Abstract:
This technical report assesses the overall quality of data collected in the first wave of NIDS-CRAM. Several quality dimensions are investigated including the importance of item non-response, measurement error and the degree to which the data on key variables adequately represent the national situation. The focus on these data-quality dimensions is motivated by several considerations. Most obvious is the usefulness of the data in informing policy. To that end this paper with its focus on item level data quality, complements an earlier technical paper that considered the issue of representivity at the level of the sample frame and unit non-response. Surveys around individual and household wellbeing are typically conducted face-to-face. The lockdown in response to Covid-19 forced researchers into new modes of data collection and a consideration of NIDS-CRAM data quality is informative in assessing the feasibility of telephone surveys for policy relevant social science research.
