The main working hypothesis in the project is that effective and appropriate strategies of attention and inclusion to Vulnerable Groups of Forcibly Displaced People (FDP) need to be tailored to their specific Vulnerability Context (VC). Defining that context and what works in it can only be achieved through the actual participation of the involved actors.

The RAISD research strategy is based on methodological triangulation with:

This allows that FDP take an active first-hand role, not only to record their experiences, but also in the design of the project’s works. Regarding the characterization of their experiences to define the VC, the project will integrate different data, including their socio-demographic profile, their narratives, or observations from cooperating staff.

The collection of data will work in two separate timing and geographic settings:

Data collection from groups of FDP that are in transit, including interviews with migrants and volunteers of non-profit organisations in the refugee camps or centres in Lebanon, Jordan and Turkey.

Gathering and managing information provided by the FDP that are in European countries.

The triangular research method, as mentioned above, will employ a combination of methodologies, techniques, sources and researchers to implement the data-production and data-analysis process.

The qualitative methodology will focus on knowing the difficulties of integration of the selected VGs (like women, children, and people with disabilities), and their specific relationship with belonging to those targets. It will include, among others, structured and semi structured interviews and focus groups. This information will subsequently allow making recommendations to policy makers, civil society, organisations and media. Based on these results, the stakeholders involved with the project will also start collaborative social inclusion actions.

The project team will analyse this information with the help of techniques from data mining, automated learning and artificial intelligence. This will include, for instance, text processing for sentiment analysis, clustering to identify VGs, and deep learning to detect discourse inconsistencies. The purpose is that researchers can use for their work on the definition of VC and TAIS (Tailored Attention and Inclusion Strategy) already pre-digested data that point out relevant patterns.

The twofold goal of this application is reducing unintended biases in the interpretation of data, and discovering unforeseen patterns.