Assessing Community Variation in Public Health Indicators
Author Information
Author(s): Arndt Stephan, Laura Acion, Kristin Caspers, Ousmane Diallo
Primary Institution: University of Iowa
Hypothesis
How much of the diversity among communities in public health indicators is due to random variation versus real differences?
Conclusion
The Pearson's chi-square measure is more effective for assessing community variation in public health indicators than the alternative method.
Supporting Evidence
- The study found that the Pearson's chi-square method provided better estimates of true community variance.
- More than 60% of community differences in the BRFSS data were attributed to random variability.
- The IYS data indicated that over 85% of community variation was due to actual differences.
Takeaway
This study looks at how different communities show varying health needs and whether those differences are real or just random chance.
Methodology
The authors compared two statistical measures of heterogeneity to assess community differences in health indicators using simulations and real data.
Potential Biases
Potential bias due to random sampling errors affecting community health indicators.
Limitations
The study may not generalize to indices measured on fewer observations, and it assumes independent observations among communities.
Participant Demographics
Participants included individuals from 99 counties in Iowa, with varying sample sizes and demographics.
Statistical Information
P-Value
0.0002
Confidence Interval
0.075 to 0.234
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website