The current study sought to expand the current understanding of the psychometric characteristics of the collective efficacy scale at the individual level and the role of collective efficacy in promoting safe, healthy community conditions.
A team of interviewers consisting of residents of the targeted neighborhoods were selected and trained to administer the field surveys (NIJ Neighborhoods Resident Survey Data, n=649). In order to ensure accuracy of the responses, the field supervisor conducted telephone validation for approximately ten to fifteen percent of the surveys. If the initial attempt to contact a resident of the household was unsuccessful, a flier (in English and Spanish) explaining the study and including contact information was left at the residence to allow residents to schedule interview times that were more convenient. Multiple additional attempts to contact a resident of the household were made at various times and days of the week. After four attempts or a refusal to participate in the interviews, a household was considered non-responsive and a replacement household was randomly selected from the sample database.
In addition to resident surveys, trained research staff conducted systematic social observations (SSOs) of street segments in selected neighborhoods. A team of at least three researchers conducted walkthroughs of these street segments to conduct the SSOs. The day of the week and time of day for each of these walkthroughs were varied to avoid systemic biases associated with the timing of the walkthrough. The researchers jointly completed the SSO instrument for each street segment that included physical and social indicators.
In order to examine a diverse cross-section of neighborhoods, eight neighborhoods in Miami-Dada County, Florida were purposely selected on the basis of demographic, socioeconomic, and criminological variables. Specifically, researchers compiled existing 2000 and 2010 Census data, projected demographic data based on the 2000 Census, homicide data from the County Medical Examiner, and calls for service data from the Miami-Dade Police Department, city of Miami Police Department, the Miami Gardens Police Department, and he Homestead Police Department. These data were used in a GIS environment to assist researchers in identifying suitable neighborhoods for inclusion in the study. Eight neighborhoods were selected for the study: Brownsville, Bunche Park, East Little Havana, Seminole Wayside Park, Kendall Hammocks Park, Ives Dairy Estates, Auberdale, and Coral Reef Park.
The sampling frame for the resident survey (n=649) was enumerated using a database obtained from Valassis Corporation containing all active mailing address known to the United States Postal Service for Miami-Dade County. From this database, a random sample of households was selected from each of the eight targeted neighborhoods.
For the systematic social observations (SSOs), street segments were operationalized as the "face block" or set of properties facing each other along a common street between two intersections. In each neighborhood, approximately 20 percent of the street segments were selected for SS0s. When selecting the segments, the neighborhoods were first divided into quadrants and street segments were selected by randomly sampling the addresses of survey respondents within each quadrant.
All residents of Miami-Dade county, Florida between January 2010 and September 2011.
The resident survey (NIJ Neighborhoods Resident Survey Data, 108 variables, n=649) incudes variables on the physical attributes of the neighborhood including the presence of vacant or abandoned lots, poor trash pick by the city, vandalism, and graffiti. Other variables ask about the social attributes including loud neighbors, homeless loitering, public drug or alcohol use, drug dealing, truancy, gun shots or gun violence and gang activity. Additional variables ask about the interactions between neighbors and how the neighbors respond to various issues in the neighborhood. Finally, residents are asked for demographic information including homeownership, sex, age, ethnic background, employment status, veteran status, primary language, marital status, and household income.
The overall response rate for the resident survey was 79.26 percent.
Several Likert-type scales.