The Importance of Probability-Based Sampling Methods for Drawing Valid Inferences


Gary Langer authored two chapters in The Palgrave Handbook of Survey Research discussing the history of probability-based sampling in survey research and its importance in drawing valid inferences.

In the first chapter, Langer cites extensive research that finds non-probability samples, including increasingly common opt-in online panels, are less accurate and less consistent in their level of accuracy than probability sample surveys. The latter chapter expands on this research in greater detail and calls for full disclosure and honest assessment of emerging techniques, including further study of appropriate as well as inappropriate uses of convenience sample data.