Despite the well-established need for specific measurement instruments to examine the relationship between neighborhood conditions and adolescent well-being outcomes, few studies have developed scales to measure features of the neighborhoods in which adolescents reside. Moreover, measures of neighborhood features may be operationalised differently by adolescents living in different levels of urban/rurality. This has not been addressed in previous studies. The objectives of this study were to: 1) establish instruments to measure adolescent neighborhood features at both the individual and neighborhood level, 2) assess their psychometric and ecometric properties, 3) test for invariance by urban/rurality, and 4) generate neighborhood level scores for use in further analysis.
Data were from the Scottish 2010 Health Behaviour in School-aged Children Survey, which included an over-sample of rural adolescents. The survey responses of interest came from questions designed to capture different facets of the local area in which each respondent resided. Intermediate data zones were used as proxies for neighborhoods. Internal consistency was evaluated by Cronbach’s alpha. Invariance was examined using confirmatory factor analysis. Multilevel models were used to estimate ecometric properties and generate neighborhood scores.
Two constructs labeled neighborhood social cohesion and neighborhood disorder were identified. Adjustment was made to the originally specified model to improve model fit and measures of invariance. At the individual level, reliability was .760 for social cohesion and .765 for disorder, and between .524 and .571 for both constructs at the neighborhood level. Individuals in rural areas experienced greater neighborhood social cohesion and lower levels of neighborhood disorder compared with those in urban areas.
The scales are appropriate for measuring neighborhood characteristics experienced by adolescents across urban and rural Scotland, and can be used in future studies of neighborhoods and health. However, trade-offs between neighborhood sample size and reliability must be considered.