Estimation and sensitivity plan
Account for sampling and uncertainty
Account for survey design when estimating both values and uncertainty. Report item-specific denominators and assess nonresponse. Prevalence describes existing conditions and is not an incidence rate.
Flow diagram
Steps
Define prevalence and association measures
Specify target population, assessment period and denominators.
Account for sampling design
Specify survey weights, strata and clusters when applicable, including design-compatible variance estimation and outcome-specific denominators.
Specify uncertainty and precision
Use methods appropriate to the design and available sample.
Assess nonresponse and definition sensitivity
Plan plausible weighting, missingness and measurement alternatives.
Document interpretation limits
Simultaneous measurement does not establish temporal or causal direction.
Estimation and sensitivity plan
End flow
Tables and supporting information
Estimation specification
Record verified calculations separately from planned inputs.
Weighted prevalence
For included respondents i, yᵢ is 1 when the condition is present and 0 otherwise; wᵢ is the analysis weight. Equal weights give the ordinary sample proportion. Weights do not automatically repair coverage or nonresponse bias. Variance estimation must respect the sampling design; missing outcomes require an explicit policy.