Powering population health research: Considerations for plausible and actionable effect sizes

Area of Science:

• Public Health
• Health Equity Research
• Intervention Science

Background:

• The Evidence for Action (E4A) program funds research on social policies' health impacts.
• Applicants frequently struggle with selecting realistic effect sizes for power and sample size calculations.
• Existing guidance on effect size selection for population health research is limited.

Purpose of the Study:

• To provide guidance on selecting realistic and actionable effect sizes for population health intervention research proposals.
• To inform power, sample size, and minimum detectable effect (MDE) calculations.
• To address a common methodological challenge faced by researchers.

Main Methods:

• Analysis of five rigorously evaluated population health interventions.
• Illustration of considerations for selecting effect sizes.
• Examination of factors influencing effect size achievement.

Main Results:

• Plausible effect sizes for population health interventions can be smaller than commonly cited guidelines.
• Achieved effect sizes depend on intervention characteristics, target population, and outcomes.
• Population health impact is influenced by intervention reach.

Conclusions:

• Adequately powered studies of interventions with small effect sizes can yield valuable population health evidence if widely implemented.
• Demonstrating effectiveness of broad-reach interventions necessitates large sample sizes.
• Realistic effect size selection is key for robust population health research design.