Generalizability and representativeness: Considerations for internet-based research on substance use behaviors.

ABSTRACT: Substance use is frequently studied using nonprobability internet-based samples. It is difficult to evaluate the utility of these samples without a clear understanding of two key concepts: generalizability and representativeness. Part 1 of this article (a) offers a particular viewpoint on the distinctions and relations between these two concepts, (b) suggests that purposive (i.e., nonprobability) samples, when used carefully, can be used to construct valid scientific generalizations, and (c) explores some analytical consequences of sampling decisions that change sample heterogeneity. Part 2 of this article explores the overlap between internet-based sampling of substance use behaviors and the concepts discussed in Part 1. Specifically, Part 2 reviews relevant literature and presents exampl

SUBMITTER: Borodovsky JT 

PROVIDER: S-EPMC10053420 | biostudies-literature | 2022 Aug

REPOSITORIES: biostudies-literature