Surveys not of people

Not surprisingly, in Public Opinion Quarterly all the surveys were based on people. In the remaining seven social science journals, sampling of events, units of government, or organizations were more common, and amounted to 30 percent of primary survey research and 13 percent of secondary survey research. In most instances, the distinction between incomplete enumeration and sampling from a population was easy to identify; if the percent of persons or units or events sampled was over a quarter of the population, the article was not counted as survey research in the analysis here.