sse.rb.2stage {RSurveillance}

Two-stage risk-based system sensitivity

Description

Calculates system sensitivity for 2 stage risk-based sampling, llowing for a single risk factor at each stage and using either binomial or hypergeometric approxiation

Usage

sse.rb.2stage(C = NA, pstar.c, pstar.u, rr.c, ppr.c, rr.u, ppr.u,
  N = NA, n, rg, se)

Arguments

C

Population size (number of clusters), NA = unknown (default)

pstar.c

cluster level design prevalence (scalar)

pstar.u

unit level design prevalence (scalar)

rr.c

cluster level relative risks (vector with length corresponding to the number of risk strata), use rr.c = c(1,1) if risk factor does not apply

ppr.c

cluster level population proportions for risk categories (vector), NA if no cluster level risk factor

rr.u

unit level relative risks (vector with length corresponding to the number of risk strata), use rr.u = c(1,1) if risk factor does not apply

ppr.u

unit level population proportions for each risk group (optional) matrix, 1 row for each cluster, columns = unit level risk groups, not required if N is provided

N

population size per risk group for each cluster, NA or matrix of N for each risk group for each cluster, N=NA means cluster sizes not provided

n

sample size per risk group for each cluster sampled, matrix, 1 row for each cluster, columns = unit level risk groups

rg

vector of cluster level risk group (index) for each cluster

se

unit sensitivity for each cluster, scalar or vector of values for each cluster, equal in length to n

Value

list of 2 elements, a scalar of population-level (surveillance system) sensitivity and a vector of cluster-level sensitivities

Examples

# examples for sse.rb.2stage
pstar.c<- 0.02
pstar.u<- 0.1
rr.c<- c(5, 1)
ppr.c<- c(0.1, 0.9)
rr.u<- c(3, 1)
se<- 0.9
n<- cbind(rep(10, 50), rep(5, 50))    
rg<- c(rep(1, 30), rep(2, 20))
ppr.u<- cbind(rep(0.2, 50), rep(0.8, 50))
N<- cbind(rep(30, 50), rep(120, 50))
C<- 500        
sse.rb.2stage(C=NA, pstar.c, pstar.u, rr.c, ppr.c, rr.u, ppr.u, N=NA, n, rg, se) 
sse.rb.2stage(C, pstar.c, pstar.u, rr.c, ppr.c, rr.u, ppr.u, N=NA, n, rg, se) 
sse.rb.2stage(C=NA, pstar.c, pstar.u, rr.c, ppr.c, rr.u, ppr.u, N, n, rg, se) 
sse.rb.2stage(C, pstar.c, pstar.u, rr.c, ppr.c, rr.u, ppr.u, N, n, rg, se)