Coverage correction of sampling weights to known population totals

Formula: w_i' = w_i * (N_target_h / hat N_h)

Usage

Covpop(y, weights, target_totals, strata = NULL)

Arguments

y: Survey variable, length n.
weights: Design weights before correction, length n.
target_totals: Known population count per post-stratum, in the order the stratum labels first appear in strata.
strata: Post-stratum label per unit, or NULL for one stratum.

Value

List with estimate (post-stratified mean of y), w_adj, factors, total, n, method.

Details

Within each post-stratum h the weights are rescaled so that they sum exactly to the known population count. By construction the adjusted weights reproduce the control totals, which is the check the method is built to satisfy.

References

Sarndal, Swensson & Wretman (1992), Model Assisted Survey Sampling, Springer, section 7.6.

Examples

Covpop(y = c(1, 2, 3, 4, 5, 6, 7, 8), weights = c(1, 2, 3, 4, 5, 6, 7, 8), target_totals = 5L)
#> $estimate
#> [1] 5.666667
#> 
#> $w_adj
#> [1] 0.1388889 0.2777778 0.4166667 0.5555556 0.6944444 0.8333333 0.9722222
#> [8] 1.1111111
#> 
#> $factors
#> [1] 0.1388889
#> 
#> $total
#> [1] 5
#> 
#> $n
#> [1] 8
#> 
#> $method
#> [1] "coverage correction of weights to known population totals"
#> 
#> attr(,"class")
#> [1] "morie_rich_result" "list"