stratified random sample

straightforward random sampling may leave out a particular class of cases. Thus, it is conceivable that a random sample drawn, say, from a list of electors will consist overwhelmingly of women with few men selected despite the fact that the sexes are equally common on the electoral list. Random sampling, theoretically, may lead to a whole range of different outcomes some of which are not representative of the population. One way of dealing with this is to stratify the population in terms of characteristics which can be assessed from the sampling frame. For example, the sampling could be done in such a way that 50% of the sample is male and the other[Page 163] 50% is female. For example, males may be selected from the list and then randomly sampled, then the process repeated for females. In this way, variations in sample sex distribution are circumvented.