Survey-weighted and small-area estimation
Household survey data are analysed with their sampling design, and small-area models borrow strength across districts to estimate subnational coverage, with credible intervals mapped beside the estimates.
The Global Centre for Risk and Innovation (GCRI)
Probability and statistics turn incomplete, delayed and biased health data into estimates with stated uncertainty. In this field they handle survey sampling, small numbers, under-reporting and shifting baselines, so that every indicator carries a range and the assumptions behind it rather than a single, overconfident number.
Health statistics is the application of probability and statistics to incomplete, delayed and biased health data so that estimates carry stated uncertainty. It handles survey sampling, small numbers, under-reporting and shifting baselines, such as the expected deaths against which excess mortality is measured.
The working screen behind this method: its structure, terms and states. Names are placeholders and values illustrative.
Health · Statistics for health evidence · Run R-6555
generalised Pareto fit
| Shape ξ | 0.28 | −0.04 | 0.48 | heavy tail if above zero |
| Scale σ | 2.3 | 1.0 | 7.5 | units of the variable |
| Exceedances | 136 | declustered |
Probability and statistics provide the mathematical framework for describing uncertainty and learning from data. They underpin estimation, testing and prediction in every risk discipline. In health and biosecurity, statistical methods quantify the likelihood and uncertainty of losses across interconnected systems. They make the evidence behind risk decisions explicit and reproducible. The canonical description is on the STEM Commons.
Household survey data are analysed with their sampling design, and small-area models borrow strength across districts to estimate subnational coverage, with credible intervals mapped beside the estimates.
Models fitted to past mortality with trend and seasonality give the expected count against which observed deaths are compared, and the baseline's uncertainty is carried into the excess.
Capture-recapture, serosurveys and multiplier methods estimate how many infections surveillance misses, so case counts are read against what they are likely to represent.
Each term as this page uses it. Where a standard-setting body defines the term, the wording follows that body.
| Term | Definition |
|---|---|
| Excess mortality | Deaths above the number expected from a baseline, estimated with a model whose assumptions are stated. |
| Under-ascertainment | The gap between true infections or cases and those detected by surveillance. |
| Survey weight | A factor applied to each survey response so the sample represents the population. |
How far each division sits from a large city, with the roads and airports that carry patients and supplies, shown for Indonesia as an example geography for this subject.
Data from the source register feeds the methods; calibration and an independent test gate the estimate and its limits; the decision stays with the competent authority.
Fig. 01Nodes carry their type, metrics and state; connectors show what feeds what; the gates are the validation steps; the timeline is one application. Design of the method's use, not a record of a run.
Datasets and records used by probability and statistics on this platform, with the role each plays and the label its values carry. Sources are listed in the source register.
| Data | Role | Label |
|---|---|---|
| DHS and MICS microdata | Design-based and model-based estimation | ESTIMATED |
| Vital registration series | Baseline modelling | OFFICIAL |
| Serosurveys | Calibration of ascertainment | OBSERVED |
| Routine reporting with completeness metadata | Adjustment for missing reports | REPORTED |
Results are checked through diagnostics, cross-validation and sensitivity to assumptions. Independent replication confirms findings.
| Check | How it is done |
|---|---|
| Design-based variance | Standard errors account for clustering, stratification and survey weights |
| Posterior predictive checks | Models must reproduce the observed data before they are used to estimate or predict |
| Sensitivity to priors and baselines | Results are recomputed under alternative assumptions and the spread is reported |
Inference depends on assumptions about independence, distribution and sampling that may not hold. Small or biased samples limit reliability.
Systems Science and Modelling stewards probability and statistics for the Consortium and reviews how it is applied across platforms.
The current service level is R0: information and enquiries. An enquiry receives a reference and names the next step; applications, payments, compute allocations and registrations open only when their team, terms and review capacity exist.