This aspect of the study divided Chicago into 343 "neighborhood clusters," each inhabited by approximately 8,000 people, and each defined by specific geographic boundaries and internally homogeneous on a variety of census indicators. Questions asked of residents were intended to elicit their views of how much informal social control, social cohesion and trust, and violence existed in their neighborhoods. The researchers found that in neighborhoods scoring high on collective efficacy, crime rates were 40 percent below those in lower scoring neighborhoods. This difference supported the study's basic premise, i.e., that crime rates are not solely attributable to individuals' aggregate demographic characteristics; rather, crime is a function of neighborhood social and organizational characteristics. The researchers found that various dimensions of social composition influence the level of neighborhood collective efficacy. In neighborhoods where concentrated poverty was high, collective efficacy was low. Ethnicity/immigration was another important dimension, because areas of ethnic and linguistic heterogeneity may have less capacity to realize common values. Where this dimension was high, collective efficacy was low. In contrast, neighborhoods where residential stability was strong also tended to be strong on collective efficacy. Implications of study findings for crime prevention efforts are briefly discussed.