Examining methodological details of neighbourhood observations and the relationship to health: a literature review.

Neighbourhood research has been growing steadily over the past two decades across the social sciences, and, in particular in the area of public health. Despite this increase, there is currently no consensus on which measure and data source researchers should use to assess neighbourhood attributes. In the past, researchers have relied on census data and household surveys to assess neighbourhood conditions, but in recent years, neighbourhood observations have become a popular alternative method for characterizing neighbourhood environments. Rooted in sociology of crime research, observations are conducted by trained observers who use a checklist to observe and rate neighbourhoods on a number of conditions such as physical (e.g., traffic volume, housing conditions) and social (e.g., presence of people, gang activity) attributes. While this methodology has been gaining momentum in recent years, notably absent from the literature is a review to examine this methodology in detail. The purpose of the present study was to examine research that has used neighbourhood observations as a method. We do so by critically reviewing 51 English language studies published from 1990 onward paying particular attention to the areas of (1) methodological rigor (i.e. how observations are carried out in the field and how data are analyzed), (2) geographical boundaries (i.e. how neighbourhoods and areas of observation are spatially defined), and (3) the relationship between neighbourhood observations and residents' health (i.e. how studies examine and analyze the link between observed neighbourhood attributes and health). We find that little attention is given to details of neighbourhood observations as a method. Further, there is wide variability in how observations are conducted and analyzed making it impossible to confidently compare findings across studies. We see this review as a first step in developing sound observational measures of neighbourhood factors and conclude by providing recommendations for researchers undertaking neighbourhood observations in the future. Copyright 2009 Elsevier Ltd. All rights reserved.