COMBINE

Measurement uncertainty in socioeconomic data may hinder effective policy making. Correcting for measurement uncertainty, could result in an improved picture of how social phenomena evolve in time. COMBINE explores whether a statistical model (a so-called hidden Markov model) could assist Statistics Netherlands (CBS) to correct for measurement uncertainty in the production of official statistics on social assistance benefits. Practically, the statistical model triangulates information from two different sources (the municipal register and the Employment register). COMBINE thus explores the feasibility of producing model-based error-corrected statistics by National Statistical Institutes. An important question is whether COMBINE will be more efficient than the current practices of data-quality improvement that are currently applied.

For more information: https://combineproject.eu/