ABSTRACT
Sampling has developed as a research method across disciplines over the 20th century. Even though some types of sampling predate written history, sampling for research is a modern phenomenon on which all scientific observation, statistical or not, has come to depend. The accuracy of quantitative research relies in large part on the ability to keep errors low, and one of the primary concerns is sampling errors. Reaching truly random samples of the general population through the Internet and phones faces difficulties, some of which are new and some are new variations on the same challenges sampling has always faced. Most telephone surveys use the household as the unit for sampling, which arose out of the fact that a landline phone is shared by the household. Web-based research also faces issues with sampling, but its low cost and ease of use makes it appealing. Its value depends on the credibility of the data, just as with traditional mail methods.
