Observação sistemática de contextos intraurbanos no Brasil: projetos BH-Viva e ELSI-Urbe

Abstract

Introduction: Intraurban studies, such as those conducted in neighborhoods, are appropriate for analyzing and predicting individuals' daily activities and their exposures to the urban environment. Systematic observation has proven to be a reliable method for obtaining information about neighborhoods. Although well-established in the literature, few studies in Latin America have used it for this purpose. To carry out this method, raters assess relevant items in sampled locations, typically in street segments, which are contiguous parts of a street delimited by intersections or crossings. Objective: To measure neighborhood characteristics using the systematic observation method in Brazilian urban contexts. Methods: The first article was based on data from in-person systematic observation, known as Systematic Social Observation (SSO), from the BH-Viva Project, conducted in two slums and their surrounding areas in a Brazilian capital in 2019. Indicators were constructed through principal component analysis. Internal consistency was verified by Cronbach's alpha, and the indicators were transformed into scales from 0 to 5. Comparisons between slums and surroundings were made using the Mann-Whitney U Test. The second article was based on data from virtual systematic observation, known as a virtual audit, from the ELSI-Urbe Project, conducted for the first time in 20 municipalities representing all Brazilian regions in 2024, within the context of the ELSI-Brazil Project, which focuses on aging research in the country. Initially, a reliability study was conducted using Fleiss' kappa. Then, indicators were constructed through principal component analysis, with internal consistency verified using Cronbach's alpha, and the indicators transformed into scales from 0 to 5. Comparisons between income quintiles were made using the Kruskal-Wallis Test. Results: In the first article, 373 street segments in 63 neighborhoods were evaluated. Streets, sidewalks, signage, and safety had higher medians in the surroundings compared to the slums. In contrast, the social interaction and neighborhood problems domains had higher medians in the slums. In the second article, the reliability study, conducted on 189 segments in eight municipalities, showed an average agreement of 0.6. The total sample included 957 segments in 378 neighborhoods from 20 municipalities. The results indicated that, within the urban infrastructure domain, the pathway accessibility subdomain, which includes street condition, sidewalk extending along the entire segment, and unobstructed sidewalk, and within the traffic domain, the markings and controls subdomain, which includes lane markings, median lanes, traffic control devices, and crosswalks, had higher medians in higher-income quintiles. On the other hand, within the urban infrastructure domain, the surface and continuity subdomain, which refers to street material, presence of sidewalk, and sidewalk even along its entire segment, had higher medians in income quintiles 2 and 3. Within the physical disorder domain, the absence of infrastructure and vegetation subdomain, which includes the condition of electrical wires, absence of trees, and absence of gardens or potted plants, had higher medians in lower-income quintiles. Conclusion: The use of a systematic and replicable data collection method, both in-person and virtual, enabled the acquisition of detailed information on the intraurban context. This procedure allows for capturing local characteristics that better reflect residents' realities. When combined with analytical strategies, this information facilitates a deeper understanding of neighborhood characteristics, with scalability potential, especially in multiple and diverse intraurban contexts in large or continental-scale countries such as Brazil.