In observational studies, selection bias will be completely removed only if the selection mechanism is ignorable, namely, all confounders of treatment selection and potential outcomes are reliably measured. Ideally, well-grounded substantive theories about the selection process and outcome-generating model are used to generate the sample of…

Developmental studies have provided mixed evidence with regard to the question of whether children consider sample size and sample diversity in their inductive generalizations. Results from four experiments with 105 undergraduates, 105 school-age children (M = 7.2 years), and 105 preschoolers (M = 4.9 years) showed that preschoolers made a higher…