| Title | Predicting children migrants\' social exclusion risk through an innovative digital tool: Application of machine learning methods to Spanish residential centres |
|---|---|
| Authors | FERNÁNDEZ PACHECO ALISES, GLORIA, AVIGNONE, TATIANA, TORRES JIMÉNEZ, MERCEDES |
| External publication | No |
| Means | Child. Youth Serv. Rev. |
| Scope | Article |
| Nature | Científica |
| JCR Quartile | 1 |
| SJR Quartile | 1 |
| Publication date | 01/08/2025 |
| ISI | 001502148300003 |
| DOI | 10.1016/j.childyouth.2025.108345 |
| Abstract | In the improvement of social care, the adaptation of services and procedures to the analysis of the best predictors of the risk of social exclusion for Unaccompanied Migrant Children plays an important role. This study aims to identify the best predictors of social exclusion among unaccompanied migrant children, to create a digital tool that integrates the complex reality of unaccompanied migrant children in shelters in Spain. Using a logistic regression model, an innovative risk assessment tool was created ad hoc for risk assessment within institutional care centres. A sample of 209 children who had been in residential care in Andalusia in 2021 was selected retrospectively. After applying a feature selection method, it was found that 24 variables are the most predictive within the educational, social, psychological or occupational areas; 8 belonged to risk factors and 16 to protective factors. Specifically, accessing intercultural mediation, having previous criminal records, bringing original documentation from their origin countries, using a sexualized vocabulary, having a good educational level and having economic resources for autonomy had a higher regression coefficient (Beta value), which means they are determinants in promoting better social inclusion. This pioneering study demonstrates that this digital tool offers a promising new method for the efficient screening of concerns for social professionals regarding social exclusion among unaccompanied migrant children in residential centres. One of the most suggestive findings of this study is the high incidence of protective factors in preventing the risk of social exclusion. Consequently, priority should be given to policies supporting protective factors related to social resources. |
| Keywords | Unaccompanied migrant children; Risk and protective factors; Social inclusion; Residential care; Predictive algorithms |
| Universidad Loyola members |