Modelling bias in combining small area prevalence estimates from multiple surveys
2011

Improving Smoking Prevalence Estimates Using Multiple Surveys

Sample size: 48 publication 10 minutes Evidence: high

Author Information

Author(s): Giancarlo Manzi, David J. Spiegelhalter, Rebecca M. Turner, Julian Flowers, Simon G. Thompson

Primary Institution: Medical Research Council Biostatistics Unit

Hypothesis

Can Bayesian hierarchical models effectively adjust for biases in smoking prevalence estimates from multiple surveys?

Conclusion

The proposed Bayesian models provide more accurate and precise estimates of smoking prevalence by adjusting for biases in commercial surveys.

Supporting Evidence

  • Bayesian models allow for additive biases in data sources.
  • Estimates from different surveys were combined to improve accuracy.
  • Correlation between sources was assessed to refine estimates.
  • Confidence intervals were adjusted to reflect bias variance.

Takeaway

This study shows how to combine information from different surveys to get better estimates of how many people smoke in small areas.

Methodology

Bayesian hierarchical models were used to adjust for biases in smoking prevalence estimates from seven surveys across local authorities.

Potential Biases

Commercial surveys may introduce biases due to unclear methodologies and sampling issues.

Limitations

The methodology of some surveys was unclear, and the estimates may be affected by various biases.

Participant Demographics

The study focused on local authorities in the east of England, with smoking prevalence estimates derived from various demographic groups.

Statistical Information

Confidence Interval

95%

Digital Object Identifier (DOI)

10.1111/j.1467-985X.2010.00648.x

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication