Measuring Violent Behavior: Effects of Study Design on Reported Correlates of Violence (From Violent Crime, Violent Criminals, P 14-33, 1989, Alan Weiner and Marvin E Wolfgang, eds. - See NCJ-118975)

• NCJ Number: 118976
• Author(s): G S Bridges; J G Weis
• Date Published: 1989
• Length: 20 pages
• Sale Source: Sage Publications, Inc
• Publication Type: Research (Applied/Empirical)
• Language: English
• Country: United States of America

Annotation

This analysis of empirical studies of violence published between 1945 and 1983 concluded that the research design can affect the study findings.

Abstract

The research began with the collection of all empirical research publications on violent behavior during the study period, using searches of automated and printed references, an issue-by-issue search of 50 journals, and a search of the automated indexing system of the Library of Congress. Further screening resulted in the listing of 516 publications, from which a 50 percent systematic sample was taken. The analysis used the study design characteristics as the independent measures and study findings relating to age, gender, and social class as the dependent measures. Results showed that a study's findings based on individuals yielded weaker race/violence, gender/violence, and social class/violence correlations than did findings based on higher levels of aggregation. In addition, findings based on offender characteristics were stronger than those based on victim characteristics or the situation of the offense. Moreover, statistical controls for other correlates of violence significantly reduced the strength of the effects of race and social class. Discussion of implications and recommended research strategies, tables, notes, and 46 references.