*1.3. Mentoring Programmes for Unaccompanied Youths*

In order to address the needs of immigrant youth, the number of mentoring programmes targeting this group have increased, especially in Europe, in the last five years after the so-called "refugee crisis" [49]. Nevertheless, there is still scarce information about the effects that the programmes specifically targeting migrant adolescents and youths have [50,51]. The meta-analyses, mostly with evidence from US programmes for general youth populations, highlight that youth mentoring interventions have a modest but significant effect on improving diverse outcomes across the behavioural, emotional, social, and academic domains [52,53]. These studies have also shown that mentoring programmes are more effective among mentored youth who have significant levels of environmental risk and among samples with greater proportions of male youths [54]. Besides, in recent years, a growing number of scholars have highlighted that mentoring, as a specific approach, can be more effective when young people are provided with the skills to recruit adults from their own networks instead of assigning participant youths to an unknown caring adult by the mentoring organisation [55–57]. This body of research has shown that some approaches to mentoring (such as youth-initiated mentoring, network engaged mentoring, or intentional mentoring) provide more enduring and emotionally supportive relationships than traditional approaches, because these programmes tend to empower the youths in deepening their existing ties and creating new ones [58]. However, as far as we know, there

is no research showing the effects of mentoring on the mental health of unaccompanied migrant youths. What we know is that mentoring programmes can widen their social networks [59] and improve a sense of belonging and hope in the receiving society among migrant children living with their families in their new context [60].

#### **2. Current Research**

This study examined the effectiveness of the mentoring programme *Referents*, initiated by the *Punt de Referència* Association in 1998 [61]. The main goal of this programme is to support young people, mainly those leaving care, who, without family networks of support, start the transition to adulthood when they turn 18 after having been under the tutelage of the Generalitat de Catalunya (the Catalan government). From 2015 on, almost all participants of the programme are former unaccompanied minors who had been in the minor protection system before turning 18 and volunteer adult mentors.

The mentoring programme looks for adult volunteers who are established in Barcelona and have already completed their transition to adulthood (while also looking for young people interested in participating). After a selection process of adult volunteers and young people, training is carried out with the mentors, where the socio-legal situation of migrant youths leaving care is explained, as well as what their task as mentors involves. Each mentoring relationship (consisting of one adult and one youth) is instructed to meet once a week during a period of six months to carry out an activity. The mentoring programme practitioners suggest starting with leisure activities such as going to museums, activities in local public services, or doing sport activities. The aim of the programme is to create a bond between the mentor and mentee that facilitates significant conversations for the young person (concerns about administrative procedures, emotional discomfort, or doubts that affect their educational and occupational path), or, in other words, the provision of different types of social support.

However, despite monitoring the development of a strong and lasting bond between mentor and mentee, the *Referents* programme differs from models that provide non-specific care to their participants, in which the mentors are encouraged to provide friendship and support in general terms. These models that are less focused on solving the specific problems of the young people consider that a close relationship with the mentor is, by itself, a corrective experience that leads to a wide range of improvements in the young person's development [62]. In the *Referents* programme, however, specific objectives for each relationship are established, thanks to exhaustive training with the mentors on the obstacles the youths need to be accompanied with through their transition to adulthood, and a strong monitoring of the relationship (with regular meetings with the mentor; the mentee; and, on some occasions, with both). It is also characterised as a well-targeted programme and is focused on solving specific problems, since the group to which it provides support is clearly defined and due to the constant coordination with the other agents that intervene in the young person's development in order to specify what the focus of the intervention is.

The specific problems that these young people try to deal with during mentoring are usually related to learning the language, getting to know new places in Barcelona, and/or meeting new people. Here are a few of the responses of young people when we asked them why they signed up for the project:

*My first idea was that I was going to meet with someone who would be older than me, and I thought that was a great idea for me. I was going to ask lots of things about Barcelona, things about Spain, to practise Spanish* . . . *This is what I was thinking* (Amadou, mentee)

*I like that they help me from many sides. Mirela (mentor) has helped me know many places in Barcelona. [* . . . *] Besides that, I have practised Spanish with her many times, and I have improved.* (Abás, mentee)

*As I said before, I felt alone in the centre and didn't know anybody from here, from Barcelona or from Spain. I wanted to meet some kind of friend, I wanted to get to know places, practise Spanish more and everything went well.* (Hassan, mentee)

#### **3. Methods**

This research followed a *sequential explanatory mixed method design* which is characterised by gathering quantitative data and analysis before carrying out the qualitative fieldwork [63]. Mixed methods designs tend to provide a more complete and holistic view of the impact an intervention has rather than solely quantitative or qualitative designs. In this sense, we assessed the effects of participation in the programme on the lives of unaccompanied migrant youths in providing them with informal support in their coming of age and improving their psychological adjustment and expected or desired educational outcomes. More precisely, we were interested in how the unaccompanied youth see themselves (i.e., self-esteem), the psychological distress they experience, their future prospects (youth hope), the resilience skills they develop, how they perceive their educational near futures, and the role that social support they received has on these outcomes. In the survey data, we assessed these psychological and educational outcomes before (Time 1) and after (Time 2) participation in the programme among mentored and non-mentored (control group) youths. After the analysis of the quantitative data, we elaborated the guidelines for the interviews and carried out the qualitative fieldwork with mentored and non-mentored youths, as well as mentors.

#### *3.1. Quantitative Data*

#### 3.1.1. Participants

Survey data were gathered from October 2018 to October 2019, coinciding with the start date of the mentoring matches of the *Referents* programme. In this period of time, the programme began three mentoring groups, with each consisting of between 10 and 15 mentoring pairs (a mentor and mentee), since this is the number set by the programme itself to ensure correct follow-up by the mentoring expert that monitors the development of the group's relationships. Therefore, the initial objective was to survey the maximum number of young migrant mentees, which, in this case, could have been 45. However, the programme did not reach the maximum number of recruits expected in each group and, in addition, were not all migrants (there were also Spanish youths who had left government tutelage without a family support network). Therefore, we aimed to recruit all mentees (*N* = 39); however, those surveyed in Time 1 were 32 youths (seven were discarded because mentoring matches had already started or because the youths refused to participate), and those in the control group (*N* = 26) were interviewed within one to two weeks later. From all these cases, we ended up with 21 youths in the mentoring group and 23 in the control group at Time 2, because we could not trace 12 of the pre-tested youths seven months later, and because the sample had a strong gender imbalance and thus two female participants were dropped from the analyses (initially, 91% of the mentored and 92.3% of the nonmentored youths were male). Our data confirm official statistics showing that 81.2% of unaccompanied minors in 2018 were from Morocco (see descriptive data below), and 97.7% were male [64]. The results, which include the two female participants, were consistent with the results obtained from the exclusively male sample (see Section 4).

Participants' ages ranged from 17 to 23 years (*M* = 18.52, *SD* = 1.50) for the mentoring group and from 17 to 19 (*M* = 18.04, *SD* = 0.37) in the control group. This difference in the age ranges was because the young people in the control group were accessible, since they were in the housing resource mentioned above, which limits their stay until the age of 21, while the mentoring programme does not set an age limit in order for young people transitioning to adulthood to access it. Most of the youths had been residing in Spain for two years at the time of the study (*M* = 2.10, *SD* = 1.00 and *M* = 2.52, *SD* = 2.19 in the mentoring and control groups, respectively). The majority came from Morocco (61.9% and 82.6%, respectively), while some others came from Algeria or Sub-Saharan countries and a

few from Latin America. When asked about their arrival to Spain, 47.6% of the mentored youths and 43.5% of the control group crossed the Mediterranean in a small boat, and 28.7% and 30.3%, respectively, were hidden in trucks. Most of the mentees lived in Barcelona city and a few in the Barcelona metropolitan area in shelter flats (71.4%), flats shared with other young people (14.3%), and some in a residence or in a rented shared apartment (14.3%). Most of participants in the mentoring (85.7%) and the control group (91.3%) had to move in the last year.

#### 3.1.2. Procedure

One of the main challenges of the fieldwork was to adequately select and follow, for more than six months, former unaccompanied minors between 18 and 23 years old. With this aim in mind, we counted on the active support and collaboration of Punt de Referència and the Catalan Federation of Residence Care Organizations (FEPA). Their technical staff contacted the participants, informed them about the purpose of the study, and scheduled appointments for data collection. Informed consent was gathered from all youths. In a few cases—those who were 17 at the moment of pre-assessment (Time 1)—we also asked for consent from their legal tutors (Catalan Government Agency).

All the surveys and interviews were conducted in Spanish. Language was not a barrier with most of the interviewees because the majority had a good knowledge of this language. With regard to their Spanish speaking level, at the beginning of the programme, 48% indicated that it was good, 30% very good, and 14% excellent (four participants mentioned that it was sufficient). Similarly, 41% reported that their understanding of Spanish was good, 32% very good, and 23% excellent (only two participants said that it was sufficient). Participants also reported having good overall reading and writing skills: 34% and 27% reported good, 36% and 34% very good, and 21% and 14% excellent skills, respectively. Only three had sufficient reading skills, five had sufficient writing skills, and one participant mentioned that he had some difficulties in reading and writing in Spanish We have tested a regression model where participation in the programme (yes vs. no), language ability in Spanish (i.e., the average score with the four aspects of language ability), and the interaction between the two variables were introduced as predictors and each outcome variable at T2 as a criterion variable in a separate model. We additionally controlled for the T1 scores in each model. We did not observe statistically significant interactions and thus moderation by language ability.

#### 3.1.3. Measures and Materials

We assessed diverse psychological outcomes, including psychological distress, selfesteem, resilience, and youth hope, at two time points. In addition, we evaluated the mentees' perceptions of their educational aspirations and expectations.

