Abstract: The article deals with drug use by young Roma as members of a marginalized population. It reflects on the use of collages and of metaphor analysis on images produced by participants within an action-research approach. Additionally, various stakeholders were involved in a participatory manner in defining research questions, priorities and strategies. In order to understand young Roma's perceptions on drugs, 23 visual focus groups and 58 life-story interviews were conducted in five European countries. In the focus groups, young people built collages to represent drugs and then explained their meanings. Metaphor analysis was used to analyze the data. With the biographical interviews, histories of consumption from drug users were gathered.

Among the trends identified are the early onset of tobacco use, exposure to consumption of alcohol by adults, underestimation of the consequences of many drugs, addiction to injecting drug use in specific areas. Gender roles are strongly associated with patterns of consumption. Roma young people also express belonging and reclaim positive social status by using certain types of drugs and by sharing consumption patterns with non-minority young people. The research process also illustrates how qualitative research can contribute to selective prevention programs.

Key words: participatory action research; visual methods; focus groups; biographical interviews; metaphor analysis; Roma youth; drug use; addiction

Health disparities derive from social inequalities; it is widely accepted that minority groups and young people living in conditions of vulnerability and exclusion have overall poorer health outcomes (CSDH, 2008) and, more specifically, they experience higher risk of drug use and abuse, with stronger negative effects (EMCDDA, 2008). This is also the case for the Roma, a marginalized ethnic community throughout Europe, with 90% of its people living in households with an income below national poverty lines, experiencing discrimination and stigma (EUROPEAN UNION AGENCY FOR FUNDAMENTAL RIGHTS [FRA] & UNITED NATIONS DEVELOPMENT PROGRAMME [UNDP], 2012). [1]

Research on the health of the Roma population in Europe is scarce, and pursuing it presents the serious drawback of the difficulty in defining the ethnic group. It is consequently very difficult to estimate the actual size of Europe's largest minority population, as the stigma associated to being Roma leads to constant underestimation in national censuses. [2]

In some countries of the European Union (especially in Eastern-European countries), the Roma have acquired the legal status of a national minority, while in other countries (e.g. Italy) their status is not clearly regulated. This group includes both nationals and migrants, some European citizens, while others are citizens of third countries, some legal residents while others are undocumented (mostly because unemployed), or engaged in temporary migration. [3]

For these reasons, Roma groups pose specific challenges for research, but also for intervention. First, researchers face "technical" challenges in choosing appropriate methods and instruments in order to reach the Roma and to produce results which acknowledge the differences within this population. Second, some specific characteristics of the group (prevailing proportion of young people, intergenerational gaps, levels of education lower than the general population and segregation, implying difficult access to the research field) need to be taken into account when designing research (ibid.). Third, researchers face ethical issues as they work with a stigmatized group whose voice is rarely heard. Research may risk reinforcing negative stigma, especially if it interprets findings as ethnic differences, without denouncing the inequality structure that perpetuates these differences. Data on persons' ethnicity is sensitive and, in some countries, gathering it is allowed not only after having gained consent, but also solely for purposes benefiting the target group. For these reasons, researchers need to bear in mind that their process should be open to the participation of the Roma, whose voice should be listened to, and that research should always aim at better social inclusion. [4]

The most comprehensive study up to date regarding the health of the Roma population in seven European countries was carried out under the coordination of FUNDACIÓN SECRETARIADO GITANO (2009) by the means of a door-to-door survey, reaching a sample of 7,604 interviewees. The study describes a predominantly young population, due to the shorter life expectancy and to the smaller decline in birth rates, as compared to the general population. There is thus a relative weight of minors and youth, with an average age of 25,1%, as compared to 40,2 in the European Union. 26,7% of the Roma population is aged between 15 and 29. This data strongly supports the need to dedicate attention to this particular age group and its specific health challenges. [5]

The results of the survey point out the precarious health of the Roma in all participating countries. Along with other tendencies, the study shows high incidence among minors of asthma, allergies, migraines and mental illnesses. In adults, migraines, high blood pressure, high cholesterol, arthritis and heart disease were reported most frequently. Adults visit the physician rarely, while there is more thorough control of children's health. Neglected medical needs were reported by 7,5% of the population, and the most frequent reason quoted for not receiving medical assistance was lack of money and of health insurance (ibid.). [6]

Most scholars attribute the wide health disparities between the Roma and the general population to the conditions of poverty in which most members of this ethnic group live (FÖLDES & COVACI, 2011; KÓSA et al., 2007; MONASTA, 2004). Research carried out in Hungary within the framework of the Roma Health Project, produced data that can be used to partly clarify the relationship between the various factors responsible for the poor health in Roma communities (FÖLDES & COVACI, 2011; VOKÓ et al., 2009). Socioeconomic status proves to be one of the main determinants, but does not fully explain differences in health. Health behaviors and barriers in access to healthcare are other factors. VOKÓ et al. (2009), in particular, investigate the effect of ethnicity as a potential determinant of health outcomes: in Roma communities in Hungary, the odds of respondents reporting very bad or bad health is 2,2 times higher than in the general population, after adjustment for gender and age, while the effect of ethnicity on health status almost disappears after adjustment for income and education. Income seems to be strongly correlated to daily smoking (the lower the income, the higher the incidence of smoking). Still, the levels of income do not account completely for the differences in health behaviors measured in the study. The authors conclude that ethnicity could be an explanatory variable for health behaviors, but they do not consider other potential factors such as segregation, exclusion from prevention programs or barriers in access to health services, nor do they consider indicators of ethnicity other than residence in a Roma settlement (which is per se a biased sample). [7]

IVANOV (2004), analyzing the reasons for the low levels of health among the Roma, focuses on some structural issues that represent barriers in access to health care. First, he examines poverty, meaning that people cannot afford services and medicines, but they are also constantly subject to stress, have poor nutrition and live in precarious conditions, which seriously affects their health situation. The second barrier is represented by health insurance policies, meaning that, in some European countries, in order to have access to public health services one has to comply with different criteria (e.g. be registered as unemployed or be qualified for social benefits). Finally another set of barriers that limit health care access for Roma is represented by segregation within medical facilities and discriminatory treatment carried out by medical personnel. [8]

Of great importance, either for the topic of health more generally but especially for the topic of drug use, is also the residential segregation many Roma experience. The concentration of deprived groups in certain areas, gradually abandoned by middle-class and working people, leads to an increased probability that dwellers will share the same trajectories of social exclusion (WILSON, 1987), with high dependency on internal resources and with probability of deviant behaviors (MANZONI, 2013). [9]

European Monitoring Centre for Drugs and Drug Addiction (EMCDDA, 2002, p.34) underlines that it is not ethnicity in itself that leads to drug use, but its consequences, more specifically, social exclusion:

"black and minority groups are vulnerable to problematic drug use in the same way as the socially excluded, white, indigenous populations are, and this applies particularly to the younger generations of some groups. It was pointed out by informants from most of the countries that participated in this study²⁾ that drug use was more prevalent amongst the younger generations of BME³⁾ groups than amongst the older generations, and many informants linked this with social exclusion." [10]

EMCDDA's (2008) report on drugs and vulnerable groups of young people also underlines that a group's exposure to social disadvantage or inequality may result in limited individual choice and lead to earlier, more frequent or more problematic drug use. [11]

The few studies concerning drug related problems among the members of Europe's largest minority, the Roma, are quite alarming and indicate significant differences as compared with the general population, in incidence and level of use, abuse and addiction. Research on drug consumption in Roma groups in Bulgaria (EMCDDA, 2009), Finland, Spain, Portugal (EMCDDA, 2002), Hungary (GEREVICH, BACSKAI, CZOBOR & SZABO, 2010), and Ireland (EMCDDA, 2008) identifies tendencies such as an early age at onset of tobacco and alcohol consumption, higher lifetime prevalence for all types of drugs, stigma and concealment of consumption. [12]

FUNDACIÓN SECRETARIADO GITANO's (2009) research on the health of the Roma also reported some aspects of drug use and drug-related problems. It showed that self-medication is common for blood pressure, pain or fever, cold or flu, birth control and with antibiotics. Tranquillizers, relaxants and sleeping pills are the seventh most frequently taken medicines without medical prescription, from the 19 types of medicines included in the questionnaire. These are used by 11,8% of the adult Roma population in the sample, percentage referring to the consumption in the two weeks before the interview. [13]

According to the survey, 44,2% of the sample were daily smokers, with men generally smoking more (59%) and women less (31%). A high percentage of smokers are in the 16 to 29 age group (45%) while the countries with the most smokers are Greece, Portugal and the Czech Republic. Smoking on a daily basis begins on average at 15,5 years (ibid.). [14]

As for alcohol, large numbers of consumers are reported among men (73,1% of Roma men drank within the year before the survey, as compared to 39,6% of the women). Alcohol consumption begins at an older age than smoking (the average is 17 years), but the recent trend shows a diminishing age of onset. Problematic consumption of alcohol was also measured, by means of one family member reporting for the entire family: 3,5% of the population and 11% of the households acknowledge having problems with alcohol or other drugs, especially in Bulgaria, Greece and the Czech Republic. These tended to be larger families than the ones without any drug consumers, to live in substandard housing or in shanty towns, to live in areas lacking health and social services, and to have poorer health (ibid.). At a European level it is acknowledged that one of the characteristics of the segregation experienced by Roma groups is related to their low access to national welfare services. This is one more reason to conclude that urgent efforts are needed to improve access and positive outcomes of health care among the Roma. [15]

The present research adds knowledge to that produced by FUNDACIÓN SECRETARIADO GITANO's study, in particular regarding the profiles and patterns of drug use, focusing on youth and extending the area of interest to substances other than alcohol and tobacco. [16]

This intention is consistent with the selective prevention approach, which underlines that only a small minority of young people experimenting with drugs develop problematic drug use, and this is linked to vulnerability factors. According to EMCDDA (2004, p.2) "selective prevention focuses on these vulnerable groups and settings because they are often not reached by the values, messages and contents of universal prevention (i.e. prevention targeting the entire juvenile population)." In Europe, some of the groups are young offenders, minority ethnic groups, school drop-outs or those at risk of early school leaving, experimenting young people and families at risk. [17]

In the particular case of minority ethnic groups, it is interesting to explore how, for young people, specific consumption patterns contribute to expressions of belonging, either to the ethnic group or to hybrid, globalized youth cultures. For young Roma, especially for Eastern-Europeans, displaying status goods and Western well-being symbols means affirming valued identities and escaping the stigma associated to belonging to a marginalized ethnic group (CONTE, MARCU & RAMPINI, 2009). [18]

PORTES and ZHOU (1993) analyze the processes of shaping identities in the second generation of immigrants in the US. They explain that the assimilation (back then seen as a positive outcome) of the second generation youth is a segmented process, following different patterns determined by various vulnerability factors (such as race, housing and segregation and occupational mobility) as well as resources (government programs, less negative prejudice for some groups, and ethnic communities' networks). Factors and resources can combine in specific situations to determine the exit of what was then called "incorporation" (p.74). Contrary to the general opinion, in some cases, remaining attached to their ethnic communities and not following the native-born ways for attaining social and economic success can actually lead to better chances of social mobility. [19]

In the same vein, NAGASAWA, QIAN and WONG (2001) show that for the second generation in the US, the greater the assimilation (measured as the length of residence in the country and the use of English in the home), the greater the likelihood of marijuana use among youth. Their findings support the classic theory of assimilation, although many differences between subgroups emerged, most importantly related to the degree to which community values are maintained. Marijuana use is viewed as normative in teen drug culture, while minority group members are encouraged to mimic antisocial behavior in ways that are normative and adaptive, leading to what was described as assimilation to the underclass (ibid., see also PORTES & ZHOU, 1993). [20]

This implies that drug consumption, like other forms of consumption, has a role in expressing belonging and distance, in shaping ethnic, gender and class identities. In this article I explore these processes with young Roma in five contexts, while explaining how local factors (segregation of communities, availability of services, young people's networks) influence consumption. In order to do so, it is necessary to understand how young people think of drugs and what images they associate with consumption and its consequences. [21]

In the following sections I will detail the methods used for conducting the study with Roma youth and other stakeholders, explaining the methodological approaches and debates in social sciences that inspired mine. I will describe the way in which the research group co-designed the process with stakeholders, who participated and how data were collected and analyzed. Afterwards, some selected results are presented. [22]

The approach used in this research aimed to produce knowledge in a participatory action-research (PAR) perspective, collecting data in a practice-oriented manner, and actively involving all the stakeholders in the research process. [23]

Research was embedded in a wider project, Addiction Prevention Within Roma and Sinti Communities (SRAP), financed by the Executive Agency for Health and Consumers of the European Union, which intended to innovate selective prevention strategies dedicated to Roma youth in six European countries. The partners in the project were local bodies (municipalities) and nongovernmental organizations, which involved professionals in the fields of addiction prevention, and of social work with the Roma to conduct fieldwork locally. Here I present the results of the research carried out in Italy, Spain, France, Slovenia and Bulgaria. [24]

One of the innovative aspects of this research consisted of already involving stakeholders in the definition phase of the research design: the objectives and targets were established according to their perceived priorities. This choice was motivated by the intention to build knowledge through collaborative processes while avoiding the imposition of institutional, theoretical or academic priorities to the research field: necessary preconditions for successful PAR, as the core aim of PAR is giving voice to marginalized groups (BERGOLD & THOMAS, 2012). [25]

Building the research process together with the target groups also helped us understand how to approach the drug "issue" in a non-stigmatizing manner, working with parents' and young people's existing interest and motivation to tackle this problem in their communities. It was done by carrying out seven initial meetings with various stakeholders (youth, professionals and decision-makers in social and health services), focused on prioritizing problems and gathering suggestions for the field data collection phase. I will detail the participants and the organization of these meetings in the following sections. [26]

As BERGOLD and THOMAS (2012) state, participation of stakeholders in the phase of research design is an important part of the process orientation approach. PAR approach also underlines the need to pay attention that the methods are appropriate to the subjects of the study. They suggest that a ʺparticipatory design processʺ is ʺan attractive and fruitful knowledge-generating option when it comes to researching the social worldʺ (§3). [27]

The results of these meetings contributed to the development of a conceptual framework that guided the construction of tools for data collection and analysis, carried out by outreach social workers and Roma mediators. The process was coordinated by a central team formed of myself and two colleagues from my research agency. [28]

Surely the voices of social workers and mediators who had a more constant role in all stages were stronger than those of the young people. In order to partly balance these perspectives, in the training we emphasized the ways in which qualitative research methods are based on listening and bringing out the points of view of young people themselves. [29]

When conducting research with a particular ethnic group, the criteria for including participants in the sample is always a matter of debate, as the boundaries of ethnicity are often blurred. There are various approaches to operationally define ethnic belonging: people's own declaration, the place where they live (in a Roma neighborhood) or their participation in and identification with a specific culture (with a lifestyle, for example). The approach to ethnicity we settled upon was that Roma refers to a galaxy of minority groups (DELL'AGNESE & VITALE, 2007), whose identities are socially and contextually constructed. This definition requires that each group should be carefully selected and described, most importantly avoiding generalization to the entire Roma population. This approach was consistent with the goals of the SRAP project: the selective prevention approach underlines that intervention has to be specifically tailored to the needs of vulnerable groups. [30]

Three types of stakeholders were consulted in seven initial meetings carried out in all five countries: Roma youth (74 participants), decision makers in institutions or NGOs (41 participants) and professionals in the field of addiction and social work with Roma (62 participants), with the specific goal of gathering indications for the research design regarding targets, contexts, recruitment and ethical issues. These participants were recruited by a process of convenience sampling (GOBO, 2013), mainly those who were already in contact with outreach workers, services and NGOs. Young people's parents were also consulted by researchers, who explained the project and asked for written consent regarding their sons' and daughters' participation. [31]

For the following stages of the research, the research team applied purposive sampling (SILVERMAN, 2005), taking into account three criteria: experience with drugs, age, and living context. On the one hand, we wanted to explore the onset of consumption at a young age, on the other hand, the history of consumption for those who were already familiar with drugs. During young age and adolescence in particular, behaviors and world views change at a very fast pace for young Roma (DANIELE, 2013). This is why we chose to divide the young population into three age groups: 11-13, 14-16 and 17-24. [32]

Wide, unsystematic differences were clear from a first analysis of the literature between local groups, both in the structural conditions (access to services, housing, educational levels, legal status, family arrangements, cultural beliefs) and in drug-related behaviors and attitudes. This limited the possibility of achieving analogue samples in all the countries. The research team carefully explored and selected the contexts where research was to be carried out, taking into account the typical situations that characterized the condition of the Roma groups in each country:

Participants were first recruited by social workers and/or Roma mediators from NGOs or institutions (public, local social services) that, in most cases, knew them and their families professionally. Subsequently, snowball sampling (GOBO, 2013) was applied to reach other participants. The existing relationship between outreach workers and participants was an important resource, given the sensitivity of the issue of drug use, often accompanied by concealment and stigma. [34]

The research team conducted 7 initial meetings, 23 visual focus groups and 58 life story interviews. Outreach social workers and cultural Roma mediators were trained in research methodology and in the use of the specific tools designed for this study by the coordination team I was guiding. My team developed the instruments and the strategy, drawing from the participatory design process. We conducted shadowing and on-site training of field researchers, guided constant exchange between field researchers, periodically summarized results and observations and gave feedback. Moreover, in the initial phase, a pilot study was conducted in order to test the methodology, with one focus group and two life-story interviews in each country. [35]

The procedure of the seven initial meetings, where participants were aware that the goal was the participatory design process (as described above in Section 2.1), consisted of a brainstorming task on the topic "Roma youth and drugs," followed by the instigation of a conceptual map. The main categories that were formed were then ranked by participants on the basis of their importance and of their accessibility for research and prevention actions. They then prioritized the problems, ordered the resources, identified the critical issues and warned about potential errors to be avoided in the research process. [36]

The visual focus groups were conducted by the research team using the technique of the collage (AWAN, 2007; REAVEY, 2011), a less common method the utility of which needs to be tested and discussed. The collage technique involved participants in the process of pasting various materials on a surface, such as newspaper clippings or parts of photographs, around a common theme, here represented by alcohol, tobacco and other drugs. Researchers interacted with participants during the execution of the task, eliciting meanings involved in the choice and placement of the images. Following the individual production, presentation and elicitation of the collages, the researcher guided a group discussion about the shared themes, in order to reach an understanding of the issues with input from all participants, interacting with one another. [37]

The method is effective in research with young people as editing visual material requires active personal and physical involvement, allowing for differences and nuances of one's own point of view to be expressed and discussed during the completion of the task itself (AWAN, 2007; MacGREGOR & MORRISON, 1995), while also having more time to reflect and express themselves, and make implicit knowledge explicit (GAUNTLETT, 2007). On the topic of drugs, it is important to access meanings that are difficult to verbalize, because of the strong social norms associated with their consumption, to allow for emotional content to be expressed and illustrated in the image making-process (GRADY, 2008) and for different, sometimes contrasting meanings to be expressed simultaneously (LANGER, 1957). The collage technique has previously been used for educational purposes and to explore attitudes, gender identity and self-image (AWAN, 2007; REAVEY, 2011). [38]

The elicitation process, usually conducted with photos, here with collages, helps researchers gain more insight into participants' points of view, by relying on the verbalization of intended meanings. This process, as HARPER (2000) suggests, can be seen as an insight into cultural explanations, a "cultural Rorschach test" (p.725). [39]

A large and diverse selection of newspapers, magazines and other sources was made available to participants, as the material might limit the final result. It is common to use this print media (such as magazines and newspapers) for the collage task, but it is important to acknowledge it can introduce bias in the direction of socially accepted or commercially promoted images and discourses. [40]

As concerns minority groups, media communication has often been found to misrepresent and to stereotype negatively; or the lack of depiction of these groups in the media constructs by omission their social identity as marginal. There haven't been many studies regarding the representation of the Roma in the media, although a few examples pointed out negative, stereotypical discourse; also notable is the lack of representation and absence of the voices of Roma and Sinti: they are pictured by the voices of others (SIGONA, 2006; WARINGO, 2005). [41]

Wide evidence showed that media representations stereotype gender roles, bodies and sexual relationships (KYRÖLÄ, 2014; WARD & HARRISON, 2005; WYKES & GUNTER, 2005). In particular, feminist readings of US media images underlined how women of color are marginalized in magazine advertisements, being depicted as to enforce negative racial stereotypes (SANCHEZ-HUCLES, HUDGINS & GAMBLE, 2005). Other scholars have also shown that ethnicity can moderate the impact of the exposure to advertisement—slim—body images, meaning that women belonging to an ethnic minority identify less with non-minority models (DeBRAGANZA & HOUSENBLAS, 2010), and that media images are critically analyzed by consumers for ethical judgments which sometimes lead to rejection (ANDERSEN & JASPER PAAS, 2012). Except for some isolated experiences (see, for example, CONTE et al., 2009), little is known about Roma youth's media and cultural consumption. [42]

It is thus essential, in the collage task, to use various kinds of visual material, in order to facilitate identification and to support the construction of images that talk about the self. The lack of representation in print of minority people and non-normative bodies is a real challenge for this research method. [43]

We used mostly mainstream printed media, but integrated it with other, more sensitive publications, for example, informative material produced by non-governmental organizations or minority media productions, where they existed. Moreover we attempted to diversify the categories of publications, using magazines and newspapers on fashion and celebrities but also on wildlife, nature, science and social issues. [44]

The 23 visual focus groups included 199 participants: 98 participants were aged 11-13, while 101 participants were 14-16 years old. The average number of participants in each focus group was 9. [45]

The life-story interviews were conducted using common guidelines between all interviewers and a non-directive style of questioning. Following the typology of interviews proposed by BICHI (2007), they were semi-structured, meaning that the questions were only partly defined by the researcher and were reformulated to research participants to better follow their line of discussion. Although the guidelines comprised a detailed list of topics to be addressed in the interviews, some were not touched while new ones were also introduced. The interviewing style was non-directive, meaning that the researcher tended to follow participant's discourse, assign priority to what he/she raised as most relevant, using fewer open-ended questions and more probes than in common in-depth interviewing (ibid.). Non-directive interviewing was particularly useful given the sensitive and stigmatized topic of the research, as it put less pressure on the participant to answer than with directive questions (even if open ended). [46]

The interview started with an open question ("can you tell me about your experience with tobacco, alcohol or other drugs?"), eliciting a first account in which the interviewee chose freely which topics to touch upon and how much detail to give. Then the interviewers used probes and open-ended questions in order to detail and complete the account, paying attention to the most relevant topics, as organized in the common guidelines. These were constructed following the conceptual framework of the study developed in the preparatory phase. [47]

Some of the difficulties in involving interviewees were silence and fear of disclosure regarding any kind of activity connected to drugs, or frequent dropouts from scheduled activities. For similar reasons it proved very difficult, if not impossible, to interview young people involved in drug dealing, therefore this remained an area that our research could not explore. Girls proved to be more reluctant to speak about their own consumption, and all perceived that fewer girls used drugs than boys. [48]

Interviews were conducted with the older youths (17-24 years old), who were purposely selected as consumers of different drugs, 37 male and 21 female. In communities where multiple patterns of consumption were present, the panel for the interviews was composed of consumers of different drugs (i.e. 2-4 persons using tobacco and/or alcohol, 1-2 persons using cannabis, 2-4 persons using other drugs). [49]

The interpretation of the data put participants' motivations, interests and perspectives, subjectively raised during the interviews, at the center of the explanation. Special attention was given to meaning-making processes, and thus to what was perceived as important or decisive for consumption by the Roma young people themselves. [50]

The results of the initial meetings with stakeholders were analyzed cross-nationally in a meeting between field researchers by building a conceptual map. Its concepts collated into five main categories. As specified above, these formed the conceptual framework that guided the entire study from the formulation of research questions, to the design of the instruments and finally the analysis. [51]

With the collage, metaphorical analysis was applied, while for the life-story interviews and the more discursive parts of the focus groups we used thematic analysis. [52]

The use of metaphorical analysis in qualitative research has been explored by SCHMITT (2005), drawing on the seminal work of LAKOFF and JOHNSON (1980), who argued that it can reconstruct social thought, language and action when passing from individual metaphorical models to collective ones. In the present research, we used the steps proposed by GAUNTLETT (2007). We listed the signifier and the signified for each metaphor, successively identifying recurrences, similarities, contrasts or other relationships between metaphors, then mapping source and target domains in order to access broader areas of meaning. In later stages, it was possible to link metaphors, by means of interpretation, to categorical knowledge and to the conceptual framework of the study. [53]

Images are polysemic (GRADY, 2008); researchers and participants might have different understandings of the elements and of the collage in its entirety. In this research, analysis has drawn both from the presentation that each participant verbally made of their own collage and from the interpretation of the image itself by the researchers. The first source of analysis comprised participants' own words and meanings regarding the images they selected and the overall product, which were privileged as compared to researchers' interpretations, following GRADY's suggestion. Some participants wrote texts, ranging from just a few words to page-long stories, next to images, which were also used as an input for metaphorical analysis. Then each of them explained what he/she intended. Afterwards, during the group interview, researchers checked their understanding with the participants. [54]

The second source of interpretation was the image composition, disposition, closeness, relative dimensions of the components, types of elements (persons, objects and their characteristics), following the argument made by BOHNSACK (2008) regarding the usefulness of analyzing the formal structure of images, a method parallel to discourse analysis for texts. In this manner it is possible to reconstruct pre-iconographic knowledge that is difficult for participants to conceptualize or verbalize. Thus, metaphorical analysis brought together the two different approaches to the analysis of images: the formal one and the symbolic one. [55]

Thematic analysis was conducted on the transcripts of the recorded interviews (BICHI, 2007). Local researchers assigned codes while re-reading the transcripts. These were either drawn from the conceptual framework built together with stakeholders in the first phase, or emerging from the text itself, as new, unforeseen topics invoked by interviewees. [56]

Themes and metaphors were then progressively categorized in order to reconstruct wider meanings which converged into the five areas of the conceptual framework developed in the participatory planning phase. The researchers therefore made sure that the priorities set by the stakeholders at the beginning of the research were followed throughout the research process. [57]

Within each area of meaning, various levels of comparison were applied, based on the interpretation of the texts and images. The context of the production of the images was taken into consideration by analyzing the commonalities and the differences between the results from each country, taking into account not only drug-related behaviors and perceptions, but also the more general social and economic situation of each local group. [58]

The results obtained from the two methods were complementary: those from the focus groups referred to younger people who had little experience with drugs, while the ones from the interviews referred to personal histories of consumption by young adults. [59]

In conclusion, the methodology in this research was innovative because it worked on a sensitive topic with a marginalized group, using a participatory design process; it tested the use of the collage and differentiated the instruments between age groups in order to look at different aspects of consumption; it used metaphorical analysis, still under-utilized in visual research. [60]

In the following section, I will present the results of the study, reflecting on the efficacy of the methodological strategy in facing the challenges derived from the target and the topic of this research. [61]

The following results offer a description of the social mechanisms by which drug consumption is understood, dealt with and maintained in the communities, and exemplified by patterns of behavior as described by young people themselves. They are not by any means statistically representative regarding substance consumption incidence within the same communities. Although the main goal was descriptive, in some instances the data allows for an interpretation of the differences and similarities between contexts in order to understand how young people make sense, contextually, of their social situation, and how this influences their substance use. The topics that emerged from the analysis are grouped in this section in three main thematic areas. The first refers to the patterns of consumption: the onset, evolution and circumstances of consumption and the types of drugs that are used. The second section describes the representations of substances that young people described. The third positions drug use in the context of young people's social networks: describing their role in the dynamics of consumption, addiction and cure. [62]

Most of the information regarding the actual consumption comes from the interviews with consumers, aged 17-24. This was due to the fact that younger participants didn't have experience with most drugs and the focus group was not a good context to elicit such personal information. [63]

Different types of drugs were known and were present in the circles that young people frequented. In the present section we will refer to the onset, frequency and contexts of consumption. One first result is that the patterns of consumption varied between different types of substances and in the different local contexts. [64]

The first pattern we identified in all local contexts refers to tobacco consumption, with a very early onset (11-12 years for boys, a few years later for girls). Tobacco use was initiated within the peer group, some young people acquire tobacco with older peers or get it from their own family. Young people recall that the families had ambivalent reactions to their tobacco use: they did not agree, but they did little to stop it. For the most part, tobacco remains the only addictive substance used, unless it is associated with occasional alcohol use. [65]

Consumption of alcohol, by boys and girls, without recourse to other drugs was described by participants in social occasions and gatherings, where most people drink, youth and adults alike, from a couple of drinks to binge-drinking. While in most contexts, the age of the first drink was13-14 years (for boys), in others, young people claimed they started drinking already from 9-10 years old (Italy), or at 11 (Bulgaria). In Spain, alcohol consumption increased during the weekends and shows a widespread pattern among young people in general. Some differences between genders referred mostly to the later onset of drinking for girls and the social pressure against girls drinking, leading to hidden consumption but also more control from the family. [66]

Cannabis was the next best known drug after alcohol and tobacco and its use was widespread, according to participants. At times young people did not identify it as a drug at all and minimized its effects and consequences. This widespread perception has been described as the normalization of cannabis consumption (PARKER, WILLIAMS & ALDRIDGE, 2007). [67]

In some cases, cannabis is used with friends outside of the Roma community (as happens often with participants in Italy) and it is an occasion to socialize with other peers. Cannabis was consumed either in groups, or individually and was described as a way to deal with anxiety, to calm down, to relax or to sleep. Generally, cannabis was considered to be harmless and participants using it didn't perceive themselves as addicted, except a few cases. In Bulgaria, for example, several participants reported using it more than three times a day, but only one female user identified herself as an addict, and has unsuccessfully tried stopping. [68]

Heroin consumption emerged as a problematic pattern in Bulgaria, especially in the closed Roma neighborhood where fieldwork was carried out (a needle exchange program is based in this large Roma neighborhood in Sofia), but was also reported in Italy and in Spain. Social workers were already aware of this, as many of these participants had sought help or used harm reduction programs. [69]

Many users in Bulgaria started using heroin before any other drug, when they were 14 or 15, and prefer it to other drugs, although sometimes they mix it with amphetamines. Italian users, instead, describe having used first marijuana and subsequently heroin, without trying or being interested in other drugs. Heroin consumption started by smoking, but with the onset of addiction, lack of money and tolerance, intravenous use became common. The frequency for those who used it was 2-3 times a day, on a continuous basis, until they ran out of money. The lack of the substance was sometimes compensated for with the use of medicines of similar effect (methadone, diazepam). Many participants in Bulgaria underlined the wide availability of heroin in their own neighborhood. [70]

Heroin users we interviewed were painfully aware of their addiction, had tried to stop, but only for short periods, and thought that the social problems affecting their communities (such as the difficulty of finding a job) were also barriers for their rehabilitation. [71]

Recreational use of other drugs was also familiar to Roma young people in our study. Cocaine use was reported in Spain and in Italy; amphetamines were mentioned in Bulgaria, benzodiazepines in Bulgaria and Spain, ecstasy and hallucinogenic mushrooms in Spain. Most of these were used at parties, with groups of friends, and most often in combination. Generally, the consequences of these substances were underestimated, but they rarely led to addiction, as claimed by participants. [72]

Poly-drug consumption was common among all consumers interviewed, especially in party settings, in combinations such as: tobacco with occasional alcohol consumption, tobacco and cannabis with alcohol, benzodiazepines with alcohol, or amphetamines with alcohol and cannabis. [73]

Except for heroin users, consumers did not perceive themselves as addicted and were not aware of the consequences. Although drugs other than heroin were widespread in all contexts, they were not tackled by services. [74]

Another salient, universal finding was that the use of most of these drugs, especially as regards the recreational use during weekends and at parties, was not exclusively within Roma environments and groups, but it was shared with non-Roma peers and widely present in other social groups. [75]

What makes Roma groups vulnerable from this point of view is the situation of unemployment, low levels of education and social exclusion, which dramatically influences consumption, addiction and the likelihood of finding professional help, especially in segregated social contexts. Young people also confirmed this link, underlining that they felt drug use helped them cope with anxiety about the problems they face in other spheres of their life, such as unemployment, discrimination or housing insecurity. [76]

Another aspect related to marginalized groups and segregated environments regards the closeness of illegal activities, leading often to easy access to drugs, as was reported in some of the places where we carried out the research. [77]

The collages and their elicitation were a particularly rich source for reconstructing the representations Roma young people reported regarding drugs. Generally, a distinction was made by young people between "drugs," which meant illicit substances, and alcohol and tobacco, seen as lighter or not drugs at all. This distinction was also somewhat induced by the task which asked them to represent their relationship with substances in three columns: one dedicated to alcohol, one to tobacco and the third to other drugs, following the so-called ATOD⁴⁾ approach. Nonetheless, when selecting images for the collage, many participants included cannabis in the category of lighter drugs by positioning it close to tobacco, not to "other drugs." The borders of the categories were stretched and bent to fit the perceptions that young people had. They sometimes even included other addictions in the collage, such as computers, gambling or even chocolate, in order to underline the psychological similarities in the feeling of need and craving, but also to point out that some addictions are not accompanied by such risk as drugs. The main types of images used to depict drugs, which will be detailed below, referred to their consequences and in particular to their licentious implications—often associated with gender norms. Drugs were also linked to glamorous life-styles or seen as culturally distant and exotic. [78]

Many drugs were depicted by illustrating health, legal or social consequences, but also by using sensorial aspects, like the pungent smell or pale skin. Old age, wrinkles, skinny or dead bodies indicated the physical conditions that young people associated with drug use. Regarding the implications of consumption, participants affirmed unspecific, almost anecdotal knowledge: tobacco provoked cancer, other drugs lead to madness, and alcohol increased vulnerability, while marijuana was considered harmless. [79]

Positive effects of drugs, such as relaxation, energy or "coolness," were sometimes quoted (by young people in Italy, France), often in a hyperbolic manner. Following expectations of gender roles, girls tended to minimize their own consumption, while boys boasted about it. [80]

In Italy, for example, younger participants saw the use of alcohol as becoming crazy, dancing, jumping while the use of other drugs as making people sad, shaky or almost ill. They relied largely on the observation of the consumption by adults they knew, as it happens often that young people are present when adults in their family drink, but also had information from media and from universal prevention campaigns, thus most of them had a far-removed, approximate image of drugs, except for alcohol.

In this collage, alcohol is represented by an image of alcoholic beverages, a party and a picture concerning the risk of alcohol consumption in relationship with unwanted pregnancies. Written phrases explain the images, referring to the loss of awareness by people at parties and clubs, who abandon themselves, "people are not aware of what they do" and "alcohol is a major cause of undesired pregnancy." The association between drugs and new risk factors is repeated with the tobacco and other drugs, adding the topic of drug use in sport: doping "helps athletes win a lot easier." [82]

In Spain, Bulgaria, France and Italy, strong differences between gender norms were evoked by participants as characteristic of their group's culture. In the visual task, images used to represent drugs (especially alcohol and illegal substances) were very often associated to the sexual promiscuity of women. Examples of girls naked, dressed in sexy outfits or holding their legs wide open, represented for collage authors the lack of inhibition associated with alcohol and consumption of illegal drugs. Furthermore, in order to represent illegal drugs, two participants in Bulgaria used photos of women they identified as sex workers, another metaphor connecting drug consumption to sexual behavior. No sexualizing images of men were used by participants, thus indicating that this issue is sensitive for girls, who constantly need to monitor their behavior in order to respond to stricter gender norms than boys. [83]

Sexy women, partying people, models and catwalks conveyed a more glamorous image of drug use. The use of different types of drugs was associated with social class, as it clearly emerged from the research carried out in Bulgaria, in the segregated neighborhood where heroin use was present. Participants described that some drugs are used mostly by rich people while others by poor people. They related marijuana, amphetamines and cocaine to a life of entertainment and luxury, as publicized in the lives of many celebrities, while heroin was linked to social degradation, illness and suicide. These representations might contribute to reinforce desirability of "trendy" drugs but also to add weight to the stigma for those using "poor people's drugs." Participants in Bulgaria observed that it is fashionable among young people to use drugs, and those who refuse to do so risk being labeled stupid or weak. [84]

Another type of representation of drug use was constructed by distancing oneself and one's group from the image of drug users, who are depicted as different, unusual, in extravagant postures or outfits. Sometimes racial difference is used for the same goal, as drugs are connected to black people or immigrants; this latter metaphor was mostly present in focus groups in Italy: women with burqa were associated by one participant in Bologna to cannabis, because the substance comes from Muslim countries, while another participant explained that drugs are characteristic of the non-Roma, called gadji. [85]

Attributing drug use, the object of negative prejudice, to other groups, is a way to reclaim positive group identities for one's own group of belonging, and to escape the stigma of being stereotyped as drug users. It is interesting to notice that these appeared mostly in segregated contexts, where the pressure for reclaiming positive group identities and cohesion is higher. This is the case in Italy, for example, where Roma and Sinti live in especially designed "camps" often far from the center of the cities and restricted exclusively to people of the same ethnicity. [86]

This section refers to the social environment of young Roma, their significant others, family and friends, and the influence of these relationships on consumption. It also refers to other social support sources such as health and social services. These topics emerged both in focus groups and interviews, where we included questions and prompts regarding community, family, couple relationships and friendships. [87]

As discussed above with the research of PORTES and ZHOU (1993) and NAGASAWA et al. (2001), community norms can have a protective effect and prevent young people from initiating or maintaining drug use. This was particularly clear in Bulgaria and Spain, where religious beliefs and norms in groups who identify as Evangelical were protective of young people's health. [88]

Many other situations were also evoked. The closure of some segregated communities makes it sometimes difficult to tackle the topic of drugs, as there is a tendency to avoid these topics with people outside the communities, in order to protect the group. This reaction constitutes a barrier in the access to health services for drug-related problems. [89]

Other young Roma live separated from their communities. In the case of those in situations of vagrancy studied in France, there is a wide distance between them, their street-based activity and the community they come from. These were drug users who can count neither on parental control nor on the social support of the extended family in case of health problems. These groups of young Roma are most at risk. [90]

When referring to the Roma, it is common to interpret behaviors as permeated or dictated by specific cultural attitudes, beliefs or norms. On the contrary, it is important to underline that young Roma people share the culture and identity with many other groups: with people living in the same country, city or neighborhood, with peers of various nationalities in school or with various youth cultures, known from the media. I will detail the role of these other group cultures in the section regarding peers. [91]

Many of the participants pointed out that adults in their communities have a negative perception of drugs, especially the illegal ones, although they are relatively permissive with alcohol and tobacco consumption. [92]

Often children are exposed to adults' consumption; sometimes family members who use drugs become role models (i.e. in Italy and Spain). In France, when prompted regarding alcohol, a participant related a story of domestic violence by an intoxicated parent. Childhood experiences of intoxicated adults also convinced some to resist the use of drugs. [93]

In Slovenia, Spain, Bulgaria, and France, participants observed that smoking at a young age is a general habit in their Roma reference group, and that alcohol consumption by boys begins very early, in the context of community celebrations. Although participants thought this was a cultural trait, it could actually be due to social exclusion, and more specifically, to the lack of prevention skills and knowledge. [94]

Roma youth in most contexts stated that they would never show themselves using illegal drugs in front of significant adults and family members, as it would be disrespectful and that they generally conceal their consumption. This attitude is stronger for girls, who wouldn't smoke or drink alcohol either. In contrast, young people display consumption with less reticence in the presence of adults who are not Roma. [95]

Little or no communication with parents about drugs is reported in all contexts; young people feel ashamed of talking about drug consumption with their family. The father is seen as a more authoritarian figure, while the mother sometimes offers help or warns children regarding the risks of consumption. Nonetheless, young people do not recognize their parents as sources of accurate knowledge regarding drugs. [96]

Parents' reactions to children's consumption would generally be negative, according to the young people. The most common reaction they imagine or experience when parents find out about their consumption is violence and rage (as was reported in Spain, Bulgaria, Italy). In other cases, parents react by expelling the consumer from the house, or forcing him or her into withdrawal. In the case of tobacco consumption, reactions are softer, and include nagging or verbally forbidding at the beginning, means of control which often prove ineffective. According to young people, few parents know that their children smoke, while with illegal drugs complete concealment was the most frequent strategy. Hiding consumption of unaccepted drugs is connected to the fear of seeking social support from the family in case of drug-related problems, despite the fact that relying on extended family and the strength of familial bonds are still considered strong cultural traits. [97]

The fear of adult persons' reactions to drug use constitutes both a risk and a protection factor for youth. On the one hand, it is a risk factor as it leads to concealment and lowers the chance of seeking familial support for dealing with drug problems. On the other hand, the control and limitations that adults exert can be somewhat effective in preventing young people from using them. Nonetheless, as I described above, parental control strategies are often ineffective, if not backed up with other preventive strategies. [98]

Siblings in most cases do not talk about drugs with their brothers and sisters, although sometimes they play a protective role (trying to prevent the younger from use), and generally oppose consumption. [99]

An important role was attributed to the peer group, which is the privileged group of consumption in all the contexts, where initiation and continuation of the habit takes place. Sometimes consumption starts with peers in school, with only Roma or with non-Roma peers. This applies to both younger and older children. Peer groups are also formed of cousins and other members of the extended family, and are the main environment where young people talk about consumption and share their knowledge regarding drugs and existing services for drug-related problems. Given the social stigma associated with their consumption, girls tend to consume only with female friends. [100]

Leisure activities were explicitly linked to consumption; this was the case in Bulgaria, but also in Spain, especially going out on weekends. Those who are excluded from work and education settings experience an extension of spare time which is usually spent in the company of peers, "hanging around." [101]

Group norms and values, shared with non-Roma peers, highly influence young Roma's consumption decisions. Two trends, frequently reported in the general young population, were identified in the interviews and focus groups: one is the normalization of cannabis use while the other is binge drinking at parties during weekends. [102]

The tendency to share patterns of drug use with non-Roma counterparts is stronger in non-segregated neighborhoods. In Spain, this tendency was described more clearly than in other countries for cannabis, recreational drug and alcohol use during weekends, while in Italy it was mostly described for cannabis. [103]

This is consistent with the theories of assimilation described above and stresses how consumption, in this case of drugs, is a process by which young people build valued identities and express group belonging. In previous works I have described how young Roma's clothing style is contextually adapted to express belonging: sometimes to the ethnic group, other times to a globalized youth culture (MARCU, 2014). As I described above, while drugs can be associated with ethnic belonging or described as culturally distant; they have a role in contextually asserting one's belonging to a minority or to a majority group. [104]

Gender norms among the Roma become stricter for married women, and marriages between young people are more frequent than in the general population, although the age of marriage is slowly increasing (ERRC, 2011; MARCU, 2014; TESĂR, 2012). In Spain it was reported that the consumption by the wife, even of tobacco only, is usually concealed. Two interviewed women recalled husbands' reactions of violence when they found out about wives' drug use. In Spain it seems women could expect a reaction of violence both from the family of origin, for younger girls, and from their husbands, for the older ones. [105]

In Bulgaria, married or engaged women used drugs with groups of female friends, friendships not shared with the husband. One participant said that the reaction of her husband when he found out that she was smoking marijuana was indifference, and even if he also used it, they did not smoke together nor did they talk about it. [106]

In the research carried out in Bulgaria, all male participants were married or engaged, except for one who was divorced. For male participants who used drugs, all their wives found it out by accident, and opposed consumption, at first trying to help, then using other means, such as threats of separation. [107]

There is generally a low access to health and drug use prevention services by Roma, and there are barriers to service use, from the point of view of Roma youngsters. [108]

All participants in our study had already had some kind of contact with public or private social services, at least with outreach workers, as the recruiting relied on NGOs. Given that this excluded those who have never had any contact with services, our group might be characterized by a better knowledge of these resources than others in the population. Nonetheless, there was generally a low level of knowledge regarding the existing services, while participation in prevention programs was almost non-existent. Some participants mentioned having received information about drugs and their health effects in school, but this was the case only in Slovenia. [109]

Services for curing addiction were mentioned and have been used by some participants, especially by older participants who were selected as drug users. Still, there were multiple barriers preventing young people from seeking help: lack of trust, the attempt to hide that consumption was present in some communities in order to protect the image of the group (Italy), the criteria to enter addiction treatment programs or the scarcity of places (Bulgaria). In other contexts, such as for street children in France or for Slovenian young people, dedicated services were not well known and young people stated that they would use general health services in case they experienced problems (such as the emergency room or the general practitioner). [110]

Outreach and harm reduction services such as providing sterile syringes and needles in Bulgaria or mobile units in Italy were well known within their target group due to their efforts to approach the users in most danger in their own environments. [111]

The action research process of the SRAP project aimed at developing selective prevention strategies while shedding light on the circumstances of drug use and addiction among Roma youth in five countries, taking into consideration the diversity of the populations, as well as their commonalities, such as, for example, living in segregated environments. Given the complexity of the population and the sensitivity of the topic, qualitative research was best suited for supporting in-depth descriptions and comparisons that provided input for a successive action of developing and testing selective prevention strategies and instruments. [112]

In order to achieve these results, there were several methodological challenges that the research team had to face. Firstly, working with a marginalized group often means that the people themselves hold little power of self-representation in the public sphere, making it crucial to pay attention to ethics and involve people in research and decision making regarding their own lives. Secondly, tackling the topic of drug use in an ethnic community with an external mandate might have produced legitimate resistance to a potentially stigmatizing knowledge production process. In this regard, we felt the need to understand how this topic was being perceived and what priority was assigned to it by the people themselves, before entering the field research phase and actually gathering data on the phenomenon. This is why we envisaged an initial consultation phase in which stakeholders—Roma (youth and their parents), social workers and health services representatives—were involved in decisions regarding the project design. Their participation at an early stage had the advantage of establishing an initial partnership, of understanding the importance given to the issue of addiction by the target population and of prioritizing the perceived problems, avoiding the imposition of research topics on the group. We were also warned regarding the possible problems, barriers and ways to tackle them in the data collection process. We saw in this phase that drug use was felt to be a problem both by Roma and by institutional stakeholders and there was a shared willingness to tackle the topic. The design stage was crucial, important decisions were taken here that influenced the rest of the process, such as research questions, target groups and field strategies. [113]

The group that interested us most was formed of young people aged 11 to 25, a range that included people with very different experiences and skills. We chose to use different research instruments with targets of different ages. Among the youngest we knew from the previous stage that there were very few consumers, so we focused on group knowledge and shared representations. We needed a flexible, fun, and visual research instrument, given the different levels of schooling, the sensitivity of the topic and the age of participants. We chose the collage, a visual construction task based on metaphorical thinking, which served as a base for individual elicitation of ideas and group interviewing. Participants associated images referring mostly to medical and bodily consequences of drug consumption (old age, pale skin, loss of weight, death), but they also associated drugs to glamour, celebrities, models and fun. They used many images of sexualized women and eccentric characters, but also talked about psychological implications such as sadness or "going crazy." [114]

Metaphorical analysis was a useful tool that helped us reconstruct meanings, and successively identify recurrent themes and links with other spheres of the participants' lives and identities. Thus we saw how drug consumption is associated to group belonging and identities as it can hold different meanings within the systems of symbols that young people use to position themselves in terms of gender, ethnicity, class and, not least, youth culture. [115]

Group discussions were not the best setting to find out about the actual use of drugs, so the team decided to work with older participants (aged 17 to 25), selected as drug users and interviewed individually with life-story interviews. In this manner we gained an in-depth understanding of their consumption history. This gave us an enormous insight into the actual progress of consumption from onset to addiction and sometimes to rehabilitation, and into the main problems and barriers Roma face in preserving their health. [116]

Although some patterns of drug use by Roma youth did not differ much from those of their counterparts in the general population (normalization of cannabis use, recreational drug use and binge drinking during weekends, for example), poverty, segregation, low access to education, employment and health services still keep them at higher risk and lead to longer-lasting and more serious consequences. [117]

Inter-country diversity was not the only different axis we had to deal with in the research process. The in-depth exploration carried out by means of biographical and visual research allowed us to point out high variability in the patterns of consumption, not only between contexts, but also within the communities included in the research. It allowed us to focus on extreme patterns, which were typical only of some places and showed different characteristics than others, such as heroin consumption by youth in one neighborhood in Bulgaria. [118]

Among the worrying trends identified in the patterns of consumption of Roma young people are the early onset of tobacco use (11-12 years old), exposure to consumption of alcohol by adults (either due to binge drinking or alcoholism), underestimation of the consequences of many drugs (starting with cannabis: widespread and considered to be harmless, also cocaine or amphetamines), presence of addiction to injecting drug use in specific areas. [119]

A perception on rooted gender differences seems to characterize young Roma's experience of the social sphere and of drug use. Different norms apply for boys and girls, the onset, frequency and context of consumption are different between them. Promiscuity is a trait easily associated with girl consumers, who suffer from stigma, and fear rejection or violent reactions from the family and community. Drug use in married couples also shows the same distinctions between gender roles: the wife may pressure the husband to quit, while when both consume, it remains solitary or carried out in groups of the same gender. [120]

The family of origin was largely expected to influence the life of young people, but we saw that relationships, communication and parental control strategies are often not adequate for creating a climate of mutual understanding and help. Fear of parental violence was often quoted by participants, which leads to hiding consumption in order to keep up with the required image of a good Roma, this being so especially for girls. [121]

The role of the peer group in the experiences linked to drug consumption remained central for young Roma, as for adolescents in general. The peer groups can be formed of non-Roma peers, where drug use comes to represent a way of integration into an ethnically mixed group of peers in youth drug cultures. It can be formed, in other cases, of Roma peers and extended family members, such as cousins, who play an important role in mediating knowledge regarding drugs. [122]

These results supported the action-research framework and allowed for the SRAP network to proceed in creating a tailored prevention program, dedicated to these communities. Although many issues and needs for intervention emerged from the research, the SRAP project tackled that of selective prevention and stakeholders' awareness raising. Next to the description of the phenomenon, the research also suggested specific instruments in order to "hook" young Roma and get them in contact with services (outreach teams, cultural mediation) as there is still a large part of the young population not in school or dropping out earlier than their non-Roma counterparts. Moreover, the prevention programs needed to be flexible enough to accommodate different characteristics of the communities and of consumption. Although tobacco, alcohol and cannabis consumption were present and had similar characteristics in all contexts (early onset, relatively accepted socially), these are also shared with some strata of the general population. As for the other drugs, there is a high variability in availability, context and intensity of consumption and, most importantly, in the meanings young people associate with using them. For example, recreational use of magic mushrooms in Spain holds different meanings and dynamics for users than injecting drug use in Bulgaria. Likewise, recreational use of amphetamines and alcohol during parties, which emerged in Bulgaria, is extremely different than heroin addiction in the same geographical context. While the first is associated with the desire to emulate "rich people's" entertainment, while minimizing the consequences, the latter is linked to feelings of hopelessness and a painful awareness of the consequences of this addiction. Prevention programs should draw from local specificities and activate contextual resources, starting with the health services system, and including the family, community and youth groups, in order to obtain long term, sustainable results. [123]

These and other conclusions informed the next phase of pilot testing of two intervention strategies: motivational interviews and life skills education (SRAP, 2013a) and of developing a handbook for prevention actions with Roma youth (SRAP, 2013b), documents and results that were widely disseminated among the stakeholders. The following phases of the project also maintained an action-research orientation by implementing cycles of intervention and data collection in order to test the efficacy of the tools and to continue raising their quality. [124]

We believe that the current research represents a good example of how reflection and practice can work together in order to develop tailored interventions while involving the target group in designing and carrying out the research. Given the initial goal of proposing new intervention programs, it was necessary to gain an in-depth understanding of the history of consumption and the representation of drugs among young people. These two goals were successfully addressed by qualitative, biographical and visual research. Methodologically, we proposed techniques to engage young people and other stakeholders in creative processes, although their involvement as decision-makers was limited to the initial phase of research design. [125]

Our conclusions underline the importance of beginning with the point of view of young people in developing intervention programs and the need to tackle broader, structural issues such as access to services, segregated living, employment, and arrangement between genders. We can see how drug use is not just about individual behavioral patterns, but intertwines with ideas about gender, class and ethnicity. In stigmatized groups, depictions of drug use work either to repudiate other groups or to build more valued social identities. This underlines the importance of working together with the communities in order to avoid further stigmatizing on an ethnic basis and of recognizing and dealing with drug issues in an open and proactive manner. [126]

The current article is based on field research carried out by the consortium of the SRAP project (Addiction Prevention Within Roma and Sinti Communities), with the scientific coordination of Codici | Social Research Agency, which I represented. The project was funded by the Health Program of the European Union between 2010 and 2013.

1) This article is based on research carried out by the consortium of the SRAP project. <back>

2) The study was carried out by asking key informants (drug services providers, researchers, community organizations, teachers, general health and social services, police, government professionals and youth services) in 16 countries (in the European Union and Norway) to respond to a questionnaire on the use of drugs by black and minority groups, on key points for policy making and on further work needed to fill the knowledge gap regarding these groups‘ consumption behaviors. <back>

4) ATOD is an acronym widely used to refer to alcohol, tobacco and other drugs. This division is based on the legal status of these substances. <back>

Andersen, Kelly & Jasper Paas, Leonard (2012). Extremely thin models in print ads: The dark sides. Journal of Marketing Communications, 20(6), 447-464, http://dx.doi.org/10.1080/13527266.2012.723027 [Date of Access: November 3, 2015].

Awan, Fatimah (2007). Young people, identity and the media: A study of conceptions of self-identity among youth in Southern England. PhD Thesis (PhD), Bournemouth University, UK. http://www.artlab.org.uk/fatimah-awan-phd.htm [Date of Access: November 3, 2015].

Bergold, Jarg & Thomas, Stefan (2012). Participatory research methods: A methodological approach in motion. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 13(1), Art. 30, http://nbn-resolving.de/urn:nbn:de:0114-fqs1201302 [Date of Access: November 3, 2015].

Bichi, Rita (2007). La conduzione delle interviste nella ricerca sociale [Conducting interviews in social research]. Roma: Carocci.

Conte, Massimo; Marcu, Oana & Rampini, Andrea (2009). Giovani rom: consumi e strategie di affermazione sociale [Roma youth: Vonsumption and social prestige strategies]. In Luca Massimiliano Visconti & Mario Enzo Napolitano (Eds.), Cross generation marketing (pp.283-302). Milano: Egea.

Commission on Social Determinants of Health [CSDH] (2008). Closing the gap in a generation: Health equity through action on the social determinants of health. Final Report of the Commission on Social Determinants of Health. Geneva, World Health Organization, http://whqlibdoc.who.int/publications/2008/9789241563703_eng.pdf?ua=1 [Date of Access: November 3, 2015]

Daniele, Ulderico (2013). Questo campo fa schifo. Un'etnografia dell'adolescenza rom fra periferie e scenari globali [This slum sucks. An ethnography of Roma adolescence between peripheries and global scenarios]. Roma: Meti.

DeBraganza, Ninoska & Housenblas, Heather (2010). Media exposure of the ideal physique on women's body dissatisfaction and mood. The moderating effects of ethnicity. Journal of Black Studies, 40(4), 700-716.

Dell'Agnese, Elena & Vitale, Tommaso (2007). Rom e sinti, una galassia di minoranze senza territorio [Roma and Sinti. A galaxy of minorities without a territory]. In Gabriella Amiotti & Alessandro Rosina (Eds.), Identità ed integrazione. Passato e presente delle minoranze nell'Europa mediterranea (pp.123-145). Milano: Franco Angeli.

European Monitoring Centre for Drugs and Drug Addiction [EMCDDA] (2002). Update and complete the analysis of drug use, consequences and correlates amongst minorities. Lisbon: EMCDDA, http://www.emcdda.europa.eu/html.cfm/index1661EN.html [Date of Access: November 3, 2015].

European Monitoring Centre for Drugs and Drug Addiction [EMCDDA] (2008). Drugs and vulnerable groups of young people. Luxembourg: Office for Official Publications of the European Communities, http://www.emcdda.europa.eu/publications/selected-issues/vulnerable-young [Date of Access: November 3, 2015].

Földes, Maria Eva & Covaci, Alina (2011). Research on Roma health and access to healthcare: State of the art and future challenges. International Journal of Public Health, 57, 37-39, http://link.springer.com/article/10.1007%2Fs00038-011-0312-2 [Date of Access: November 3, 2015].

Fundacion Secretariado Gitano (2009). Health and the Roma community, analysis of the situation in Europe: Bulgaria, Czech Republic, Greece, Portugal, Romania, Slovakia, Spain. Madrid: Fundacion Secretariado Gitano, http://ec.europa.eu/justice/discrimination/files/roma_health_en.pdf [Date of Access: November 3, 2015].

Gauntlett, David (2007). Creative explorations. New approaches to identities and audiences. Oxford: Routledge.

Gerevich, Josef; Bacskai, Erika; Czobor, Pal & Szabo, Jozsef (2010). Substance use in Roma and non-Roma adolescents. The Journal of Nervous and Mental Disease, 198(6), 432-436.

Gobo, Giampietro (2013). Descrivere il mondo. Teoria e pratica del metodo etnografico in sociologia [Describing the world. Theory and practice of the ethnographic method in sociology]. Roma: Carocci.

Grady, John (2008). Visual research at the crossroads. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 9(3), Art. 38, http://nbn-resolving.de/urn:nbn:de:0114-fqs0803384 [Date of Access: November 3, 2015].

Harper, Douglas (2000). Reimagining visual methods: Galileo to neuromancer. In Norman K. Denzin & Yvonna S. Lincoln (Eds.), The handbook of qualitative research (2nd ed., pp.717-733). Thousand Oaks, CA: Sage.

Ivanov, Ivan (2004). Reflections on the access of Roma to health care. European Roma Rights Centre, http://www.errc.org/cikk.php?cikk=2067 [Date of Access: November 3, 2015].

Kósa, Zsigmond; Széles, György; Kardos, László; Kósa, Karolina; Németh, Renáta; Országh, Sándor; Fésüs, Gabriella; McKee, Martin; Ádány, Róza & Vokó, Zoltán.(2007). A comparative health survey of the inhabitants of Roma settlements in Hungary. American Journal of Public Health, 97(5), 853-859.

Kyrölä, Katariina (2014). The weight of images: Affect, body image and fat in the media. Burlington, VT: Ashgate.

Lakoff, George & Johnson, Mark (1980). Metaphors we live by. Chicago, IL: University of Chicago Press.

Langer, Susanne (1957). Philosophy in a new key. A study in the symbolism of reason, rite, and art (3rd. ed.). Cambridge, MA: Harvard University Press.

MacGregor, Brent & Morrison. David E. (1995). From focus groups to editing groups. A new method of reception analysis. Media, Culture and Society, 17(1), 141-150.

Manzoni, Chiara (2013). L'uscita dal campo e dalla baraccopoli: le carriere abitative dei rom, tra vincoli strutturali e strategie individuali [Exiting the camps and the slums: housing pathways of Roma between structural constraints and individual strategies]. PhD Thesis (PhD), Università degli Studi di Milano Bicocca, Italy.

Marcu, Oana (2014). Malizie di strada. Una ricerca-azione con giovani rom romeni [Street flirts. An action research with Romanian Roma youth]. Milan: Franco Angeli.

Monasta, Lorenzo (2004). The health of foreign Romani children in Italy: Results of a study in five camps of Roma from Macedonia and Kosovo. Roma Rights Quarterly, 3-4, 46-55.

Nagasawa, Richard; Qian, Zhenchao & Wong, Paul (2001). Theory of segmented assimilation and the adoption of marijuana use and delinquent behavior by Asian Pacific youth. The Sociological Quarterly, 42(3), 351-372.

Parker, Howard; Williams, Lisa & Aldridge, Judith (2007). The normalisation of "sensible" recreational drug use: Further evidence from the North West longitudinal study. In Paul Manning (Ed.), Drugs and popular culture. Drugs, media and identity in contemporary society (pp.74-97). Cullompton: Willan Publishing.

Portes, Alejandro & Zhou, Min (1993). The new second generation: Segmented assimilation and its variants. Annals of the American Academy of Political and Social Science, 530, 74-96.

Reavey, Paula (2011). Visual methods in psychology. Using and interpreting images in qualitative research. Hove: Psychology Press.

Sanchez-Hucles, Janis; Hudgins, Patrick & Gamble, Kimberly (2005). Reflection and distortion: Women of color in magazine advertisements. In Ellen Cole & Jessica Henderson Daniel (Eds.), Featuring females. Feminist analyses of media (pp.185-197). Washington, DC: American Psychological Association.

Tesăr, Cătălina (2012). Becoming Rom (male), becoming Romni (female) among Romanian cortorari Roma: On body and gender. Romani Studies, 2(22), 113-140.

Vokó, Zoltan; Csépe, Peter; Németh, Renáta; Kósa, Karolina; Kósa, Zsigmond; Széles, György & Ádány, Róza (2009), Does socioeconomic status fully mediate the effect of ethnicity on the health of Roma people in Hungary? Journal of Epidemiological Community Health, 63, 455-460.

Ward, Monique & Harrison, Kristen (2005). The impact of media use on girls' beliefs about gender roles, their bodies, and sexual relationships: A research synthesis. In Ellen Cole & Jessica Henderson Daniel (Eds.), Featuring females. Feminist analyses of media (pp.3-23). Washington, DC: American Psychological Association.

Wilson, William Julius (1987). The truly disadvantaged: The inner city, the underclass, and public policy. Chicago, IL: University of Chicago Press.

Oana MARCU, Ph.D., is lecturer at the Faculty of Social and Political Sciences at the Catholic University of Milan, working in the field of social research methodology. Her main interests are action research and qualitative methods, in particular ethnography and visual methods. The main topics she works on are Roma youth, migration and intersectionality. She teaches social research methodology and is an associate of Codici Social Research Agency, with whom she has conducted various action-research projects with Roma and Sinti.

Sociology Department
Università Cattolica di Milano
L.go Gemelli n. 1 20123, Milan, Italy

Marcu, Oana (2016). Using Participatory, Visual and Biographical Methods with Roma Youth [126 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 17(1), Art. 5,
http://nbn-resolving.de/urn:nbn:de:0114-fqs160155.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Creative Commons Attribution 4.0 International License