Using Mixed Methods to Better Sample Migrant Populations: Surveying the Venezuelan Diaspora in Colombia and Peru

Résumé fourni par la source

Reaching migrant and diaspora populations faces methodological hurdles arising from their mobility and clustered settlement patterns, as well as the frequent absence of comprehensive sampling frames. Adaptive Cluster Sampling (ACS) offers a probabilistic but flexible design for surveying hard-to-reach populations, but implementing ACS requires making methodological decisions that depend on the context. We employ a mixed-methods approach to implement ACS to survey Venezuelan migrants living in Colombia and Peru in 2024. To do so, we use administrative records, qualitative expert interviews, and small-scale pilot studies to identify settlement clusters, refine definitions of households and respondent eligibility, and set threshold triggers and stopping rules for sample expansion. These initial studies inform our implementation of multistage ACS in two large-scale surveys. Our mixed-method formative phase proved indispensable for identifying high-density neighborhoods and adapting fieldwork procedures to local contexts. Even though our two studies sampled the same migrant population, successful cross-national sampling required calibrating ACS parameters to country-specific densities. In Peru, larger census zones served as the optimal sampling unit, whereas in Colombia, smaller city blocks were more effective due to the greater concentration of Venezuelan households. We conclude that mixed-methods approaches can enhance coverage, efficiency, and feasibility in adaptive designs for surveying high-mobility and hard-to-reach populations, particularly in urban settings where traditional probability sampling methods are impractical.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.