Some low budget designs for sampling in m-dimensional subsets
(m = 1, 2, 3, ... , k-1) of some k-dimensional populations are proposed. Elements
from the selected subset of the population (instead of the whole population)
are drawn by one of the classical and well known sampling techniques like the
Simple Random Sampling or the Systematic one, etc. The final budget comes down
over 10 times. Some illustrative examples are given.