A Reliable, Feasible Method to Observe Neighborhoods at High Spatial Resolution.

Area of Science:

• Environmental health
• Urban planning
• Public health research

Background:

• Systematic social observation (SSO) traditionally uses street-level measurements, with virtual methods enhancing feasibility.
• Higher spatial resolution may better reveal neighborhood health impacts, but virtual parcel-level data reliability is unproven.

Purpose of the Study:

• Introduce and validate a novel parcel-level SSO methodology.
• Assess the reliability and feasibility of virtual data collection using Google Street View at the parcel level.

Main Methods:

• Compared inter-rater and intra-rater reliability of four SSO methods: in-person and virtual at both street and parcel levels.
• Utilized 5 raters observing 42 neighborhoods in 2016.

Main Results:

• All four SSO methods demonstrated substantial inter-rater agreement (>70% of items).
• Parcel-level virtual observation showed high intra-rater agreement (>75% for most items).
• Physical decay was the only factor showing higher agreement with direct vs. virtual observation.

Conclusions:

• Google Street View is a reliable and feasible tool for virtual SSO at the parcel level.
• This validated parcel-level virtual method can enhance the assessment of social determinants of health.
• Improved assessment may contribute to understanding and reducing health disparities.