Presenting the Shiny app

We want to integrate information from two independent surveys run by Eustat (Basque Government) that gather social and labour market information about the Basque population, namely:

1. The Living Conditions Survey, in which people are interviewed every 5 years to inform about individual and family dimentions of life such as health status, working conditions, type of household, social relations, level of instruction, income level, and so on.

2. The Population in Relation to Activity Panel, the primary mechanism by which the labour market classification is derived for the Basque population.

• Samples:

1. around 5000 people in the Living Conditions Survey,
2. around 12000 people in The Population in Relation to Activity Panel.

• Period of time: last quarter of 2009.

• My research project involves using statistical matching or data fusion methods in order to combine information from distint sources in the best (statistically sound) possible way.

• By processing distinct microdata files which have common information about the same population of interest (but on different units), data fusion methods produce a synthetic file potentially containing information about all the units and all specific variables, from all surveys.

• A synthetic file was obtained containing all items from the Living Conditions Survey and the Labour Force Segment.

• So now it is possible to study living conditions for each labour force segment.

• But this is possible only if the synthetic file was derived in a correct way...

So let's explore tables with easiness and care!

• Statistical matching is not easy to explain.

• If you actually 'see' the results, it should be much easier to understand and communicate.

The chosen visualization method is the balloon plot, where the value of each pairs of categories (in this case, % of people in each segment, having each condition) is mapped to the area or radius of a balloon.

In this way, differences in patterns between distinct segments are highlighted.

• Susanne Rässler, Statistical Matching: A Frequentist Theory, Practical Applications and Alternative Bayesian Approaches. Lecture Notes in Statistics. Springer.