Ontology highlight
ABSTRACT: Background
Recruiting hard-to-reach populations for health research is challenging. Web-based platforms offer one way to recruit specific samples for research purposes, but little is known about the feasibility of online recruitment and the representativeness and comparability of samples recruited through different Web-based platforms.Objective
The objectives of this study were to determine the feasibility of recruiting a hard-to-reach population (pregnant smokers) using 4 different Web-based platforms and to compare participants recruited through each platform.Methods
A screener and survey were distributed online through Qualtrics Panel, Soapbox Sample, Reddit, and Amazon Mechanical Turk (mTurk). Descriptive statistics were used to summarize results of each recruit
SUBMITTER: Ibarra JL
PROVIDER: S-EPMC5928332 | biostudies-literature | 2018 Apr
REPOSITORIES: biostudies-literature