One of the more recent topics in transportation policy making and project prioritization in DOT's across the United States is how to improve quality of life. This quality of life focus is an intersection of public health and transportation and has many different branches including traffic safety, pollution, active transportation, equity, and accessibility to resources. The topic among these that seems to receive the least attention is equitable access to resources. However, with the proven impact that equitable access to resources has on physical and emotional well-being, as well as economic opportunity it is important to ensure that all populations, including minority and low income populations have the same accessibility as those in the majority and high income populations. Transportation and public health researchers have constructed several quantitative measures to define accessibility, but the most frequent methods used lack the ability to include quality measures of the resource or a variable distance threshold. Therefore, it is important to use a more inclusive definition of accessibility that we can develop using a utility-based accessibility model. This paper develops a logit model with different variables and parameters in order to identify which variables are important and should be included in accessibility measures. The variables come from the impedance attributes for the different modes of travel and the location attributes that were obtained from a NEMS-S (Nutrition Environments Measures Survey in Stores) survey conducted in three counties in Utah. The parameters were developed using LBS data and identifying flows from home block groups to grocery stores, and then creating synthetic trips with non-chosen alternatives. After the utility model is created, we are able to use those utility values to create different logit models, each having a different set of variables included. These logit models were analyzed to find the variables that have a significance on the model. From the models it was found that cost of goods, as well as availability of quality goods have a significant impact on which grocery stores were chosen in the model. In addition, other variables such as size of store and mode of travel have a significance as well. From the models identified, it was found that in order to have an accurate accessibility model it is important to include these variables that are found to be significant in grocery store choice. These results can help with developing transportation policies and help to improve equitable access in cities.





Date Submitted


Document Type





public health, accessibility, transportation policy



Included in

Engineering Commons