– selection of cases because they are interesting to study (e.g. the famous Kinsey study of sexual behavior)

– inherent difficulties of the sampling process (e.g. undercoverage of transient populations in population samples)

– selection of cases that yield an interesting result (e.g. repeated modifications and iterations of a model to data, until something of interest emerges from one of the models – this verges of research fraud)

– Self-selection in response to surveys (people who respond may differ systematically from those who do not)