Assessing exposure to violence using multiple informants: application of hierarchical linear model

Area of Science:

• Criminology
• Developmental Psychology
• Sociology

Background:

• Exposure to violence (ETV) is a significant concern in child development.
• Understanding demographic risk factors is crucial for effective intervention.
• Discrepancies in reporting ETV between children and parents can complicate research.

Purpose of the Study:

• To assess how demographic risk factors influence children's exposure to violence (ETV).
• To examine how these effects differ based on the informant (child vs. parent).
• To evaluate the utility of a multivariate approach integrating data from multiple sources.

Main Methods:

• Utilized data from the Project on Human Development in Chicago Neighborhoods (PHDCN).
• Collected self-reports from 1880 children (ages 9, 12, 15) and 1776 parents.
• Employed a two-level hierarchical linear model (HLM) with multivariate outcomes.

Main Results:

• Parents consistently reported lower levels of ETV compared to child self-reports.
• Associations between age, gender, parent education, and ETV were stronger in child self-reports.
• HLM analysis revealed significant interactions between risk factors and informants.

Conclusions:

• Integrating data from multiple informants provides a more comprehensive understanding of ETV.
• Demographic risk factor effects on ETV are informant-dependent.
• Hierarchical linear modeling is effective for analyzing multi-informant data and handling missing data.