Challenges and Lessons Learned in the Co-Creation of a Community Survey to Assess Neighborhood Collective Efficacy

Background: Democratizing knowledge requires social work researchers to collaborate with the communities experiencing the issues we are attempting to understand and address. This paper discusses community engagement strategies utilized in a 5-year, multi-stakeholder-informed randomized cluster-controlled trial testing a collective efficacy anti-violence intervention in ten intervention and ten comparison neighborhoods. Our project addresses the gap in community experiences of violence and assets that currently exists in publicly available and commonly relied upon data by illustrating how a community survey can be participatory developed, administered, disseminated, and interpreted.

This study addresses the stark increase and racial disparities in community violence in our region. Homicide rates increased by 43% in the city from 2019 to 2021, largely reversing declining trends of previous years. Despite making up only 6% of the population, Black men are victims in 66% of annual homicides, and most are between the ages of 18 and 34. Homicide victimization rates are 50 times higher among young Black males than the U.S. average.

This paper discusses strategies used to engage community partners in the design, administration, and dissemination of a community survey that captures concepts commonly missing from available Census and other public data sources.

Methods: The intervention consists of three phases: (1) community organization and mobilization; (2) violence prevention training for youth and adults; and (3) a community-based violence prevention project. The community survey is conducted in phase 1 and phase 3. Survey measures capture exposure to violence, the impact of exposure to violence, fear of crime, collective efficacy, neighborhood activism, values/norms regarding prevention and perceptions of police. We will discuss how original grassroots distribution methods evolved to using the CDC modified Casper method, which includes two stage cluster sampling where, for each neighborhood (i.e., sampling frame), residences are clustered within Census block groups, from which the sample is drawn.

Results: The community survey included areas prioritized and was pilot tested by community partners in the first two intervention neighborhoods. The survey was distributed through community partners, organizations, and events, purposely targeting each part of the neighborhood. The initial results were shared throughout the training to engage participants in understanding how survey respondents were experiencing violence and ways to prevent it. For example, results related to exposure to violence and the impact on mental health were shared along with community crime statistics, engaging participants in sharing lived experiences as they reflected on the data. Results regarding neighborhood collective efficacy were discussed along with strategies for building relationships and intervening in non-violent restorative ways.

Implications: Despite robust recruitment efforts, response rates ranged from 30-60% of the total 100 targeted responses in each neighborhood. Community partners were engaged in discussing the limitations of this approach and ways we could draw on existing resources and train community partners and team members to use door-to-door methods to collect data from a stratified random sample of residents using Casper. This approach gives everyone in the community an equal chance to participate, increasing representativeness and generalizability, and increases awareness about and engagement in the intervention.