Statistical matching techniques in order to plan interventions on socioeconomic weakness: An Italian case

Abstract

The techniques for integrating data from multiple sources have the objective of identifying records relating to similar or equal units. Moreover, they estimate the joint distribution of several variables observed on different data files and merge information records.As an example, the estimation of the family risk of poverty, based on “objective” indicators (such as income or recourse to indebtedness), is independent from the awareness of those directly involved and therefore is correctly used in order to plan effective interventions against poverty. However, the researchers could usefully investigate also on the “subjective” perceptions of the standard of living and the causes that lead to socioeconomic hardships of families, but often such perceptions are referred to sub-samples of the previous population (or to different samples of it). In this work, we describe a useful model to integrate data between two surveys (Eurostat EU-SILC and Lifestyles survey) through a Statistical Matching method (hot deck distance). The Cluster Analysis is an essential component of the chosen technique of data integration, as well as of the following descriptive analysis.