Caritas Clause Samples

Caritas. The Caritas dataset allows to investigate a segment of the poor population, which can be hardly detected by official statistics. According to Istat 11, people in absolute poverty accounted for about 2,9% of people living in the Centre of Italy in 2008. Anyhow such estimate is obtained taking into account data from the Household budget survey i.e. interviewing people with a stable housing. On the contrary, Caritas data concern people asking for the Caritas services, be they resident or homeless people. As a consequence part of poor people captured by the Caritas dataset is not covered by the Istat absolute poverty index. Unfortunately, the weak quality of the data (see § 2.3) affects the statistical analysis. Moreover the accessed dataset contains only part of the actually collected data. People seeking assistance in the Caritas centres are typically foreign (about 70%) and female people. They are 42 years old on average, with foreigners younger than Italians (respectively 30 and 57 years old on average). Women live generally in a household (with relatives or friends) whereas a significant part of male people (23%) live on their own. Most of people, especially if stranger, asks for goods and services(food, cloth, furniture, taking a bath etc.). Italians more frequently seek an economic support. 11 As defined by Istat (see: Istat ▇▇▇▇▇▇ e Norme, “La ▇▇▇▇▇▇ ▇▇▇▇▇ povertà assoluta” del 22 Aprile 2009, ▇▇▇▇://▇▇▇.▇▇▇▇▇.▇▇/dati/catalogo/20090422_00/) The JC and RA data sources seem to cover populations similar to the Silc population. As a consequence we expect to find some of the Pi-Silc sampled individuals in the administrative archives. The exact matching (or record linkage) is the technique used to identify and pick up such units. Record linkage is a technique which compares records contained in two files A and B, in order to determine pairs of records referred to the same population unit. The A and B files are supposed to contain identical units that have to be found according to an identifier (like the social security number) or a set of identifying variables (k variables) present in both files. Record linkage between two files is very simple provided that each record in both files contains the same identifier and this identifier is recorded without errors. In this case the problem is solved by simply picking out the records (if any) with the same identifier value. This procedure is known as exact matching.