Evaluating interventions and policies for robbery prevention in a matched case-control study of convenience stores.
Landsittel-DP; Hendricks-A; Amandus-HE; Malcan-J
Am J Epidemiol 1998 Jun; 147(11)(Suppl):S28
Previous research on convenience store robbery has indicated that robbers selectively choose their target from stores within some radius, based on characteristics of the store and surrounding area. In this study we proposed, and illustrated the use of a prospective matched case-control design to directly address the robber's selection of a target store in evaluating preventive interventions and policies. This approach provided an innovative, and statistically powerful design for studying the problem of convenience store robbery. Four hundred cases were selected by identifying convenience store robberies in the three largest metropolitan areas in Virginia. For each robbery, three controls were randomly selected from other convenience stores (not robbed on the same day) within a 2-mile radius. Using conditional logistic regression, and methods to adjust for correlated observations, we identified which policies and interventions were significantly I associated with robbery. This study overcame many limitations of previous studies by collecting data on a wide range of variables. The advantages and limitations of this, and other methodologies are discussed.
Epidemiology; Demographic-characteristics; Injury-prevention; Safety-measures; Case-studies; Retail-workers; Grocery-stores
American Journal of Epidemiology