A web survey on an elusive population: a focus on indicators to manage data collection process

Abstract

In 2021, the Italian National Statistical Institute (Istat) started the study to carry out a survey about labor discrimination of LGB (lesbians, gays and bisexuals) persons without legal relationship2. The survey population is extremely sensitive and hard-to-count mainly for two reasons: it is (self) defined by sexual orientation and there is no list to count and identify the initial population. Moreover, labor discrimination is a topic generally underrepresented in any kind of subgroups of Italian population.
The survey fieldwork has been carried out from January to the end of May in 2022.
The project team chose the web Respondent Driven Sample (w-RDS) method as the most suitable way to sample the population. The web questionnaire has been developed following “a privacy by design” approach to preserve the privacy of respondents.
The main efforts of data collection design were devoted to the implementation of a system of ad hoc indicators able to: 1. monitor the quality of data collection process (how to define and calculate the response rate without an initial population?); 2. evaluate the goodness of the sample (how many and which respondents can guarantee the data quality criteria?); 3. allow decisions regarding the change of data collection techniques or data design in itinere.
The paper addresses methodological and organizational aspects of data collection design, focusing on system of monitoring survey indicators, for an experimental survey on LGB adult population resident in Italy. Lesson learned could be useful for RDS implementation in future surveys conducted by National Statistical Institute.