Measuring the relationship between spatial configuration, diversity and user behavior: A Post Occupancy Evaluation study in Istanbul’s peripheral districts

Post Occupancy Evaluation (POE) is a robust tool to systematically evaluate the effects of design decisions on spatial performance and to identify the relationship between the space and its users. Although there is a growing body of POE research on complex buildings, such as hospitals and education spaces, studies on the POE evaluation of public open spaces are limited. More importantly, few studies have investigated public squares designed at the periphery and how they are used. This study aims to identify the extent to which spatial configuration of public squares is related to users’ behavior (i.e., modes/distances of access, level of satisfaction). For this purpose, we focused on four peripheral urban squares located in Istanbul, Turkey. The methodology applied in the study includes a synthesis of three types of expertise: 1) behavioral mapping of urban squares (through the analysis of patterns of use based on direct observation), 2) cognitive evaluation of spaces based on perceived factors (through user questionnaires), and 3) quantifying urban public spaces objectively (through the methodology of space syntax and urban morphology). The results identify associations between objective characteristics of public spaces designed at the peripheral districts, patterns of use and users’ perception of these areas, to a certain extent. For example, the variety and intensity of activities within the square as well as the length of occupancy are highest for Avcılar square, which is most integrated within its urban surroundings with reduced average block size. An important finding is the association between the average street connectivity levels of these squares with their pedestrian catchment areas. In other words, the more integrated a public urban space is with its surroundings (800mt buffers), its users will access this space on foot from a larger distance. Based on these findings, spatial configuration as measured by space syntax measures appears to be an explanatory