How to Use EDA to Investigate the Relationship Between Urban Green Spaces and Public Health

Exploratory Data Analysis (EDA) is a crucial first step in any data analysis project, including when examining the relationship between urban green spaces and public health. It involves using statistical and visualization techniques to understand the patterns, trends, and potential correlations within the data. By applying EDA to this topic, we can identify key relationships between green spaces in urban areas and various health outcomes, such as physical activity, mental well-being, and overall public health. Here’s how you can use EDA to investigate this relationship:

1. Data Collection and Preprocessing

Before diving into the analysis, you need to gather data from various sources. For investigating urban green spaces and public health, you might need the following datasets:

• Urban Green Space Data: This could include the number of parks, green areas, or the density of greenery in a city. Sources like local government agencies, urban planning databases, or satellite imagery (such as from Google Earth or OpenStreetMap) can be helpful.

• Public Health Data: This might include data on health outcomes, such as rates of physical activity, mental health statistics, or disease prevalence. You can find such datasets from health organizations, national surveys, or local hospitals.

• Demographic Data: Information about the population such as age, gender, income, and social factors can also impact both the availability of green spaces and health outcomes. This data can often be obtained from census reports.

Once you have gathered the data, preprocess it by cleaning any missing values, removing duplicates, and ensuring that all datasets are aligned in terms of geographic areas and timeframes. You might need to aggregate or join different datasets, depending on their structure.

2. Univariate Analysis: Understanding Individual Variables

Start by conducting univariate analysis to understand the distribution of individual variables.

• For Urban Green Spaces:

  • Plot the distribution of the number or area of parks or green spaces per district or neighborhood.

  • Visualize the density of green spaces (e.g., number of square meters of green space per capita).

  Visualization techniques like histograms or box plots can reveal whether green spaces are evenly distributed across the city or if some areas are underserved.

• For Public Health Data:

  • Examine the distribution of health indicators such as the rate of physical activity, mental health conditions, or chronic diseases.

  • Use histograms to look at the spread of variables like average walking distance, incidence of depression, or cardiovascular diseases across different neighborhoods.

Univariate analysis helps to see the basic properties of the data and prepares you for more complex relationships.

3. Bivariate Analysis: Examining Relationships

Next, conduct bivariate analysis to explore how urban green spaces correlate with public health outcomes.

• Green Spaces vs. Physical Activity:

  • You can use scatter plots or correlation matrices to explore how the proximity or size of green spaces correlates with physical activity levels. For example, do people living closer to green spaces engage in more physical activities like walking, cycling, or running?

  • Conduct a t-test or ANOVA to compare the mean physical activity levels between neighborhoods with high and low green space density.

• Green Spaces vs. Mental Health:

  • Use scatter plots to analyze the relationship between green space access and mental health conditions, such as the prevalence of anxiety or depression.

  • Overlay health data on maps to visualize whether areas with more green spaces have lower reported cases of mental health issues.

• Green Spaces vs. Chronic Diseases:

  • Investigate the association between green space access and chronic diseases like diabetes, hypertension, or respiratory illnesses using regression analysis.

Visualization Examples:

• Scatter Plots: To visualize relationships between continuous variables (e.g., green space area vs. physical activity rate).

• Heatmaps: To show correlations between different variables across geographic regions.

• Bar Charts: To compare health outcomes between areas with different levels of green space.

4. Multivariate Analysis: Considering Multiple Factors

EDA becomes even more insightful when you include multiple variables. Urban green spaces are just one of many factors that influence public health. Other variables like income level, education, access to healthcare, and air quality also play crucial roles.

• Multivariate Regression: Use multivariate regression models to examine how green space affects public health, controlling for other variables such as socioeconomic status or environmental factors. This approach will help isolate the effect of green spaces from other confounding variables.

• Cluster Analysis: Perform clustering to group areas with similar health outcomes and green space characteristics. This could highlight regions where green space has a more significant impact on health, or where it is less important due to other mitigating factors.

5. Geospatial Analysis

Since both green spaces and public health are geographically distributed, geospatial analysis is an essential component of EDA in this context.

• Geospatial Visualization: Use GIS (Geographic Information Systems) tools to visualize the geographic distribution of green spaces and health outcomes. Overlay these layers on city maps to identify areas that are underserved or have poor health outcomes despite having green spaces.

• Spatial Correlation: Conduct spatial autocorrelation analysis to determine if areas with more green spaces tend to have better health outcomes in neighboring regions. Tools like Moran’s I statistic can quantify spatial dependencies in the data.

• Accessibility Analysis: Evaluate the accessibility of green spaces by analyzing the distance from residential areas. You could use buffers or heat maps to assess whether people living closer to parks and green spaces experience better health outcomes.

6. Time Series Analysis (if applicable)

If you have time-series data, such as changes in green space availability or health outcomes over time, you can apply time-series analysis to see how these variables change over time and if any trends emerge.

For example:

• Track how public health outcomes (such as physical activity levels or mental health statistics) evolve as new parks or green spaces are developed.

• Assess whether there is a time-lag effect, where health improvements take a few years after the establishment of green spaces.

7. Hypothesis Testing

Based on the EDA, you might generate hypotheses about the relationship between green spaces and public health. For instance:

• Hypothesis 1: Increased access to urban green spaces is associated with higher levels of physical activity.

• Hypothesis 2: Urban areas with more green spaces have lower rates of mental health issues.

You can test these hypotheses using statistical techniques like chi-square tests, t-tests, or ANOVA, depending on the nature of your data.

8. Reporting and Interpretation

Finally, summarize the key findings from your EDA and interpret the results in the context of urban planning and public health.

• Insight Generation: Highlight significant trends and correlations, such as whether access to green spaces appears to reduce the risk of certain diseases or improve mental well-being.

• Policy Implications: Based on the results, suggest recommendations for policymakers. For example, if the data shows that areas with more green spaces have better public health outcomes, advocate for the expansion of green spaces in underserved urban areas.

• Limitations and Future Work: Acknowledge any limitations in your analysis, such as incomplete data or confounding variables, and suggest areas for future research or more detailed studies.

Conclusion

Using EDA to investigate the relationship between urban green spaces and public health can provide valuable insights that inform both urban planning and public health policies. Through careful data collection, preprocessing, and the application of various statistical and visualization techniques, you can uncover patterns and correlations that are critical for improving the quality of life in cities. By focusing on both individual variables and the interactions between multiple factors, EDA allows for a comprehensive exploration of how urban green spaces influence public health.