Conferences

In statistical practice systematic sampling is used in many modifications due to its simple handling. In addition, systematic
sampling may provide efficiency gains if it is well adjusted to the structure of the population under study. However, if systematic sampling
is based on an inappropriate picture of the population a high decrease of efficiency, i.e. a high increase in variance may result by changing
from simple random sampling to systematic sampling. In the context of 2-stage designs systematic sampling so far seems often in use for
subsampling within the primary units. In contrast to this practice we propose to randomize the order of the primary units to select thereafter
systematically a number of primary units and, then, to draw secondary units by simple random sampling without replacement within the primary
units selected. To compute the design effect the inclusion probabilities of second order must be known. First findings concerning the design
effect will be presented.