PracTools (version 1.6)

nDomain: Compute a simple random sample size for an estimated mean or total for a domain

Description

Compute a simple random sample size using either a target coefficient of variation, $C{V}_0(d)$, or target variance, $V_0(d)$, for an estimated mean or total for a domain.

Usage

nDomain(CV0d=NULL, V0d=NULL, S2d=NULL, ybarUd=NULL, N=Inf, CVpopd=NULL, Pd, est.type)

Value

numeric sample size

Arguments

CV0d

target value of coefficient of variation of estimated domain mean or total

V0d

target value of variance of estimated domain mean or total

S2d

unit (population) variance for domain units

ybarUd

population mean of target variable for domain units

N

number of units in full finite population (not just the domain population)

CVpopd

unit (population) coefficient of variation for domain units

Pd

proportion of units in the population that are in the domain

est.type

type of estimate; allowable values are "mean" or "total"

Author

Richard Valliant, Jill A. Dever, Frauke Kreuter

Details

If CV0d is the desired target, then the unit CV, CVpopd, or the domain population mean and variance, ybarUd and S2d, must also be provided. If V0d is the constrained value, then ybarUd must be also be included in the function call. CV0d will then be computed as sqrt(V0d)/ybarUd.

References

Valliant, R., Dever, J., Kreuter, F. (2018, sec. 3.5.2). Practical Tools for Designing and Weighting Survey Samples, 2nd edition. New York: Springer.

See Also

nCont, nLogOdds, nProp, nPropMoe, nWilson

Examples

nDomain(CV0d=0.05, N=Inf, CVpopd=1, Pd=0.5, est.type="total")
nDomain(CV0d=0.05, N=Inf, CVpopd=1, Pd=0.5, est.type="mean")
nDomain(V0d=50, ybarUd=50, S2d=100, N=Inf, Pd=0.5, est.type="total")
nDomain(CV0d=0.05, ybarUd=50, S2d=100, N=Inf, Pd=0.5, est.type="total")
nDomain(CV0d=0.05, ybarUd=50, S2d=100, N=Inf, Pd=0.5, est.type="mean")