This paper reflects on an experience of integrating qualitative and quantitative approaches to sampling, data collection, data types, and analysis in a study of destitution in Ethiopia.It focuses on three "qualitative" elements within the household questionnaire: (1) a holistic selfassessment

of household dependence; (2) household-level recall questions to estimate trends in

the absence of time-series data; and (3) proportional piling (a quantification technique widely used in participatory fieldwork) to estimate income diversification. The paper concludes that the pragmatic combination of qualitative and quantitative elements was successful, and indeed essential, in addressing the study’s policy-oriented research questions: however, tensions and trade-offs remain.