Respondents and non respondents to population and housing census: some strategies for data collection design in the era of low response rate and high response burden. An application of a decision tree model

Abstract

The Italian National Institute of Statistics (Istat) has adopted the methodology of permanent Population and Housing Census in 2018, which is planned as an annual cycle. This study aims to identify the profile of respondents and non-respondents to the Census, in order to put respondents at the centre of the design and management of data collection and to develop specific strategies focused on subsets of the sample population. The methodology used was the 'Tree decision model', which was able to present the relative weight of the included variables on the independent variable through a good and direct visual layout. The first results confirm the relevance of socio-demographic variables, such as the age structure, education level, and occupational status of the sampled household, and present some new analytical insights capable of describing first the profile of non-respondents and then the profile of respondents by channel.