Comparing alternative methods of measuring cumulative risk based on multiple risk indicators: Are there differential...

Area of Science:

• Child Development
• Psychometrics
• Public Health

Background:

• Cumulative risk (CR) measurement is crucial for understanding child development.
• Existing methods for assessing CR have varying conceptual and methodological assumptions.
• Identifying optimal CR measurement approaches is essential for accurate risk assessment.

Purpose of the Study:

• To examine and compare alternative methods for measuring cumulative risk (CR).
• To assess the conceptual and methodological assumptions of different CR measurement approaches.
• To provide guidance for researchers in selecting appropriate CR measurement techniques.

Main Methods:

• Compared individual risk measurement approaches (dichotomous, proportion, z-scores).
• Evaluated composite CR measures: observed score, variable-centered (reflective, formative), and person-centered (latent class, latent profile analysis).
• Utilized data from 169 low-income families assessing 12 risk indicators and children's externalizing problems.

Main Results:

• Variable-centered cumulative risk (CR) composites showed larger effect sizes on children's externalizing problems.
• Observed summary score CR indices and person-centered methods yielded smaller effect sizes.
• The study proposed a decision tree to assist researchers in method selection.

Conclusions:

• Variable-centered approaches offer a more sensitive measure of cumulative risk (CR) impact.
• The choice of CR measurement method significantly influences the observed effect size on child outcomes.
• Further research should consider the implications of different CR measurement strategies in diverse populations.