang, Y. Walkability Assessment in a Rapidly Urbanizing City and Its Relationship with Residential Estate Value. Sustainability 2019, 11, 2205. [Google Scholar] [CrossRef]
  26. Kim, E.J.; Won, J.; Kim, J. Is Seoul Walkable? Assessing a Walkability Score and Examining Its Relationship with Pedestrian Satisfaction in Seoul, Korea. Sustainability 2019, 11, 6915. [Google Scholar] [CrossRef]
  27. Cerin, E.; Sit, C.H.P.; Barnett, A.; Johnston, J.M.; Cheung, M.-C.; Chan, W.-M. Ageing in an ultra-dense metropolis: Perceived neighbourhood characteristics and utilitarian walking in Hong Kong elders. Public Health Nutr. 2014, 17, 225–232. [Google Scholar] [CrossRef]
  28. Kim, E.J.; Kim, Y.-J. A Reliability Check of Walkability Indices in Seoul, Korea. Sustainability 2020, 12, 176. [Google Scholar] [CrossRef]
  29. Towne, S.D.; Won, J.; Lee, S.; Ory, M.G.; Forjuoh, S.N.; Wang, S.; Lee, C. Using walk score™ and neighborhood perceptions to assess walking among middle-aged and older adults. J. Community Health 2016, 41, 977–988. [Google Scholar] [CrossRef] [PubMed]
  30. Fonseca, F.; Ribeiro, P.J.G.; Conticelli, E.; Jabbari, M.; Papageorgiou, G.; Tondelli, S.; Ramos, R.A.R. Built environment attributes and their influence on walkability. Int. J. Sustain. Transp. 2022, 16, 660–679. [Google Scholar] [CrossRef]
  31. De Vos, J.; Lättman, K.; van der Vlugt, A.-L.; Welsch, J.; Otsuka, N. Determinants and effects of perceived walkability: A literature review, conceptual model and research agenda. Transp. Rev. 2023, 43, 303–324. [Google Scholar] [CrossRef]
  32. Rivera-Navarro, J.; Bonilla, L.; Gullón, P.; González-Salgado, I.; Franco, M. Can we improve our neighbourhoods to be more physically active? Residents’ perceptions from a qualitative urban health inequalities study. Health Place 2022, 77, 102658. [Google Scholar] [CrossRef] [PubMed]
  33. Kim, S.; Park, S.; Lee, J.S. Meso- or micro-scale? Environmental factors influencing pedestrian satisfaction. Transp. Res. Part D Transp. Environ. 2014, 30, 10–20. [Google Scholar] [CrossRef]
  34. Adkins, A.; Dill, J.; Luhr, G.; Neal, M. Unpacking walkability: Testing the influence of urban design features on perceptions of walking environment attractiveness. J. Urban Des. 2012, 17, 499–510. [Google Scholar] [CrossRef]
  35. Camhi, S.M.; Troped, P.J.; Garvey, M.; Hayman, L.L.; Must, A.; Lichtenstein, A.H.; Crouter, S.E. Associations between Walk Score and objective measures of physical activity in urban overweight and obese women. PLoS ONE 2019, 14, e0214092. [Google Scholar] [CrossRef] [PubMed]
  36. Wasfi, R.A.; Dasgupta, K.; Orpana, H.; Ross, N.A. Neighborhood walkability and Body Mass Index trajectories: Longitudinal Study of Canadians. Am. J. Public Health 2016, 106, 934–940. [Google Scholar] [CrossRef] [PubMed]
  37. Cole, R.; Dunn, P.; Hunter, I.; Owen, N.; Sugiyama, T. Walk Score and Australian adults’ home-based walking for transport. Health Place 2015, 35, 60–65. [Google Scholar] [CrossRef] [PubMed]
  38. Brown, S.C.; Pantin, H.; Lombard, J.; Toro, M.; Huang, S.; Plater-Zyberk, E.; Perrino, T.; Perez-Gomez, G.; Barrera-Allen, L.; Szapocznik, J. Walk Score®: Associations with purposive walking in recent Cuban immigrants. Am. J. Prev. Med. 2013, 45, 202–206. [Google Scholar] [CrossRef] [PubMed]
  39. Koohsari, M.J.; Sugiyama, T.; Hanibuchi, T.; Shibata, A.; Ishii, K.; Liao, Y.; Oka, K. Validity of Walk Score® as a measure of neighborhood walkability in Japan. Prev. Med. Rep. 2018, 9, 114–117. [Google Scholar] [CrossRef] [PubMed]
  40. Khder, H.M.; Mousavi, S.M.; Khan, T.H. Impact of Street’s Physical Elements on Walkability: A Case of Mawlawi Street in Sulaymaniyah, Iraq. Int. J. Built Environ. Sustain. 2016, 3, 18–26. [Google Scholar] [CrossRef]
  41. Moura, F.; Cambra, P.; Gonçalves, A.B. Measuring walkability for distinct pedestrian groups with a participatory assessment method: A case study in Lisbon. Landsc. Urban Plan. 2017, 157, 282–296. [Google Scholar] [CrossRef]
  42. Papageorgiou, G.; Maimaris, A.; Efstathiadou, T.; Balamou, E. Evaluating Attitudes on the Quality of Service of Pedestrian Networks. WIT Trans. Built Environ. 2017, 176, 35–41. [Google Scholar] [CrossRef]
  43. Arisar, U.; Talpur, M.A.H.; Shar, B.; Ali, M.; Khoso, A. Influence of design characteristics on walkability A study on MUET Campus Jamshoro. Int. J. Eng. Technol. 2018, 2, 13–16. [Google Scholar]
  44. Harun, N.Z.; Nashar, A.; Bachok, S. Walkability factors for a campus street. Plan. Malays. 2020, 18. [Google Scholar] [CrossRef]
  45. Naharudin, N.; Salleh, A.H.; Halim, M.A.; Latif, Z.A. Conceptual Framework for Walkability Assessment for Pedestrian Access to Rail Transit Services by using Spatial-MCDA. IOP Conf. Ser. Earth Environ. Sci. 2020, 540, 012023. [Google Scholar] [CrossRef]
  46. Manzolli, J.A.; Oliveira, A.; Neto, M.D. Evaluating Walkability through a Multi-Criteria Decision Analysis Approach: A Lisbon Case Study. Sustainability 2021, 13, 1450. [Google Scholar] [CrossRef]
  47. Darmawan, A.M.; Rahmi, D.H. Quality of walkability in Peunayong, Banda Aceh. Built Environ. Stud. 2021, 2, 43–50. [Google Scholar] [CrossRef]
  48. Loo, B.P.Y. Walking towards a happy city. J. Transp. Geogr. 2021, 93, 103078. [Google Scholar] [CrossRef]
  49. Fonseca, F.; Fernandes, E.; Ramos, R. Walkable Cities: Using the Smart Pedestrian Net Method for Evaluating a Pedestrian Network in Guimarães, Portugal. Sustainability 2022, 14, 10306. [Google Scholar] [CrossRef]
  50. Sukor, N.S.A.; Fisal, S.F.M. Safety, Connectivity, and Comfortability as Improvement Indicators of Walkability to the Bus Stops in Penang Island. Eng. Technol. Appl. Sci. Res. 2020, 10, 6450–6455. [Google Scholar] [CrossRef]
  51. Hassan, D.K.; Elkhateeb, A. Walking experience: Exploring the trilateral interrelation of walkability, temporal perception, and urban ambiance. Front. Arch. Res. 2021, 10, 516–539. [Google Scholar] [CrossRef]
  52. Kim, E.J.; Jin, S. Walk Score and Neighborhood Walkability: A Case Study of Daegu, South Korea. Int. J. Environ. Res. Public Health 2023, 20, 4246. [Google Scholar] [CrossRef] [PubMed]
  53. Koohsari, M.J.; Sugiyama, T.; Lamb, K.E.; Villanueva, K.; Owen, N. Street connectivity and walking for transport: Role of neighborhood destinations. Prev. Med. 2014, 66, 118–122. [Google Scholar] [CrossRef] [PubMed]
  54. Ye, Y.; Jia, C.; Winter, S. Measuring Perceived Walkability at the City Scale Using Open Data. Land 2024, 13, 261. [Google Scholar] [CrossRef]
  55. Herrmann-Lunecke, M.G.; Mora, R.; Vejares, P. Perception of the built environment and walking in pericentral neighbourhoods in Santiago, Chile. Travel Behav. Soc. 2021, 23, 192–206. [Google Scholar] [CrossRef]
  56. Zumelzu, A.; Estrada, M.; Moya, M.; Troppa, J. Experiencing Public Spaces in Southern Chile: Analysing the Effects of the Built Environment on Walking Perceptions. Int. J. Environ. Res. Public Health 2022, 19, 12577. [Google Scholar] [CrossRef] [PubMed]
  57. Kim, M.-J. Understanding the determinants on household electricity consumption in Korea: OLS regression and quantile regression. Electr. J. 2020, 33, 106802. [Google Scholar] [CrossRef]
  58. Yuan, H.; Golpelwar, M. Testing subjective well-being from the perspective of social quality: Quantile regression evidence from Shanghai, China. Soc. Indic. Res. 2013, 113, 257–276. [Google Scholar] [CrossRef]
  59. Gim, T.H.T. Quantile regression on the nonlinear relationship between land use and trip time. Pap. Reg. Sci. 2021, 100, 1055–1078. [Google Scholar] [CrossRef]
  60. Binder, M.; Coad, A. From Average Joe’s happiness to Miserable Jane and Cheerful John: Using quantile regressions to analyze the full subjective well-being distribution. J. Econ. Behav. Oragan. 2011, 79, 275–290. [Google Scholar] [CrossRef]
  61. McCord, M.J.; Davis, P.T.; Bidanset, P.; McCluskey, W.; McCord, J.; Haran, M.; MacIntyre, S. House prices and neighbourhood amenities: Beyond the norm? Int. J. Hous. Mark. Anal. 2018, 11, 263–289. [Google Scholar] [CrossRef]
  62. Perez-Sanchez, V.R.; Serrano-Estrada, L.; Marti, P.; Mora-Garcia, R.-T. The What, Where, and Why of Airbnb Price Determinants. Sustainability 2018, 10, 4596. [Google Scholar] [CrossRef]
  63. Giambona, F.; Porcu, M. Student background determinants of reading achievement in Italy. A quantile regression analysis. Int. J. Educ. Dev. 2015, 44, 95–107. [Google Scholar] [CrossRef]
  64. Petscher, Y.; Logan, J.A.R. Quantile regression in the study of developmental sciences. Child Dev. 2014, 85, 861–881. [Google Scholar] [CrossRef]
  65. Jun, H.-J.; Hur, M. The relationship between walkability and neighborhood social environment: The importance of physical and perceived walkability. Appl. Geogr. 2015, 62, 115–124. [Google Scholar] [CrossRef]
  66. Rani, K.; Boora, A.; Gr, B.; Parida, M. Which Factors Affect “Walkability” of Pedestrians on Sidewalk in Indian cities? Proc. East. Asia Soc. Transp. Stud. 2018, 11, 1–19. [Google Scholar]
  67. Bornioli, A.; Parkhurst, G.; Morgan, P.L. Affective experiences of built environments and the promotion of urban walking. Transp. Res. Part A Policy Pract. 2019, 123, 200–215. [Google Scholar] [CrossRef]
  68. Chung, U.-K. A Basic Study on the Development of Pedestrian Friendly City in Daegu; Daegu Gyeongbuk Development Institute: Daegu, Republic of Korea, 2018. [Google Scholar]
  69. Walk Score. Walk Score Methodology. Available online: https://www.walkscore.com/methodology.shtml (accessed on 27 May 2022).
  70. D-Data Hub. Available online: https://data.daegu.go.kr/open/main.do (accessed on 13 May 2022).
  71. Financial Supervisory Service of Korea. Available online: https://www.fcsc.kr/ (accessed on 25 May 2022).
  72. Road Name Address. Available online: https://business.juso.go.kr/addrlink/main.do (accessed on 1 June 2022).
  73. BigData MarketC. Available online: https://www.bigdata-culture.kr/bigdata/user/data_market/detail.do?id=33660160-404b-11eb-af9a-4b03f0a582d6 (accessed on 1 June 2022).
  74. Spatial Information Open Platform. Available online: https://www.vworld.kr/dtmk/dtmk_ntads_s001.do (accessed on 27 May 2022).
  75. Koenker, R.; Bassett, G. Regression quantiles. Econometrica 1978, 46, 33–50. [Google Scholar] [CrossRef]
  76. Dall’Oglio, A.M.; Rossiello, B.; Coletti, M.F.; Caselli, M.C.; Ravà, L.; Di Ciommo, V.; Orzalesi, M.; Giannantoni, P.; Pasqualetti, P. Developmental evaluation at age 4: Validity of an Italian parental questionnaire. J. Paediatr. Child Health 2010, 46, 419–426. [Google Scholar] [CrossRef]
  77. Dall’Oglio, I.; Nicolò, R.; Di Ciommo, V.; Bianchi, N.; Ciliento, G.; Gawronski, O.; Pomponi, M.; Roberti, M.; Tiozzo, E.; Raponi, M. A Systematic Review of Hospital Foodservice Patient Satisfaction Studies. J. Acad. Nutr. Diet. 2015, 115, 567–584. [Google Scholar] [CrossRef]
  78. Nguyen, M.C.; Gabbe, S.G.; Kemper, K.J.; Mahan, J.D.; Cheavens, J.S.; Moffatt-Bruce, S.D. Training on mind-body skills: Feasibility and effects on physician mindfulness, compassion, and associated effects on stress, burnout, and clinical outcomes. J. Posit. Psychol. 2020, 15, 194–207. [Google Scholar] [CrossRef]
  79. Singh, R. Factors affecting walkability of neighborhoods. Procedia Soc. Behav. Sci. 2016, 216, 643–654. [Google Scholar] [CrossRef]
  80. Barton, J.; Hine, R.; Pretty, J. The health benefits of walking in greenspaces of high natural and heritage value. J. Intergr. Environ. Sci. 2009, 6, 261–278. [Google Scholar] [CrossRef]
  81. Villanueva, K.; Knuiman, M.; Nathan, A.; Giles-Corti, B.; Christian, H.; Foster, S.; Bull, F. The impact of neighborhood walkability on walking: Does it differ across adult life stage and does neighborhood buffer size matter? Health Place 2014, 25, 43–46. [Google Scholar] [CrossRef] [PubMed]
  82. Barros, A.P.; Martínez, L.M.; Viegas, J.M. How urban form promotes walkability? Transp. Res. Procedia. 2017, 27, 133–140. [Google Scholar] [CrossRef]
Sustainability 16 07074 g001
Figure 1. Study areas and the respondents’ home locations.
Figure 1. Study areas and the respondents’ home locations.
Sustainability 16 07074 g001
Sustainability 16 07074 g002
Figure 2. Quantile regression estimates for the walk score and perceptions of the built environment. Note: Vertical axes show coefficient estimates of independent variables in the quantile regression model. Horizontal axes depict quantiles of the dependent variable. The dashed black line denotes the quantile regression coefficient estimation. The grey area shows the 95% confidence intervals of the coefficients. The solid red horizontal line indicates the OLS coefficient, and the red dashed line embodies the 95% confidence interval of the OLS estimation.
Figure 2. Quantile regression estimates for the walk score and perceptions of the built environment. Note: Vertical axes show coefficient estimates of independent variables in the quantile regression model. Horizontal axes depict quantiles of the dependent variable. The dashed black line denotes the quantile regression coefficient estimation. The grey area shows the 95% confidence intervals of the coefficients. The solid red horizontal line indicates the OLS coefficient, and the red dashed line embodies the 95% confidence interval of the OLS estimation.
Sustainability 16 07074 g002
Table 1. The elements that create a walkable environment and suggestions of such elements for the present study.
Table 1. The elements that create a walkable environment and suggestions of such elements for the present study.
Element[40][41][42][43][44][45][46][47][48][49]Elements Used in This Study
Accessibilityx x
Attractiveness x
Coexistence x
Comfortxxx xxxxxxx
Commitment x
Connectivity xxxxxxx xx
Conspicuousness x xxx
Conveniencexxxx xxxxxx
Conviviality x xxx
Ease x
Mobility x
Safetyx xxx xxxx
Table 2. Explanation of the components of the walk score and data source a.
Table 2. Explanation of the components of the walk score and data source a.
CategoryTotal CountWeightData Source
Access to amenities
Grocery stores13[70]
Restaurants100.75, 0.45, 0.25, 0.25, 0.225, 0.225, 0.225, 0.225, 0.2, 0.2
Shopping50.5, 0.45, 0.4, 0.35, 0.3
Coffee21.25, 0.75
Banks11[71]
Parks11[72]
Schools11
Books11[73]
Entertainment11[70]
Pedestrian friendliness
Intersection density
(intersections per square miles)		Over 200: no penalty
150–200: 1% penalty
120–150: 2% penalty
90–120: 3% penalty
60–90: 4% penalty
Under 60: 5% penalty		[74]
Average block length
(in meters)		Under 120 m: no penalty
120–150 m: 1% penalty
150–165 m: 2% penalty
165–180 m: 3% penalty
180–195 m: 4% penalty
Over 195 m: 5% penalty
a We obtained the data by requesting them from the relevant data source. Source: Summarized based on [69].
Table 3. Measurement of the variables and descriptive statistics.
Table 3. Measurement of the variables and descriptive statistics.
Variable(s)DescriptionFrequency (%)Mean (SD)
Dependent variable
Perceived neighborhood walkabilityContinuous: 0–100 76.10 (17.57)
Independent variables
Walk score aContinuous: 0–100 77.73 (9.24)
Perceptions of the built environment
ComfortGreen spacesCategorical: 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree 3.57 (1.07)
Natural scenery 2.96 (1.22)
Street cleanliness 3.16 (1.04)
Odors and smoke 2.59 (1.02)
Noise level 2.85 (1.10)
ConnectivityMultiple alternative routes 3.77 (0.98)
Sidewalk connections 3.54 (1.05)
Pedestrian obstacles 2.52 (1.03)
ConvenienceVarious facilities 4.03 (0.92)
Sidewalk width 3.49 (1.12)
Sidewalk conditions 3.61 (0.98)
Hills and stairs 2.44 (1.09)
Pedestrian shelters 3.29 (1.17)
SafetyPedestrian segregation 3.69 (1.06)
Crosswalk and traffic lights 3.81 (0.87)
Traffic volume 3.86 (0.91)
Security facilities 3.64 (0.92)
Control variables
Individual characteristicsGenderBinary: 0 = male
1 = female		139 (37.5)
232 (62.5)
AgeContinuous: Age 34.80 (14.46)
Perceived distance of neighborhood unitCategorical: 1 = less than 400 m radius (5 min walk), 2 = 400–800 m radius (10 min walk),
3 = 800 m–1.6 km radius (20 min walk),
4 = 1.6 km–2.4 km radius (30 min walk), 5 = more than 2.4 km radius (more than 30 min walk)		 3.11 (1.00)
Car ownershipBinary: 0 = no
1 = yes		213 (57.4)
158 (42.6)
a We calculated this variable directly based on GIS; we collected details on the other variables from the survey.
Table 4. Results of the quantile regression model for perceived neighborhood walkability.
Table 4. Results of the quantile regression model for perceived neighborhood walkability.
VariablesQuantile Regression
Q10Q25Q50Q75Q90
Walk score0.412 **
(0.201)		0.192
(0.176)		−0.071
(0.086)		−0.022
(0.054)		−0.092 *
(0.049)
Perceptions of the built environment
ComfortGreen spaces5.371 ***
(1.703)		2.647 **
(1.220)		2.306 ***
(0.722)		3.010 ***
(0.545)		1.743 *
(0.906)
Street cleanliness−1.921
(1.590)		0.035
(1.054)		0.150
(0.745)		0.154
(0.596)		0.490
(0.667)
Odors and smoke−1.095
(1.469)		−1.624
(1.065)		−2.480 ***
(0.634)		−2.249 ***
(0.502)		−1.600 *
(0.941)
ConnectivityMultiple alternative routes6.458 ***
(1.886)		5.472 ***
(1.267)		3.611 ***
(0.815)		3.817 ***
(0.796)		2.625 ***
(0.760)
ConvenienceSidewalk width3.664 **
(1.432)		1.550
(1.084)		1.272
(0.824)		0.861
(0.635)		1.149
(0.895)
Hills and stairs−2.826 *
(1.615)		−3.132 ***
(0.984)		−1.546 **
(0.736)		−0.742
(0.532)		−0.665
(0.685)
SafetyTraffic volume−3.872 ***
(1.212)		−1.190
(1.226)		−0.007
(0.749)		−0.163
(0.506)		−0.102
(0.765)
Security facilities1.795
(2.080)		−0.382
(1.429)		1.794 *
(0.942)		−0.197
(0.841)		0.085
(0.961)
Individual characteristics
Gender (ref. male)−0.486
(3.283)		0.122
(1.973)		1.026
(1.377)		0.255
(0.963)		1.600
(1.158)
Age0.105
(0.101)		−0.090
(0.059)		−0.080
(0.051)		0.006
(0.056)		0.027
(0.050)
Perceived distance of neighborhood unit3.723 **
(1.586)		0.484
(1.039)		1.090
(0.687)		0.220
(0.552)		0.412
(0.691)
Car ownership (ref. no)−3.989
(4.203)		−1.939
(2.179)		−0.625
(1.656)		0.543
(0.978)		0.469
(1.144)
(Intercept)−20.744
(14.382)		38.971 **
(15.389)		59.497 ***
(8.340)		66.572 ***
(5.901)		78.482 ***
(6.919)
R20.2600.1880.1850.1630.167
*** p < 0.01, ** p < 0.05, * p < 0.1.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Jin, S.; Kim, E.J. Correlation of the Walk Score and Environmental Perceptions with Perceived Neighborhood Walkability: The Quantile Regression Model Approach. Sustainability 2024, 16, 7074. https://doi.org/10.3390/su16167074

AMA Style

Jin S, Kim EJ. Correlation of the Walk Score and Environmental Perceptions with Perceived Neighborhood Walkability: The Quantile Regression Model Approach. Sustainability. 2024; 16(16):7074. https://doi.org/10.3390/su16167074

Chicago/Turabian Style

Jin, Suin, and Eun Jung Kim. 2024. "Correlation of the Walk Score and Environmental Perceptions with Perceived Neighborhood Walkability: The Quantile Regression Model Approach" Sustainability 16, no. 16: 7074. https://doi.org/10.3390/su16167074

APA Style

Jin, S., & Kim, E. J. (2024). Correlation of the Walk Score and Environmental Perceptions with Perceived Neighborhood Walkability: The Quantile Regression Model Approach. Sustainability, 16(16), 7074. https://doi.org/10.3390/su16167074

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop