Gap Analysis of Research Literature on Issues Related to Street-Involved Youth

4. A Vulnerable Population

4. A Vulnerable Population

This survey of antecedent family background, and the interrelatedness of the effects of poverty, health and types of victimization among street-involved youth clearly shows how these factors intersect to make street-involved youth vulnerable to serious violence. The potential and actual violence emanating from life on the street is all encompassing. Street-involved youth frequently find themselves in extremely risky situations as they try to find ways to survive, eat, sustain their substance abuse, and sleep in a safe place. Their situation can be complicated and exacerbated by factors such as race, gender, and sexual orientation. They become, as Brannigan and Caputo (1993:96) remark, subjects of research interest and of control. This control is seen as crucial because the general public comes to view their disheveled and unkempt appearance and their unpredictable behavior as dangerous, as potentially violent. At the same time, because they are young people outside the reach of caring adults, they are decidedly vulnerable to the violence of the streets where they live.

And above all, speaking as the authors of this review, it must be remembered that these are youth, which by definition means individuals who have not reached adult levels of cognitive development or mature levels of judgement. They may want to be "grown-ups," and they certainly face challenges that the majority of adults have never faced. However, street-involved youth are by virtue of their age particularly ill-equipped to overcome the stark nature of their present life – while carrying the burdens that most research shows they carry from the untenable home life from which they are fleeing.

4.1 Research Methodologies

This section will briefly describe the key research methodologies using some of the sources included in the annotated section of the bibliography. A number of the researchers made significant efforts to include the direct reporting by street youth of their experiences and views of what kinds of programming would be of use to them for leaving the street. The studies tended to focus on attempting to develop definitions of street youth and typologies of the youth themselves and the kinds of circumstances and activities that made up their day-to-day lives.

Of particular interest is Brannigan and Caputo's (1993) research design for the Runaways and Street Youth Project conducted in Ottawa and Saskatoon by Caputo, Weiler, and Kelly (1994b, 1994c). The research design was the product of attempts to conceptualize street involved youth in such away that the multidimensional nature of this "varied and highly mobile" (Brannigan and Caputo 1993:3) population could be reflected with clarity. Characteristics of the target population meant that a classification scheme could not be developed with mutually exclusive conceptual categories. For example, young people of different ages might be involved in overlapping or "…very different types of behaviour" (p.3). Even making a decision about what "…age range to include in a viable definition of runaways and street youth" (p.5) was problematic, since the literature on street-involved youth has included youngsters from their pre-teens to individuals in their late 20s and even early 30s. The problem of age range is compounded by variable legal definitions of youth used by various community, provincial and federal jurisdictions.

Brannigan and Caputo (p.6) note two other problems in research design and methodology. First, returning to the challenges of the age range of youth, comparisons of the actions of two individuals as much as eight years apart in age presents substantial difficulties for analysis. They (p.6) ask, "how can questions of responsibility, choice or intent be decided for individuals in such very different stages of their lives and with such different amounts of power and resources at their disposal?"

The second problem relates to the "fluid and mobile nature of the street population [which] makes estimates of its size and composition virtually impossible" (p.6).

Without a single, accepted and appropriate definition of runaways and street youth, the authors (1993:53, Fig.1) attempted to construct a "schematic overview" that took into account the relationship among types of street youth, antecedents, consequences and institutional responses and services. The authors (p.54) point out that "…this schematic overview can do little more than highlight the complexity of the problem of …runaways and street youth."

One of Brannigan and Caputo's (1993:103, Fig.2) schematic models is utilized in Caputo, Weiler, and Kelly's (1994c) analysis of "…[Saskatoon's] response to runaways and street youth." The study uses the two major, intersecting dimensions of this schematic model to characterize four groups of youth according to, first, the amount of time that young people spend on the street, from only occasionally to most of their time, and second, degree of involvement in conventional activities or in dangerous and risky street activities. The primary methodology for this study involved interviews with staff in a variety of youth service agencies, with a wide range of young people either associated with the street or in high school, and with representatives of community groups. The interviews addressed the characteristics of the street-involved population, the operation of various aspects of the social system that involved interaction with street youth, and the operations of the youth service system as a whole (Caputo, Weiler, and Kelly 1994c:1).

Community and youth participation were essential components of the project, and community groups were consulted about problems facing runaways and street youth. Notable features of the methodology included, first, a youth conference where young people from various parts of the community came together to consider and discuss issues arising from the community consultation, and second, a "delphi conference" (p.10), involving representatives from agency and community groups and young people who met to discuss the results of the research project and to plan future strategies.

The research team associated with the Runaways and Street Youth Project used an ethnographic approach to interviewing, and this same qualitative approach is used by Sobo et al. (1997) in their study of AIDS misconceptions among runaway adolescents in the United States. A similar, qualitative "phenomenological approach" is used by Kurtz et al. (2000) and Lindsey et al. (2000) in two studies designed to investigate the external and internal resources that runaway and homeless youth utilized "…to make successful transitions to adulthood." Thus, semi-structured, face-to-face interviews and self-administered questionnaires are used to elicit information, personal experiences, attitudes and ideas from the respondents. The research team also used focus groups with peer educators and social service providers to obtain data.

Self-reporting instruments were also employed by McCarthy and Hagan (1992) to examine the extent to which "foreground causes" such as unemployment, hunger and food scarcity, and shelter have direct impact on criminal activities such as stealing food, serious theft, and prostitution. They employed bivariate statistical techniques to test the strength of relationships between pairs of variables.

It is worth noting that McCarthy and Hagan (1992), Moon et al. (2000), and Greene and Ringwalt (1998), among others, used food, food vouchers, restaurant coupons, and cash to pay respondents for interviews.

More formally structured interviews were used by Baron (1999) and Baron and Hagan (1998) in their studies of street youth in Edmonton. The interviews were administered to samples of male homeless street youth. Both studies employed a variety of criminological perspectives to examine the intersection of street subculture with substance use, in the first case, and criminal activity and violence, in the second. Both analyses employed multivariate techniques.

A very large project in the United States (Greene and Ringwalt 1998) drew upon three nationally representative surveys of female street youth to compare pregnancy rates. A "purposive sampling strategy" (p.372) was used. This entailed selection of sites in 10 American cities where high concentrations of street youth were expected. Staff of local street outreach programs and local police departments aided in the identification of locations and selection of the most appropriate times to contact street youth, but none of the surveys included data collection from the outreach staff or police.

There are no "best practices" that can be identified in the various methodologies reviewed for this study. There is a strong theme of respect for and involvement of the youth themselves, with which we as researchers can only agree. This approach is compatible with – and may to a great degree rely upon – an ethnographic, or at least highly qualitative, research strategy. This allows for the youth's own experience to come to the fore, for them to define their fears and their strengths, the risks they undergo, the decisions they make, and their successes in overcoming obstacles. An ethnographic, highly consultative approach does not, however, preclude quantitative data and statistical analysis – if on a modest scale. That is, if a large enough number of youths can be interviewed, if enough detail of their background and current activities can be elicited, and if the research coverage is systematic and comprehensive in scope, then there are enough "cases" to allow for at least univariate and bivariate analysis. For example, length of time on the street can be compared with extent of use of drugs, or various means of getting money[3].


[3] We cannot refrain from mentioning a study done by one of the authors which was able to cross-tabulate such data as length of time on the street and use of drugs. It is too early for inclusion in this review, but the methodological approach was rewarding. (Stephenson, Study of Vancouver Street Youth and the Reconnect Program, 1989)

Date modified: