We use Demographic and Health Survey data to evaluate the impact of random spatial displacements on analyses that involve assigning covariate values from ancillary areal and point feature data. We introduce a method to determine the maximum probability covariate (MPC), and compare this to the naive covariate (NC) selection method with respect to obtaining the true covariate of interest. The MPC selection method outperforms the NC selection method by increasing the probability that the correct covariate is chosen. Proposed guidelines also address how characteristics of ancillary areal and point features contribute to uncertainty in covariate assignment.
Influence of demographic and health survey point displacements on point-in-polygon analyses
Warren, J. L., Perez-heydrich, C., Burgert, C. R., & Emch, M. E. (2016). Influence of demographic and health survey point displacements on point-in-polygon analyses. Spatial Demography, 4(2), 117-133. https://doi.org/10.1007/s40980-015-0015-z
To contact an RTI author, request a report, or for additional information about publications by our experts, send us your request.
Multifaceted risk for non-suicidal self-injury only versus suicide attempt in a population-based cohort of adults
Long-term effects of a diet supplement containing Cannabis sativa oil and Boswellia serrata in dogs with osteoarthritis following physiotherapy treatments
Tailoring off the shelf global evidence with local implementation research can boost action on overweight and obesity