Reliability and One-Year Stability of the PIN3 Neighborhood Environmental Audit in Urban and Rural Neighborhoods.

ABSTRACT

Reliable and stable environmental audit instruments are needed to successfully identify the physical and social attributes that may influence physical activity. This study described the reliability and stability of the PIN3 environmental audit instrument in both urban and rural neighborhoods. Four randomly sampled road segments in and around a one-quarter mile buffer of participants' residences from the Pregnancy, Infection, and Nutrition (PIN3) study were rated twice, approximately 2 weeks apart. One year later, 253 of the year 1 sampled roads were re-audited. The instrument included 43 measures that resulted in 73 item scores for calculation of percent overall agreement, kappa statistics, and log-linear models. For same-day reliability, 81% of items had moderate to outstanding kappa statistics (kappas ≥ 0.4). Two-week reliability was slightly lower, with 77% of items having moderate to outstanding agreement using kappa statistics. One-year stability had 68% of items showing moderate to outstanding agreement using kappa statistics. The reliability of the audit measures was largely consistent when comparing urban to rural locations, with only 8% of items exhibiting significant differences (α < 0.05) by urbanicity. The PIN3 instrument is a reliable and stable audit tool for studies assessing neighborhood attributes in urban and rural environments.