Table 3. Key Questions and Answers for the MRP Methodology, BRFSS 2013, Massachusetts BRFSS 2013, and Boston BRFSS 2010/2013.
Questions | Answers |
---|---|
Where can I find additional information on the methodology used in small area estimation? | A summary of small area estimation and the MRP approach can be found in references 5 and 10 in this article. |
What surveys can be used for the approach? | State BRFSS or other local health surveys with hierarchical structure and spatial identifier. |
Can the approach be used to generate estimates for other areas, such as rural areas? | Yes. The approach can be used to generate estimates for any target small geographic area. |
Can the models be used to evaluate the effectiveness of the local public health interventions? | The estimates are generated based on the multilevel models, which include covariates obtained from the source survey. Unless the survey provides such information on local interventions, the model is not able to predict intervention effectiveness. |
Can the model be used to track the changes at the local level over time? | The methods in this study are not designed for assessing trends. |
Has the methodology been evaluated for accuracy? | The model was evaluated in comparison with direct estimates from local health survey at the county and city levels. Please refer to correlation results in the findings. |
Where can I find additional information about the methodology application? | Please refer to the website www.cdc.gov/500cities. For common questions and answers, please refer to https://www.cdc.gov/500cities/faqs/index.htm. For specific questions, please contact 500cities@cdc.gov. |
Abbreviation: MRP, multilevel regression and poststratification.