Applying Geospatial Connectivity: A Multidimensional Analysis of Poverty Defined Census Blocks and Food Deserts within Deland, Florida

Applying Geospatial Connectivity: A Multidimensional Analysis of Poverty Defined Census Blocks and Food Deserts within Deland, Florida A PAPER SUBMITTED FOR COMPLETION OF SENIOR RESEARCH FOR THE COLLEGE OF ARTS AND SCIENCES STETSON UNIVERSITY BY Justin Pierre Baumann IN PARTIAL FULFILLMENT OF THE REQRUIEMENTS FOR THE DEGREE OF BACHELOR OF SCIENCE ENVIROMENTAL SCIENCE ADVISOR Dr. J. Anthony Abbott, Ph.D. & Dr. J. Evans, Ph.D. May 2016 ii Table of Contents List of Illustrations ……………………………………………………………………… iii List of Tables…………………………………………………………………………..… iv Acknowledgments…………………………………………………………………..…… v Abstract………………………………………………………………………………..… vi Introduction ……………………………………………………………………….…...…1 Literature Review…………………………………………………………………….… 1- 3 Study Area.…………………………………………………………………………….… 3 Methodology……………………………………………………………………..……… 4- 7 Data Analysis & Results ……………………………………………………………… 8-10 Conclusions & Discussion…………………………………………………………… 10 – 11 Appendix A – Geoprocessing Guidelines……………………………………………12-13 Works Cited…………………………………………………………………………….14-15 iii List of Illustrations Map 1: Network Analysis of Grocery Stores Based On Ten-Mile Distance Parameters..6 Map 2: Displays The Number Of Persons Per Block Group In Poverty………………..7 Graph 1: Shows the number of block groups in poverty vs. the distance of block groups to grocery locations……………………………………………………………………………...9 Graph 2: Shows the number of block groups in poverty vs. the distance of block groups to convenience locations.……………………………………………………………………….9 iv List of Tables Table 1: Fields/Data Utilized…………………………………………………………… 5 Table 2: Results of the Excel Correlation Analysis……………………………………8 Table 3: Shows the mean distance amongst all centroid block groups needed to travel for food access…………………………………………………………………………………8 v Acknowledgments Thank you to my advisor Dr. Tony Abbott for his meticulous editing, patience, and sincere regard for the emotional turmoil experienced by his senior research students. Thank you to my advisor Dr. Jason Evans for aiding me in the specifics of my geo-processing and working with me to understand the steps to perform to yield the most accurate results. Thank you to my parents and grandparents for supporting me throughout my undergraduate experience. If it weren’t for your sacrifices I would never have been able to attend Stetson University and receive the insightful education that I have at this institution. Thank you to Dr. John Jett for being my boss & mentor for the last 2+ years. The wisdom you have passed on and the discussions we have had will always be a memory I sincerely hold onto through my life. vi Abstract Increasing our understanding of the non-static issue of food insecurity is crucial during a time of perceived abundance and emerging food deserts. Food deserts are areas of limited food access or options within and are not conducive to maintaining a healthy dietary life-style. Using geographic analysis programs, such as ArcMap, commercial food locations (grocery vs. convenience) were geographically studied against low-socioeconomic households via Tiger census tract block group data within the focal city of Deland, Fl. Connectivity was studied by conducting a network analysis to conceptualize the distance requirements of block groups to reach points of food access. Block groups with the highest rates of poverty were additionally identified to calculate a correlation between percent of a population in poverty and their access to commercial food sources. I ran two correlation analyses for each commercial variable. The first compared the percent of poverty within a block group against the each type of commercial data. The second compared the percent of the poverty within a block group to the total land area of a census block group. Both of my correlation analyses indicated a weak negative correlation between the variables. These results contradicted my original hypothesis and support that Deland is not a food desert. Introduction Understanding and quantifying access to food will continue to become a focal field of study for both the public health and geography academic fields. This is because understanding how people interact within their food environments is a multi-dimensional issue that arcs across multiple academic fields including health, geography, economics, politics, city-planning, and even has environmental connotations. Research into the field has slowly developed since the early 1990s as geographic information systems have become increasingly powerful as tools to understand the both positive and negative correlations between the emergence of food deserts and primarily socioeconomic status & race/ethnicity. Contemporary research has primarily focused on both larger population/high density metropolitan areas and small population/low density rural area, which has caused a gap to develop within the academic and planning literature, which does not address issues of the “average” American town/city. This gap leads to many American citizens being left unsure if they are classified as being within a food desert due to the definitions not being multi-faceted enough to envelope multiple city designs and geographic locations. While issues of the larger academic literature are not the primary focus of this study, it is however the hope of this researcher, that this research can aid in studies of smaller cities that do not meet the either metropolitan or rural classification standards primarily dictated by the United States Department of Agriculture. This study focuses on understanding the connection between socioeconomic status and the distance requirement for access to both low quality and high quality foods and how these variables can cause an area to be labeled a food dessert within Deland, Florida. It is hypothesized that Deland will be a food dessert due to its lack of significant access to quality food based on a preliminary appraisal of the city’s layout. Literature Review Defining Food Deserts The USDA (United States Department of Agriculture) subjectively defines distance parameters to delineate identified census tracts as either “low-access community” or “low-access rural.” Tracts that have 33% of their population reside more than one-mile away from a supermarket are considered urban food deserts, with greater than ten-miles being the requirement for rural food deserts. While this provides a framework for identification, a gap exists with the lack of a middle ground between one and ten mile classifications, which would constitute suburban/in-development areas. Within USDA-defined food deserts there is a distinct over representation of racial and ethnic minorities with social class disparities developing inadequate access to food (Wright et al. 2016). Food deserts are typically areas demographically characterized by lower education and higher concentrations of ethnic minorities with lower access to quality foods, in the absence of grocery/super markets, in favor of convenience stores, which sell 2 lower quality/higher fat quantity foods (Larsen et al. 2009). Prominent policymakers have worked to expound upon the USDA definition via the 2008 Farm Bill, which defined a food desert as “an area in the United States with limited access to affordable and nutritious food, particularly such an area composed of predominantly lower-income neighborhoods and communities” (Food, Conservation, and Energy Act of 2008, p. 2039). Geographic and political studies have intertwined within an issue that is multi-faceted involving ethnicity, socioeconomic status, and distance requirements. Theories of Food Desert Formation Researchers have proposed multiple theories within the field to explain how food deserts emerged in the United States. One key theory proposes that chain supermarkets financially choked out smaller neighborhood “mom & pop” grocery stores, putting them out of business, and causing a reduction in access (Wright et al. 2016). Other researchers have argued that food deserts developed within inner-city areas due to the median incoming dropping when higher socioeconomic groups migrated to emerging suburban areas during the 1960s-80s. This same epoch marked the development of supermarket disparity based on income differences with upper-income/primarily white suburban communities having twice as many supermarkets compared to lower-income/African-American communities (Shaffter 2002). Demographics Food insecurity, which is normally thought to be an issue of the under-privileged or “third-world” countries, has since the mid 1990s been identified as an issue within “first-world” countries. It does not only define a lack of food, but specifies if consuming this food is healthy and does not directly lead to negative medical effects. Public health and political officials have been the proponents for research into “obesogenic environments,” more commonly known as food desserts (Sohi et al. 2014). Areas that experience food insecurity are known as food deserts and can be identified based on: (1) access to supermarkets, (2) income/socioeconomic status, (3) racial/ethnic disparities, (4) differences in chain versus non-chain stores (Walker et al. 2010). Areas that are affected by food insecurity are also correlated with high crime rates, lower levels of education, and higher densities of ethnic minorities (Powell et al. 2006). Racial disparity amongst food desert census tracts is supported by research from Morland et al. (2002) who conducted a study of Atherosclerosis Risk in communities and determined that predominantly white census tracts had five times greater access to supermarkets compared to tracts of predominantly African Americans. This is congruent with the African-American population having an obesity rating of 44.1% with network analysis and access studies supporting lack access to supermarkets in favor of small convenience stores (Powell et al. 2007; Bower et al. 2014; Zenk et al. 2015). The economics of food deserts convey that there is a statistically significant correlation between lower income and great exposure in relation to food access and unhealthy fast foods (Hilmers et al. 2012). Hilmers study of academic literature revealed an evident pattern amongst neighborhoods of economic disadvantage/minority populations and access to unhealthy foods that lead to chronic diet-related illness. Low socioeconomic standing is correlated to lack of education and unemployment, in addition 3 to crime rates, lack of transportation, low return on investment as a business, and reduced budgetary availability (increased number of mouths to feed or loss of income) being identified as minor variables (Wang et al. 2003; Powell et al 2006; Bergstrand et al. 2014). This leads to areas of high poverty having less buying power, compared to other tracts, which causes a reduction in access and availability leading to increases in prices and a reduced number of easily accessible locations (Powell et al. 2007). Psychological Variables Research by Donald Rose (1999) identified that sudden economic stressors can drastically impact the decision making processes of individuals within identified food deserts tracts. Due to low resilience, should a loss of employment or taking on another mouth to feed occur, individuals controlling food consumption would forego their personal health in favor of others; this is due to the non-static conditions of food desert populations. Research into the psychology of network analysis limitations revealed the importance of personal vehicle ownership due to the unreliability of social aid programs to provide assistance. Respondents to a qualitative poll revealed that “self-reliance” was the most important factor, even if displacement occurred with a time vs. financial cost-analysis (Conveney 2009). GIS Applications Related To Study: Proximity, Connectivity, and Network Analysis Early food deserts studies did not utilize GIS technologies, but instead used maps to study area as a general concept with the number of shops per geographic unit being compared against changes in socioeconomic variables per geographic unit (Wrigley et al., 2002; Cummins & Macintyre 1999; Guy et al., 2004). However, as of Donkin et al. (1999) GIS technologies have become standard to compare distance of commercial food locations over road networks to multiple variables, with general technique and focus altering per study (McEntee & Agyeman 2010). GIS studies are crucial in accessibility studies due to their ability to show graphical information and complex spatial statistical analysis. Donkin et al. (2000) and Guy and David (2004) demonstrate that for many food access studies authors need to create original data (typically within GIS) and geographically reference it. This allows for statistical analysis to be performed based on the created data and correlation analysis to be run based on access variables such as distance, socioeconomic status, and race against different classifications of food retail locations. Study Area Deland was selected based primarily because it houses Stetson University, the host institution for this research. Supporting attributes for the selection include, Deland’s local government actively contributing to geographic studies and hosts small library of free usable data. Additionally, the city represents a variety of socio-economic, racial, political, and historical backgrounds spread throughout a unique city zoning landscape. Finally, Deland does meet the criteria of being a larger metropolitan area nor a rural town, which allows it to represent and address the spatial gap within the literature. 4 Methodology Data Mining and Pre-Processing: The following methodology works to provide an outline for the creation and utilization of a network analysis within the ESRIs ArcMap. This is to determine the distance of individual census block groups, which intersect with the registered commercial data geographically to illustrate if a correlation exists between economic status and distance from food retail providers of differing food quality Creation of Base Map This project requires multiple maps to be generated to provide feasible ease of data manipulation. However, while multiple maps will be generated, they are not mutually exclusive, instead they are separated for readability and not for functionality. Essentially, different portions of the data need to be created, manipulated, and tested against one another for their final forms to be brought together within two active frames in ArcMap. Spatially Delineating Deland A general city boundary for Deland was downloaded from the Volusia County GIS site, which provided all of the administrative city boundaries within the county. Next, on the U.S. Census Bureau’s website (https://catalog.data.gov/dataset/tiger-line-shapefile-2014-state-florida-current-block-group-state-based-shapefiles), state files are available that display all of the block groups that are within a state. The file “tl_2014_12_bg” was downloaded, extracted, and pulled into ArcMap, where a clip function was performed based on a the general administrative boundary to create the “Deland_bg_clip.shp. This file served as the basic boundary for the majority of my project and would later be expanded into another file based on an intersect function to create the file “Deland_intersect_boundary” that includes all block groups, within the city of Deland that are within the network analysis and congruent with a registered NAICS (North American Industry Classification System) store. Locating and Geospatially Generating Commercial Data Acquiring commercial data was the most difficult aspect of this project due to financial and logistical reasons. However, the Volusia County Library System pays for a subscription to a commercial database known as InfoUSA that displays registered NAICS codes. The limitations of this database, and other commercial databases, are that not all food-based commercial businesses register with the NAICS system. I searched for two specific NAICS code numbers: 445110 (Supermarkets and Other Grocery Stores) and 445120 (Convenience Stores). My pull from the system provided a total of 52 results, of which 45 were viable results with the business still being in operation currently. Addresses were provided with each registered business and these were each searched for within Google EarthTM (GE), verified, and put into two separate folders within GE, which were based on whether they were a registered grocery or convince store. Each folder was then exported from GE as a .kmz file, which is compatible with 5 ArcMap and allows for the geo-referenced points to easily become point shapefiles and displayable. Acquiring Census Data Census data for this project was acquired from a larger Florida Tiger file known as “2007-2011 Block Group Data – Florida.” This file contained all socioeconomic data for the state of Florida delineated by the block group level. The key pieces of data needed for further analysis were: Table ID Name C17002 Ratio of Income To Poverty Level In The Past 12 Months Universe: Population for whom poverty status is determined C17002e3 .50 to .99 (Under Poverty) C17002e2 .50 (Under Poverty) C17002e1 Ratio of Income To Poverty Level For Whom Poverty Status is Determined B01001e1 Sex By Age – Universe – Total Population Table 1: Fields/Data Utilized These data fields were selected out by their attribute names from the larger Tiger File and clipped based on a Deland shapefile. In addition to these aforementioned fields a custom float field was created to generate the value “percent of population in poverty.” This was calculated with (C17002e2_C17002e3)/C7002e1. This value allows for the percent of the population within each block group to be calculated. Street Network Volusia county provides an integrated shapefile (http://www.volusia.org/gis/data/streets.htm) that contains both the local, state, and federally registered and maintained roads. This file was clipped to the Deland level so that connectivity could be studied. Coordinate System Georeferenced data was initially in GCS_North_American_1983 and was re-projected into Albers Conical Equal Area [Florida Geographic Data Library]. These were kept as individual shapefiles, due to my preference to manage data this way. This projection was chosen due to the state plane system being as most accurate projection available. Geo-processing Network Analysis Two network analysis were performed within ArcMap (the spatial analyst extension needs to be enabled to perform this analysis) to generate service area values based on the Deland road network. Service areas are geographically identified areas 6 within a region that encompasses all accessible streets. Network datasets are used to calculate transportation models and output values that calculate service areas and distance values. For the purpose of this project this was utilized to calculate the distance to the closest grocery and convenience store from the centroid of each census block. Additionally, the analysis accounted for both road and commercial location intersection for blocks that are not officially recognized within the legal boundary of Deland, Florida. This means that some blocks are not considered within the legal boundary of the city, however, the businesses, i.e. commercial data, within these blocks are registered to the city. A network analysis was performed for both the identified grocery and convenience stores against the Volusia street network (See Appendix A for specific geoprocessing instructions & Map 1 for an reference image). Map 1: Network analysis of grocery stores based on ten-mile distance parameters Displaying Census Data The Tiger File, which contained the fields in Table 1, was clipped to only show Volusia County and populated with centroids (See Appendix A for specific geoprocessing instructions). The remaining census data was joined with the centroid data and compared against the road network via a select by geography and intersect function within Deland to display the census blocks that are within the city of Deland’s road network, census blocks, and commercial locations. This final file is known as the “Deland_intersect_boundary” and represents the conglomeration of data created by this project (See Map 2). 7 This final data set was then used to perform a spatial join (See Appendix A for specific geo-processing) to include the commercial data (done two separate times for both the grocery and convenience data respectively). Once the commercial data was added a final field within the attribute table was calculated to create a distance function via including the network analysis data: (Frombreak + Tobreak)/2, which shows the average distance from the centroid of a block group to their nearest grocery and convenience store respectively. This data was then exported as a dBASE table to excel for a correlation analysis to be executed and final values for this project were calculated. Map 2: Displays the number of persons per block group below in poverty 8 Data Analysis & Results A simple correlation analysis was performed within Microsoft excel to quantify the relationship between the both sets of commercial data and the percent of the population in poverty within a block group. Test Output Grocery Store Locations vs. Percent of Population In Poverty -0.17648 Convenience Location vs. Percent of Population in Poverty -0.23746 Table 2: Results of the Excel Correlation Analysis My findings indicated that there is not a strong correlation between the location of either a grocery store or convenience store to higher rates of poverty within a census block. These initial correlation values, while negative, are not strong negatives and holistically indicated that there is a weak negative correlation between the variables. The preliminary statistical analysis led to the additional performance of a linear aggression analysis to illustrate the data and calculated r2 value. Graph 1 and 2 each display the number of block groups in poverty vs. the distance of the centroid of that block group to either grocery or convenience commercial locations respectively. Both graphs illustrates the negative correlation between each variable i.e. that as poverty increases the distance/access to food decreases (See Table 3). Table 3: Shows the mean distance amongst all centroid block groups needed to travel for food access. Both the correlation analysis and the linear regression analysis indicate a negative correlation between each of the variables. As the quantity of individuals within poverty increases, the distance needed to travel via the road network decreases. However, it is still interesting to note that while the variables are inversely correlated, the mean distance needed to travel is still significantly greater for grocery locations compared to convenience locations, with a travel distance difference of 1.28 miles. Commercial Data Mean Distance (miles) Convenience Location 1.66 Grocery Location 2.94 9 Graph 1: Shows the number of block groups in poverty vs. the distance of block groups to grocery locations. Graph 2: Shows the number of block groups in poverty vs. the distance of block groups to convenience locations. y = -2.1162x + 3.3797 2.1162x + 3.3797 R² = 0.0311R² = 0.0311 R² = 0.0311 0.0000 0.0000 1.0000 1.0000 2.0000 2.0000 3.0000 3.0000 4.0000 4.0000 5.0000 5.0000 6.0000 6.0000 7.0000 7.00007.0000 0.0000 0.0000 0.1000 0.1000 0.2000 0.2000 0.3000 0.3000 0.4000 0.4000 0.5000 0.5000 0.6000 0.6000 Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Number of Grocery Locations Distance (miles) Distance (miles) Distance (miles) Distance (miles) Distance (miles) Distance (miles) Distance (miles) Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Number Of Block Groups In Poverty vs. Distance of Block Group To Grocery Location Block Group To Grocery Location Block Group To Grocery LocationBlock Group To Grocery LocationBlock Group To Grocery LocationBlock Group To Grocery Location Block Group To Grocery LocationBlock Group To Grocery LocationBlock Group To Grocery LocationBlock Group To Grocery LocationBlock Group To Grocery LocationBlock Group To Grocery LocationBlock Group To Grocery LocationBlock Group To Grocery Location Block Group To Grocery LocationBlock Group To Grocery LocationBlock Group To Grocery Location Block Group To Grocery Location Block Group To Grocery LocationBlock Group To Grocery Location y = -3.7773x + 2.444 3.7773x + 2.444 R² = 0.0564R² = 0.0564 R² = 0.0564 0 1 2 3 4 5 6 7 8 9 10 0.0000 0.0000 0.1000 0.1000 0.2000 0.2000 0.3000 0.3000 0.4000 0.4000 0.5000 0.5000 0.6000 0.6000 Number of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience LocationsNumber of Convenience Locations Distance (miles)Distance (miles) Distance (miles)Distance (miles) Distance (miles) Distance (miles) Distance (miles) Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Number Block Groups In Poverty vs. Distance of Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location Block Group To Convenience Location 10 This contradicts my original hypothesis that Deland would be a food desert due to its high rate of poverty and its city/demographic profile of not being within the urban metropolitan/rural town categorization. Conclusions & Discussion My data analysis section indicates a weak negative correlation between commercial food locations and low-income census tracts, which is a good outcome for the city of Deland. This project set out with the larger goal of proving that Deland was a food desert that is existing outside of the scope of the current literature. This was based on Deland’s odd city categorization and with lack of significant prior literature focusing on smaller towns/cities. My hypothesis was proven incorrect which indicates, even in a minor way, that Deland is not a food desert, at least not within the USDA defined guidelines, with access to food being feasible for poverty stricken census blocks within Deland’s connectivity system i.e. road network. This does give me hope that within the foreseeable future there will be new amendments to the parameters that work to classify food desert categories. However, while this is a good sign for Deland, this does not indicate that other cities within the same size are witnessing the same level of success with their poverty populations. The majority of contemporary academic literature focuses on the subject of food deserts or food access in relation to poverty are channeling their focus on larger metropolitan cities or there opposite extreme: small rural towns. This leads to an enormous gap within the literature, which is unfortunate when there is a significant portion of the American population not living within either of the two focal categories. While this gap does exist, I believe that the research taken from within this project is capable of streamlining future food access studies, which could allow for a greater number of studies to be completed and increase our understanding. More variables could be factored into supplemental or additional studies that could focus on different perspectives of the issue, such as public health officials tracking the correlation between access to fast food locations and hypertension amongst different census tracts. The possibilities and applications for more refined geospatial studies of food access are near limitless and will only become more prominent as focus continues to shift towards understanding these areas and eradicating them. One of the biggest caveats to food access studies currently is the lack of a definition that incorporates the differences in distance between populations of different city sizes and geographic locations. This leads to the recommendation that the USDA create a definition for a city that does fall within the middle of their current 1 vs. 10 mile distance parameters. Additionally, continuing to incorporate the public health field will only strengthen the field of food access studies holistically due to the correlations between consistent low-quality food consumption and long-term health effects, such as experienced by patients who are over-weight or obese. 11 This study can benefit from additional studies that calculate correlations between the poverty levels and their access to food within other cities that have similar profiles to Deland. This would allow for the methodology described within this paper to be reapplied, improved, or deemed executable as a template for future food access studies. 12 Appendix A – Geoprocessing Guidelines Creating and Performing a Network Analysis After finding all the necessary data perform the following steps: 1. Create a network data set (within ArcCatalog) a. R-click “Streets_volusia” – create network data set b. Allow model turns c. Leave connectivity d. No elevation model e. Attributes: i. Units: Miles ii. Time function: 1. Usage: cost 2. Units: minutes iii. Data type: double (change to a constant) 1. = 0 iv. No driving/routing directions f. Click “accept” g. The network data set has been created 2. Performing Network Analysis for each set of commercial data. a. Make sure network analyst extension is on b. Open ArcToolbox i. Network analyst 1. Server a. Generate service areas ii. Facilities – Points iii. Calculate – within road distance breaks iv. Network Data Set: Create a nested shape file 1. For X commercial data v. Optional: Allow U-turns c. Best methods within the selectable Attributes (there are various ways to do this): i. Separate by service area’s specific to each location ii. Merge polygons with multiple facilities 1. No overlap 2. Rings Creating and Populating Centroids 1. Create two additional float fields within the census block group data attribute table and populate Name one X for Latitude and the other Y for Longitude. 2. To calculate field geometry for each centroid via the Block Group I.D field and join them with a spatial join based on the network data set. Final Spatial Join 1. Target Feature – intersected centroids 13 2. Join features – grocery locations (repeat for convenience locations) 14 Works Cited Bergstrand, K., B., Mayor, B., Brumback, and Y., Zhang. 2014. Assessing the relationship between social vulnerability and community resilience to hazards. Social Indicators Research 122 (2): 391-409. Bower, K., R., Thorpe Jr., C., Rohde, and D., Gaskin. 2014. The intersection of neighborhood racial segregation, poverty, and urbanicity and its impact on food store availability in the United States. Preventative Medicine 58: 33-39. Coveney, J. and L., O’Dwyer. 2009. Effects of mobility and location of food access. Health and Place 15(1): 45-55. Cummins, S., & Macintyre, S. 1999. The location of food stores in urban areas: a case study in Glasgow. British Food Journal, 101(7), 545-553. Donkin, A.J., Dowler, E.A., Stevenson, S.J., Turner, S.A. 1999. Mapping access to food at a local level. British Food Journal 3(1): 554. Donkin, A.J., Dowler, E.A., Stevenson, S.J., Turner, S.A. 2000. Mapping access to food deprived area: the development of price and availability indices, Public Health Nutrition, 3 (1): 31-38. Guy, C.M. & David, G. 2004. Measuring physical access to ‘healthy foods’ in areas of social deprivation: a case study in Cardiff. International Journal of Consumer Studies, 28 (3): 222-224 Guy, C.M., Clarke, G., Eyre, H. 2004. Food retail change and the growth of food deserts: a case study of Cardiff. International Journal of Retail & Distribution Management, 32 (17): 72-88. Hilmers, A., Hilmers, D. C., & Eikenberry, N. 2006. Fruit and vegetable access in four low-income food desert communities in Minnesota. Agriculture and Human Values, 23, 371-383. Larsen, K., J., Gilliland. 2009. A Farmers’ market in a food desert: Evaluating impacts on the price and availability of healthy food. Health and Place 15 (4): 1158-1162. McEntee, J., and J., Agyeman. 2010. Towards the development of a GIS method for identifying rural food deserts: Geographic Access in Vermont, USA. Applied Geography 30 (1): 165-176. Morland, K., Wing, S., & Roux, A.D. 2002. “The contextual effect of the local food environment on residents’ diets: the atherosclerosis risk in communities study. American Journal of Public Health, 92(11): 1761-1767. Powell, L., S., Slater, D., Mirtcheva, Y., Bao, and F., Chaloupka. 2007. Food store availability and neighborhood characteristics in the United States. Preventative Medicine 44 (3): 189-195. Rose, D. 1999. Economic determinants and dietary consequences of food insecurity in the United States. The Journal of Nutrition 129(2s): 517s-520s. Shaffer, A. 2002. The Persistence of L.A.’s Grocery Gap: The Need for a New Food Policy and Approach to Market Development. UEP Faculty & UEPI Staff Scholarship. Retrieved 25 April 2016. (http://scholar.oxy.edu/uep_faculty/16/) Sohi, I., Bell, B., Liu, J., Battersby, S., and Liese, A. 2014. Difference in food environment perceptions and spatial attributes of food shopping between residents of low and high food access areas. Journal of Nutrition Education and Behavior 46 (4): 241-249. 15 US Congress, 2008, Us Congress Food, conservation, and energy act of 2008, 110th congress, Us Government Printing Office (2008). Walker, R. E., C. Keane, J. Burke. 2010. Disparities and access to healthy food in the United States: A review of food deserts literature. Health & Place 16 (5): 876-884. Wang, H., F., Qiu, B., Swallow. 2014. Can community gardens and farmers’ markets relieve food desert problems? A study of edmonton, Canada. Applied Geography 55: 127-137. Wright, J. D., A. Donley, M. Gulatieri. 2016. Food Deserts: What is the Problem? What is the Solution? Social Science and Public Policy, 53 (1): 171-181. Wrigley, N., C. Guy, M. Lowe. 2002. Urban regeneration, social inclusion and large store development: the seacroft development in context. Urban Studies, 39 (2002): 2101-2114. Zenk, S. N., Schulz, A. J., Israel, B. A., James, S. A., Bao, S., & Wilson, M. L. 2005. Neighborhood racial composition, neighborhood poverty, and the spatial accessibility of supermarkets in Metropolitan Detroit. American Journal of Public Health, 95 (4), 660-667.