The point estimate of the proportion was 14.3 percent (standard error 1.5),
which imply that the population of interest here does not qualify to be a rare
population according to the definition by Kish (1987). However it is about seven
percentage units less than the unweighted proportion. The interpretation is that
real estates with fishing rights has been selected with an higher probability than
the real estates without fishing rights. This pattern is emphasized looking only at
the sixth group of real estates, for which only the total area of the real estate was
known and the 2Pps design used as sampling design. There the proportions were
24 and 18 percent, respectively. Hence, this indicates that the design of the survey
has worked as intended, and that a fixed-size ps design as the 2Pps design was a
good choice for the problem at hand.