Readme file
SERIES
C
 Applied
Statistics
Efficient survey sampling of households via Gaussian quadrature, by
C. Arndt, J. Kozlitina and P. V. Preckel
Journal of the Royal Statistical Society, Series
C, Applied Statistics, Volume 55
(2006), 355 - 364
The file "GQ_data.txt" contains the data set that was used for
illustration in the paper. There are 6510 observations (households), and
for each we have the following variables recorded: (the data are in space-delimited
format)
"area" - 0 = rural, 1 = urban
"educ" - highest level of education attained by any of the household
members, years
"prov" - 0 = northern, 1 = southern
"size" - no. of household members
"empl" - no. of employed family members
"depend" - number of dependents
"roof" - type of roof in the house
"inc" - daily expenditure on food, per adult equivalent unit (AEU)
"food" - daily per capita expenditure on food, including own production
"total" - total daily per capita expenditure on food and non-food
related items
PROGRAMS:
All computations are performed using the GAMS modelling language:
******************
GAMS IDE 2.0.20.0
www.gams.com
******************
The program "GQ_samp" performs GQ subsampling and Monte Carlo
resampling. The file "GQ_sens" contains additional code for sensitivity
analysis discussed in the paper.
ADDRESS FOR CORRESPONDENCE:
Channing Arndt
Ministry of Planning and Development
Av. Ahmed Sekou Touré nº 21, 4º Andar
Maputo, Mozambique
Telephone: +258 82 314 6920
Fax: +258 21 492 625
E-mail: channingarndt@gmail.com
Datasets (arndt.zip,
size - 172KB) |
