Comparing Random with Non-Random Sampling Methods

摘要

Although most people doing survey work would prefer to use random methods when drawing their samples, it is rarely practical. Instead they use a method involving every nth member of the population. This study compares the two methods. It was found that as long as the attribute being sampled is randomly distributed among the population the two methods give essentially the same results. However, if the attribute is not randomly distributed among the population the two methods give radically different results. In some instances the every nth method gives much better inferences about the population than do the random methods. In other instances it gives much worse inferences. (Author)