COORDINATED SAMPLES FOR COMPLEX
SURVEYS
Signe Balina1 and Janis Lapins2
1
University of Latvia, Latvia
e-mail: signe@lanet.lv
2
Bank of Latvia, Latvia
e-mail: Janis.Lapins@bank.lv
Abstract
In a lecture we are discussing how samples of two household surveys can be coordinated
(keeping in mind that for each of the surveys several different goals have to be reached) in order
to ensure rational usage of the survey resources. The main ideas, problems and their solutions are
illustrated by case studies of two surveys carried out by the Central Statistical Bureau of Latvia –
Household Budget Survey and Labour Force Survey.
1 Introduction
Rational usage of survey resources is a natural demand when planning a sample survey. Since
for different surveys we can use the same sampling frames, interviewer network, technical
facilities, etc., one can try to find more economic allocation of total survey resources by
harmonizing sample designs of two (or more) surveys. However, the reduction of the total
survey resources has not to be considered as the only task – the priority demand is to have a
high quality output for each of the surveys. In a lecture we share with our experience in rede-
signing two household surveys Latvia – Household Budget Survey and Labour Force Survey.
2 Topics considered
We plan to consider the following main topics:
Sample design of and problems detected in the old HBS and LFS,
Moving/switch to continuous Labour Force Survey,
Usage of a single (mobile) interviewer network for HBS and LFS,
Usage of Population Census 2000 areas as PSU,
Stratification and sample allocation,
Sampling of PSU (systematic probabilities proportional to size sampling),
Development of rotation scheme of PSU,
Rotation of households,
Estimation of the sampling error (dependent random groups technique).
References
Lapins, J.,and E. Vaskis (1996). The new household budget survey in Latvia. Statistics in Transition, 2, 7, 1085-
1102.
Lapins, J., Vaskis, E., Priede, Z., and Balina, S. (2002) Household Surveys in Latvia. Statistics in Transition, 5, 4,
617-641 (see also: “Design and implementation of household surveys in Latvia. Household Surveys in
Developing and Transition Countries: Design, Implementation and Analysis, UN, Ch. 25, 2003, 26 –
30”).