Case vs households

UNHCR regularly conducts quantitative surveys of its population-of-concern. Its registration database, proGres, often serves as the sampling frame from which a systematic or random sample is drawn – which is, in principle, a statistically robust approach. However, the units in the proGres list are cases, which are different from a household as defined in standard survey methodology.

Ignoring this discrepancy between the sampling unit and the analytical unit and attempting to simply carry out the interview at the case level can lead to improper measurement for household-level and individual-level data.

This document presents three options for dealing with this problem, outlining considerations for which option is best suited to the context where the data collection exercise is planned.