? Do you notice anything interesting about the tables?
Strata are designed in such way they are homogeneous and non-overlapping with respect to under study characteristic such that ? Ni=N
? Is there anything noteworthy about the mean?
The population mean is basically stratum means weighted arithmetic mean in the stratified sampling case. Strata are designed in such a way that they should be within homogeneous. Population total or population mean are estimated with higher precision through providing weights form obtaining each stratum estimate.
? Is there anything noteworthy when comparing the differences between smaller and larger sample size?
Stratified sampling main objective of any estimation problem is to find a population estimator parameter that can handle population salient features. If with respect to the characteristic of under study the population is homogeneous, then simple random sampling method will generally yield sample which are homogeneous and in turn, mean of the sample will act as a population mean good estimator. For increasing an estimator of the precision, there is a requirement to use sampling of the scheme that can decrease heterogeneity in the particular.