Modeling Spatial Knowledge of Cities in South Carolina

A number of studies of spatial knowledge have asked people to recall lists of city names. These studies have used the frequency of a city on lists from a reference location as a surrogate measure of the knowledge acquired for that city by the typical person living at that location. Previous studies done with children in Sweden and college students in the United States found acquired knowledge related to the gravity model variables of population and distance. This study considered knowledge acquired for cities in South Carolina from three locations in the state. A number of formal models were used to predict acquired knowledge of cities and consider the effects of gravity model variables, a spatial competition variable, and the respondent's home location on variation in knowledge. Results demonstrated that a neural network model accounted for more variance than other models. Population and distance were found to be the most important variables explaining spatial knowledge.