Sweden {stratification}

The MU284 Population of Sweden Municipalities from Sarndal et al. (1992)

Description

This data set comes from Sarndal et al.'s book (1992), Appendix B. It contains different variables that describe 284 municipalities in Sweden.

Usage

Sweden

Format

A data frame with 284 observations on the following 11 variables.

id

Identifier running from 1 to 284

P85

1985 population (in thousands)

P75

1975 population (in thousands)

RMT85

Revenues from the 1985 municipal taxation (in millions of kronor)

CS82

Number of Conservative seats in municipal council

SS82

Number of Social-Democratic seats in municipal council

S82

Total number of seats in municipal council

ME84

Number of municipal employees in 1984

REV84

Real estate values according to 1984 assessment (in millions of kronor)

REG

Geographic region indicator

CL

Cluster indicator (a cluster consists of a set of neighboring municipalities)

Details

In this package, REV84 is used as a stratification variable and RMT85 as a survey variable.

Source

Sarndal, C. E., Swensson, B. and Wretman, J. (1992). Model Assisted Survey Sampling. Springer Verlag, New York.

References

Rivest, L.-P. (2002). A generalization of the Lavallee and Hidiroglou algorithm for stratification in business surveys. Survey Methodology, 28(2), 191-198.

Examples

X <- Sweden$REV84
Y <- Sweden$RMT85

# Study of the relationship between X and Y
plot(log(X), log(Y))
# Extreme values are omitted for a more robust estimation
keep <- X/Y>quantile(X/Y,0.03)&X/Y<quantile(X/Y,0.97)
plot(log(X)[keep], log(Y)[keep])
reg<-lm( log(Y)[keep]~log(X)[keep] )
summary(reg)

# Stratification assuming X=Y
nomodel <- strata.LH(x=X, CV=0.05, Ls=3, alloc=c(0.5,0,0.5), takeall=1, model="none")
nomodel
var.strata(nomodel, y=Y) # The target CV is not reached

# Stratification taking into account a loglinear model between X and Y, 
# using the estimated parameters values
model <- strata.LH(x=X, CV=0.05, Ls=3, alloc=c(0.5,0,0.5), takeall=1, model="loglinear",
        model.control=list(beta=reg$coef[2], sig2=summary(reg)$sigma^2, ph=1))
model
var.strata(model, y=Y) # The target CV is reached