Perspective of senior citizens regarding intersectional equity on walkability

With the current trend of exploring walkability measures using objective approach, it is essential not to overlook subjective perspectives. Intersectionality examines how heterogenous members of a specific group might have different experiences depending on several aspects of their identities, including gender, age, sexuality, ethnicity, religion, and other social factors. This study explores the influence of age on walkability perceptions, revealing that senior citizens have distinct perspectives on this matter compared to the average population. The study involved 392 senior citizens who rated various variables affecting their walkability. These ratings were then analyzed using Random Forest Classifier to identify key variables that predict their willingness to walk. The findings indicate that walkability is positively associated with senior citizens’ willingness to walk when certain variables are present, such as adequate lighting, evenness of pedestrian walks, short travel time to recreational facilities, availability of benches or resting spaces, the presence of family or friends during walks, the presence of security officers, availability of pedestrian walks, and the availability of social activities and gatherings. This study contributes to the understanding of intersectional equity in pedestrian facilities and provides considerations to its future development by including the perspective of senior age individuals.


Introduction
Walkability measures the capabilities of a built environment to enable walking condition by being easily traversable, compact, physically-enticing, or safe [1].Many variables can affect a neighborhood's walkability, such as residential density, street connectivity, traffic conditions, and aesthetics.Increased walkability can encourage people to use public transportation, which may become a solution to high traffic issues, whilst also bringing a positive trend to public health and economy [2].
The measure of walkability can be explored from an objective and subjective standpoint.The trend of current study of walkability measures tend to use a more objective standpoint.However, the subjective standpoint of the pedestrian is just as important in determining walkability [3].Environments that have high walkability score does not necessarily have high pedestrian density.The walking behavior and neighborhood characteristics also promote public's motivation to walk.Measures of walkability needs to consider more socioeconomic factors so that environments with high walkability will be able to promote more walking trips [4].
Intersectionality is a concept that examines how heterogenous members of a specific group might have different experiences depending on several aspects of their identities [5].Intersectionality includes gender, age, sexuality, ethnicity, religion, and other social locations.These aspects are interrelated and influence one another to create unique social experiences for each individual members of society [6].Pedestrians' perspective on walkability can be affected by their intersectionality.People of different gender, age, and cultures can have different perspectives on what they consider as a walkable environment [7].Intersectional equity for walkability is achieved by considering the different walking experiences of the pedestrians of a built environment.An ideal walkable environment should be designed for diversity so that it becomes beneficial for all [8].
World Health Organization (WHO) defines individuals aged 60 and above as senior citizens [9].This age group often faces specific health and mobility challenges that can limit their ability to engage in physical activities like walking [10].Therefore, it is essential to focus on the development of infrastructure and services tailored to support senior citizens in maintaining a healthy and active lifestyle [3].To achieve this goal effectively, it becomes essential to consider the perspectives of senior citizens regarding walkability variables [8].By doing so, walking trips among senior citizens can be better promoted and encouraged.
Departing from that idea, this study is developed to fulfill two aims.Firstly, it seeks to comprehensively analyze the variables that influence the perspective of senior citizens regarding walkability in urban environments.By examining these factors, we aim to gain a deeper understanding of the specific elements that are significant to senior citizen considering their walking experiences.Secondly, this study aims to prioritize these variables, shedding light on which factors carry greater influence in shaping the walkability perceptions of senior citizens.By achieving these objectives, this study aims to provide valuable insights into the unique needs and priorities of senior citizens in urban planning, hence contributing to the development of more age-friendly and inclusive communities.

Sample
Senior citizens of Bandung will be the primary source of this study.The study is focused on the perspectives of senior citizens towards walkability on Braga Street, who have taken walks on Braga Street for the last 30 days of the data collection period.The whole length of Braga Street is applicable to this study.The surveyor is required to first identify senior citizens who fit into the criteria before approaching them to fill in the survey questionnaire.
Braga Street is chosen due to its abundance of recreational facilities such as shops and restaurants, also for its close vicinity to shopping malls and parks.Braga Street is also known as one of the historic sites in Bandung City [11].Its status as a recreation spot provides higher population of pedestrians [12].Higher population of pedestrians also means that there is higher chance of finding senior citizens who walk in the area.It is important to select a location with many pedestrians since the number of senior citizens who travels by walking can be quite low.

Data Collection
Data collection of this study is done using an online survey via Google Forms.The survey questionnaire is separated into four parts.The first part of the questionnaire asks about the respondents' characteristics, such as their gender, age, and travel behavior.The second part of the questionnaire consists of filter questions that made sure that the respondents are senior citizens who have walked on Braga Street within 30 days and have not filled in the questionnaire before.The third part of the survey asks the respondents to rate 16 different variables of walkability on a scale of 1 to 5. The final part of the survey consists of a statement question used for the Random Forest Classifier model used for this study.
The variables of walkability for this study are taken from walkability variables listed by previous studies.The studies referenced are Global Walkability Index [13], Field Walkability Survey Parameters adjusted to Asian cities [14], Walkability Audit Elements [15], and Facilitators of Neighborhood Activity [16].The walkability variables and its sources can be seen in Table 1.

Data Analysis
The survey result will be split into training data and test data for the random forest classifier.The training data will be used to build and train the random forest classification model.The test data will be used to test the accuracy and precision of the model.
The training data will be separated into 8 bootstrapped datasets in order to build 8 decision trees that contains 2 variables each.Then, the Gini Index of each variable can be calculated to decide the root node and branch node of each decision tree.The decision tree will then be used to make a prediction on senior citizens' willingness to walk based of majority voting between 8 decision trees.The accuracy and precision of the model is measured using a Confusion Matrix.Accuracy is the ratio between the number of correct predictions to the total number of samples, whereas precision is the ratio between the number of correctly classified positive sample to the total number of samples classified as positive [17].
Table 1.Walkability variables for senior citizens.

Parameter
Variable Source

Availability of walking paths
Availability of pedestrian walks Field Walkability Survey Parameters [14] Path size Width of pedestrian walks Walkability Audit Elements [15] Quality and maintenance of walking path surface

Evenness of pedestrian walks
Global Walkability Index [13] Availability of crossings Availability of crossings Global Walkability Index [13] Security Presence of security officer Global Walkability Index [13] Security Adequate lighting Global Walkability Index [13] Cleanliness Cleanliness of pedestrian facilities Global Walkability Index [13] Aesthetics Aesthetic of pedestrian facilities Walkability Audit Elements [15] Amenities Verdancy and adequate shading Walkability Audit Elements [15] Connectivity Travel distance to nearest recreational facilities Global Walkability Index [13] Connectivity Travel time to nearest recreational facilities Global Walkability Index [13] Connectivity Availability of public transportation Global Walkability Index [13] Connectivity, disability infrastructure

Accessibility for bicycles and mobility aids
Global Walkability Index [13] Amenities

Availability of benches or other resting space
Field Walkability Survey Parameters [14] Incentive of social interaction

Availability of social activities and gatherings
Facilitators of Neighborhood Activity [16] Incentive of social interaction

Results
The survey was conducted on June 17th to June 29th 2023.From this survey, a total of 392 responses can be used for data analysis.The respondents were comprised of majority females (58.67%) genderwise, and 60-64 years old (54.08%) in age group.Moreover, all of the senior citizen respondents walk at least once a month, with more than 70% of them actually revealed to walk at least once a week.The most popular destinations among them are around their neighborhood (31.38%), parks (29.85%), and shopping districts (28.57%).Finally, this study unveils that the most preferred mode of transportation of the senior citizens are by car (34.95%) and motorcycle (28.57%), in contrast to non-motorized mode such as walking (12.76%) and bicycle (4.08%).Surprisingly, 14.54% of the respondents prefer using local public transportation, angkot, which has various available routes around the city and relatively cheaper.These descriptions reflect typical characteristics of senior age individuals who are enjoying their retirement and spending their time doing leisure activities.More details on the respondents' characteristics can be seen in Table 2.
In the section about walkability, there are 16 variables that has been rated by the respondents in the survey.The average rating of each variable is then counted and sorted from highest to lowest.The variable with the highest mean is availability of pedestrian walks at 4,2347 and the variable with the lowest mean is availability of social activities and gatherings at 3,7959.The overall mean for all 16 variables is at 4,0886.The list of variables, the distribution of the rating and the average/mean of each variable can be seen in Table 3.The respondents were also asked to rate the current walkability of Braga Street on a scale of 1 to 5, with 1 being very poor and 5 being excellent.Out of 392 respondents, there were no respondents that rated the walkability of Braga Street at 1.There were 7 respondents who gave a rating of 2, 54 respondents gave a rating of 3, 106 respondents gave a rating of 4, and 222 respondents gave a rating of 5, bringing the mean of the rating to 4,40.
The data is then separated into 292 responses for the training data and 100 responses for the test data.The training data is then bootstrapped into 8 datasets to build the decision trees.The decision trees with its variables can be seen in Table 4, which is later used as a foundation to predict senior citizens' willingness to walk more often.The prediction of training data resulted in 237 true positive (TP), 0 true negative (TN), 31 false positive (FP), and 24 false negative (FN), which leads to 81,2% accuracy and 88,4% precision.Moreover, model estimation on the test data generates 83 TP, 0 TN, 11 FP, and 6 FN, with 83% accuracy and 88,3% precision produced.
The higher the model's accuracy, the better it is at making a prediction.Hence, with the test data result, the model is deemed to be accurate with satisfactory precision, meaning the result is ideal.The and precision of the model can be improved further, but it is not recommended since it can cause overfitting.

Discussion
The most popular walking destinations amongst the respondents of this study are parks at 117 responses (29,8%), shopping districts at 112 responses (28,6%), and there are 123 responses (31,4%) that stated that they often walk around their neighborhood.This result is consistent with the findings of previous studies that stated that senior citizens tend to walk to destinations that are close to their neighborhood [10].This finding suggests that walkable environments need to be close to parks and shops to appeal more towards senior citizens.Shopping areas and neighborhood parks need to provide accessible walking paths for senior citizens to encourage more walking trips amongst senior citizens within the area.This suggestion takes into consideration that age-friendly infrastructures can encourage walking as a physical activity amongst senior citizens [3].The walkability of Braga Street was rated quite high by the respondents, with the average rating at 4,40.This means that the respondents consider the pedestrian facilities of Braga Street to have satisfactory performance in support their walking activities.However, the high rating does not guarantee that the senior citizens are willing to walk.Amongst 392 respondents, 362 stated that they will walk more often if their walkability preferences are satisfied and 30 respondents stated that they would not walk more often even if their preferences are satisfied.This finding shows that high rating towards walkability does not guarantee high pedestrian density [4].
In the questionnaire, the respondents were asked to rate 16 variables by how much each variable affects walkability based on their personal experiences.Availability of pedestrian walk is rated the highest by the respondents of this study.This shows that availability of pedestrian walk has high priority as a variable of walkability.This finding is consistent with Field Walkability Survey Parameters that set this variable with the highest weight for walkability measure in Asian cities [14].However, some variables that were highly rated, such as: cleanliness of pedestrian facilities that has second highest rating, followed by availability of benches and other resting space, adequate lighting, and availability of public transportation are not put at high priority by previous studies [14,15].This shows the difference between walkability measures with objective standpoint with the subjective views of pedestrians.This supports the notion that subjective measures must also be considered for walkability measurement since the priorities of objective measures and subjective measures can be different [1].
Random Forest Classifier is utilized to differentiate variables that are better at identifying senior citizens who are willing to walk more often from senior citizens who are not.The model suggests that adequate lighting, evenness of pedestrian walks, travel time to nearest recreational facilities, availability of benches or other resting space, presence of family or friends during walks, presence of security officer, availability of pedestrian walks, and availability of social activities and gatherings are variables that are better at predicting senior citizens' willingness to walk compared to the other variables.Adequate lighting, evenness of pedestrian walks, and availability of benches or other resting space can also help with physical limitations and mobility issues faced by senior citizens.The benefits of these variables may have affected their ability to predict senior citizens' willingness to walk.This finding suggests that minimizing barriers related to physical limitations and mobility issues can motivate senior citizens to do more physical activities, including walking [16].

Conclusion
The variables that are considered to affect walkability typically concerns safety, comfort, and aesthetic parameters of pedestrian facilities.The significance of each parameter to walkability varies with pedestrians' perspective and affected by their intersectionality.Age as factor of intersectionality can affect perspective on walkability, thus senior citizens can have unique perspectives on walkability variables compared to the average pedestrians.
Most senior citizens do not see walking as their main mode of transportation and tend to take walking trips for recreational purposes or for its health benefits.Walkability variables that are highly rated by senior citizens are typically beneficial to help with their health and mobility issues common with old age, such as availability of benches and other resting space can help individuals with mobility disadvantages and adequate lighting can help individuals with eyesight issues.
The findings from the model stated that variables that are rated higher by senior citizens are not necessarily better at predicting their motivation to walk.For example, evenness of pedestrian walks is better at predicting senior citizens' motivation to walk compared to cleanliness of pedestrian facilities despite being rated lower by the respondents.Furthermore, this study reveals that the variables that are better at predicting senior citizens' motivation to walk are adequate lighting, evenness of pedestrian walks, travel time to nearest recreational facilities, availability of benches or other resting space, presence of family or friends during walks, presence of security officer, availability of pedestrian walks, and availability of social activities and gatherings.
This study explores how age influences perceptions of walkability, revealing that senior citizens have distinct perspectives on this matter compared to the average population.It emphasizes the need for urban planning and walkability assessments to consider age-related factors, advocating for a more inclusive and age-sensitive approach.Ultimately, the study aims to contribute to the development of healthier, more supportive, and inclusive communities for senior citizens and all residents.

Recommendations for future works
Considering the findings and limitations of this study, several recommendations can be proposed for future research in the field of walkability assessment.Firstly, there is an opportunity to delve deeper into the intersectionality of age with other factors, such as gender, ethnicity, or socioeconomic status, to comprehensively understand how these intersections influence pedestrians' perceptions of walkability.This could provide a more nuanced view of how various aspects of identity interact with age to shape individual perspectives.
Secondly, to enhance the understanding of pedestrian density among senior citizens, future studies could consider the implementation of pedestrian counting stations in the study area.This data collection method could provide valuable insights into the concentration of senior pedestrians in specific areas.Subsequently, researchers could explore whether there is a correlation between areas with high senior pedestrian density and the travel behavior of senior citizens, as well as how this density relates to their

Table 4 .
Decision Trees with Root Nodes, Branch Nodes and Gini Index.