Web Survey Bibliography

Title Non-probability Sampling
Year 2016
Access date 10.03.2017
Abstract A sample is a subset of a population and we survey the units from the sample with the aim to learn about the entire population. However, the sampling theory was basically developed for probability sampling, where all units in the population have known and positive probabilities of inclusion. This definition implicitly involves randomization, which is a process resembling lottery drawing, where the units are selected according to their inclusion probabilities. In probability sampling the randomized selection is used instead of arbitrary or purposive sample selection of the researcher, or, instead of various self-selection processes run by respondents. Within this context, the notion of non-probability sampling denotes the absence of probability sampling mechanism. In this chapter we first reflect on the practice of non-probability samples. Second, we introduce probability sampling principles and observe their approximate usage in the non-probability setting and we also discuss some other strategies. Third, we provide a closer look at two contemporary – and perhaps also the most exposed – aspects of non-probability sampling: online panels and weighting. Finally, we summarize recommendations for deciding on probability–nonprobability sampling dilemmas and provide concluding remarks.
Year of publication2016
Bibliographic typeBook section

Web survey bibliography (8060)